Medicaid and Medicare clinical and economic research | 3M US
Geographic Variation in Hospital Admission Rates in the Medicare Population 3M Clinical and Economic Research Richard F. Averill, MS Richard L. Fuller, MS Ronald E. Mills, PhD January 2021 Table of Contents Executive Summary....................................................................................................................2 Introduction ................................................................................................................................ 3 Potentially Preventable Admissions (PPAs) ..............................................................................3 Risk Adjusting PPAs ...................................................................................................................5 Interrelationship of PPAs with Other Quality Performance Measures .....................................5 Risk Adjusting PPRs and Hospital Admissions from the ED ......................................................6 National and Best Practice Norms .............................................................................................7 PPA Financial Impact .................................................................................................................8 Data ............................................................................................................................................. 9 PPA Results by Risk Categories .................................................................................................9 PPA Results by Geographic Region ........................................................................................ 13 PPA Frequency ........................................................................................................................ 16 Overlap Between PPAs, PPRs and ED Admits......................................................................... 17 Summary and Conclusions...................................................................................................... 18 References............................................................................................................................... 19 Appendix A: Bibliography of Articles and Reports on PPAs, CRGs, PPRs, APR DRGs........... 20 Appendix B: Potentially Preventable Admissions (PPAs)....................................................... 40 Appendix C: Description of CRG Logic .................................................................................. 48 Appendix D: Overlap among PPAs, PPRs and ED Admits....................................................... 50 Appendix E: PPA %(A-E)/E and $(A-E) by CBSA ..................................................................... 51 Geographic Variation in Hospital Admission Rates in the Medicare Population 1 Executive Summary Health care expenditures continue to steadily increase with hospital stays making up about onethird of health care expenditures. A well-functioning delivery system within a managed care plan or a geographic region should be able to minimize the need for hospitalizations. In this 3M Clinical and Economic Research report, the Potentially Preventable Admissions (PPAs) methodology was used to identify hospital admissions that may be potentially preventable. If there are an excess number of PPAs compared to a national norm within a managed care plan or geographic region, it is likely the excess PPAs represent hospital admissions that could be avoided if the delivery system functioned effectively. The study used a random data sample of five percent of Medicare fee-for-service (FFS) beneficiaries contained in the Medicare Standard Analytic Files for calendar year 2017 and 2018. The data from 2017 was used to determine the burden of chronic disease for each beneficiary and to risk adjust PPA rates in the 2018 data. After excluding FFS beneficiaries who were not enrolled in part A and B for the full three-year period, 1,388,114 beneficiaries remained in the analysis database. These beneficiaries experienced 379,814 hospital admissions of which 85,974 were considered a PPA (22.6 percent of admissions). Extrapolated to the entire Medicare population, the 85,974 PPAs represent $33.3 billion in annual FFS Medicare expenditures. Based on a risk-adjusted national norm, the analysis found considerable PPA performance variation across census regions, states and Core Based Statistical Areas (CBSAs). Across states, PPA performance compared to the risk-adjusted national norm varied from 32.66 percent below expected for Hawaii to 47.62 percent above expected for Arkansas. A best practice PPA norm was determined using 40 percent of the CBSAs with the best PPA performance that had at least 1,500 beneficiaries in the analysis data. To achieve PPA best practice performance nationally, overall PPA performance would need to improve by 14.85 percent which would result in an annual reduction in Medicare expenditures of $4.3 billion (12.9 percent of the $33.3 billion in PPA expenditures). If there are an excess number of Potentially Preventable Admissions (PPAs) compared to a national norm within a managed care plan or geographic region, it is likely the excess PPAs represent hospital admissions that could be avoided if the delivery system functioned effectively. 9.0 percent of the PPAs were followed by one or more potentially preventable readmissions and 30.5 percent of PPAs were low severity medical admission through the emergency department (ED). Across states, PPA performance was correlated with readmission performance (r=0.694) and ED admission performance (r=0.566). The interdependence of PPAs with potentially preventable readmission and low severity medical admissions through the ED can provide useful insights for targeted quality improvement initiatives aimed at reducing PPAs. Geographic Variation in Hospital Admission Rates in the Medicare Population 2 PPA performance can be an effective measure of delivery system performance within a managed care plan or geographic region. The extent of PPA performance variation indicates that there are PPA performance improvement opportunities in many geographic areas. The $4.3 billion annual Medicare expenditure reduction gained through PPA best practice provides an achievable PPA quality improvement target. Introduction Health care expenditures represent about 18 percent of the US gross domestic product and are steadily increasing. Hospital stays make up about one-third of healthcare expenditures.1 To the extent that hospital care can be shifted to the outpatient setting or avoided altogether, the cost of health care can be reduced.2 Studies have documented not only that preventable hospitalizations exist, but that they can be reduced by specific interventions. For example, guidelines implemented in nursing homes have been shown to decrease the rate of hospital admissions.3 Patients with chronic obstructive pulmonary disease (COPD) who were provided a higher level of continuity of care have been shown to have a significantly lower likelihood of avoidable hospitalizations.4 The objective of this report is to determine for Medicare beneficiaries the extent of geographic variation in the rate of hospital admissions that are potentially preventable and to quantify the financial impact of excess potentially preventable hospital admissions. Potentially Preventable Admissions (PPAs) A well-functioning delivery system within a managed care plan or geographic region should be able to minimize the need for hospitalizations. Potentially Preventable Admissions (PPAs) are hospital admissions that can often be avoided. The occurrence of an excess number of PPAs is indicative of an ineffective delivery system. Of course, not every PPA can be prevented. But if there are an excess number of PPAs compared to national benchmarks within a managed care plan or geographic region, it is likely that the excess PPAs represent hospital admissions that could be avoided if the delivery system functioned more effectively. There are six broad categories of PPAs: · Admissions for chronic disease management that could potentially have been managed in the outpatient setting (e.g., asthma) · Admissions for acute diseases that could potentially have been managed in the outpatient setting (e.g., viral pneumonia) · Admissions for a procedure that could be done in an outpatient setting (e.g., cardiac catheterization for non-acute disease such as atherosclerosis) · Admissions for a procedure for which there is a less invasive alternative procedure (e.g., percutaneous coronary angioplasty with a stent instead of coronary bypass surgery5) · Admissions for a procedure that research has shown to be prone to overuse (e.g., spinal procedures for back pain6) · Admissions that could potentially have been avoided for residents of a residential care facility such as a skilled nursing facility (e.g., trauma due to a fall) The most prevalent PPAs will be for medical management of chronic and acute diseases. These hospital admissions may result from hospital or ambulatory care inefficiency, lack of adequate access to outpatient care, or inadequate coordination of ambulatory care services. In many cases, PPAs are for flare-ups of chronic conditions (e.g., heart failure) for which adequate monitoring and Geographic Variation in Hospital Admission Rates in the Medicare Population 3 follow-up, such as proper medication management, could have avoided the need for hospitalization. As such, the occurrence of high rates of PPAs within a managed care plan or geographic region may represent a failure of the ambulatory care delivery system. The most prevalent PPAs will be for medical management of chronic and acute diseases. These hospital admissions may result from hospital or ambulatory care inefficiency, lack of adequate access to outpatient care, or inadequate coordination of ambulatory care services. The PPAs associated with medical management of chronic and acute diseases are more comprehensive than the U.S. Department of Health and Human Services (HHS) Agency for Healthcare Research and Quality (AHRQ) list of ambulatory care sensitive conditions initially defined in the 1980s, and more comprehensive than the list of AHRQ Preventable Quality Indicators (PQIs).7 PPAs focus on potentially preventable hospital admissions and exclude admissions that are not considered preventable. The PPA methodology takes three factors into consideration when assessing the potential preventability of PPAs for medical management of chronic and acute disease: the length of the required care coordination time period, acuteness of the reason for admission and living arrangement at the time of admission. Length of the required care coordination time period Potential preventability is assessed relative to the care given in the immediate period preceding a hospital admission (months). Conditions that require an extended period of coordinated and integrated care are not considered potentially preventable. For example, an admission for chronic renal failure is not considered a PPA because it is not preventable unless appropriate care has been given for several years before the admission making it difficult to judge potential preventability based solely on the care given in the immediate period preceding the admission. Acuteness of the reason for admission Preventability is also assessed based on the relative acuteness of the reason for the admission. For example, an admission for a cardiac catheterization is considered potentially preventable for patients with a diagnosis of coronary atherosclerosis, but not preventable for patients with an acute myocardial infarction or unstable angina. The rate of PPAs is risk adjusted for the complexity of the patient population whereas the AHRQ PQIs do not include any risk adjustment. For example, the AHRQ PQIs include all patients admitted with diabetes irrespective of the severity of the patient. A diabetic who is diet controlled has a different probability of hospital admission as compared to a diabetic patient who is on dialysis, thereby necessitating that any comparison of admission rates be risk adjusted. Living arrangement at the time of admission Medicare beneficiaries living in residential care facilities such as a skilled nursing facility (SNF) or nursing home generally are expected to be receiving a higher level of coordinated care than beneficiaries living at home. Many conditions such as fever, urinary tract infections, metabolic disturbances and pneumonia can often be managed in a residential care facility, thereby avoiding the need for hospitalization. Other conditions such as diseases of the skin and injuries due to falls should be more readily avoided in a residential care facility setting. In determining whether an Geographic Variation in Hospital Admission Rates in the Medicare Population 4 admission is potentially preventable, PPAs apply a broader list of conditions that are considered potentially preventable when a beneficiary is living in a residential care facility. Appendix A contains PPA research articles and studies using PPAs and Appendix B contains a more detailed description of the PPA methodology. By assessing potential preventability based on the length of the required care coordination time period, acuteness of the reason for admission and living arrangement at the time of admission and by risk adjusting PPA rates, the hospital admissions included in the PPAs can be more comprehensive than the AHRQ ambulatory care sensitive conditions and the AHRQ PQIs. A comprehensive evaluation of potentially preventable admissions can provide a more complete assessment of the continuity of care and of the functioning of the health care delivery system within a managed care plan or geographic region. Risk Adjusting PPAs Clinical Risk Groups (CRGs) are a categorical clinical model that uses historical claims data to assign beneficiaries to a single mutually exclusive category that defines a beneficiary's chronic disease burden.8 The CRGs (Version 2.1) are composed of 332 base CRGs that describe the beneficiary's most significant chronic conditions and explicit severity levels that distinguish differences in disease burden due to severity of illness resulting in 1,414 individual CRGs. The individual CRGs are aggregated into nine health statuses ranging from catastrophic to healthy. Status 1 Healthy Status 2 History of Acute Disease e.g., Chest Pain Status 3 Single Minor Chronic Disease e.g., Migraine Status 4 Minor Chronic Disease in Multiple Organ Systems e.g., Migraine and BPH Status 5 Single Dominant or Moderate Chronic Disease e.g., CHF Status 6 - Dominant or Moderate Chronic Disease in Multiple Organ Systems, e.g., CHF, COPD Status 7 - Dominant Chronic Disease in Three or More Organ Systems, e.g., CHF, COPD, DM Status 8 - Malignancy, Under Active Treatment, e.g., Lung Cancer Status 9 - Catastrophic Conditions, e.g., Major Organ Transplant Based on the severity levels of the chronic conditions that comprise each status, beneficiaries in the nine statuses are assigned a severity level between one and six resulting in 53 aggregated CRG risk categories. Six of the aggregated CRGs in statuses 1 and 2 relate to pregnancy and delivery. Because this report analyzed Medicare data, the pregnancy and delivery CRGs were very low volume and were excluded from the analysis, resulting in the 47 CRG risk categories that were utilized to risk adjust the PPAs. The CRGs are a transparent system with a definition manual available for inspection. Appendix A contains CRG research articles and studies using CRGs and Appendix C contains a more detailed description of the CRG methodology. Interrelationship of PPAs with Other Quality Performance Measures PPAs represent an evaluation of hospital admitting performance within a population and reflect the impact of adequate access to ambulatory care and/or the adequate coordination of ambulatory care services. Readmissions and hospital admissions through the ED primarily reflect the performance of hospitals and have a direct impact on PPA performance. While, in general, Geographic Variation in Hospital Admission Rates in the Medicare Population 5 managed care plans primarily focus on and are measured on population management performance, they are highly dependent on hospital performance to achieve better overall population PPA performance. Because of this interdependence, managed care plans will often provide incentive plans to hospitals to improve hospital admission performance. Managed care plans must understand and quantify the impact of hospital performance on population performance to develop an effective incentive plan for hospitals. The interrelationship between PPAs and hospital readmissions and admissions through the ED was examined using the following two performance measures. Potentially Preventable Readmissions (PPRs) Potentially Preventable Readmissions (PPRs) are return hospitalizations within 30 days following a prior hospitalization.9 PPRs may result from deficiencies in the process of care (readmission for a surgical wound infection) or inadequate post-discharge follow-up (prescription not filled) rather than unrelated events that occur post discharge (broken leg due to trauma). Readmissions may result from actions taken or omitted during the initial hospital stay, such as incomplete treatment or poor care of the underlying problem, or from poor coordination of services at the time of discharge and afterwards, such as incomplete discharge planning or inadequate access to care. The admissions considered at risk for a PPR and the clinical circumstances under which a subsequent readmission is considered potentially preventable are specified in the PPR methodology logic. The PPR designation is assigned to any admission that was followed by one or more potentially preventable readmissions during the 30 days following a hospital discharge. Appendix A contains PPR research articles and studies using PPRs. Hospital Admissions from the ED The ED Admit measure identifies hospital admissions that are a low-severity medical admission from the ED. Patients that died, those admitted for surgical procedures, those admitted for conditions that are inherently high risk (e.g., AMI), those at high severity and those covered by medical necessity considerations (e.g., behavioral health) are excluded from the ED Admit measure. High severity is defined using the APR DRG methodology as discussed below. The ED visits that are not excluded are the at-risk population for the ED Admit measure. For the at-risk ED visits, the ED Admit rate is the sum of ED visits that were admitted divided by the sum of ED visits that were admitted plus the ED visits that were not admitted. Risk Adjusting PPRs and Hospital Admissions from the ED All Patient Refined Diagnosis Related Groups (APR DRGs) are a categorical clinical model composed of base categories (base APR DRGs) that are subdivided into four severity of illness subclasses.10 These subclasses are unique to each base APR DRG and are based on the extent of physiologic decompensation or organ system loss of function. The four severity of illness subclasses are numbered sequentially from 1 to 4 indicating respectively, minor, moderate, major, and extreme severity of illness. The combination of the base APR DRGs and the four severity of illness subclasses constitute a system of patient risk classes. The APR DRG based risk classes are exhaustive and mutually exclusive resulting in a patient being assigned to one and only one risk class. APR DRG assignment is computed both at the time of admission and at the time of discharge. At the time of admission, the Present on Admission (POA) indicator field on the UB04 is used to exclude complications and other conditions that arise after admission from the APR DRG assignment. The POA indicator became a required reporting element for both principal and Geographic Variation in Hospital Admission Rates in the Medicare Population 6 secondary diagnoses on hospital claims submitted for payment after October 1, 2007.11 The APR DRGs and severity of illness subclasses are used for performance reporting in five U.S. states and as the basis of payment adjustments in 30 states. The APR DRG methodology is a transparent system with a full definition manual. For the ED Admit measure, all hospital admissions from the ED are assigned to an APR DRG. The APR DRG is used to identify the at-risk ED visits (e.g., exclude surgical admissions). Admissions assigned to APR DRG severity of illness level 3 and 4 are considered high severity and are excluded from the ED Admit measure. Any comparison of PPR and ED Admit rates requires that the rates be risk adjusted. The APR DRGs are used to risk adjust PPR rates and ED Admit rates. PPRs use the APR DRG assigned at the time of discharge from the hospital admission that initiated the subsequent readmission. For ED Admit, the APR DRG assigned at the time of hospital admission from the ED is used. Overlap Between PPAs, PPRs and ED Admit An admission can simultaneously be a PPA, PPR and an ED Admit. If an admission is both a PPA and a PPR (i.e., a PPA that initiates a sequence of one or more readmissions), the admission is still considered a PPA because it is likely associated with a lack of adequate access to ambulatory care and/or the adequate coordination of ambulatory care services and is a population performance issue. However, the subsequent readmissions that follow the PPA/PPR are not eligible to be a PPA because those readmissions are more likely to be associated with the care and follow-up provided by the hospital and therefore reflect a hospital performance issue. Identification of a PPA as an ED Admit does not affect the PPA or PPR assignment, but the overlap does provide useful insight into the source of some PPAs. National and Best Practice Norms Each Medicare beneficiary is assigned to a CRG risk class based on their disease burden, which is determined from claims history data for the year preceding the year in which PPAs are assigned, as illustrated in Figure 1. Figure 1: CRG and PPA assignment periods Assign CRGs Base Year Assign PPAs Evaluation Year Within each CRG risk class a PPA relative weight is computed that reflects the PPA rate (frequency) and the case mix (relative costliness) of the PPAs being admitted. Thus, a higher weight for a CRG risk class can be the result of a high rate of occurrence of PPAs or that the mix of PPAs being admitted is of higher severity and therefore more costly. To determine the severity of the mix of PPAs within a CRG risk class, each PPA is assigned to an APR DRG and the standard APR DRG relative resource weights are used to measure the case mix (relative costliness) of the PPAs in the CRG risk class. Geographic Variation in Hospital Admission Rates in the Medicare Population 7 National Norm A national norm is calculated by summing the APR DRG relative resource weights for all PPAs identified in the evaluation year within a CRG risk category--and across all beneficiaries assigned to the CRG risk category for the base year--and computing the mean value per beneficiary (referred to as the PPA national norm value). The end result is that each CRG risk class is assigned a PPA relative weight that can be used to compute expected PPA performance. The expected PPA value (E) for any subset of beneficiaries is the number of beneficiaries in each CRG risk category times the PPA norm value for the CRG risk category and summed overall CRG risk categories (indirect rate standardization). For any subset of beneficiaries, such as beneficiaries in a specific geographic region, the PPA actual value in a CRG risk category is computed by summing all the APR DRG relative resource weights of the PPAs for beneficiaries assigned to the CRG risk category. By summing all the PPA relative resource weights across all beneficiaries across all CRG risk categories, the actual value (A) is determined. The actual value (A) represents good performance if (A-E) is negative (A<E) and poor performance if (A-E) is positive (A>E). The %(A-E)/E is the percent that actual performance is below expected (%(A-E)/E is negative) or above expected (%(A-E)/E is positive). Best Practice Norm In addition to the national PPA norm, this report also determined a "best practice" norm. Using the metropolitan areas identified in the Core Based Statistical Areas (CBSAs) from the Office of Management and Budget, PPA performance across metropolitan areas was examined. Using the national norm, the (A/E) for each CBSA with at least 1,500 beneficiaries is used to determine the subset of CBSAs with the best PPA performance and that constitutes 40 percent of the beneficiaries in the Medicare FFS population sample included in the analysis. This subset of CBSAs is referred to as the PPA best practice CBSAs. For the PPA best practice CBSAs, the overall A/E is computed. The A/E ratio for the PPA best practice CBSAs is less than one and is a measure of the level of relative performance achieved by PPA best practice hospitals. For example, an A/E ratio of 0.8 for the PPA best practice CBSAs means that in these CBSAs, the PPA performance is 20 percent (1 - 0.8) lower than would be expected compared to all CBSAs. The value of the PPA relative weight in each CRG risk category in the PPA national norm is multiplied by the A/E ratio for the PPA best practice CBSAs to create a PPA best practice norm. Rather than selecting an arbitrary performance percentile as a best practice norm, using a PPA best practice norm created in this way represents a performance level that is actually being achieved in a substantial number of geographic areas and represents an achievable performance improvement level. PPA Financial Impact A PPA financial conversion factor is computed based on allowed Medicare payments (the amount actually paid by Medicare). The financial conversion factor is used to express PPA actual performance (A) and PPA expected performance (E) in financial terms so that the financial impact of a PPA performance difference (A-E) can be determined. By comparing the financial impact of PPAs at the level of each clinically meaningful CRG risk category, the clinical and financial aspects of care are linked, which can facilitate behavior change and performance improvement initiatives. The Medicare savings estimated in this report is conservative because it is based solely on the (AE) difference. Thus, the underlying rate of PPAs as measured by E is accepted as a baseline level of underlying quality performance and only the PPA (A-E) difference is viewed as the basis for potential savings. The magnitude of the PPA (A-E) differences is directly related to the level of variation in PPAs across geographic regions. The greater the variation in PPAs across geographic Geographic Variation in Hospital Admission Rates in the Medicare Population 8 regions, the greater the opportunity for savings. Thus, if there is little variation in PPAs across geographic regions, this analysis will conclude there is little opportunity for improvement and savings, essentially accepting the status quo as an acceptable level of performance. Data The study used data in the Medicare Standard Analytic Files (Limited Data Set (LDS)) for calendar year 2017 and 2018. The LDS files contain 100 percent of Medicare fee-for-service (FFS) claims data for inpatient, outpatient, skilled nursing facilities and home health agencies. The LDS carrier file contains Medicare FFS claims data for professional providers, including physicians, physician assistants, clinical social workers, and nurse practitioners for a random sample of five percent of Medicare beneficiaries. The LDS Master Beneficiary Summary File (MBSF) contains enrollment data on all Medicare beneficiaries enrolled in or entitled to Medicare within a given calendar year. By comparing the financial impact of PPAs at the level of each clinically meaningful CRG risk category, the clinical and financial aspects of care are linked, which can facilitate behavior change and performance improvement initiatives. To identify the burden of chronic disease and to assign CRGs, it was necessary to build a complete longitudinal record of all FFS claims for each Medicare beneficiary. Because the LDS carrier file was limited to a five percent sample of Medicare beneficiaries, the data in this study was limited to the beneficiaries in the LDS carrier file. The carrier file is a sample across all types of beneficiaries including beneficiaries in Medicare Advantage plans. To create a sample of FFS beneficiaries, the following edits were applied: · Exclude beneficiaries who were not enrolled in both Part A and B for the full two-year time period (i.e., newly enrolled, disenrolled or reported died) · Exclude beneficiaries who were enrolled in a managed care plan for one or more months · Exclude beneficiaries who were enrolled in hospice Calendar year 2017 was used to assign the CRG to each beneficiary and calendar year 2018 was used to assign the PPAs to each beneficiary. After these exclusions were applied, there were 1,388,114 beneficiaries in the analysis data. PPA Results by Risk Categories For the 1,388,114 beneficiaries, there were 379,841 hospital admissions of which 85,974 were a PPA (22.6 percent of admissions). Based on each beneficiary's claim history from 2017, beneficiaries were assigned to one of 47 CRG risk categories. Beneficiaries in each CRG risk category who had a PPA were identified using the 2018 data. Beneficiaries assigned to CRG status 3-9 all had at least one chronic disease. Table 1 contains summary data by CRG risk category for the beneficiaries with at least one chronic disease. Geographic Variation in Hospital Admission Rates in the Medicare Population 9 Table 1: PPA data by CRG risk category for beneficiaries with at least one chronic disease CRG Status 3 Single Minor Beneficiaries Chronic Disease Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight 4 Minor Chronic Beneficiaries Disease in Admissions Multiple Organ PPAs Systems PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight 5 Single Beneficiaries Dominant or Admissions Moderate PPAs Chronic PPAs/1,000 Disease Days/PPA PPA APR CMI PPA Weight PPA $ Weight 6 Dominant or Beneficiaries Moderate Admissions Chronic PPAs Disease in PPAs/1,000 Multiple Organ Days/PPA Systems PPA APR CMI PPA Weight PPA $ Weight 7 Dominant Beneficiaries Chronic Admissions Disease in PPAs Three or More PPAs/1,000 Organ Systems Days/PPA PPA APR CMI PPA Weight PPA $ Weight 8 Malignancy Beneficiaries under Active Admissions Treatment PPAs PPAs/1,000 Days/PPA 9 Catastrophic Conditions PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight 1 65,271 5,328 747 11.44 3.74 1.3272 0.0165 201.23 29,906 2,416 311 10.40 3.43 1.4411 0.0172 209.77 188,238 23,015 3,628 19.27 3.65 1.2144 0.0259 315.88 131,904 20,941 4,066 30.83 3.78 1.3460 0.0456 556.14 27,445 8,245 2,010 73.24 4.49 1.0812 0.0935 1,140.33 3,205 803 129 40.25 3.86 1.0897 0.0563 686.63 977 233 34 34.80 3.00 0.8330 0.0612 746.40 2 15,539 1,885 256 16.47 3.90 1.2890 0.0247 301.24 15,467 1,359 144 9.31 3.36 1.3754 0.0172 209.77 92,835 15,360 2,829 30.47 3.79 1.2213 0.0425 518.33 116,473 26,619 5,564 47.77 4.01 1.2248 0.0649 791.52 31,652 16,226 4,592 145.08 4.28 1.0696 0.1808 2,205.04 3,975 1,648 308 77.48 4.18 1.0559 0.0882 1,075.69 2,379 862 172 72.30 4.27 1.1204 0.0867 1,057.39 Severity Level 3 4 5 6 21,184 2,583 352 16.62 3.38 1.3251 0.0227 276.85 51,829 10,742 2,094 40.40 4.27 1.2527 0.0536 653.71 98,172 28,813 6,157 62.72 3.97 1.2000 0.0860 1,048.86 17,434 12,481 3,459 198.41 4.42 1.0456 0.2389 2,913.62 4,156 2,405 477 114.77 4.29 1.0547 0.1285 1,567.19 2,664 1,430 300 112.61 4.32 1.0206 0.1555 1,896.48 6,120 877 125 20.42 4.18 1.0413 0.0274 334.17 19,500 6,041 1,235 63.33 4.49 1.1958 0.0838 1,022.02 78,452 30,694 7,238 92.26 4.14 1.1119 0.1186 1,446.45 14,431 13,170 3,778 261.80 4.64 1.0709 0.3355 4,091.76 2,170 1,928 451 207.83 4.62 0.9649 0.2163 2,637.99 3,702 3,394 853 230.42 5.45 1.2429 0.3259 3,974.68 5,672 2,129 453 79.87 4.75 1.0994 0.1092 1,331.80 54,610 28,315 7,003 128.24 4.29 1.0491 0.1584 1,931.85 13,916 15,760 4,538 326.10 4.74 1.0341 0.4241 5,172.32 542 602 138 254.61 4.41 0.9819 0.3363 4,101.51 7,341 7,406 1,557 212.10 4.46 1.2299 0.3259 3,974.68 304 84 15 49.34 3.00 0.9293 0.1092 1,331.80 36,839 27,623 7,045 191.24 4.59 1.0353 0.2442 2,978.26 17,017 27,664 7,997 469.94 5.25 1.0858 0.6722 8,198.15 7,037 13,885 3,135 445.50 4.84 1.2568 0.7411 9,038.46 Geographic Variation in Hospital Admission Rates in the Medicare Population 10 There is a 50-fold difference in the number of PPAs per 1000 beneficiaries across CRG risk category ranging from 9.31 to 469.94. The bed days per PPA (average LOS) varies from 3.00 days to 5.45 with a general increase at the higher statuses and severity levels. Using standard relative resource weights available with the APR DRGs, a case mix index (CMI) was computed for the PPAs in each CRG risk class. While the APR DRG CMI varied from 0.83 to 1.44, the APR DRG CMI tended to be higher for the lower statuses and lower severity levels. This means that at the higher CRG statuses and severity levels, beneficiaries tend to be more frequently admitted for less serious conditions and tend to stay longer in the hospital than expected. This pattern may reflect a tendency to treat beneficiaries with a high chronic illness burden more conservatively resulting in more admissions and bed days for less serious reasons for admission. In general, PPAs tend to be admissions for less serious conditions. The overall APR DRG CMI across all admissions is 1.31 and for PPAs is 1.14. This is expected because most surgical admissions and most admissions for extreme acute events like an AMI are not a PPA. The PPA relative weight for each CRG risk category reflects the combined impact of the frequency of admission and the relative costliness of the PPAs being admitted. The relative expected costliness of PPAs in each CRG risk category is determined by multiplying the PPA relative weight by the financial conversion factor of $12,196. The product of the number of admissions in each CRG risk category and the PPA relative expected costliness for the CRG risk category summed over all CRG risk categories determines the expected PPA cost for any subset of beneficiaries. At the higher CRG statuses and severity levels, beneficiaries tend to be more frequently admitted for less serious conditions and tend to stay longer in the hospital than expected. This pattern may reflect a tendency to treat beneficiaries with a high chronic illness burden more conservatively resulting in more admissions and bed days for less serious reasons for admission. Table 2 contains summary data by CRG risk category for beneficiaries who do not have a chronic disease. Status 1 is for beneficiaries who are healthy and have no significant acute diseases in their history. Healthy nonusers with no significant contact with the health care system and healthy beneficiaries who had a mention of a chronic disease in their history but no subsequent treatment (potentially a rule out diagnosis) are assigned to separate CRGs. Across these three healthy Status 1 CRG categories, the PPAs per 1,000 varied from 11.23 to 19.02. There are four CRG risk categories in Status 2 for beneficiaries with a history of acute disease. The four significant acute CRG risk categories are for beneficiaries with significant acute disease, multiple or reoccurring significant disease, major trauma or major acute disease and significant acute disease with a mention of a chronic disease in their history but no subsequent treatment. Across these four significant acute Status 2 CRG categories, the PPAs per 1,000 varied from 13.79 to 22.43. While the variation in PPA/1,000 for status 1 and 2 was modest, status 1 and 2 had 199,756 of the beneficiaries (14.4 percent) and 2,784 of the PPAs (3.2 percent). Geographic Variation in Hospital Admission Rates in the Medicare Population 11 Table 2: PPA data by CRG risk category for beneficiaries with no chronic diseases CRG Status 1 Healthy 1 Healthy Nonuser 1 Healthy with Unconfirmed Chronic Disease 2 Multiple or Reoccurring Significant Acute Disease 2 Significant Acute Disease 2 Major Trauma Or Major Acute Disease 2 Significant Acute Disease With Unconfirmed Chronic Disease Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight Beneficiaries Admissions PPAs PPAs/1,000 Days/PPA PPA APR CMI PPA Weight PPA $ Weight 67,482 4,881 758 11.23 4.20 1.2185 0.0155 189.04 79,613 6,475 1,086 13.64 5.16 1.3370 0.0196 239.04 20,921 2,314 398 19.02 4.44 1.1859 0.0281 342.71 6,187 446 86 13.90 3.69 1.1484 0.0164 200.01 13,124 1,119 86 13.90 3.69 1.1484 0.0178 217.09 2,363 387 53 22.43 3.68 0.8607 0.0238 290.26 10,066 1,253 222 22.05 3.72 1.2039 0.0295 359.78 Geographic Variation in Hospital Admission Rates in the Medicare Population 12 PPA Results by Geographic Region PPA %(A-E)/E and $(A-E) by Census Region Table 3 contains the PPA %(A-E)/E and $(A-E) by census region for the national norm and best practice norm. Across census regions the PPAs/1,000 beneficiaries ranged from 42.9 for the mountain census region to 70.32 for the east south central census region. The %(A-E)/E with the national norm ranged from 13.20 percent below expected for the mountain census region to 6.6 percent above expected for the east north central census region. The %(A-E)/E with the best practice norm ranged from 0.30 percent below expected for the mountain census region to 22.43 percent above expected for the east north central census region. To achieve best practice across all regions, overall PPA performance would need to improve by 14.85 percent, which would generate $154.5 million in reduced Medicare expenditures. The 1,388,114 beneficiaries in the analysis data represent 3.59 percent of the 38,665,082 Medicare FFS beneficiaries in 2018.12 Extrapolating the reduction in Medicare expenditures from these beneficiaries to the full Medicare FFS population results in an estimated annual reduction of Medicare expenditures of $4.3 billion, assuming PPA performance is improved by the 14.85 percent needed to achieve best practice nationally. It is important to keep in mind the $4.3 billion represents a reduction in expenditures from achieving best practice $(A-E). The 85,974 PPAs represent $1.2 billion in Medicare expenditures ($A) which extrapolated to the full Medicare FFS population is $33.3 billion. While the $33.3 billion reflect Medicare expenditures associated with PPAs, only the $4.3 billion reduction is viewed as achievable in the short term. Approximately onethird of Medicare beneficiaries are enrolled in a Medicare Advantage (MA) Plan. The PPA performance in MA plans may differ from Medicare FFS so MA plan beneficiaries are not included in the estimated PPA reduction in Medicare expenditures. Table 3: PPA %(A-E)/E and $(A-E) by Census Region Region Count Benef Count PPAs PPAs per 1000 Benef %(A-E)/E PPA Nat Norm %(A-E)/E PPA BP Norm PPA $(A-E) Nat Norm (000) PPA $(A-E) BP Norm (000) New England ME, VT, NH, CT, MA, RI 78,205 5,059 64.69 -2.08 12.46 -1,364 7,121 Middle Atlantic South Atlantic NY, NJ, PA FL, GA, SC, NC, VA, WV, DC, MD, DE 174,276 11,568 305,134 19,113 66.38 62.64 1.04 0.70 16.04 15.66 1,592 1,880 21,456 36,532 E North Central IL, WI. MI, IN. OH 212,275 14,618 68.86 6.60 22.43 12,525 37,050 E South Central KY, TN, AL, MS 97,793 6,877 70.32 3.41 18.77 2,993 14,342 W South Central W North Central Mountain TX, OK, AR, LA MN, IA, MO, KS, NE, SD, ND AZ, NM, UT, CO, NV, WY, ID, MT 148,401 100,994 96,064 9,885 6,237 4,043 66.61 61.76 42.09 4.02 1.84 -13.20 19.47 16.96 -0.30 5,387 1,532 -9,629 22,698 12,311 -194 Pacific CA, OR, WA, HI, AK 174,972 8,574 49.00 -10.65 2.62 -14,915 3,199 TOTAL 1,388,114 85,974 61.94 0.00 14.85 0 154,514 Geographic Variation in Hospital Admission Rates in the Medicare Population 13 PPA %(A-E)/E and $(A-E) by State Table 4 contains the PPA %(A-E)/E and $(A-E) by state for the national norm and best practice norm. The PPAs/1,000 beneficiaries ranged from 24.9 for Hawaii to 77.1 for Louisiana. The %(AE)/E with the national norm ranged from 41.37 percent below expected for Hawaii to 28.54 percent above expected for Arkansas. The %(A-E)/E with the best practice norm ranged from 32.66 percent below expected for Hawaii to 47.62 percent above expected for Arkansas. Table 4: PPA %(A-E)/E and $(A-E) by State State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware DC Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Count Benef 23,675 3,451 28,123 18,214 117,877 19,223 14,634 6,878 2,475 92,161 38,527 4,573 8,253 59,705 33,376 19,369 16,784 24,386 20,150 8,616 32,329 36,477 45,595 14,387 18,922 30,072 6,784 11,224 11,432 9,480 44,306 9,431 73,425 48,553 4,079 48,376 21,991 16,311 56,545 3,902 29,366 5,079 30,810 88,046 8,651 5,096 43,229 Count PPAs 1,641 108 1,175 1,255 6,418 720 911 440 158 6,331 2,483 114 301 4,366 2,399 1,008 1,075 1,787 1,554 401 1,987 2,712 3,139 808 1,435 2,231 270 617 648 487 2,933 373 4,876 2,867 224 3,382 1,419 674 3,759 275 1,526 274 2,014 5,657 348 273 2,506 PPAs per 1000 Benef 69.31 31.30 41.78 68.90 54.45 37.46 62.25 63.97 63.84 68.70 64.45 24.93 36.47 73.13 71.88 52.04 64.05 73.28 77.12 46.54 61.46 74.35 68.85 56.16 75.84 74.19 39.80 54.97 56.68 51.37 66.20 39.55 66.41 59.05 54.92 69.91 64.53 41.32 66.48 70.48 51.96 53.95 65.37 64.25 40.23 53.57 57.97 %(A-E)/E PPA Nat Norm -0.17 -22.55 -9.66 28.54 -6.40 -20.59 -7.54 -1.29 -6.02 1.26 1.07 -41.37 -19.87 9.42 10.64 -9.03 5.24 3.80 9.29 -16.10 2.61 6.59 4.04 -7.88 14.26 14.78 -20.96 -5.47 0.08 -18.81 1.66 -27.40 2.46 -3.47 -11.25 11.97 2.26 -11.05 -1.25 7.04 -4.25 0.06 -0.75 -1.34 -8.57 -6.66 3.57 %(A-E)/E PPA BP Norm 14.65 -11.05 3.76 47.62 7.51 -8.80 6.20 13.37 7.93 16.30 16.08 -32.66 -7.96 25.66 27.07 4.48 20.87 19.21 25.52 -3.64 17.85 22.42 19.49 5.80 31.23 31.83 -9.23 8.57 14.94 -6.75 16.76 -16.62 17.68 10.86 1.93 28.60 17.45 2.16 13.42 22.94 9.96 14.92 13.99 13.31 5.00 7.20 18.95 $(A-E) PPA Nat Norm -36,868 -530,889 -2,079,309 4,221,112 -6,273,495 -2,982,488 -999,917 -76,723 -139,515 1,079,425 368,458 -1,462,204 -1,243,747 4,889,734 3,197,630 -1,398,225 705,514 855,654 1,795,574 -1,093,884 717,605 2,070,316 1,738,087 -948,116 2,377,027 3,991,268 -997,135 -471,139 7,174 -1,338,470 654,102 -1,955,110 1,568,093 -1,476,760 -349,646 5,287,637 445,818 -1,322,841 -630,071 239,345 -1,003,090 2,124 -202,701 -1,075,957 -556,272 -241,529 1,270,442 $(A-E) PPA BP Norm 2,727,524 -226,531 705,052 6,133,685 6,410,243 -1,109,843 715,885 691,636 160,014 12,123,494 4,806,034 -1,005,195 -434,202 11,604,882 7,082,246 604,782 2,445,295 3,768,621 4,293,812 -215,527 4,275,276 6,133,272 7,298,814 607,135 4,531,692 7,482,111 -382,088 643,024 1,226,875 -418,308 5,745,736 -1,032,630 9,802,926 4,018,381 52,197 10,998,705 2,996,599 224,711 5,907,083 678,753 2,045,445 476,951 3,313,891 9,273,909 282,673 227,206 5,875,288 Geographic Variation in Hospital Admission Rates in the Medicare Population 14 State Washington West Virginia Wisconsin Wyoming Count Benef 32,760 11,616 25,223 4,167 Count PPAs 1,260 815 1,332 208 PPAs per 1000 Benef 38.46 70.16 52.81 49.92 %(A-E)/E PPA Nat Norm -22.06 10.55 -12.61 6.16 %(A-E)/E PPA BP Norm -10.49 26.97 0.36 21.92 $(A-E) PPA Nat Norm -5,325,206 1,139,722 -2,588,152 177,600 $(A-E) PPA BP Norm -2,204,467 2,535,972 64,910 550,501 Figure 2 is a U.S. map with the %(A-E)/E for the national norm by state color coded as follows: Green: Yellow: Orange: Red: %(A-E)/E >10% below expected 12 states %(A-E)/E 0-10% below expected 16 states %(A-E)/E 0-10% above expected 17 states %(A-E)/E >10% above expected 6 states Figure 2: PPA %(A-E)/E performance by state Wide PPA performance variation is not only across states but also within states. The state of residency of the beneficiary data was used to assign beneficiaries to a state in Table 4. Using the metropolitan areas identified in the Core Based Statistical Areas (CBSAs) from the Office of Management and Budget, Appendix E contains PPA %(A-E)/E and $(A-E) for the national norm and best practice for each CBSA with at least 1,000 beneficiaries in the analysis database. Some CBSAs encompass multiple states. For example, the Philadelphia metropolitan area encompasses parts of New Jersey, Delaware and Maryland. When a CBSA encompassed more than one state, the CBSA in Appendix E was assigned to the primary state associated with the CBSA (the Philadelphia metropolitan area was assigned to Pennsylvania). Geographic Variation in Hospital Admission Rates in the Medicare Population 15 Table 5 contains the seven largest CBSAs in Florida. The PPA performance of the Miami and Tampa CBSAs is relatively consistent with the overall Florida state performance of 1.26 percent above expected for the PPA national norm. However, the Orlando and Jacksonville CBSAs have PPA performance well above expected (18.54 and 8.08 percent above expected for the PPA national norm) while the North Port/Sarasota, Cape Coral/Fort Myers and Deltona/Daytona Beach CBSAs have PPA performance well below expected (26.00, 9.78 and 23.22 percent below expected for the PPA national norm) Table 5: PPA %(A-E)/E and $(A-E) for the seven largest CBSAs in Florida CBSA Florida Miami-Fort Lauderdale-West Palm Beach Tampa-St. Petersburg-Clearwater Orlando-Kissimmee-Sanford Jacksonville North Port-Sarasota-Bradenton Cape Coral-Fort Myers Deltona-Daytona Beach-Ormond Beach Count Benef 92,161 16,543 11,547 7,859 6,999 6,535 4,825 3,563 Count PPAs 6,331 1,383 982 749 580 320 299 199 PPAs per 1000 Benef 68.70 83.57 85.02 95.31 82.81 49.00 61.93 55.78 %(A-E)/E PPA Nat Norm 1.26 2.95 2.38 18.54 8.08 -26.00 -9.78 -23.22 %(A-E)/E PPA BP Norm 16.30 18.24 17.58 36.14 24.13 -15.01 3.62 -11.82 PPA $(A-E) PPA $(A-E) Nat Norm BP Norm 1,079,425 483,221 278,229 1,428,600 528,718 -1,372,219 -394,988 -733,138 12,123,494 2,600,979 1,790,407 2,425,079 1,374,349 -689,799 127,275 -324,908 PPA Frequency Table 6 contains the APR DRG assigned to the 23 PPAs comprising at least one percent of the PPAs. As expected, the highest volume PPAs are for medical management of chronic diseases (heart failure, COPD) and acute diseases (non-bacterial pneumonia and urinary tract infections). Spinal procedures and cardiac catheterization were the most frequent PPAs for admissions related to procedures. Table 6: APR DRG of the 23 PPAs comprising at least one percent of the PPAs APR DRG of PPA 194 HEART FAILURE 139 OTHER PNEUMONIA 140 CHRONIC OBSTRUCTIVE PULMONARY DISEASE 463 KIDNEY & URINARY TRACT INFECTIONS 304 DORSAL & LUMBAR FUSION PROC EXCEPT FOR CURVATURE OF BACK 720 SEPTICEMIA & DISSEMINATED INFECTIONS 192 CARDIAC CATHETERIZATION FOR NON-CORONARY CONDITIONS 249 OTHER GASTROENTERITIS, NAUSEA & VOMITING 113 INFECTIONS OF UPPER RESPIRATORY TRACT 254 OTHER DIGESTIVE SYSTEM DIAGNOSES 422 HYPOVOLEMIA & RELATED ELECTROLYTE DISORDERS 199 HYPERTENSION 53 SEIZURE 166 CORONARY BYPASS W/O AMI OR COMPLEX PDX 420 DIABETES 198 ANGINA PECTORIS & CORONARY ATHEROSCLEROSIS 144 RESPIRATORY SIGNS, SYMPTOMS & MINOR DIAGNOSES 175 PERCUTANEOUS CORONARY INTERVENTION W/O AMI 383 CELLULITIS & OTHER SKIN INFECTIONS 351 MUSCULOSKELETAL SYSTEM & CONNECTIVE TISSUE DIAGNOSES 203 CHEST PAIN 141 ASTHMA Count 14,780 10,196 8,436 7,530 3,899 2,914 2,823 2,739 2,199 2,055 1,992 1,842 1,838 1,709 1,447 1,321 1,312 1,292 990 964 943 858 Percent 17.2 11.9 9.8 8.8 4.5 3.4 3.3 3.2 2.6 2.4 2.3 2.1 2.1 2.0 1.7 1.5 1.5 1.5 1.2 1.1 1.1 1.0 Geographic Variation in Hospital Admission Rates in the Medicare Population 16 Overlap Between PPAs, PPRs and ED Admits Of the 1,388,114 beneficiaries, 227,138 had one or more hospital admissions (16.4 percent), resulting in a total of 379,841 admissions. 17,032 of those admissions were considered potentially preventable readmissions based on PPRs and were not eligible to be a PPA, resulting in 362,809 admissions being eligible to be a PPA. Of the 379,841 admissions: · 85,974 admissions are a PPA (22.6 percent) · 25,370 are an admission that initiated one or more readmissions (6.7 percent) · 69,611 are low severity medical admissions from the ED (18.3 percent) Any admission can simultaneously be a PPA, PPR and/or an ED Admit. The Venn diagram in Appendix D contains the details of the overlap among PPAs, PPRs and ED admits. Of the 85,974 PPAs: · 42,314 are only a PPA (49.2 percent) · 7,739 are a PPA that is followed by one or more PPRs (9.0 percent) · 26,243 are both a PPA and an ED Admit (30.5 percent) · 1,992 are a PPA, an ED Admit and are followed by one or more PPRs (2.3 percent) The overlap among PPAs, PPRs and ED Admits is important because it can help focus quality improvement efforts. For example, the 7,739 PPAs that are followed by one or more PPRs results in 9,982 additional readmissions, which substantially increases the downstream impact of the initial PPA. Quality improvement initiatives need to focus on preventing the PPA admission as well as preventing the subsequent readmissions. The 26,243 admissions that are both a PPA and an ED Admit require a focus on the admission criteria in the ED for low severity medical admissions. Since the overlap between PPAs and low severity medical admissions through the ED (30.5 percent) is so substantial, any quality improvement initiative directed at lowering the hospital PPA rate in a population of Medicare beneficiaries should evaluate ED admission performance and practices. To illustrate the relationship between PPAs, PPRs and ED Admits, the %(A-E)/E was computed for PPRs and ED Admits for 2018 using all hospitalization from 2018 and not just the hospitalizations in the 5 percent analysis data. Table 6 contains the %(A-E)/E for PPAs, PPRs and ED Admits by state. The state assignment for PPAs may differ slightly from the state assignment for PPRs and ED Admits; for PPAs the state is assigned based on the state of residence of the beneficiary, but for PPRs and ED Admits the state is assigned based on the location of the hospital. The %(A-E)/E for PPAs, PPRs and ED Admits across states are correlated as follows: PPAs PPAs PPRs PPRs ED Admits ED Admits Correlation 0.694 0.566 0.763 The significant correlation among PPA, PPR and ED Admit performance across states indicates that hospital performance on PPRs and ED Admits impacts the PPA performance of the Medicare population in a state. PPA, PPR and ED Admit performance provide insights into the functioning Geographic Variation in Hospital Admission Rates in the Medicare Population 17 of the health care delivery system in a state and illustrates the interdependence of these performance measures. Summary and Conclusions The 1,388,114 beneficiaries in the analysis database had 379,841 hospital admissions of which 85,974 were a PPA (22.6 percent of admissions). The 85,974 PPAs represent $33.3 billion in annual FFS Medicare expenditures. If PPA best practice was achieved nationally, overall PPA performance would need to improve by 14.85 percent, which would result in an annual reduction in Medicare expenditures of $4.3 billion (12.9 percent of the $33.3 billion in PPA expenditures). There was significant PPA performance variation across census regions, states and CBSAs. Across states, PPA performance based on a national norm varied from 32.66 percent below expected for Hawaii to 47.62 percent above expected for Arkansas. Nine percent of the PPAs were followed by one or more readmissions (PPRs), 30.5 percent of PPAs were a low severity medical admission through the ED and 2.3 percent of PPAs were a low severity medical admission through the ED that was followed by one or more readmissions. Across states PPR and ED Admit performance (%(A-E)/E) was associated with PPA performance with a PPA/PPR performance correlation of 0.694 and a PPA/ED Admit performance correlation of 0.566. PPA performance is an effective measure of delivery system performance in a managed care plan or geographic region. The extent of PPA performance variation across states indicates there are PPA performance improvement opportunities in many geographic areas. The interdependence of PPAs with PPRs and ED Admits can provide useful insights for targeted quality improvement initiatives aimed at reducing PPAs. Geographic Variation in Hospital Admission Rates in the Medicare Population 18 References 1 U.S. Agency for Health Care Research and Quality. Statistical Brief #261: National Inpatient Hospital Costs: The Most Expensive Conditions by Payer, 2017. Rockville, MD: AHRQ, 2020. https://www.hcup-us.ahrq.gov/reports/statbriefs/sb261-Most-Expensive-Hospital-Conditions2017.jsp 2 U.S. Agency for Health Care Research and Quality. Statistical Brief #259: Characteristics and Costs of Potentially Preventable Inpatient Stays, 2017. Rockville, MD: AHRQ, 2020. https://hcup-us.ahrq.gov/reports/statbriefs/sb259-Potentially-Preventable-Hospitalizations2017.jsp 3 J.G. Ouslander and S.M. Handler. Consensus-Derived Interventions to Reduce Acute Care Transfer (INTERACT) - Compatible Order Sets for Common Conditions Associated with Potentially Avoidable Hospitalizations. Journal of the American Medicaid Directors Association. 2015 Jun 1; 16 (6): 524-6 4 I.P. Lin, S.C., Wu, and S.T. Huang. Continuity of care and avoidable hospitalizations for chronic obstructive pulmonary disease (COPD). Journal of the American Board of Family Medicine. 2015 Mar-Apr; 28 (2): 222-30 5 Smith PK, et al. Selection of surgical or percutaneous coronary intervention provides differential longevity benefit. Annals of Thoracic Surgery. 2006;82:1420-28. Weintraub WS, et al. 6 Deyo RA, Mirza SK. The case for restraint in spinal surgery: does quality management have a role to play? Eur Spine J. 2009 Aug; 18 Suppl 3:331-7; Deyo RA, Mirza SK, Turner JA, Martin BI. Overtreating chronic back pain: time to back off? Journal of the American Board of Family Medicine. 2009 Jan-Feb; 22(1):62-8. 7 U.S. Agency for Health Care Research and Quality. AHRQ Prevention Quality Indicators Overview. Rockville, MD: AHRQ, 2020. http://www.qualityindicators.ahrq.gov/Modules/pqi_resources.aspx. 8 Hughes, Averill, Eisenhandler, Goldfield, Muldoon, Neff, Gay. (2003). Clinical Risk Groups (CRGs): A Classification System for Risk-Adjusted Capitation-Based Payment and Managed Care. Medical Care, 42(1). 9 Goldfield, McCullough, Hughes, Tang, Eastman, Rawlins, Averill. (2008). Identifying Potentially Preventable Readmissions. Health Care Financing Review, 30(1) 10 Averill, Goldfield, Muldoon, Steinbeck, Grant. (2002). A Closer Look at All Patient Refined DRGs. Journal of the American Health Information Management Association, 73(1). 11 Houchens RL, Elixhauser A, Romano PS. How often are potential patient safety events present on admission? The Joint Commission Journal on Quality and Patient Safety, 2008;34(3):154-163. doi:10.1016/s1553-7250(08)34018-5 12 Centers for Medicare and Medicaid Services. 2018 Medicare Enrollment. Washington, D.C. 2018. https://www.cms.gov/research-statistics-data-systems/cms-program-statistics/2018medicare-enrollment-section Geographic Variation in Hospital Admission Rates in the Medicare Population 19 Appendix A: Bibliography of Publicly Available Articles and Reports on PPAs, CRGs, PPRs, APR DRGs All articles and reports are publicly available and are listed in chronological order. The opinions and conclusions in these articles and reports are solely those of the authors. Potentially Preventable Admissions (PPAs) Articles, Reports, and Book Chapters Fuller RL, Clinton S, Goldfield NI, Kelly WP. Building the affordable medical home. J Ambul Care Manage. 2010;33(1):71-80. Goldfield N, Kelly W, Patel K. Potentially Preventable Events: an actionable set of measures for linking quality improvement and cost savings. Qual Manage Health Care. 2012;21(4):213-219. Millwee B, Goldfield N, Averill R, Hughes J. Payment system reform: one state's journey. J Ambul Care Manage. 2013;36(3):199-208. Medicare Payment Advisory Commission. Feasibility of measuring population-based outcomes: potentially preventable admissions and emergency department visits. Online Appendix 3A in Report to the Congress: Medicare and the Health Care Delivery System. Washington, DC: MedPAC, June 2014. 3M Health Information Systems. The 3M Value Index Score: Measurement and Evidence. Murray, UT: 3M HIS, 2015. Bernstein AB. Potentially Preventable Events: Comparing Medicaid and Privately Insured Populations. Presentation to the Medicaid and CHIP Payment and Access Commission. Washington, DC: MACPAC, Dec. 15, 2015. Minnesota Department of Health. An Introductory Analysis of Potentially Preventable Health Care Events in Minnesota. St. Paul. MN: MNDOH, 2015. Minnesota Department of Health. An Introductory Analysis of Potentially Preventable Health Care Events in Minnesota: Supplemental Technical Information. St. Paul. MN: MNDOH, 2015. DuBard CA. Key Performance Indicators of Cost and Utilization for Medicaid Recipients Enrolled in Community Care of North Carolina. N C Med J. 2016;77(4):297-300. Quinn K, Weimar D, Gray J, Davies B. Thinking about clinical outcomes in Medicaid. J Ambul Care Manage. 2016;39(2). Florida Agency for Health Care Administration. Analysis of Potentially Preventable Healthcare Events of Florida Medicaid Enrollees: July 2015 to June 2016. Tallahassee, FL: AHCA, Winter 2017. Florida Agency for Healthcare Administration. Analyzing Potentially Preventable Healthcare Events of Florida Medicaid Enrollees. Tallahassee, FL: AHCA, Spring 2017. Myers & Stauffer. Cost Effectiveness Study Report for Mississippi Coordinated Access Network (MississippiCAN). Report to the Mississippi Division of Medicaid. Windsor, CT: Myers & Stauffer, 2017. Geographic Variation in Hospital Admission Rates in the Medicare Population 20 University of Florida, Institute for Child Health Policy. Texas Medicaid Managed Care and CHIP Smmary of Activities and Trends in Healthcare Quality. Tallahassee, FL: ICHP, 2017. Florida Agency for Health Care Administration. Analysis of Potentially Preventable Healthcare Events of Florida Medicaid Enrollees 2015-2016 and 2016-2017. Tallahassee, FL: AHCA, Winter 2018. North Carolina Department of Health and Human Services. Plan for Implementation of Hospital Quality Outcomes Program and PHP Quality Outcomes Program. Report to the Legislature. Raleigh, ND: NCDHHS, Sept. 28, 2018. Fuller RL, Goldfield NI, Hughes JS, McCullough EC. Nursing home compare star rankings and the variation in potentially preventable emergency department visits and hospital admissions. Popul Health Manage. Epub ahead of print. July 30, 2018. Millwee B, Goldfield N, Turnipseed J. Achieving improved outcomes through value-based purchasing in one state. Am J Med Qual. 2018;33(2):162-171. Millwee B, Quinn K, Goldfield N. Moving toward paying for outcomes in Medicaid. J Ambul Care Manage. 2018;41(2):88-94. Florida Agency for Health Care Administration. Analysis of Potentially Preventable Healthcare Events of Florida Medicaid Enrollees 2015-2016 and 2016-2017. Tallahassee, FL: AHCA, Winter 2018. Texas External Quality Review Organization. Quality, Timeliness, and Access to Health Care for Texas Medicaid and CHIP Recipients: Summary of Activities Calendar Year 2017. Austin, TX: Texas EQRO, n.d. Websites Pennsylvania Department of Human Services. Hospital Assessment Initiative. Fiscal Year (FY) 2018-2019 Hospital Quality Incentive (HQI) Program Statewide Results. Web document at https://www.dhs.pa.gov/providers/Documents/Hospital%20Assessment%20Initiative/c_29243 5.pdf. [Accessed May 18, 2020] Pennsylvania Department of Human Services. Hospital Assessment Initiative. Hospital Quality Incentive (HQI) Program State Fiscal Year (SFY) 2017-2018 Q&As. Web document available at https://www.dhs.pa.gov/providers/Documents/Hospital%20Assessment%20Initiative/c_26664 7.pdf. [Accessed May 18, 2020] 3M Health Information Systems. 3M Patient Classification Methodologies. Webpage: www.3m.com/his/methodologies. Accessed Sept. 28, 2020 Texas Health and Human Services Commission. www.thlcportal.com. Accessed 2020 Superior Health Plan. www.superiorhealthplan.com/providers/resources/providerprograms/3m-his.html. Accessed 2020 Superior Health Plan. 3M Health Information. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 195046-3M-HIS-Resource-Guide-P-508-03202019.pdf Geographic Variation in Hospital Admission Rates in the Medicare Population 21 Superior Health Plan. 3M HIS Prospective Dashboard User Guide. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 173928-3M-HIS-Dashboard-Training-P-05312018.pdf. Clinical Risk Groups (CRGs) Articles, Reports, and Book Chapters Goldfield N, Averill R, Eisenhandler J, Hughes JS, Muldoon J, Steinbeck B, Bagadia F. The prospective risk adjustment system. J Ambul Care Manage. 1999;22(2):41-52. National Association of Children's Hospitals and Related Institutions. Summary Description of Clinical Risk Groups (CRGs). Washington, DC: NACHRI; 2000. Medicare Payment Advisory Commission. Report to the Congress: Improving Risk Adjustment in Medicare. Washington, DC: MedPAC, November 2000. Goldfield N, Averill R, Eisenhandler J. Payment and provider profiling of episodes of illness of clinical illnesses involving rehabilitation. J Outcome Meas. 2000;4(3):706-720. Majeed A, Bindman AB, Weiner JP. Use of risk adjustment in setting budgets and measuring performance in primary care I--how it works. BMJ 2001;323:604607. Bethell C, Read D. Approaches to Identifying Children and Adults with Special Health Care Needs: A Resource Manual for State Medicaid Agencies and Managed Care Organizations. Report to CMS. Available at www.childhealthdata.org. 2002. Neff JM, Sharp VL, Muldoon J, Graham J, Popalisky J, Gay JC. Identifying and classifying children with chronic conditions using administrative data with the Clinical Risk Group classification system. Ambul Pediatr. 2002;2(1):71-79. Averill RF, Goldfield NI, Eisenhandler J, Muldoon JH, Hughes JS, Neff JM, Gay JG, Gregg LW, Gannon DE, Shafir BV, Bagadia FA, Steinbeck BA. Development and evaluation of Clinical Risk Groups. In: Goldfield N, Delivering High-Quality Cost-Effective Health Care to All: The Scientific and Political Ingredients for Success. Northampton, MA: Artichoke Publications, 2004. Goldfield N, Eisenhandler J, Gay G, McCullough E, Bao M, Neff J, Muldoon J, Hughes J, Mills R. Development of an episode of illness classification for population management using pharmacy data. Dis Manag. 2004;5(3). Hughes JS, Averill RF, Eisenhandler J, Goldfield NI, Muldoon J, Neff JM, Gay JC. Clinical Risk Groups (CRGs): a classification system for risk-adjusted capitation-based payment and health care management. Med Care. 2004;42(1):81-90. Neff JM, Sharp VL, Muldoon J, Graham J, Myers K. Profile of medical charges for children by health status group and severity level in a Washington State health plan. HSR. 2004;39(1):73-90. Berlinguet M, Preyra C, Dean S. Comparing the Value of Three Main Diagnostic Based Risk Adjustment Systems. Ottawa: ON: Canadian Health Services Research Foundation, 2005. Geographic Variation in Hospital Admission Rates in the Medicare Population 22 Neff JM, Sharp VL, Popalisky J, Fitzgibbon T. Using medical billing data to evaluate chronically ill children over time. J Am Care Manage. 2006; 29(4):283-290. Maine Health Information Center. Children in Out-of-Home Placement in New Hampshire Health Status, Utilization, Payments, and Preventive Visits, State Fiscal Year 2007. (Concord, NH: DHHS, 2009) Bernstein RH. New arrows in the quiver for targeting case management: high-risk versus highopportunity case identification. J Ambul Care Manage. 2007;30(1):39-51. Alberta Health Quality Council. 2009 Measuring and Monitoring for Success. Calgary, AB: AHQC, 2009. Kelly WP, Wendt SW, Vogel BB. Guiding principles for payment system reform. J Ambul Care Manage. 2010;33(1):29-34. Neff JM, Clifton H, Park KJ, Goldenberg C, Popalisky J, Stout JW, Danielson BS. Identifying children with lifelong chronic conditions by using hospital discharge data. Acad Pediatr. 2010;10(6):417-423. Eisenhandler J, Averill R, Vertrees J, Quain A, Switalski J. A Comparison of the Explanatory Power of Two Approaches to the Prediction of Post Acute Care Resources Use. Report to CMS. Wallingford, CT: 3M Health Information Systems, 2011. New Hampshire Department of Health and Human Services. New Hampshire Medicaid Annual Report State Fiscal Year 2010. Concord, NH: DHHS, 2011. Children's Hospital Association. Defining Children with Medical Complexities. Alexandra, VA: CHA, 2013. 3M Health Information Systems. The Impact of Disability Measures on Expected Medicare Payments and Expected Provider Charges for Event-Based Episodes that Include Post-Acute Care. Salt Lake City, UT: 3M HIS, 2013. Medicare Payment Advisory Commission. Approaches to bundling payment for post-acute care. Chapter 3 in Report to the Congress: Medicare and the Health Care System. Washington, DC: MedPAC, June 2013. Onpoint Health Data. Children's Health Insurance Programs in New Hampshire: Access, Prevention, Care Management, Utilization, & Payments (State Fiscal Year 2011). Report to DHHS. Concord, NH: DHHS, 2013 Schone E, Brown RS. Risk Adjustment: What Is the Current State of the Art, and How Can It Be Improved? Princeton, NJ: Robert Wood Johnson Foundation, 2013 Vertrees J, Averill R, Eisenhandler J, Quain A, Switalski J, Gannon D. The Ability of Event-Based Episodes to Explain Variation in Charges and Medicare Payments for Various Post Acute Service Bundles. Report to MedPAC. Wallingford, CT: 3M Health Information Systems, 2013. Vigen G, Coughlin S, Duncan I. Measurement and Performance Healthcare Quality and Efficiency: Resources for Healthcare Professionals. Third update. Society of Actuaries, 2013. Berry J, Hall M, Hall DE, Kuo DZ, Cohen E, Agrawal R, Mandl KD, Clifton H, Neff J. Inpatient growth and resource use in 28 children's hospitals. JAMA Pediatrics. 2013;167(2):170-177. Geographic Variation in Hospital Admission Rates in the Medicare Population 23 Fuller R, Goldfield N, Averill R, Eisenhandler J, Vertrees J. Adjusting Medicaid managed care payments for changes in health status. Med Care Res Rev. 2013;70(1):68-83. Lion KC, Rafton SA, Shafii J, Brownstein D, Michel E, Tolman M, Ebel BE. Association between language, serious adverse events, and length of stay among hospitalized children. Hosp Pediatr. 2013;3(3): 219-225. https://doi.org/10.1542/hpeds.2012-0091 Vertrees J, Averill R, Eisenhandler, J, Quain, A, Switalski J. Bundling Post-Acute Care Services into MS-DRG Payments. Medicare Medicaid Res Rev. 2013;3(3):E1-E19 3M Health Information Systems. The 3M Value Index Score: Measurement and Evidence. Murray, UT: 3M HIS, 2015. Johnson TL, Brewer D, Estracio R, Vlasimsky T, Durfee MJ, Thompson KR, Everhart RM, Rinehart DJ, Batal H. Augmenting predictive modeling tools with clinical insights for care coordination. eGEMs (Generating Evidence & Methods to Improve Patient Outcomes). 2015;3(1). North Carolina Community Care Networks, Inc. Clinical Program Analysis. Report to the North Carolina Department of Health and Human Services. Raleigh, NC: NCCC, 2015 Berry JG, Hall M, Cohen E, O'Neill M, Feudtner C. Ways to identify children with medical complexity and the importance of why. J Pediatr. 2015;167(2):229-237. HSR. 20014;39(1):73- DuBard CA, Jacobsen Vann JC, Jackson C. Conflicting readmission rate trends in a high-risk population: implications for performance measurement. Popul Health Manag. 2015;18:351357 Jackson C, Shahsahehi M, Wedlake T, DuBard CA. Timeliness of outpatient follow-up: an evidence-based approach for planning after hospital discharge. Ann Fam Med. 2015:13(2):155122. Jones C, Finison K, McGraves-Lloyd, Tremblay T, Mohlman MK, Tanzman B, Hazard M, Maier, Samuelson J. Vermont's community-oriented all-payer medical home model reduces expenditures and utilization while delivering high-quality care. Popul Health Manag. 2015. DOI: 10.1089/pop.2015.0055. Neff JM, Clifton H, Popalisky J, Zhou C. Stratification of children by medical complexity. Acad Pediatr. 2015;15(2):191-196. Pfister DG, Rubin DM, Elkin EE, Neill US, Duck E, Radzyner M, Bach PB. Risk adjusting survival outcomes in hospitals that treat patients with cancer without information on cancer stage. JAMA Oncol. 2015;1(9):1303-1310. Quinn K. The 8 basic payment methods in health care. Ann Intern Med. 2015;163(4):300-306. Florida Agency For Healthcare Administration. Analyzing the Disease Burden of Florida Medicaid Enrollees Using Clinical Risk Groups. Tallahassee, FL: AHCA, Winter 2016. Hileman G, Steele S. Accuracy of Claims-Based Risk Scoring Models. Schaumburg, IL: Society of Actuaries, 2016. DuBard CA. Key Performance Indicators of Cost and Utilization for Medicaid Recipients Enrolled in Community Care of North Carolina. N C Med J. 2016;77(4):297-300. Geographic Variation in Hospital Admission Rates in the Medicare Population 24 Fuller RL, Goldfield N. Paying for on-patent pharmaceuticals: limit prices and the emerging role of a pay for outcomes approach. J Ambul Care Manage. 2016;39(2):143-149. Fuller RL, Goldfield N. Response to commentaries on "Paying for on-patent pharmaceuticals: limit prices and the emerging role of a pay for outcomes approach". J Ambul Care Manage. 2016;39(2):155-156. Fuller RL, Hughes JS, Goldfield NI. Adjusting population risk for functional health status. Popul Health Manage. 2016;19(2):136-144. Gareau S, Lopez-De Fede A, Loudermilk BL, Cummings TH, Hardin JW, Picklesimer AH, Crouch E, Covington-Kolb S. Group prenatal care results in Medicaid savings with better outcomes: a propensity score analysis of CenteringPregnancy participation in South Carolina. Matern Child Health J. 2016;20(7):13841393. Juhnke C,Bethge S, Mühlbacher AC. A review on methods of risk adjustment and their use in integrated healthcare systems. Int J Integr Care. 2016;16(4):118 Mohlman MK, Tanzman B,Finison K, Pinettte M, Jones C. Impact of medication-assisted treatmernt for opioid addiction on Medicaid expenditures and health services utilization rates in Vermont. J Subst Abuse Treat. 2016;67: 914 Finison K , Mohlman M, Jones C, Pinette M, Jorgenson D, Kinner A, Tremblay T, Gottlieb D. Risk-adjustment methods for all-payer comparative performance reporting in Vermont. BMC Health Serv Res. 2017;17. Bednar WR, Axene JW, Liliedahl RL. An Analysis of End-of-Life Costs for Terminally Ill Medicare Fee-for-Service (FFS) Cancer Patients. Schaumburg, Society of Actuaries, 2018. Fuller RL, Goldfield NI, Hughes JS, McCullough EC. Nursing home compare star rankings and the variation in potentially preventable emergency department visits and hospital admissions. Popul Health Manage. Epub ahead of print. July 30, 2018. Averill RF, Fuller RL, Mills RE. Financial Impact of Geographic Variation in Hospital Quality Performance in Medicare. Murray, UT: 3M Health Information Systems, 2019. Connecticut Department of Social Services. Connecticut State Innovation Model Operational Plan Award Year 4. Hartford, CT: DSS, 2019. Vermont Agency of Human Services. Annual Report on The Vermont Blueprint for Health. Report to the Legislature. Burlington, VT; Agency of Human Services, 2020 Vermont Agency of Human Services. Community Health Profiles [webpage]. https://blueprintforhealth.vermont.gov/community-health-profiles. Accessed Aug. 17, 2020. Andrews AL, Bettenhausen J, Hoefgen E, Richardson T,Macy ML; Zima BT, Colvin J; Hall M; Shah SS, Neff NM, Auger KA. Measures of ED Utilization in a National Cohort of Childen. Am J Manag Care. 2020;26(6):267-272. 3M Health Information Systems. 3M Patient Classification Methodologies. Webpage: www.3m.com/his/methodologies. Accessed Sept. 28, 2020 Geographic Variation in Hospital Admission Rates in the Medicare Population 25 Vermont Agency of Human Services. Hub and Spoke Profiles [webage]. Annual Report on The Vermont Blueprint for Health. Report to the Legislature. Burlington, VT; Agency of Human Services, 2020 Superior Health Plan. 3M Health Information. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 195046-3M-HIS-Resource-Guide-P-508-03202019.pdf Superior Health Plan. 3M HIS Prospective Dashboard User Guide. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 173928-3M-HIS-Dashboard-Training-P-05312018.pdf. Potentially Preventable Readmissions (PPRs) Articles, Reports, and Book Chapters Medicare Payment Advisory Commission. Payment policy for inpatient readmissions. Chapter 5 in Report to the Congress: Promoting Greater Efficiency in Medicare. Washington, DC: MedPAC, June 2007. Goldfield N, McCullough E, Hughes J, Tang A, Eastman B, Rawlins L, Averill R. Identifying potentially preventable readmissions. Health Care Financ Rev. 2008;30(1):75-91. Feudtner C, Levin JE, Srivastava R, Goodman DM, Slonim AD, Sharma V, Shah SS, Pati S, Fargason C Jr, Hall M. How well can hospital readmission be predicted in a cohort of hospitalized children? A retrospective, multicenter study. Pediatrics. 2009;123(1):286-293. Utah Department of Health. Potentially Preventable Hospital Readmissions. Health Status Update. Salt Lake City: Utah DOH,2010. Vest JR, Gamm LD, Oxford BA, Gonzalez MI, Slawson KM. Determinants of preventable readmissions in the United States: a systematic review. Implement Sci. 2010;5:88. Utah Department of Health. Readmissions to Utah Hospitals, 2005-2007. Salt Lake City,UT: 2010 Fuller RL, Clinton S, Goldfield NI, Kelly WP. Building the affordable medical home. J Ambul Care Manage. 2010;33(1):71-80. Goldfield N. Strategies to decrease the rate of preventable readmission to hospital. CMAJ. 2010;182(6):538-539. Boutwell AE, Jencks SF. It's Not Six of One, Half Dozen the Other: A Comparative Analysis of 3 Rehospitalization Measurement Systems for Massachusetts. AcademyHealth Annual Research Meeting; Seattle, WA. 2011. Eisenhandler J, Averill R, Vertrees J, Quain A, Switalski J. A Comparison of the Explanatory Power of Two Approaches to the Prediction of Post Acute Care Resources Use. Report to CMS. Wallingford, CT: 3M Health Information Systems, 2011. Goldfield N. How important is it to identify avoidable hospital readmissions with certainty? CMAJ. 2011;183(7):e368-369. Geographic Variation in Hospital Admission Rates in the Medicare Population 26 Barrett M, Raetzman S, Andrews R. Overview of Key Readmission Measures and Methods. 2012. HCUP Methods Series Report #2012-04. Rockbille, MD: AHRQ, 2012. Fuller R, Goldfield N, Averill R, Hughes J. Inappropriate use of payment weights to risk adjust readmission rates. Am J Med Qual. 2012;27(1):341-344. Goldfield N, Kelly W, Patel K. Potentially Preventable Events: an actionable set of measures for linking quality improvement and cost savings. Qual Manage Health Care. 2012;21(4):213-219. Texas Health and Human Services. Potentially Preventable Readmissions in the Texas Medicaid Population, State Fiscal Year 2012. Austin, TX: HHSC, 2013. Texas Health and Human Services Commission. Potentially Preventable Readmissions in the Texas Medicaid Population, State Fiscal Year 2012. Austin, TX: HSSC, 2013. Averill R, Goldfield N, Hughes JS. Medicare payment penalties for unrelated readmissions require second look. Healthc Financ Manage. 2013(October):96-98. Berry JG, Toomey SL, Zaslavsky AM, Jha AK, Nakamura MM, Klein DJ, Feng JY, Shulman S, Chiang VW, Kaplan W, Hall M, Schuster MA. Pediatric readmission prevalence and variability across hospitals. JAMA. 2013;309(4):372-380. Davies S, Saynina O, Schultz E, McDonald KM, Baker LC. Implications of metric choice for common applications of readmission metrics. Health Serv Res. 2013;48:19781995. Fuller RL, Atkinson G, McCullough EC, Hughes JS. Hospital readmission rates: the impacts of age, payer, and mental health diagnoses. J Ambul Care Manage. 2013;36(2). Millwee B, Goldfield N, Averill R, Hughes J. Payment system reform: one state's journey. J Ambul Care Manage. 2013;36(3):199-208. Quinn K, Davies B. Potentially Preventable Readmissions in Rhode Island. Cranston, RI: Xerox State Healthcare, 2014. McCoy KA, Bear-Pfaffendof K, Foreman JK, Daniels T, Zabel EW, Grangaard LJ, Trevis JE, Cummings KA. Reducing avoidable hospital readmissions effectively: a statewide campaign. Jt Comm J Qual Patient Saf. 2014;40(5):198-204. Stratis Health. RARE Campaign Exceeds Goals, Prevents 7,975 Avoidable Hospital Readmissions in Minnesota [news release]. Available at http://www.stratishealth.org/news/20140617.html. Accessed Jan. 28, 2020 3M Health Information Systems. The 3M Value Index Score: Measurement and Evidence. Murray, UT: 3M HIS, 2015. Minnesota Department of Health. An Introductory Analysis of Potentially Preventable Health Care Events in Minnesota. St. Paul. MN: MNDOH, 2015. Minnesota Department of Health. An Introductory Analysis of Potentially Preventable Health Care Events in Minnesota: Supplemental Technical Information. St. Paul. MN: MNDOH, 2015. North Carolina Community Care Networks, Inc. Clinical Program Analysis. Report to the North Carolina Department of Health and Human Services. Raleigh, NC: NCCC, 2015 Geographic Variation in Hospital Admission Rates in the Medicare Population 27 Borzecki AM, Chen Q, Restuccia J, Mull HJ, Shwartz M, Gupta K, Hanchate A, Strymish J, Rosen A. Do pneumonia readmissions flagged as potentially preventable by the 3M PPR software have more process of care problems? A cross-sectional observational study. BMJ Qual Saf. 2015;24:753-763. DuBard CA, Jacobsen Vann JC, Jackson C. Conflicting readmission rate trends in a high-risk population: implications for performance measurement. Popul Health Manag. 2015;18:351357 Fuller RL, Atkinson G, Hughes JS. Indications of biased risk adjustment in the Hospital Readmission Reduction Program. J Ambul Care Manage. 2015;38(1):39-47. Gay JC, Agrawal R, Auger KA, Del Beccaro MA, Eghtesady P, Fieldston ES, Golias J, Han PD, McClead R, Morse RB, Neuman ML, Simon HK, Tejedor-Sojo J, Teufel RJ, Harris JM, Shah SS. Rates and impact of potentially preventable readmissions at children's hospitals. J Pediatr. 2015;166(3):615-619.e5 Jackson C, Shahsahehi M, Wedlake T, DuBard CA. Timeliness of outpatient follow-up: an evidence-based approach for planning after hospital discharge. Ann Fam Med. 2015:13(2):155122. Soong C, Bell C. Identifying preventable readmissions: an achievable goal or waiting for Godot? BMJ Qual Saf 2015;24:741743. doi:10.1136/bmjqs-2015-004484 DuBard CA. Key Performance Indicators of Cost and Utilization for Medicaid Recipients Enrolled in Community Care of North Carolina. N C Med J. 2016;77(4):297-300. Goldfield N, Averill R, Fuller R, Hughes J. Misinterpretation of meaning and intended use of potentially preventable readmissions. BMJ Qual Saf. 2015;25(3):2078. Lagoe R, Kronenberg P, Littau S. Readmissions by hospital inpatient service at the community level. Intern Med Rev. 2016;2.10.18103/imr.v2i9.234. Nakagawa K, Ahn HJ, Taira DA, Miyamura J, Sentel TL. Ethnic comparison of 30-day potentially preventable readmissions after stroke in Hawaii. Stroke. 2016;47:2611-2617 Quinn K, Weimar D, Gray J, Davies B. Thinking about clinical outcomes in Medicaid. J Ambul Care Manage. 2016;39(2). Florida Agency for Health Care Administration. Analysis of Potentially Preventable Healthcare Events of Florida Medicaid Enrollees: July 2015 to June 2016. Tallahassee, FL: AHCA, Winter 2017. Florida Agency for Healthcare Administration. Analyzing Potentially Preventable Healthcare Events of Florida Medicaid Enrollees. Tallahassee, FL: AHCA, Spring 2017. Medicare Payment Advisory Commission. Health Care Spending and the Medicare Program: A Data Book (June 2017). Washington, DC: MedPAC, 2017. Medicare Payment Advisory Commission. Hospital inpatient and outpatient services. Chapter 3 in Report to the Congress: Medicare Payment Policy. Washington, DC: MedPAC, March 2017 Myers & Stauffer. Cost Effectiveness Study Report for Mississippi Coordinated Access Network (MississippiCAN). Report to the Mississippi Division of Medicaid. Windsor, CT: Myers & Stauffer, 2017. Geographic Variation in Hospital Admission Rates in the Medicare Population 28 University of Florida, Institute for Child Health Policy. Texas Medicaid Managed Care and CHIP Summary of Activities and Trends in Healthcare Quality. Tallahassee, FL: ICHP, 2017. Burns & Associates. External Quality Review of Indiana's Care Programs: Hoosier Healthwise, Hoosier Care Connect and HIP 2.0 Review Year Calendar 2016. Report to the Indiana Office of Medicaid Policy and Planning. Phoenix, AZ: Burns & Associates, 2018. Florida Agency for Health Care Administration. Analysis of Potentially Preventable Healthcare Events of Florida Medicaid Enrollees 2015-2016 and 2016-2017. Tallahassee, FL: AHCA, Winter 2018. Medicare Payment Advisory Commission. Mandated report: The effects of the Hospital Readmissions Reduction Program. Chapter 1 in Report to the Congress: Medicare Payment Policy. (Washington, DC: MedPAC, June 2018) North Carolina Department of Health and Human Services. Plan for Implementation of Hospital Quality Outcomes Program and PHP Quality Outcomes Program. Report to the Legislature. Raleigh, ND: NCDHHS, Sept. 28, 2018. Texas Department of State Health Services. Potentially Preventable Readmissions in Texas: Calendar Year 2016 Report. Austin, TX: DSHS, 2018. Fuller RL, Hughes JS, Goldfield NI, Averill RF. Will hospital peer grouping by patient socioeconomic status fix the Medicare hospital readmission reduction program or create new problems? Jt Comm J Qual Patient Saf. 2018;44:177-185. McCoy RG, Peterson SM, Borkenhagen LS, Takahashi PY, Thorsteinsdottir B, Chandra A, Naessens JM. Which readmissions may be preventable? Lessons learned from a posthospitalization care transitions program for high-risk elders. Med Care. 2018;56(8):693 700. Millwee B, Goldfield N, Turnipseed J. Achieving improved outcomes through value-based purchasing in one state. Am J Med Qual. 2018;33(2):162-171. Millwee B, Quinn K, Goldfield N. Moving toward paying for outcomes in Medicaid. J Ambul Care Manage. 2018;41(2):88-94. Mississippi Division of Medicaid. Quality Incentive Payment Program Potentially Preventable Readmissions Methodology Supplement. Jackson, MS: Mississippi Division of Medicaid, 2019. Available at https://medicaid.ms.gov/wp-content/uploads/2020/01/MS-QIPP-ReadmissionsMethodology-Supplement-2019-09.pdf New York Department of Health. DSRIP PAOP Meeting June 24, 2019. Presentation, available at https://www.health.ny.gov/health_care/medicaid/redesign/dsrip/paop/meetings/2019/docs/ 2019-06-24_pm-ff.pdf. Averill RF, Fuller RL, Mills RE. Financial Impact of Geographic Variation in Hospital Quality Performance in Medicare. Murray, UT: 3M Health Information Systems, 2019. Burns & Associates. External Quality Review of Indiana's Care Programs: Hoosier Healthwise, Hoosier Care Connect and HIP Review Year Calendar 2017. Report to the Indiana Office of Medicaid Policy and Planning. Phoenix, AZ: Burns * Associates, 2019. Geographic Variation in Hospital Admission Rates in the Medicare Population 29 Florida Agency for Health Care Administration. Analysis of Potentially Preventable Healthcare Events of Florida Medicaid Enrollees 2015-2016 and 2016-2017. Tallahassee, FL: AHCA, Winter 2018. Medicare Payment Advisory Commission. The effects of the Hospital Readmissions Reduction Program. Chapter 1 in Medicare and the Health Care Delivery System. Report to Congress. Washington, DC: MedPAC, June 2018. New York Department of Health. Delivery System Reform Incentive Payment (DSRIP) Amendment Request. Albany, NY: NYDOH, Sept. 17, 2019. Calsolaro V, Antognoli R, Pasqualetti G, Okoye C, Aquilini F, Cristofano M, Briani S, Monzani F. 30-day potentially preventable hospital readmissions in older patients: clinical phenotype and health care related risk factors. Clin Interv Aging. 2019;14:18511858. Mississippi Division of Medicaid. DOM to phase in quality incentive payment program (QIPP) for hospitals. MS Medicaid Provider Bulletin. 2019;25(3):pp. 1-2 New York Department of Health. Hospital Inpatient Potentially Preventable Readmission (PPR) Rates by Hospital (SPARCS): Beginning 2009 [webpage]. https://healthdata.gov/dataset/hospital-inpatient-potentially-preventable-readmission-pprrates-hospital-sparcs-beginning. Accessed Aug. 14, 2020. Maryland Health Services Cost Review Commission. Final Recommendation for the Readmission Reduction Incentive Program for Rate Year 2022. Baltimore, MD: HSCRC, March 2020 Averill RF, Fuller RL, Mills RE. Geographic Variation in Hospital Quality Performance in Medicare by Disease and Procedure Categories. Supplement to the report: Financial Impact of Geographic Variation in Hospital Quality Performance in Medicare. Murray, UT: 3M Health Information Systems, 2020. Zafar SN, Shah AA, Nembhard C, Wilson LL, Habermann EB, Raoof M, Wasif N. Readmissions after complex cancer surgery: analysis of the Nationwide Readmissions Database. J Oncol Pract. 2018;14(6):e335-345 Lindsey M, Patterson W, Ray K, Roohan P. Potentially Preventable Hospital Readmissions among Medicaid Recipients with Mental Health and/or Substance Abuse Health Conditions Compared with All Others: New York State, 2007. Statistical Brief No. 3. Albany, NY: NY Department of Health,n.d. Patterson W, Lindsey M. Potentially Avoidable Hospitalizations: New York State Medicaid Program, 2009. Statistical Brief #6. Albany, NY: NY Department of Health, n.d. New York Department of Health. DSRIP Stories of Meaningful Change in Patient Health. Albany, n.d. Available at: www.health.ny.gov/health_care/medicaid/redesign/dsrip/2019/docs/stories.pdf. Texas External Quality Review Organization. Quality, Timeliness, and Access to Health Care for Texas Medicaid and CHIP Recipients: Summary of Activities Calendar Year 2017. Austin, TX: Texas EQRO, n.d. Geographic Variation in Hospital Admission Rates in the Medicare Population 30 Websites 3M Health Information Systems. 3M Patient Classification Methodologies. Webpage: www.3m.com/his/methodologies. Accessed Sept. 28, 2020 Florida Agency for Health Care Administration--consumer information. www.floridahealthfinder.gov. Accessed 2020 New York Department of Health--consumer information. https://health.data.ny.gov/. Accessed 2020 Ohio Department of Medicaid Modernize Hospital Payments. https://medicaid.ohio.gov/RESOURCES/Reports-and-Research/-Modernize-HospitalPayments. Accessed 2020 Texas Department of State Health Services--readmissions. www.dshs.texas.gov/thcic/hospitals/Potentially-Preventable-Readmission-Reports/. Accessed 2020 Texas Health and Human Services Commission. www.thlcportal.com. Accessed 2020 Superior Health Plan. 3M Health Information. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 195046-3M-HIS-Resource-Guide-P-508-03202019.pdf Superior Health Plan. 3M HIS Prospective Dashboard User Guide. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 173928-3M-HIS-Dashboard-Training-P-05312018.pdf. All Patient Refined Diagnosis Related Groups (APR DRG) Articles, Reports, and Book Chapters Jones P. A case study in APR DRGs: the Greater Southeast Community Hospital Experience. Manage Care Q. 1994;2(3):48-56. Averill RF, Muldoon JH, Vertrees JC, Goldfield NI, Mullin RL, Finneran EC, Zhang MC, Steinbeck B, Grant T. The evolution of case mix measurement using Diagnosis Related Groups. In: Goldfield N. Physician profiling and risk adjustment. 2nd ed. Gaithersburg, MD: Aspen; 1999. p. 391-454. Franklin PD, Legault JP. Using data to evaluate hospital inpatient mortality. J Nurs Care Qual. 1999;14(1):55-66. Muldoon J. Structure and performance of different DRG classification systems for neonatal medicine. Pediatrics. 1999;103(1 Suppl E):302-18. Goldfield N, Averill R. On "Risk-adjusting acute myocardial infarction mortality: are APR DRGs the right tool?" Health Serv Res. 2000;34(7):1491-1495; discussion 1495-1498. Romano PS, Chan BK. Risk-adjusting acute myocardial infarction mortality: are APR DRGs the right tool? Health Serv Res. 2000;34(7):1469-1489 Averill RF, Goldfield NI, Muldoon J, Steinbeck BA, Grant TM. A closer look at All-Patient Refined DRGs. J AHIMA. 2002;73(1):46-49. Geographic Variation in Hospital Admission Rates in the Medicare Population 31 Lorenzoni I, Cisbani I, Manzoli I, Fantini MP. The evaluation of neonatal case mix using Medicare DRG and APR DRG classification systems. Italian Journal of Pediatrics. 2002;28:225-229. Fantini MP, Cisbani L, Manzoli L, Vertrees J, Lorenzoni I. On the use of administrative databases to support planning activities. The case of the evaluation of neonatal casemix in the EmiliaRomagna region using DRG and APR DRG classification systems. Eur J Public. 2003;13(2):138145. Shen Y. Applying the 3M All Patient Refined Diagnosis Related Groups Grouper to measure inpatient severity in the VA. Med Care. 2003;41(6 Suppl):II103-10 Zhan C, Miller MR. Excess length of stay, charges, and mortality attributable to medical injuries during hospitalization. JAMA. 2003;290(14):1868-1874. Sedman AB, Bahl V, Bunting E, Bandy K, Jones S, Nasr SZ, Schulz K, Campbell DA. Clinical redesign using All Patient Refined Diagnosis Related Groups. Pediatrics. 2004;114;975-969. Fontaine P, Licoppe C, D'Andrea R. International-Refined (IR-DRG) versus 3M All Patient Refined DRG (APR DRG) to describe and predict costs of patients in 42 Belgium hospitals. Proceedings, WHO Family of International Classifications, Tokyo Meeting. http://www3.who.int/icd/ tokyomeeting/documentlist (June 2005), P2-9. Medicare Payment Advisory Commission. Physician-Owned Specialty Hospitals. Report to Congress. Washington, DC: MedPAC, March 2005. Davis MP, Walsh D, LeGrand SB, Lagman Rl, Harrison SB, Rybicki L. The financial benefits of acute inpatient palliative medicine: an inter-institutional comparative analysis by All Patient Refined-Diagnosis Related Group and case mix index. J Support Oncol. 2005;3(4):313-316. Pirson M, Martins D, Jackson T, Dramaix M, Leclercq P. Prospective casemix-based funding, analysis and financial impact of cost outliers in All-Patient Refined Diagnosis Related Groups in three Belgian general hospitals. Eur J Health Econ. 2006;7(1):55-65. Pirson, M., Dramaix, M., Leclercq, P., Jackson, T.: Analysis of cost outliers within APR-DRGs in a Belgian general hospital: two complementary approaches. Health Policy. 2006;76(1):1325. Wynn BO, Scott M. Evaluation of Severity-adjusted DRG Systems: Addendum to the Interim Report. Santa Monica, CA: RAND, 2007. Fay MD, Jackson DA, Vogel BB. Implementation of a severity-adjusted diagnosis-related groups payment system in a large health plan: implications for pay for performance. J Ambul Care Manage. 2007;30(3):211-217. Hayes KJ, Pettengill J, Stensland J. Getting the price right: Medicare payment rates for cardiovascular services. Health Aff (Millwood). 2007;26(1):124-136. Baram D, Daroowalla F, Garcia R, Zhang G, Chen JJ, Healy E, Riaz SA, Richman P. Use of the All Patient Refined-Diagnosis Related Group (APR-DRG) Risk of Mortality score as a severity adjustor in the medical ICU. Clin Med Circ Respirat Pulm Med. 2008;2:1925. Baram D, Daroowalla F, Garcia R, Zhang G, Chen JJ, Healy E, Riaz SA, Richman P. Use of the All Patient Refined-Diagnosis Related Group (APR-DRG) Risk of Mortality score as a severity adjustor in the medical ICU. Clin Med Insights Circ Respir Pulm Med. 2008;2:(1-25). Geographic Variation in Hospital Admission Rates in the Medicare Population 32 Quinn K. New directions in Medicaid payment methods for hospital care. Health Aff (Millwood). 2008;27(1):269-80. Talsma A, Bahl V, Campbell D. Exploratory analyses of the "failure to rescue" measure: evaluation through medical record review. J Nurs Care Qual. 2008;2(3):202-210. Averill R, McCullough E, Hughes J, Goldfield N, Vertrees J, Fuller R. Redesigning the Medicare inpatient PPS to reduce payments to hospitals with high readmission rates. Health Care Financ Rev. 2009;30(4):1-15. Feudtner C, Levin JE, Srivastava R, Goodman DM, Slonim AD, Sharma V, Shah SS, Pati S, Fargason C Jr, Hall M. How well can hospital readmission be predicted in a cohort of hospitalized children? A retrospective, multicenter study. Pediatrics. 2009;123(1):286-293. Kernisan LP, Lee SJ, Boscardin WJ, Landefeld CS, Dudley RA. Association between hospitalreported Leapfrog safe practices scores and inpatient mortality. JAMA. 2009;301(13):13411348. Kozower BD, Ailawadi G, Jones DR, Pates RD, Lau CL, Kron IL, Stukenborg GJ. Predicted risk of mortality models: surgeons need to understand limitations of the University HealthSystems Consortium models. J Am Coll Surg. 2009;209(5):551-556 Lavernia CJ, Laoruengthana A, Contreras JS, Rossi MD. All-Patient Refined Diagnosis-Related Groups in primary arthroplasty. J Arthroplasty. 2009 Sep;24(6 Suppl):19-23. Goldfield N. The evolution of diagnosis-related groups (DRGs): from its beginnings in case-mix and resource use theory, to its implementation for payment and now for its current utilization for quality within and outside the hospital. Qual Manage Health Care. 2010;19(1)3-16. Kelly WP, Wendt SW, Vogel BB. Guiding principles for payment system reform. J Ambul Care Manage. 2010;33(1):29-34. Shahian M, Wolf RE, Iezzoni LI, Kirle L, Normand ST. Variability in the measurement of hospitalwide mortality rates. New Engl J Med. 2010;363(26):2530-2539. Puget Sound Health Alliance. 2011 Report: Use of Resources in High-Volume Hospitalizations. https://wahealthalliance.org/wpcontent/uploads/2013/11/puget_sound_health_alliance_resource_use_report_2011.pdf Mills R, Butler R, McCullough E, Bao M, Averill R. Impact of the transition to ICD-10 on Medicare inpatient hospital payments. Medicare Medicaid Res Rev. 2011;2(2);E1-E13. Quinn K, Davies B. Variation in Payment for Hospital Care in Rhode Island. Report to the Office of Health Insurance Commissioner. Cranston, RI: Xerox State Healthcare; 2012. Myers RP, Hubbard JN, Shaheen AAM, Dixon E, Kaplan GG. Hospital performance reports based on severity adjusted mortality rates in patients with cirrhosis depend on the method of risk adjustment. Ann Hepatol. 2012;11(4):526-535 Shine D. Risk-adjusted mortality: problems and possibilities. Comput Math Methods Med Iezzoni, LI. Coded data from administrative sources. In Iezzoni LI, ed., Risk Adjustment for Measuring Healthcare Outcomes. 4th ed. Chicago: Health Administration Press, 2013 Geographic Variation in Hospital Admission Rates in the Medicare Population 33 Vertrees J, Averill R, Eisenhandler J, Quain A, Switalski J, Gannon D. The Ability of Event-Based Episodes to Explain Variation in Charges and Medicare Payments for Various Post Acute Service Bundles. Report to MedPAC. Wallingford, CT: 3M Health Information Systems, 2013. Vigen G, Coughlin S, Duncan I. Measurement and Performance Healthcare Quality and Efficiency: Resources for Healthcare Professionals. Third update. Society of Actuaries, 2013. Xerox State Healthcare. Medi-Cal DRG Project Policy Design Document. Report to the California Department of Health Care Services. Atlanta: Xerox, 2013. Berry JG, Toomey SL, Zaslavsky AM, Jha AK, Nakamura MM, Klein DJ, Feng JY, Shulman S, Chiang VW, Kaplan W, Hall M, Schuster MA. Pediatric readmission prevalence and variability across hospitals. JAMA. 2013;309(4):372-380. Mull HJ, Chen Q, O'Brien WJ, Shwartz M, Borzecki AM, Hanchate A, Rosen AK. Comparing 2 methods of assessing 30-day readmissions: what is the impact on hospital profiling in the Veterans Health Administration? Med Care. 2013;51(7):589-96. Pirson M, Schenker L, Martins D, Duong D, Chale JJ, Leclerq P. What can we learn from international comparisons of costs by DRG? Eur J Health Econ. 2013;14(1):67-73. Vertrees J, Averill R, Eisenhandler, J, Quain, A, Switalski J. Bundling Post-Acute Care Services into MS-DRG Payments. Medicare Medicaid Res Rev. 2013;3(3):E1-E19 Averill R, Fuller R. Low-cost outliers as alternatives to the two-midnight rule. Healthc Financ Manage. 2014(December) McCullough EC, Sullivan C, Banning P, Goldfield NI, Hughes JS. Challenges and benefits of adding laboratory data to a mortality risk adjustment method. Qual Manage Health Care. 2011;20(4):253-262. Quinn K. After the revolution: DRGs at age 30. Ann Intern Med. 2014;160:426-429. Quinn K, Davies B. Applicability of Hospital-Specific Relative Value (HSRV) DRG Weights. Memorandum to California Department of Health Care Services. West Sacramento, CA: Xerox State Healthcare, 2015. Mellinger JL, Richardson CR, Mathur AK, Volk ML. Variation among United States hospitals in inpatient mortality for cirrhosis. Clin Gastroenterol Hepatol. 2015;13(3):577-584. Mills R, Bulter R, Averill R, McCullough E, Fuller R, Bao, M. The impact of the transition to ICD10 on Medicare inpatient hospital payments. J AHIMA. 2015(February). Quinn K. The 8 basic payment methods in health care. Ann Intern Med. 2015;163(4):300-306. Villwock JA, Goyal P. Early versus delayed treatment of primary epistaxis in the United States. Int Forum Allergy Rhinol. 2014;4:6975. Wissoker D, Garrett B. Designing a Unified Prospective Payment System for Postacute Care. Contractor report. Washington, DC: MedPAC, 2016 Averill RF, Fuller RL. Implementing a site-neutral PPS. Healthc Financ Manag. 2016(April). Fuller RL, Averill RF, Muldoon JH, Hughes JS. Comparison of the properties of regression and categorical risk-adjustment models. J Ambul Care Manage. 2016;39(2):157-165. Geographic Variation in Hospital Admission Rates in the Medicare Population 34 Fuller RL, Averill RF, Muldoon JH, Hughes JS. Response to commentaries on "Comparison of the properties of regression and categorical risk-adjustment models." J Ambul Care Manage. 39(2):175-177. doi:10.1097/JAC.0000000000000147. Leyenaar JK, Ralston SL, Shieh M, Pekow PS, Mangione-Smith R, Lindenauer PK. Epidemioology of pediatric hospitalizations at general hospitals and freestanding children's hospitals in the United States. J Hosp Med. 2016;11(11):743-749. Medicaid and CHIP Payment and Access Commission. Comparing Medicaid Hospital Payment Across States and to Medicare. Washington, DC: MACPAC, 2017. California Department of Health Care Services. Review of SFYs 2013-14 and 2014-15 Utilization and Payment. Sacramento, CA: DHCS, 2017. Navigant Inc. Arkansas DRG Conversion Plan. Report to the Arkansas Department of Human Services. Chicago: Navigant, 2017. Alaska Department of Health and Social Services. AK DHSS Annual Medicaid Reform Report FY 2018. Anchorage, AK: DSS, 2018. Fuller R. An Analysis of Real Price Effects Resulting from Charge Setting Practices in the US Hospital Sector. Highland, MD: Jayne Koskinas Ted Giovanis Foundation for Health and Policy, 2018. Marks T, Gifford K, Perlin S, Byrd M, Beger T. Factors Affecting the Development of Medicaid Hospital Payment Policies--Findings from Structured Interviews in Five States. Report to MACPAC. Lansing, MI: HMA, 2018. Medicaid and CHIP Payment and Access Commission. State Medicaid Payment Policies for Inpatient Hospital Services. Available at https://www.macpac.gov/publication/macpacinpatient-hospital-payment-landscapes/ Fuller RL, Hughes JS, Goldfield NI, Atkinson G. Are we confident of across-hospital mortality comparisons? Am J Med Qual. 2018;33(6):662-664. McCormick PJ, Lin HM, Deiner SG, Levin MA. Validation of the All Patient Refined Diagnosis Related Group (APR-DRG) risk of mortality and severity of illness modifiers as a measure of perioperative risk. J Med Syst. 2018;42(5):81. Deschepper M. Using standard available hospital-wide data in the interpretation and prediction of outcome indicators. Doctoral dissertation, Ghent University. Faculty of Medicine and Health Sciences; 2019. Averill RF, Fuller RL, Mills RE. Financial Impact of Geographic Variation in Hospital Quality Performance in Medicare. Murray, UT: 3M Health Information Systems, 2019. Medicare Payment Advisory Commission. The effects of the Hospital Readmissions Reduction Program. Chapter 1 in Medicare and the Health Care Delivery System. Report to Congress. Washington, DC: MedPAC, June 2018. U.S. Agency for Health Care Research and Quality. AHRQ Quality Indicators: Quality Indicator Empirical Methods. Rockville, MD: AHRQ, 2019. Geographic Variation in Hospital Admission Rates in the Medicare Population 35 Fuller RL, Hughes JS, Atkinson G, Aubry BS. Problematic risk adjustment in National Healthcare Safety Network Measures. Am J Med Qual. 2019:1-8. Lawrence YR, Golan T, Urban D, Hammer L, Amit U, Catane R, Bar J, Goldstein J, Symon Z, Urban G. Effect of hospital volume on mortality rates amongst neutropenic cancer patients within the United States. J Clin Onc.2016;34:15_sup 6600\ Souza J, Santos JV, Canedo VB, Betanzos A, Alves D, Freitas A. Importance of coding comorbidities for APR-DRG assignment: focus on cardiovascular and respiratory diseases. Health Inf Manag. 2019; doi: 10.1177/1833358319840575. [Epub ahead of print] Averill RF. Fuller RL, Mills RE. Surgical Mortality as a Measure of Hospital Quality. Murray, UT: 3M Health Information Systems, 2020. Fuller R, Hughes J. DNR orders known at the time of admission can improve hospital mortality ratings [abstract]. HSR. 2020;55(51):96 Websites Washington Health Alliance. Inpatient Spending Trends in Washington State (February 2020). Webpage: https://www.wacommunitycheckup.org/highlights/inpatient-spending-trends-inwashington-state-february-2020/. Accessed Sept. 28, 2020. Washington Health Alliance. Variation of Pricing for Inpatient Treatments in Washington State. 2019. webpage: https://www.wacommunitycheckup.org/highlights/variation-of-pricing-forinpatient-treatments-in-washington-state/. Accessed Sept. 28, 2020. Illinois DRG Pricing Calculator. https://www.illinois.gov/hfs/MedicalProviders/hospitals/hospitalratereform/Pages/default.asp x. Accessed 2020 Montana Medicaid Inpatient Pricing Calculator. https://medicaidprovider.mt.gov/01#186035117-fee-schedules---hospital---apr-drg. Accessed 2020 RI Medicaid APR-DRG Pricing Calculator. http://www.eohhs.ri.gov/ProvidersPartners/GeneralInformation/ProviderDirectories/Hospitals. aspx. Accessed 2020 3M Health Information Systems. 3M Patient Classification Methodologies. Webpage: www.3m.com/his/methodologies. Accessed Sept. 28, 2020 Arizona Health Care Cost Containment System. AZ APR-DRG Pricing Calculator FY 2020. Available at: www.azahcccs.gov/PlansProviders/RatesAndBilling/FFS/APRDRGrates.html Colorado Department of Health Care Policy and Financing. Inpatient Hospital Payment. [Webpage]. https://www.colorado.gov/pacific/hcpf/inpatient-hospital-payment. Accessed Aug. 14, 2020 Connecticut Department of Social Services. Medicaid Hospital Reimbursement. Webpage: www.ctdssmap.com/CTPortal/Hospital%20Modernization/tabId/143/Default.aspx. Accessed Sept. 28, 2020. Geographic Variation in Hospital Admission Rates in the Medicare Population 36 District of Columbia Department of Health Care Finance. Rates and Reimbursements. Webpage: https://dhcf.dc.gov/page/rates-and-reimbursements. Accessed Aug. 22, 2020. Indiana Department of Health. Hospital Discharge Data [webpage]. www.in.gov/isdh/20624.htm Minnesota Department of Human Services. Payment Methodology for Inpatient Hospitals. Webpage: https://mn.gov/dhs/partners-and-providers/policies-procedures/minnesota-healthcare-programs/provider/types/payment-methodology-for-inpatient-hospitals.jsp. Accessed Sept. 28, 2020 Mississippi Division of Medicaid. Inpatient Hospital Payment Method for Mississippi Medicaid [webpage]. https://medicaid.ms.gov/providers/reimbursement/. Accessed Aug. 14, 2020. Texas Medicaid and Healthcare Partnership. Acute Care Hospital Reimbursement [webpage]. http://www.tmhp.com/resources/rate-and-code-updates/acute-care-hospital-reimbursement. Accessed Oct. 29, 2020.. Washington HealthCareCompare [webpage]. https://www.wahealthcarecompare.com/. Accessed Aug. 17, 2020. Wisconsin Department of Health Services. ForwardHealth Rates and Weights [webpahe]. https://www.forwardhealth.wi.gov/WIPortal/Tab/42/icscontent/Provider/Medicaid/hospital/d rg/drg.htm.spage#. Accessed Aug. 14, 2020. California Department of Health Care Services. https://www.dhcs.ca.gov/provgovpart/Pages/DRG.aspx. Accessed 2020 Florida Agency for Health Care Administration--consumer information. www.floridahealthfinder.gov. Accessed 2020 Illinois Department of Healthcare and Family Services . www.illinois.gov/hfs/MedicalProviders/MedicaidReimbursement/Pages/DRGHICalcuWorkshe et.aspx. Accessed 2020 New York Department of Health--consumer information. https://health.data.ny.gov/. Accessed 2020 New York Department of Health--Medicaid. https://www.health.ny.gov/facilities/hospital/reimbursement/apr-drg/. Accessed 2020 Indiana Medicaid Diagnosis-Related Group Inpatient Reimbursement. https://www.in.gov/medicaid/providers/669.htm. Accessed 2020 Ohio Department of Medicaid Hospital Payment Policy. https://medicaid.ohio.gov/Provider/ProviderTypes/HospitalProviderInformation/HospitalPaym entPolicy. Accessed 2020 North Carolina Community Care Networks, Inc. Clinical Program Analysis. Report to the North Carolina Department of Health and Human Services. Raleigh, NC: NCCC, 2015 Berry JG, Hall M, Cohen E, O'Neill M, Feudtner C. Ways to identify children with medical complexity and the importance of why. J Pediatr. 2015;167(2):229-237. HSR. 20014;39(1):73- DuBard CA, Jacobsen Vann JC, Jackson C. Conflicting readmission rate trends in a high-risk population: implications for performance measurement. Popul Health Manag. 2015;18:351357 Geographic Variation in Hospital Admission Rates in the Medicare Population 37 Jackson C, Shahsahehi M, Wedlake T, DuBard CA. Timeliness of outpatient follow-up: an evidence-based approach for planning after hospital discharge. Ann Fam Med. 2015:13(2):155122. Jones C, Finison K, McGraves-Lloyd, Tremblay T, Mohlman MK, Tanzman B, Hazard M, Maier, Samuelson J. Vermont's community-oriented all-payer medical home model reduces expenditures and utilization while delivering high-quality care. Popul Health Manag. 2015. DOI: 10.1089/pop.2015.0055. Neff JM, Clifton H, Popalisky J, Zhou C. Stratification of children by medical complexity. Acad Pediatr. 2015;15(2):191-196. Pfister DG, Rubin DM, Elkin EE, Neill US, Duck E, Radzyner M, Bach PB. Risk adjusting survival outcomes in hospitals that treat patients with cancer without information on cancer stage. JAMA Oncol. 2015;1(9):1303-1310. Quinn K. The 8 basic payment methods in health care. Ann Intern Med. 2015;163(4):300-306. Florida Agency For Healthcare Administration. Analyzing the Disease Burden of Florida Medicaid Enrollees Using Clinical Risk Groups. Tallahassee, FL: AHCA, Winter 2016. Hileman G, Steele S. Accuracy of Claims-Based Risk Scoring Models. Schaumburg, IL: Society of Actuaries, 2016. DuBard CA. Key Performance Indicators of Cost and Utilization for Medicaid Recipients Enrolled in Community Care of North Carolina. N C Med J. 2016;77(4):297-300. Fuller RL, Goldfield N. Paying for on-patent pharmaceuticals: limit prices and the emerging role of a pay for outcomes approach. J Ambul Care Manage. 2016;39(2):143-149. Fuller RL, Goldfield N. Response to commentaries on "Paying for on-patent pharmaceuticals: limit prices and the emerging role of a pay for outcomes approach". J Ambul Care Manage. 2016;39(2):155-156. Fuller RL, Hughes JS, Goldfield NI. Adjusting population risk for functional health status. Popul Health Manage. 2016;19(2):136-144. Gareau S, Lopez-De Fede A, Loudermilk BL, Cummings TH, Hardin JW, Picklesimer AH, Crouch E, Covington-Kolb S. Group prenatal care results in Medicaid savings with better outcomes: a propensity score analysis of CenteringPregnancy participation in South Carolina. Matern Child Health J. 2016;20(7):13841393. Juhnke C,Bethge S, Mühlbacher AC. A review on methods of risk adjustment and their use in integrated healthcare systems. Int J Integr Care. 2016;16(4):118 Mohlman MK, Tanzman B,Finison K, Pinettte M, Jones C. Impact of medication-assisted treatmernt for opioid addiction on Medicaid expenditures and health services utilization rates in Vermont. J Subst Abuse Treat. 2016;67: 914 Finison K , Mohlman M, Jones C, Pinette M, Jorgenson D, Kinner A, Tremblay T, Gottlieb D. Risk-adjustment methods for all-payer comparative performance reporting in Vermont. BMC Health Serv Res. 2017;17. Geographic Variation in Hospital Admission Rates in the Medicare Population 38 Bednar WR, Axene JW, Liliedahl RL. An Analysis of End-of-Life Costs for Terminally Ill Medicare Fee-for-Service (FFS) Cancer Patients. Schaumburg, Society of Actuaries, 2018. Fuller RL, Goldfield NI, Hughes JS, McCullough EC. Nursing home compare star rankings and the variation in potentially preventable emergency department visits and hospital admissions. Popul Health Manage. Epub ahead of print. July 30, 2018. Averill RF, Fuller RL, Mills RE. Financial Impact of Geographic Variation in Hospital Quality Performance in Medicare. Murray, UT: 3M Health Information Systems, 2019. Connecticut Department of Social Services. Connecticut State Innovation Model Operational Plan Award Year 4. Hartford, CT: DSS, 2019. Vermont Agency of Human Services. Annual Report on The Vermont Blueprint for Health. Report to the Legislature. Burlington, VT; Agency of Human Services, 2020 Vermont Agency of Human Services. Community Health Profiles [webpage]. https://blueprintforhealth.vermont.gov/community-health-profiles. Accessed Aug. 17, 2020. Andrews AL, Bettenhausen J, Hoefgen E, Richardson T,Macy ML; Zima BT, Colvin J; Hall M; Shah SS, Neff NM, Auger KA. Measures of ED Utilization in a National Cohort of Childen. Am J Manag Care. 2020;26(6):267-272. 3M Health Information Systems. 3M Patient Classification Methodologies. Webpage: www.3m.com/his/methodologies. Accessed Sept. 28, 2020 Vermont Agency of Human Services. Hub and Spoke Profiles [webage]. . Annual Report on The Vermont Blueprint for Health. Report to the Legislature. Burlington, VT; Agency of Human Services, 2020 Superior Health Plan. 3M Health Information. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 195046-3M-HIS-Resource-Guide-P-508-03202019.pdf Superior Health Plan. 3M HIS Prospective Dashboard User Guide. Available at https://www.superiorhealthplan.com/content/dam/centene/Superior/Provider/PDFs/SHP_20 173928-3M-HIS-Dashboard-Training-P-05312018.pdf. Geographic Variation in Hospital Admission Rates in the Medicare Population 39 Appendix B: Potentially Preventable Admissions (PPAs) This Appendix gives an overview of the Potentially Preventable Admissions (PPAs), a methodology that can be used to determine the amount of variability in hospital admissions and to estimate the potential magnitude of avoidable hospitalizations. Assign APR DRG Each inpatient admission is assigned to an All Patient Refined Diagnosis Related Group (APR DRG). APR DRGs classify patients according to their reason for admission and severity of illness.1 APR DRGs assign patients to a `base APR DRG' that is determined either by the principal diagnosis, or, for surgical patients, the most important surgical procedure performed in an operating room. The base APR DRG represents the underlying reason for the hospital admission and is used in the PPA logic to identify patients that had candidate PPA events. Determine if reason for the inpatient admission is an ambulatory care sensitive condition Hospital admissions make a large contribution to rising healthcare costs. To the extent that hospital care can be shortened, shifted to the outpatient setting, or eliminated altogether, the cost of healthcare can be reduced. PPAs are hospital admissions for conditions and procedures that could potentially have been dealt with in the outpatient setting. Many PPAs result from inefficiency, lack of adequate access to outpatient care, or inadequate coordination of ambulatory care services. In many cases PPAs are for flare-ups of chronic conditions (e.g., asthma) for which adequate monitoring and follow-up, such as proper medication management, could have avoided. As such, the occurrence of high rates of PPAs within a region or a healthcare system may represent a failure of the ambulatory care system. Inadequate care leading to preventable hospitalizations can occur among individuals living at home and not participating in an integrated delivery system, or among those cared for in a longer term primary care relationship, such as a capitation-based program, accountable care organization, or medical home. Integrated delivery systems should be better able to provide adequate access and coordination over a period of several years and therefore could be expected to have an impact on the rate of hospitalizations for long-term complications, such as chronic renal failure, vision loss, and vascular disease in diabetic patients. In the absence of such long-term arrangements, only acute complications (e.g. asthma) or potentially preventable interventions (such as back procedures for disc rupture) that would not have required years of good quality care might be expected to be preventable. Studies have documented not only that preventable hospitalizations exist, but also that they can be reduced by specific interventions. Guidelines implemented in nursing homes have been shown to decrease the rate of hospital admissions.2 Patients with COPD with higher continuity of care have been shown to have a significantly lower likelihood of avoidable hospitalization.3 The PPA list of 3MTM All Patient Refined Diagnosis Related Groups (APR DRGs) considered as ambulatory sensitive conditions is more comprehensive than the Agency for Healthcare Research and Quality (AHRQ) list of Prevention Quality Indicators (PQIs). PPAs focus on the potentially preventable aspect and are a fairer representation because they exclude those admissions that are not preventable without years of coordinated and integrated care. For example, surgery for Geographic Variation in Hospital Admission Rates in the Medicare Population 40 vascular complications of diabetes (e.g., amputations) are not included because they are not preventable unless appropriate care is given for several years before the admission. These surgeries, in particular, consume significant dollars but neither newly initiated managed care nor can hospitals under any circumstances be held responsible for these procedures. The rate of PPAs are adjusted for the complexity of the patient population whereas the AHRQ Prevention Quality Indicators (PQIs)4 include all patients admitted with diabetes irrespective of the severity of the patient. It is clear that a diabetic who is diet controlled has a different probability of hospital admission as compared to a diabetic patient who is on dialysis. There are two aspects of the increased comprehensiveness of the PPAs: a larger number of diagnoses that are similar to each other (the PPAs consist of similar diagnoses within a specific APR DRG) and a longer list of conditions (e.g., spinal surgery, which is frequently avoidable with medical treatment). In addition, PPAs are more comprehensive than the PQIs in large part because of advances in our understanding of the role coordinated care can play in avoiding admissions together with an appreciation of the fact that the preventability of these admissions should be adjusted for the overall burden of illness of the individual patient. Further, as described below, a focus on identifying excess PPAs by comparing risk adjusted rates of PPAs across providers allows a wider range of conditions to be identified as a PPA. PPA-based initiatives are readily suited for scaling should healthcare entities, such as Accountable Care Organizations (ACOs), with the full responsibility for coordination and preventive services become more commonplace. Further, 3M PPAs evaluate at the diagnosis codes within an APR DRG as preventable or not preventable. For example, cardiac catheterization is considered potentially preventable for patients with a diagnosis of coronary atherosclerosis, but not preventable for patients with an acute myocardial infarction or unstable angina. In summary, the following PQIs have limited utilization in the PPA algorithm: · Long term diabetes complications. While in the long term these conditions could be included in the PPA list, they should not be included in the initial efforts as decrease in these admissions pertain to care that has occurred for years (not just one) before this admission. These are considered to be potentially preventable in settings of integrated or accountable care. · Lower extremity amputation among patients with diabetes. Same as previous indicator. These are considered to be potentially preventable in settings of integrated or accountable care. · Perforated appendix. While perforated (vs. non perforated) appendix does represent an issue pertaining to access to appropriate outpatient services, there are few dollars that can be saved here as these individuals would have had, in any event, an appendectomy. · Low birth weight infants. The empirical data is not as well developed for this indicator.5 This indicator could be added over time and is especially relevant for outcomes management for Medicaid Managed Care organizations, medical homes/accountable care organizations that provide prenatal care. The following PPAs are not included in the Prevention Quality Indicators (PQI) list. A summary of the literature providing support for the inclusion of each PPA is appended. · Seizures. Recent studies have shown that non-adherence to medication, which could be corrected for many individuals with closer follow-up and better education, appears to be Geographic Variation in Hospital Admission Rates in the Medicare Population 41 associated with serious outcomes, increased utilization and costs of inpatient and Emergency Department (ED) services.6 · Migraines. This is an infrequent cause for hospitalization and can be often avoided with appropriate prophylactic and timely therapeutic interventions. · Cardiac catheterization. Many researchers have documented that, "cardiac catheterization is substantially underused among higher-risk patients with acute myocardial infarction (AMI) with appropriate indications but overused among patients with inappropriate indications."7 In addition, most of the time when appropriate, these procedures can be done on an outpatient basis. · Chest pain and abdominal pain. For both chest and abdominal pain there is considerable variation in practice patterns with respect to the necessity of hospitalizations. Many consensus document have been published for the most appropriate evaluation approach for both of these conditions; particularly chest pain.8 · Back procedures for discogenic pain. There is considerable variation in practice patterns for this procedure. It is clear that many of these procedures could be avoided altogether. There is little evidence that treatment interventions work for most individuals with this very common illness.9 · Sickle cell anemia crisis. Recent literature reports on variation in readmissions and the impact of interventions. These interventions include establishment of a dedicated outpatient clinic for adults (Lanzkron et al) and educational interventions (Shahine et al). While the most recent article by Lanzkron from May 2015 was done with adult patients who had sickle cell, an article by Raphael et al from 2013 documented the positive impact of a day hospital on a pediatric sickle cell population.10 · Mental Health and Substance Abuse (MH/SA) disorders. There is ample evidence indicating that adequate outpatient services decreases hospital use.11 We are not including MH/SA admissions for initial inclusion in the PPA list as very often these patients are only admitted once. We are including these APR DRGs for another potentially preventable event, Potentially Preventable Readmissions (PPRs). However MH/SA admissions are included in settings of integrated or accountable care. · Coronary angioplasty, Coronary Artery Bypass Grafts (CABG), other types of angioplasties and grafts. For more than a quarter century, there has been extensive documentation of the variation in practice patterns in these procedures. Among many others, Saleh, Hannan and Ting documented this variation in 200512. Most recently, research on this variation has focused on socioeconomic disparities and the importance of risk adjustment. In addition, a recent article published in 2014 documented the impact that providing data can have on the performance on angioplasties. The same article noted that there are ongoing significant opportunities for improvement.13 Determine if patient was admitted from a residential nursing care facility Additional assignment criteria specific to patients admitted directly from a residential nursing care facility for identifying patients with candidate PPA events. Fever, chest pain, heart disease (mainly heart failure), mental status changes, gastrointestinal bleeding, urinary tract infections, metabolic disturbances, pneumonia, diseases of the skin, and injuries due to falls have been identified as reasons for potentially preventable events. Researchers argue that some of these conditions, such as urinary tract infections, could be more appropriately treated in the nursing home. Other conditions, such as those related to falls or pneumonia may have been avoided by preventing the adverse health event itself. Decreasing potentially preventable events may reduce healthcare Geographic Variation in Hospital Admission Rates in the Medicare Population 42 costs, lessen trauma or complications resulting from medical treatment for nursing home residents, and improve quality of care. Refer to the APR DRG section of this manual for a list of residential nursing care facility sensitive conditions. Patients admitted directly from a residential nursing care facility and assigned to an APR DRG that is on the list of APR DRG residential nursing care facility sensitive conditions are identified as PPA candidates. Residential nursing care facilities are designated as one of the following places of service: SNF, nursing home, Intermediate Care Facility/Individuals with Intellectual Disabilities, residential substance abuse treatment facility, psychiatric residential treatment center, comprehensive inpatient rehabilitation facility. Refer to the place of service section of this manual for detailed logic for residential nursing care facility identification. Determine if the patient was part of an Integrated Delivery System (IDS) PPAs now include additional criteria for patients that belong to an Integrated Delivery System. Reducing potentially preventable admissions to hospitals using Integrated Delivery Systems using a bundled approach is another opportunity for better care coordination and lower spending. It is designed to encourage accountability for cost and quality across a spectrum of care. With bundled payments, fewer potentially preventable admissions will result due to improved transitions between healthcare settings. Providers will need to carefully consider the correct post acute care that their patients would benefit from without compromising patient care. This method eliminates incentive to provide more services that increase revenue and result in fragmented care. A significantly greater number of hospitalizations are potentially preventable if the population is managed by a well-established Integrated Health Delivery System. Coronary artery bypass grafts (CABG) and other vascular interventions such as lower extremity revascularization and lower extremity amputations from peripheral vascular disease. These are considered potentially preventable for two reasons in established integrated delivery systems. First, with population healthcare management and, for example, better diabetes control, fewer vascular interventions are needed. Secondly, for many vascular interventions such as CABGs it is well documented that a percentage of these interventions are completely avoidable.14 Coronary artery bypass grafts and percutaneous cardiac interventions are therefore considered potentially preventable, although whether the patient care organization responsible is judged to be providing inadequate care will depend on how its rates compare with peer organizations. Similarly, mental health admissions are considered potentially preventable, with assessments of quality depending on the rates of those admissions. Established vs. newly formed Integrated Delivery System (IDS) One cannot expect a newly formed IDS to provide coordinated care when first established. For example, one should not expect coordinated care for the chronically mentally disabled in the early stages of a newly formed IDS. In addition, certain complications of chronic illnesses, such as the vascular complications of diabetes, cannot be addressed without years of coordinated care. On the other hand, with expert coordinated care one should expect lower rates of complications from many chronic illnesses. As a consequence, a newly formed IDS should refer to the General Population PPA preventability status. Potentially preventable admissions (PPA) output Potentially Preventable Admissions (PPA) contain a number of outputs including risk status and reason. There are two risk (R) statuses for PPA: At Risk Potentially Preventable (RP) and At Risk Not Potentially Preventable (RN). Inpatient admissions identified as potentially preventable are Geographic Variation in Hospital Admission Rates in the Medicare Population 43 assigned a single reason that best conveys the cause of the PPA assignment. Further detail on the rationale associated with each reason is provided at the end of this section. Potentially Preventable (RP) 21 Potentially Preventable PPA Reasons 0 Not Potentially Preventable 2 Other Facility Quality Indicator 3 Other Facility Patient Safety 4 Other Facility Potential Area of Overuse 5 Other Facility Coordination Mental Health 12 Other Facility Coordination Substance Abuse 7 Other Facility Primary Care Accessibility/Coordination/Management 8 Primary Care Accessibility/Outpatient Coordination/Management 10 Potential area of overuse 50 Established Integrated Delivery System Potential area of Overuse 51 Established Integrated Delivery System Primary Care Accessibility Coordination/Management 52 Established Integrated Delivery System Coordination Substance Abuse 53 Established Integrated Delivery System Coordinated Mental Health For PPA, there are specific APR DRGs that require additional code level detail to determine the potential preventability of an admission. For these APR DRGs, the principal diagnosis is required to make a final determination. If the principal diagnosis for the claim is not considered potentially preventable, the claim will be returned with a status of RN. If the principal diagnosis is considered potentially preventable, the claim will be returned with a status of RP and the relevant reason assigned. Additionally, for APR DRGs that require code level detail, a PPA may not be assigned in some cases due to diagnosis specific age criteria. If the principal diagnosis is potentially preventable but is associated with specific age criteria, the admission is not considered potentially preventable if the patient's age falls within that range. In this case, the claim will be returned with a status of RN. Grouper assignment to one of the following APR DRGs is not compatible with PPA and will output an error return (RX): APR DRG 955 Principal diagnosis invalid as discharge diagnosis APR DRG 956 Ungroupable While Potentially Preventable Admissions are assigned to categories, it should be emphasized that there is cross-over and that some PPAs can belong to more than one category. Some PPAs fit nicely into single category. For example, potentially preventable surgical procedures for back pain secondary to disc rupture clearly belong to the Potential Overuse category, while a hospital admission from a nursing home for trauma clearly belongs to the Patient Safety category. Other PPAs do not fit so clearly into a single category. For example, pulmonary edema/respiratory failure is categorized as a potentially preventable nursing home quality indicator, but could also represent an opportunity for improvement in coordination. Geographic Variation in Hospital Admission Rates in the Medicare Population 44 Several categories of preventable admissions are labeled as applying to Integrated Delivery Systems that could be expected to implement practices and procedures to optimize care for more complex illnesses. Severe mental health conditions, for example, can be difficult to manage by themselves, and can make care for other coexisting chronic illness much more difficult than usual, and can benefit from coordinated care delivered by integrated systems. (Ultimately, we would like to see all individuals become members of Integrated Delivery Systems that link behavioral and physical care together not separately as is too often the case today.) Patients with chest pain can be difficult to deal with in a cost-effective manner, and their care can benefit from a greater degree of coordination and clear communication, so that many such patients can be appropriately treated in an outpatient setting. · Outpatient Coordination Management. Providing medical care for chronic illness is often complex, and failure to deal with complexity with a coordinated approach to care can result in a preventable admission. Patients require multiple resources, treatments, and providers that, in many healthcare settings, are not integrated into a coherent system of care. This fragmentation puts patients with serious or multiple chronic illnesses at risk of experiencing inadequate quality of care and makes their healthcare expenditures substantially higher than for those who have minor or no chronic conditions. Outpatient Coordination and Management refers to services such as case management that serve to streamline these complex services and in so doing improve outcomes and decrease potentially preventable admissions. For example, there is a great deal of literature documenting the positive impact of case management services on hospital admissions for heart failure. · Potential Overuse. Potentially unnecessary healthcare (overutilization, overtreatment) is healthcare provided for conditions and in situations for which its effectiveness has not been proved, or for which evidence has shown a lack of effectiveness. Similarly, overtreatment refers to unnecessary medical interventions. These can include treatment of a self-limited condition, or extensive treatment for a condition that requires only limited treatment. Over diagnosis, when patients are given a diagnosis that will cause no symptoms or harm, can lead to overtreatment. · Primary Care Accessibility. Primary care accessible services can be manifested by short waiting times for urgent needs, extended service hours, around-the-clock telephone or electronic access to a member of the care team, and alternative methods of communication such as email and telephone care. The medical home practice is responsive to patients' preferences regarding access. With accessible services, infections of the upper respiratory tract which can develop into pneumonia can be effectively treated in the outpatient setting. · Quality Indicator. There are circumstances when where the patient has received treatment influences the preventability. If a patient has received treatment in a nursing/psych or rehab facility within 30 days prior to the admission being evaluated for a PPA, then some conditions may be a result of a quality deficit from that previous facility. As per the Institute of Medicine definition quality is defined as the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge. Septicemia or infection in the blood could result from deficiencies in several different categories and thus is placed in the general quality category. · Patient Safety. Patient Safety refers to the reporting, analysis, and prevention of medical error that often leads to adverse healthcare events. Trauma that occurs in the nursing home clearly represents a patient safety issue. Geographic Variation in Hospital Admission Rates in the Medicare Population 45 Proper application of Potentially Preventable Admissions For potentially preventable event measures to be effectively and fairly used in performance reporting and/or pay for performance programs, the measurement tools, scoring methodology, program design and program applications must meet a number of core requirements. The classification methodologies underlying the measurement tools must be clinically precise, comprehensive, have a uniform and consistent structure, and be transparently available to affected providers. The tools must generate information at multiple levels: individual provider, service line, major diagnostic category and at the hospital or health system level. Comparative provider performance must be risk adjusted to account for the severity of patient illness and patient chronic illness burden. Providers should not be evaluated on a case-by-case basis, but via a rate-based approach which motivates providers to achieve performance levels being achieved by their peers. Inpatient expenditures can be reduced by: (1) communicating actionable riskadjusted comparative performance information to providers, and (2) by creating financial incentives focused on reducing the rates of excess potentially preventable admissions.15 The state agency must involve providers and other stakeholders in program design. Finally, patients and their families should be meaningfully engaged in care decisions. References 1. Averill RF, Goldfield NI, Muldoon J, Steinbeck BA, Grant TM. A closer look at All-Patient Refined DRGs. J AHIMA. 2002;73(1):46-49. 2. J.G. Ouslander and S.M. Handler. Consensus-Derived Interventions to Reduce Acute Care Transfer (INTERACT) - Compatible Order Sets for Common Conditions Associated with Potentially Avoidable Hospitalizations. Journal of the American Medicaid Directors Association. 2015 Jun 1; 16 (6): 524-6. 3. I.P. Lin, S.C., Wu, and S.T. Huang. Continuity of care and avoidable hospitalizations for chronic obstructive pulmonary disease (COPD). Journal of the American Board of Family Medicine. 2015 Mar-Apr; 28 (2): 222-30. 4. AHRQ Prevention Quality Indicators http://www.qualityindicators.ahrq.gov/Modules/pqi_resources.aspx. Last accessed October 2016. 5. E.L. Hannan, K. Cozzens, Z. Samadashvili, G. Walford, A.K. Jacobs, D.R. Holmes ,Jr., N.J. Stamato, S. Sharma, F.J. Venditti, I. Fergus, and S.B. King 3rd. Appropriateness of coronary revascularization or patients without acute coronary syndromes. Journal of the American College of Cardiology. 2012 May 22; 59 (21): 1870-6. 6. R.E. Faught, J.R. Weiner, A. Guérin, M.C. Cunnington, and M.S. Duh. Impact of nonadherence to antiepileptic drugs on health care utilization and costs: findings from the RANSOM study. Epilepsia. 2009 Mar; 50 (3): 501-9. 7. D.T. Ko, Y. Wang, D.A. Alter, J.P. Curtis, S.S. Rathore, T.A. Stukel, F.A. Masoudi, J.S. Ross, J.M. Foody, and H.M. Krumholz. Regional variation in cardiac catheterization appropriateness and baseline risk after acute myocardial infarction. Journal of the American College of Cardiology. 2008 Feb 19; 51 (7): 716-23. D.T. Ko, J.S. Ross, Y. Wang, and H.M. Krumholz. Determinants of cardiac catheterization use in older Medicare patients with acute myocardial infarction. Circulation: Cardiovascular Quality and Outcomes. 2010 Jan 1; 3 (1): 54-62. 8 . E.A. Amsterdam, J.D. Kirk, D.A. Bluemke, D. Diercks, M.E. Farkouh, J.L. Garvey, M.C. Kontos, J. McCord, T.D. Miller, A. Morise, L.K. Newby, F.L. Ruberg, K.A. Scordo, and P.D. Thompson; on Geographic Variation in Hospital Admission Rates in the Medicare Population 46 behalf of the American Heart Association Exercise, Cardiac Rehabilitation, and Prevention Committee of the Council on Clinical Cardiology, Council on Cardiovascular Nursing, and Interdisciplinary Council on Quality of Care and Outcomes Research. Testing of Low-Risk Patients Presenting to the Emergency Department with Chest Pain. A Scientific Statement from the American Heart Association. Circulation. 2010 Jul 26. I.E. Hawthorn. Abdominal pain as a cause of acute admission to hospital. Journal of the Royal College of Surgeons of Edinburgh. 1992 Dec; 37 (6): 389-93. T. Bjerkeset, S. Havik, K.E. Aune, and A. Rosseland. Acute abdominal pain as cause of hospitalization. Journal of the Norwegian Medical Association. 2006 Jun 8; 126 (12): 1602-4. 9. Deyo RA, Mirza SK. The case for restraint in spinal surgery: does quality management have a role to play? Eur Spine J. 2009 Aug; 18 Suppl 3:331-7; Deyo RA, Mirza SK, Turner JA, Martin BI. Overtreating chronic back pain: time to back off? J Am Board Fam Med. 2009 Jan-Feb; 22(1):628. 10. M. Givens, C. Rutherford, G. Joshi, and K. Delaney. Impact of an emergency department pain management protocol on the pattern of visits by patients with sickle cell disease. Journal of Emergency Medicine. 2007 Apr; 32 (3): 239-43. M.J. Frei-Jones, J.J. Field, and M.R. DeBaun. Multi-modal intervention and prospective implementation of standardized sickle cell pain admission orders reduces 30-day readmission rate. Pediatric Blood Cancer. 2009 Sep; 53 (3): 4015. S. Lanzkron, C.P. Carroll, P. Hill, M. David, N. Paul, and C. Haywood Jr. Impact of a dedicated infusion clinic for acute management of adults with sickle cell pain crisis. American Journal of Hematology. 2015 May; 90 (5): 376-80. J.L. Raphael, T.L. Rattler, M.A. Kowalkowski, D.C. Brousseau, B.U. Mueller, and T.P. Giordano. Association of care in a medical home and health care utilization among children with sickle cell disease. Journal of the National Medical Association. 2013 Summer; 105 (2): 157-65. 11. S. dosReis, E. Johnson, D. Steinwachs, C. Rohde, E.A. Skinner, M. Fahey, A.F. Lehman. Antipsychotic treatment patterns and hospitalizations among adults with schizophrenia. Schizophrenia Research. 2008 Apr; 101 (1-3): 304-11. 12. S.S. Saleh, E.L. Hannan, and L. Ting. A multistate comparison of patient characteristics, outcomes, and treatment practices in acute myocardial infarction. American Journal of Cardiology. 2005 Nov 1; 96 (9): 1190-6. 13. D.H Howard and Y.C. Shen. Trends in PCI volume after negative results from the COURAGE trial. Health Services Research. 2014 Feb; 49 (1): 153-70. 14. Smith PK, et al. Selection of surgical or percutaneous coronary intervention provides differential longevity benefit. Annals of Thoracic Surgery. 2006;82:1420-28.Weintraub WS, et al. Comparative effectiveness of revascularization strategies. New England Journal of Medicine. 2012;366:1467-76. Chan PS, et al. Appropriate of coronary percutaneous intervention. Journal of the American Medical Association. July 2011;306 (1):53-61. 15. Goldfield N, Kelly WP , Patel K. Potentially Preventable Events: An Actionable Set of Measures for Linking Quality Improvement and Cost Savings. Q Manage Health Care. 2012. 21 (4):213-219. Geographic Variation in Hospital Admission Rates in the Medicare Population 47 Appendix C: Description of CRG Logic Clinical Risk Groups (CRGs) are a categorical clinical model that uses historical claims data to assign individuals to a single mutually exclusive category that defines an individual's chronic disease burden (Hughes, 2004). Each CRG is composed of a base CRG that describes the patient's most significant chronic conditions and two to six explicit severity levels that distinguish differences in disease burden due to severity of illness (e.g., a patient with diabetes and congestive heart failure at severity level 3). The CRG logic follows the logical progression of a disease. The CRG assignment process is as follows: Phase 1: Categorize diagnoses and procedures · All diagnoses are assigned to an MDC (Major Diagnostic Category) · Within each MDCs diagnoses are assigned to one of 557 EDCs (Episode Diagnostic Categories) · All procedures are assigned to one of 640 EPCs (Episode Procedure Category) · Each EDC is categorized as dominant chronic, moderate chronic, minor chronic, chronic manifestation, significant acute or minor acute · Only one diagnosis from an inpatient admission is needed to establish an EDC · Two diagnoses from different days are needed to establish an EDC for outpatient visits except for diagnoses for selected conditions and diagnosis codes which are in fact procedures (e.g., history of a heart transplant) · For inpatient services diagnoses from physician and other professional claims are not used (i.e., only the hospital claim is used). · Diagnoses from "other" providers (e.g., ambulances, freestanding laboratory, etc.) are not used. · Some diagnosis codes create multiple EDCs. (e.g., the diabetic neuropathy code creates both the chronic disease EDC for diabetes and the chronic manifestation EDC for diabetic neuropathy EDC). · Conditionality rules are also applied and affect diagnosis or severity assignment: Persistence and recurrence rules (e.g., hypertension must persist over a period of time to be considered an established diagnosis) Demographic (e.g., congestive heart failure among children vs. adults) · The temporal relationship between EDCs and EPCs is used to establish final EDCs EDCs can cause other EDCs to be "ignored" Acquired hemiplegia removes stroke from contributing to the severity of illness rating EPCs can cause EDC and EPCs to be "ignored" Angioplasty removes Angina from the severity logic Kidney transplant causes renal dialysis to be removed from the severity logic Phase 2: Identify chronic illnesses and specify their severity of illness · Each MDC with a chronic EDC will be assigned a PCD (Primary Chronic Disease) Geographic Variation in Hospital Admission Rates in the Medicare Population 48 · Only one PCD can be assigned per MDC. If there is more than one EDC within an MDC, the PCDs will be selected in hierarchical order within the MDC (e.g., dominant chronic EDCs selected before moderate chronic EDCs) · Some chronic EDCs cannot become PCDs if a certain other EDC is present (e.g., skin ulcers cannot be a PCD if diabetes is present) · After a PCD is selected it is assigned a severity of illness level · The severity level assignment for each PCD is establish by the presence of related conditions (e.g., skin ulcers in a diabetic) Phase 3: Assign the CRG · Assignment to one of 272 base CRGs based on the combination of PCDs that are present · The highest volume diseases or combinations of diseases are assigned a unique base CRG, for example: Diabetes Diabetes with CHF Diabetes with CHF and COPD · All CRGs are assigned to one of nine hierarchical health statuses ranging from catastrophic to healthy Status 1 Healthy Status 2 History of Acute Disease e.g., Chest Pain Status 3 Single Minor Chronic Disease e.g., Migraine Status 4 Minor Chronic Disease in Multiple Organ Systems e.g., Migraine and BPH Status 5 Single Dominant or Moderate Chronic Disease e.g., Diabetes Status 6 - Dominant or Moderate Chronic Disease in Multiple Organ Systems, e.g., Diabetes, and COPD Status 7 - Dominant Chronic Disease in Three or More Organ Systems, e.g., CHF, Diabetes, and COPD Status 8 - Malignancy, Under Active Treatment, e.g., Lung Cancer Status 9 - Catastrophic Conditions, e.g., Major Organ Transplant · Assignment is done from most serious (catastrophic) to least serious (healthy) · Each base CRG is subdivided into discrete severity subclasses based on the severity levels of the PCDs The CRGs (Version 2.1) are composed of 332 base CRGs that describes the individual's most significant chronic conditions and explicit severity levels that distinguish differences in disease burden due to severity of illness resulting in 1,414 individual CRGs. A more detailed description of CRGs is available at: https://apps.3mhis.com/docs/Groupers/ Clinical_Risk_Grouping_CRG/methodology_overview/grp401_crg_v2.1_meth_overview.pdf References Hughes JS, Averill RF, Eisenhandler J, Goldfield NI, Muldoon J, Neff JM, Gay JC. Clinical Risk Groups (CRGs): a classification system for risk-adjusted capitation-based payment and health care management. Med Care. 2004;42(1):81-90. Geographic Variation in Hospital Admission Rates in the Medicare Population 49 Appendix D: Overlap among PPAs, PPRs and ED Admits Any admission can simultaneously be a PPA, PPR and/or an ED Admit. The following Venn diagram shows the overlap among PPAs, PPRs and ED admits. PPR ED Admit AF B D C E G PPA The table below contains the counts for the subsets of admissions in the Venn diagram. Venn Subsets C, D, E G A, D, F, G B, E, F, G A B C D E F G Empty Subset A-G + Empty Description With a PPA With a PPR With a ED Admit PPR Only ED Admit Only PPA Only PPA and PPR Only PPA and ED Admit Only PPR and ED Admit Only PPA and PPR and ED Admit No PPA, PPR or ED Admit Total Admissions Admissions 85,974 25,370 69,611 15,129 40,866 53,984 5,747 24,251 2,502 1,992 235,370 379,841 Percent of Admissions 22.6 6.7 18.3 4.0 10.8 14.2 1.5 6.4 0.7 0.5 62.0 100.0 Geographic Variation in Hospital Admission Rates in the Medicare Population 50 Appendix E: PPA %(A-E)/E and $(A-E) by CBSA State Alabama Alabama Alabama Alabama Alabama Alabama Alabama Alabama Alabama Alaska Alaska Alaska Arizona Arizona Arizona Arizona Arizona Arizona Arkansas Arkansas Arkansas Arkansas Arkansas California California California California California California California California California California California California California California California California California CBSA Birmingham-Hoover, AL Huntsville, AL Montgomery, AL Mobile, AL Daphne-Fairhope-Foley, AL Tuscaloosa, AL Florence-Muscle Shoals, AL Rural Alabama Aggregate small CBSAs Anchorage, AK Rural Alaska Aggregate small CBSAs Phoenix-Mesa-Scottsdale, AZ Tucson, AZ Prescott, AZ Lake Havasu CityKingman, AZ Rural Arizona Aggregate small CBSAs Little Rock-North Little Rock-Conway, AR Fayetteville-SpringdaleRogers, AR-MO Fort Smith, AR-OK Rural Arkansas Aggregate small CBSAs Los Angeles-Long BeachAnaheim, CA San Francisco-OaklandHayward, CA San Diego-Carlsbad, CA Riverside-San BernardinoOntario, CA Sacramento--Roseville-Arden-Arcade, CA San Jose-Sunnyvale-Santa Clara, CA Oxnard-Thousand OaksVentura, CA Fresno, CA Bakersfield, CA Santa Maria-Santa Barbara, CA Stockton-Lodi, CA Salinas, CA San Luis Obispo-Paso Robles-Arroyo Grande, CA Santa Rosa, CA Visalia-Porterville, CA Chico, CA Redding, CA Count Benef 3,829 2,163 1,410 1,331 Count PPAs 282 147 122 103 PPAs per 1000 Benef 73.74 68.14 86.66 77.66 %(A-E)/E PPA Nat Norm -3.58 -8.07 18.91 -4.30 $(A-E) PPA Nat Norm -127,929 -157,852 237,018 -56,609 %(AE)/E PPA BP Norm 10.74 5.58 36.57 9.91 1,084 1,072 54 49.53 108 100.38 -30.13 27.52 -282,387 283,178 -19.75 46.45 1,002 4,471 7,009 1,820 967 668 73 72.61 293 65.59 553 78.90 69 37.86 37 37.98 44 65.62 5.28 -8.84 6.94 -35.74 -25.45 17.05 44,489 -346,696 437,483 -467,532 -152,912 77,852 20.91 4.70 22.82 -26.20 -14.38 34.43 15,542 3,968 2,166 868 55.85 225 56.82 130 59.94 -13.45 -3.92 1.80 -1,644,575 -112,109 27,964 -0.59 10.35 16.92 1,712 789 3,832 98 57.29 51 64.43 222 58.06 -13.94 5.53 -5.31 -193,833 32,490 -152,172 -1.16 21.20 8.75 4,175 299 71.73 11.91 388,651 28.53 2,043 1,545 4,490 6,947 146 71.23 114 73.75 469 104.47 605 87.02 3.80 -1.99 60.47 28.23 64,895 -28,288 2,155,664 1,623,434 19.21 12.56 84.30 47.28 31,567 2,655 84.11 8.68 2,587,016 24.82 14,178 8,500 812 57.24 520 61.15 -9.04 -5.13 -983,679 -343,080 4.47 8.95 8,090 563 69.57 -4.97 -358,693 9.15 7,090 393 55.47 -16.59 -954,344 -4.21 5,096 245 48.06 -20.09 -750,832 -8.22 3,486 3,223 2,427 181 51.81 212 65.65 169 69.74 -23.27 -3.02 -6.60 -667,941 -80,311 -145,915 -11.87 11.38 7.27 2,425 2,251 2,249 122 50.17 120 53.20 121 53.69 -14.72 -22.82 -10.84 -256,123 -431,905 -179,023 -2.06 -11.36 2.40 2,116 2,038 1,977 1,949 1,840 69 32.70 91 44.52 163 82.47 124 63.84 96 52.29 -39.48 -29.80 11.74 -6.17 -8.87 -550,588 -469,674 208,937 -99,851 -114,170 -30.50 -19.37 28.33 7.76 4.67 $(A-E) PPA BP Norm 333,868 94,984 399,050 113,722 -161,200 416,247 153,467 160,558 1,252,932 -298,408 -75,219 136,907 -63,125 257,941 229,069 -14,100 108,452 218,359 810,631 285,994 155,060 2,616,613 2,366,882 6,439,549 423,317 520,903 575,192 -210,765 -267,561 -296,763 263,754 139,858 -31,171 -187,229 34,538 -370,282 -265,882 439,045 109,279 52,309 Geographic Variation in Hospital Admission Rates in the Medicare Population 51 State California California California California California California California California Colorado Colorado Colorado Colorado Colorado Colorado Connecticut Connecticut Connecticut Connecticut Connecticut Delaware Delaware Delaware District of Columbia Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida Florida CBSA Santa Cruz-Watsonville, CA Modesto, CA Vallejo-Fairfield, CA Merced, CA Yuba City, CA Eureka-Arcata-Fortuna, CA Rural California Aggregate small CBSAs Denver-AuroraLakewood, CO Colorado Springs, CO Fort Collins, CO Boulder, CO Rural Colorado Aggregate small CBSAs Hartford-West HartfordEast Hartford, CT Bridgeport-StamfordNorwalk, CT New Haven-Milford, CT Norwich-New London, CT Aggregate small CBSAs Salisbury, MD-DE Dover, DE Rural Delaware Washington-ArlingtonAlexandria, DC-VA-MDWV Miami-Fort LauderdaleWest Palm Beach, FL Tampa-St. PetersburgClearwater, FL Orlando-KissimmeeSanford, FL Jacksonville, FL North Port-SarasotaBradenton, FL Cape Coral-Fort Myers, FL Deltona-Daytona BeachOrmond Beach, FL Palm Bay-MelbourneTitusville, FL Port St. Lucie, FL Naples-Immokalee-Marco Island, FL Pensacola-Ferry PassBrent, FL Lakeland-Winter Haven, FL Ocala, FL The Villages, FL Crestview-Fort Walton Beach-Destin, FL Punta Gorda, FL Homosassa Springs, FL Sebastian-Vero Beach, FL Gainesville, FL Count Benef Count PPAs 1,740 57 1,583 84 1,521 75 1,224 89 1,118 72 1,093 31 2,571 147 6,725 352 6,741 347 2,801 124 1,527 67 1,039 54 2,390 157 4,707 182 4,391 301 4,009 219 3,514 300 1,260 99 970 65 3,928 224 1,300 90 7 0 24,047 16,543 11,547 7,859 6,999 6,535 4,825 3,563 3,413 3,002 2,951 2,621 2,583 2,346 1,777 1,733 1,726 1,447 1,430 1,288 1,538 1,383 982 749 580 320 299 199 277 240 196 150 202 170 108 111 101 119 77 109 PPAs per 1000 Benef %(A-E)/E PPA Nat Norm $(A-E) PPA Nat Norm %(AE)/E PPA BP Norm 32.69 53.05 49.31 72.42 64.03 -40.20 -23.96 -23.23 6.94 -8.31 -466,326 -322,706 -276,851 70,151 -79,103 -31.32 -12.67 -11.83 22.82 5.31 28.44 57.10 52.37 -50.09 -2.06 -17.14 -380,468 -37,571 -888,236 -42.68 12.49 -4.83 51.53 44.25 43.70 52.25 65.62 38.71 -21.68 -28.80 -20.34 -6.34 10.28 -35.62 -1,173,007 -611,402 -207,753 -44,811 178,247 -1,229,291 -10.05 -18.23 -8.51 7.57 26.65 -26.06 68.61 -10.05 -410,576 3.31 54.61 85.32 78.24 67.36 57.05 69.09 0.00 -20.08 9.02 -1.93 -4.67 -17.78 -10.81 -100.00 -671,103 302,425 -23,684 -39,066 -591,131 -132,751 -6,824 -8.22 25.21 12.63 9.48 -5.57 2.44 -100.00 63.96 83.57 85.02 95.31 82.81 49.00 61.93 55.78 81.08 79.81 66.32 57.39 78.24 72.37 60.74 63.78 58.46 82.35 53.72 84.24 -1.19 -224,977 2.95 483,221 2.38 278,229 18.54 8.08 1,428,600 528,718 -26.00 -1,372,219 -9.78 -394,988 -23.22 -733,138 5.55 9.44 177,539 252,106 5.94 133,884 -24.73 -602,727 -0.87 -1.59 -13.61 -21,673 -33,480 -207,287 -15.06 -23.47 15.97 -21.38 14.47 -239,083 -377,437 200,156 -254,793 167,243 13.49 18.24 17.58 36.14 24.13 -15.01 3.62 -11.82 21.23 25.69 21.67 -13.55 13.85 13.02 -0.78 -2.45 -12.11 33.19 -9.71 31.47 $(A-E) PPA BP Norm -316,334 -148,560 -122,773 200,862 44,015 -282,263 198,781 -218,029 -473,537 -336,912 -75,670 46,592 402,499 -783,021 117,590 -239,058 736,091 134,843 69,022 -161,324 26,060 -5,942 2,229,621 2,600,979 1,790,407 2,425,079 1,374,349 -689,799 127,275 -324,908 590,941 597,338 425,215 -287,583 299,799 238,575 -10,290 -33,860 -169,508 362,191 -100,716 316,719 Geographic Variation in Hospital Admission Rates in the Medicare Population 52 State Florida Florida Florida Florida Georgia Georgia Georgia Georgia Georgia Georgia Georgia Hawaii Hawaii Idaho Idaho Idaho Illinois Illinois Illinois Illinois Illinois Illinois Illinois Indiana Indiana Indiana Indiana Indiana Indiana Indiana Iowa Iowa Iowa Iowa Iowa Iowa Kansas Kansas Kansas Kansas Kentucky Kentucky Kentucky Kentucky Louisiana Louisiana Louisiana Louisiana Louisiana CBSA Panama City, FL Tallahassee, FL Rural Florida Aggregate small CBSAs Atlanta-Sandy SpringsRoswell, GA Augusta-Richmond County, GA-SC Savannah, GA Columbus, GA-AL Macon, GA Rural Georgia Aggregate small CBSAs Urban Honolulu, HI Aggregate small CBSAs Boise City, ID Rural Idaho Aggregate small CBSAs Chicago-Naperville-Elgin, IL-IN-WI Peoria, IL Rockford, IL Ottawa-Peru, IL Springfield, IL Rural Illinois Aggregate small CBSAs Indianapolis-CarmelAnderson, IN Evansville, IN-KY South Bend-Mishawaka, IN-MI Fort Wayne, IN Terre Haute, IN Rural Indiana Aggregate small CBSAs Omaha-Council Bluffs, NE-IA Des Moines-West Des Moines, IA Davenport-Moline-Rock Island, IA-IL Cedar Rapids, IA Rural Iowa Aggregate small CBSAs Wichita, KS Topeka, KS Rural Kansas Aggregate small CBSAs Louisville/Jefferson County, KY-IN Lexington-Fayette, KY Rural Kentucky Aggregate small CBSAs New Orleans-Metairie, LA Lafayette, LA Shreveport-Bossier City, LA Baton Rouge, LA Lake Charles, LA Count Benef 1,215 1,064 2,319 3,010 Count PPAs 131 94 240 256 PPAs per 1000 Benef 108.20 88.33 103.70 84.99 %(A-E)/E PPA Nat Norm 50.53 21.44 32.80 10.21 $(A-E) PPA Nat Norm 538,196 202,369 724,333 288,930 %(AE)/E PPA BP Norm 72.88 39.48 52.52 26.57 $(A-E) PPA BP Norm 675,920 324,407 1,009,888 655,008 16,932 1,090 64.39 -7.40 -1,062,320 6.35 794,401 2,798 1,425 1,334 1,053 4,697 11,002 2,851 1,734 2,224 1,347 5,454 239 85.41 60 42.22 65 48.75 126 119.54 411 87.56 879 79.90 118 41.43 53 30.45 108 48.67 69 51.26 255 46.83 24.90 -37.45 -35.78 54.91 14.49 3.87 -38.03 -46.95 -23.42 -7.88 -25.39 581,149 -439,277 -441,881 544,184 634,660 399,913 -884,221 -570,006 -403,798 -72,075 -1,059,726 43.45 -28.16 -26.24 77.92 31.49 19.30 -28.83 -39.07 -12.05 5.79 -14.30 882,876 -287,611 -282,188 672,321 1,201,123 1,734,475 -583,620 -413,034 -180,902 46,126 -519,954 39,412 2,205 1,705 1,120 1,014 4,682 11,303 3,192 120 133 53 92 375 932 81.00 54.29 78.29 47.50 90.25 80.18 82.46 13.11 -22.88 6.16 -26.94 24.07 12.56 16.30 4,511,393 -433,189 94,463 -239,259 216,510 510,749 1,592,797 29.90 -11.43 21.93 -16.09 42.49 29.27 33.57 8,962,364 -188,409 292,737 -124,428 332,831 1,036,702 2,856,651 8,320 1,795 596 71.65 177 98.47 -2.33 32.08 -173,145 523,521 12.18 51.69 789,326 734,550 1,475 1,455 1,143 3,097 10,953 138 93.70 137 94.19 99 86.67 263 84.83 836 76.37 30.26 21.01 5.30 21.85 4.99 391,588 290,229 60,838 574,546 485,252 49.61 38.99 20.94 39.95 20.59 558,904 468,805 209,195 914,525 1,741,549 4,183 278 66.40 0.81 27,055 15.77 461,581 2,829 147 52.07 -17.37 -377,550 -5.09 -96,441 2,086 1,380 6,446 6,613 3,353 1,702 3,442 5,532 149 71.50 68 49.41 367 56.92 383 57.91 221 65.99 106 62.25 263 76.26 383 69.27 9.49 -34.76 -11.91 -10.63 0.74 -9.59 18.00 2.19 157,685 -443,128 -604,864 -555,495 19,954 -137,044 488,325 100,058 25.75 -25.07 1.17 2.64 15.71 3.84 35.52 17.36 372,498 -278,296 51,945 120,237 366,276 47,747 839,136 691,420 6,103 1,929 7,098 7,564 3,355 2,558 473 77.58 127 65.85 589 82.95 585 77.32 306 91.24 197 77.01 4.75 -5.53 7.22 3.77 15.43 0.06 261,952 -90,737 483,381 259,360 499,090 1,417 20.31 8.50 23.14 19.18 32.57 14.92 974,756 121,316 1,349,361 1,148,150 917,277 311,897 2,382 2,366 1,169 204 85.63 158 66.61 94 80.11 7.30 -15.24 7.32 169,299 -345,524 77,880 23.24 -2.65 23.25 469,046 -52,332 215,489 Geographic Variation in Hospital Admission Rates in the Medicare Population 53 State CBSA Louisiana Louisiana Louisiana Maine Maine Maine Maine Maryland Maryland Maryland Maryland Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Massachusetts Michigan Michigan Michigan Michigan Michigan Michigan Michigan Michigan Minnesota Minnesota Minnesota Minnesota Mississippi Mississippi Mississippi Mississippi Mississippi Missouri Missouri Missouri Missouri Missouri Montana Montana Nebraska Nebraska Nebraska Nevada Nevada Nevada Nevada New Hampshire Houma-Thibodaux, LA Rural Louisiana Aggregate small CBSAs Portland-South Portland, ME Bangor, ME Rural Maine Aggregate small CBSAs Baltimore-ColumbiaTowson, MD Hagerstown-Martinsburg, MD-WV Rural Maryland Aggregate small CBSAs Boston-CambridgeNewton, MA-NH Worcester, MA-CT Springfield, MA Barnstable Town, MA Pittsfield, MA Rural Massachusetts Aggregate small CBSAs Detroit-Warren-Dearborn, MI Grand Rapids-Wyoming, MI Lansing-East Lansing, MI Flint, MI Ann Arbor, MI Kalamazoo-Portage, MI Rural Michigan Aggregate small CBSAs Minneapolis-St. PaulBloomington, MN-WI Duluth, MN-WI Rural Minnesota Aggregate small CBSAs Jackson, MS Gulfport-BiloxiPascagoula, MS Tupelo, MS Rural Mississippi Aggregate small CBSAs St. Louis, MO-IL Kansas City, MO-KS Springfield, MO Rural Missouri Aggregate small CBSAs Rural Montana Aggregate small CBSAs Lincoln, NE Rural Nebraska Aggregate small CBSAs Las Vegas-HendersonParadise, NV Reno, NV Rural Nevada Aggregate small CBSAs Manchester-Nashua, NH Count Benef 1,146 2,323 5,048 Count PPAs 102 212 469 PPAs per 1000 Benef 88.79 91.26 92.97 %(A-E)/E PPA Nat Norm 13.11 12.40 18.63 $(A-E) PPA Nat Norm 143,846 285,344 898,961 %(AE)/E PPA BP Norm 29.91 29.10 36.25 $(A-E) PPA BP Norm 285,712 582,771 1,522,791 2,930 1,048 3,354 1,311 140 47.80 54 51.50 207 61.70 68 51.98 -22.82 -28.20 -8.29 -11.44 -504,944 -258,526 -228,137 -107,320 -11.36 -17.54 5.33 1.72 -218,803 -139,988 127,699 14,027 15,554 1,258 80.88 13.70 1,848,829 30.59 3,593,460 1,565 743 2,265 99 62.95 50 67.51 136 59.96 -16.48 -5.61 -18.98 -237,096 -36,347 -387,898 -4.08 8.41 -6.94 -51,081 47,453 -123,595 24,121 4,022 3,433 2,484 1,281 65 697 1,773 281 222 179 65 2 61 73.50 69.90 64.80 71.90 50.92 28.32 87.37 4.07 3.66 -4.08 6.67 -23.65 -56.24 37.90 845,354 121,112 -115,340 136,300 -246,419 -28,855 204,134 19.52 19.06 10.17 22.52 -12.31 -49.74 58.38 3,531,781 548,814 250,363 400,330 -111,685 -22,221 273,771 17,843 1,756 98.43 17.31 3,160,121 34.73 5,521,220 2,932 2,164 1,783 1,420 1,375 5,223 12,954 209 71.22 121 56.00 183 102.52 72 50.59 78 56.85 409 78.34 824 63.61 -4.67 -26.34 17.05 -27.30 -13.18 14.49 -12.89 -124,673 -528,414 324,771 -329,037 -144,718 631,375 -1,487,511 9.49 -15.40 34.43 -16.51 -0.29 31.49 0.04 220,753 -268,996 571,041 -173,209 -2,735 1,194,965 4,273 7,327 1,102 2,287 5,946 2,825 468 63.90 86 77.87 148 64.62 357 60.08 231 81.80 -10.49 18.39 1.25 -11.87 16.64 -668,938 162,584 22,225 -586,772 402,053 2.81 35.97 16.28 1.22 33.96 155,848 276,880 252,412 52,395 714,467 2,053 1,093 5,165 6,225 11,743 8,499 1,829 6,300 7,147 2,731 4,049 1,691 3,005 3,035 193 77 422 535 1,066 716 126 567 473 118 194 92 186 170 93.78 70.33 81.66 85.97 90.82 84.24 69.13 90.05 66.23 43.19 47.93 54.61 61.89 55.97 23.62 3.07 13.14 17.97 14.64 20.50 -2.20 32.59 -7.78 -24.88 -17.05 -6.95 -0.78 -13.45 448,637 27,899 597,340 994,123 1,660,795 1,485,706 -34,657 1,700,690 -487,267 -476,510 -486,338 -84,115 -17,944 -321,895 41.98 18.37 29.94 35.49 31.66 38.40 12.33 52.28 5.91 -13.73 -4.73 6.87 13.95 -0.59 694,247 145,515 1,185,215 1,709,471 3,127,773 2,422,569 169,220 2,375,447 322,221 -228,902 -117,417 72,379 277,671 -12,377 6,782 2,187 233 2,207 2,257 492 72.58 110 50.46 11 45.97 162 73.48 142 63.01 2.95 -20.26 -28.25 15.27 -0.58 171,783 -341,985 -51,426 262,008 -10,136 18.23 -8.42 -17.59 32.39 14.18 925,828 -123,734 -27,884 483,874 215,425 Geographic Variation in Hospital Admission Rates in the Medicare Population 54 State CBSA New Hampshire New Hampshire New Hampshire New Hampshire New Jersey New Jersey New Jersey New Jersey New Mexico New Mexico New Mexico New York New York New York New York New York New York New York New York New York New York North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Carolina North Dakota North Dakota Ohio Ohio Ohio Ohio Ohio Ohio Ohio Ohio Claremont-Lebanon, NHVT Concord, NH Rural New Hampshire Aggregate small CBSAs Allentown-BethlehemEaston, PA-NJ Atlantic City-Hammonton, NJ Trenton, NJ Aggregate small CBSAs Albuquerque, NM Rural New Mexico Aggregate small CBSAs New York-Newark-Jersey City, NY-NJ-PA Albany-SchenectadyTroy, NY Buffalo-CheektowagaNiagara Falls, NY Rochester, NY Syracuse, NY Utica-Rome, NY Binghamton, NY Kingston, NY Rural New York Aggregate small CBSAs Charlotte-ConcordGastonia, NC-SC Virginia Beach-NorfolkNewport News, VA-NC Raleigh, NC Myrtle Beach-ConwayNorth Myrtle Beach, SCNC Asheville, NC Greensboro-High Point, NC Winston-Salem, NC Durham-Chapel Hill, NC Hickory-LenoirMorganton, NC Wilmington, NC Fayetteville, NC New Bern, NC Rocky Mount, NC Rural North Carolina Aggregate small CBSAs Rural North Dakota Aggregate small CBSAs Cleveland-Elyria, OH Cincinnati, OH-KY-IN Columbus, OH Dayton, OH Youngstown-WarrenBoardman, OH-PA Toledo, OH Akron, OH Canton-Massillon, OH Count Benef Count PPAs PPAs per 1000 Benef %(A-E)/E PPA Nat Norm $(A-E) PPA Nat Norm %(AE)/E PPA BP Norm $(A-E) PPA BP Norm 1,867 1,107 550 1,682 101 53.88 42 38.08 28 51.03 83 49.15 -6.93 -38.09 -13.72 -16.32 -91,366 -316,271 -54,422 -196,675 6.89 -28.89 -0.91 -3.90 79,084 -208,901 -3,127 -40,869 4,637 412 88.93 14.77 647,207 31.81 1,213,790 1,747 1,610 1,809 2,776 852 5,780 168 96.20 169 105.15 118 65.02 128 46.22 34 39.66 262 45.37 23.74 39.77 -12.93 -24.13 -28.85 -28.58 393,211 587,459 -213,071 -497,787 -167,120 -1,279,654 42.11 60.52 0.00 -12.87 -18.29 -17.97 607,401 778,473 -32 -231,101 -92,223 -700,650 80,297 5,938 73.96 3.72 2,597,689 19.12 11,626,310 3,426 225 65.66 -8.04 -239,727 5.62 146,006 2,952 2,579 2,484 1,484 1,260 1,140 2,331 7,449 198 67.13 151 58.68 166 66.84 114 76.72 62 49.53 79 69.71 159 68.32 459 61.63 -14.04 -19.90 1.76 6.72 -22.60 8.11 1.20 -12.63 -394,766 -458,601 35,001 87,428 -222,193 72,708 23,045 -809,208 -1.27 -8.01 16.87 22.57 -11.10 24.17 16.23 0.35 -31,203 -160,663 292,287 255,672 -95,049 188,618 271,206 19,394 9,587 645 67.25 -1.52 -121,001 13.11 911,389 8,630 4,508 668 77.41 226 50.21 5.97 458,935 -29.16 -1,136,093 21.71 -18.64 1,453,028 -632,285 4,359 2,997 2,374 2,196 2,166 1,887 1,797 1,624 1,092 1,005 4,746 12,950 1,507 1,498 8,433 7,763 6,168 3,067 2,440 2,401 2,243 1,518 227 51.99 209 69.64 157 66.27 205 93.42 120 55.37 141 74.67 129 71.82 139 85.69 94 85.92 55 54.72 352 74.15 983 75.91 82 54.23 70 46.57 736 87.27 612 78.85 453 73.46 209 68.11 187 76.60 245 101.94 158 70.62 140 92.55 -16.74 7.24 -15.74 23.93 -18.96 1.27 3.37 7.97 19.94 -28.88 2.11 3.09 -8.03 -23.92 15.23 7.38 -0.80 -10.55 4.19 37.70 -8.53 29.06 -555,748 171,910 -358,379 483,098 -342,163 21,523 51,245 125,341 190,223 -272,413 88,724 359,146 -87,053 -267,525 1,186,322 513,393 -44,609 -300,340 91,615 817,296 -180,242 385,818 -4.38 23.17 -3.23 42.33 -6.92 16.31 18.71 24.01 37.75 -18.32 17.27 18.40 5.63 -12.62 32.34 23.33 13.93 2.74 19.66 58.15 5.05 48.23 -126,483 478,828 -63,947 744,129 -108,801 240,920 248,154 328,586 313,578 -150,469 632,185 1,862,860 53,087 -122,928 2,193,468 1,412,313 675,713 67,892 374,508 1,097,603 92,845 557,480 Geographic Variation in Hospital Admission Rates in the Medicare Population 55 State CBSA Ohio Ohio Oklahoma Oklahoma Oklahoma Oklahoma Oregon Oregon Oregon Oregon Oregon Oregon Oregon Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Pennsylvania Rhode Island South Carolina South Carolina South Carolina South Carolina South Carolina South Carolina South Carolina South Carolina South Dakota South Dakota South Dakota Tennessee Tennessee Tennessee Tennessee Tennessee Tennessee Tennessee Tennessee Texas Rural Ohio Aggregate small CBSAs Oklahoma City, OK Tulsa, OK Rural Oklahoma Aggregate small CBSAs Portland-VancouverHillsboro, OR-WA Eugene, OR Medford, OR Salem, OR Bend-Redmond, OR Rural Oregon Aggregate small CBSAs Philadelphia-CamdenWilmington, PA-NJ-DEMD Pittsburgh, PA Scranton--Wilkes-Barre-Hazleton, PA Lancaster, PA Harrisburg-Carlisle, PA York-Hanover, PA Reading, PA Erie, PA ChambersburgWaynesboro, PA Rural Pennsylvania Aggregate small CBSAs Providence-Warwick, RIMA Greenville-AndersonMauldin, SC Columbia, SC Charleston-North Charleston, SC Hilton Head IslandBluffton-Beaufort, SC Spartanburg, SC Florence, SC Rural South Carolina Aggregate small CBSAs Sioux Falls, SD Rural South Dakota Aggregate small CBSAs Nashville-Davidson-Murfreesboro--Franklin, TN Memphis, TN-MS-AR Knoxville, TN Chattanooga, TN-GA Kingsport-Bristol-Bristol, TN-VA Clarksville, TN-KY Rural Tennessee Aggregate small CBSAs Dallas-Fort WorthArlington, TX Count Benef 2,599 14,764 6,195 4,527 4,345 6,311 Count PPAs 273 1,299 412 339 376 479 PPAs per 1000 Benef 105.13 87.99 66.51 74.84 86.64 75.85 %(A-E)/E PPA Nat Norm 36.64 16.34 -8.89 1.32 17.92 3.63 $(A-E) PPA Nat Norm 893,580 2,224,955 -490,519 53,795 697,651 204,765 %(AE)/E PPA BP Norm 56.93 33.61 4.64 16.37 35.43 19.02 $(A-E) PPA BP Norm 1,208,895 3,985,816 222,676 581,076 1,201,094 933,179 5,438 1,463 1,384 1,143 1,113 1,033 5,823 297 54.70 126 86.14 72 52.33 58 50.39 40 35.79 55 53.45 286 49.10 -14.28 48.00 -9.13 -20.58 -35.73 1.03 -19.80 -604,350 498,496 -88,747 -181,997 -270,101 6,832 -861,128 -1.55 69.98 4.36 -8.79 -26.18 16.03 -7.90 -57,140 632,767 36,929 -67,648 -172,355 93,010 -298,911 28,551 6,393 3,479 2,469 2,300 2,092 1,971 1,120 1,033 2,649 10,476 7,654 4,410 4,272 3,956 1,817 1,603 1,310 1,930 3,701 1,190 1,422 2,353 2,126 451 257 143 186 149 159 73 88 152 757 629 269 282 250 115 91 80 98 244 84 82 134 74.46 70.52 73.98 57.76 81.04 71.27 80.88 65.22 85.42 57.25 72.30 82.18 60.96 65.92 63.21 63.02 57.06 60.92 50.73 65.89 70.26 57.45 57.02 1.77 -2.22 -1.53 -18.90 11.44 -1.01 5.75 -8.93 16.36 -18.13 -3.46 10.93 -7.29 2.88 -6.23 23.47 -19.91 -18.34 -25.91 -3.04 20.01 -3.32 -4.03 450,854 -124,990 -48,699 -405,230 233,370 -18,486 105,683 -87,348 151,284 -409,624 -331,110 755,779 -257,817 96,281 -202,624 265,477 -277,308 -218,625 -417,505 -93,172 170,025 -34,253 -68,707 16.88 12.30 13.10 -6.85 27.99 13.69 21.45 4.60 33.64 -5.98 10.88 27.40 6.48 18.16 7.69 41.81 -8.02 -6.21 -14.90 11.36 37.83 11.03 10.22 3,745,115 602,130 363,482 -127,929 497,124 219,019 343,410 39,141 270,871 -117,528 906,135 1,650,013 199,453 527,937 217,885 411,727 -97,206 -64,500 -209,123 303,453 279,882 99,009 151,739 6,119 5,914 4,020 2,787 1,475 1,111 4,372 8,239 20,062 486 381 308 209 98 83 351 583 1,506 79.42 64.39 76.54 74.98 66.67 74.57 80.20 70.79 75.09 9.10 -9.46 6.16 0.67 -11.72 1.03 6.91 -0.46 -2.64 494,237 -485,423 217,810 16,903 -159,199 10,337 276,323 -32,600 -498,342 25.30 3.98 21.93 15.62 1.39 16.04 22.78 14.33 11.82 1,196,670 177,810 674,828 344,236 16,451 139,647 793,492 891,352 1,941,539 Geographic Variation in Hospital Admission Rates in the Medicare Population 56 State Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Texas Utah Utah Utah Utah Utah Utah Vermont Vermont Vermont Virginia Virginia Virginia Virginia Virginia Virginia Virginia Washington Washington Washington Washington Washington Washington Washington Washington Washington Washington West Virginia West Virginia West Virginia West Virginia Wisconsin Wisconsin Wisconsin CBSA Houston-The WoodlandsSugar Land, TX San Antonio-New Braunfels, TX Austin-Round Rock, TX El Paso, TX McAllen-EdinburgMission, TX Killeen-Temple, TX Beaumont-Port Arthur, TX Corpus Christi, TX Tyler, TX Amarillo, TX Brownsville-Harlingen, TX Lubbock, TX Longview, TX Waco, TX Rural Texas Aggregate small CBSAs Salt Lake City, UT Ogden-Clearfield, UT Provo-Orem, UT St. George, UT Rural Utah Aggregate small CBSAs Burlington-South Burlington, VT Rural Vermont Aggregate small CBSAs Richmond, VA Roanoke, VA Lynchburg, VA Charlottesville, VA BlacksburgChristiansburg-Radford, VA Rural Virginia Aggregate small CBSAs Seattle-Tacoma-Bellevue, WA Spokane-Spokane Valley, WA Kennewick-Richland, WA Bremerton-Silverdale, WA Olympia-Tumwater, WA Yakima, WA Port Angeles, WA Bellingham, WA Rural Washington Aggregate small CBSAs Huntington-Ashland, WVKY-OH Charleston, WV Rural West Virginia Aggregate small CBSAs Milwaukee-WaukeshaWest Allis, WI Madison, WI Green Bay, WI Count Benef Count PPAs PPAs per 1000 Benef %(A-E)/E PPA Nat Norm $(A-E) PPA Nat Norm %(AE)/E PPA BP Norm $(A-E) PPA BP Norm 15,102 1,212 80.26 8.48 1,156,171 24.59 2,918,059 7,788 6,053 1,692 449 57.66 429 70.83 98 58.11 -17.31 9.21 -19.95 -1,146,230 440,999 -298,869 -5.03 25.43 -8.06 -289,834 1,060,112 -105,187 1,591 1,582 1,547 1,351 1,184 1,165 1,157 1,105 1,082 1,006 9,249 14,744 2,686 1,933 1,099 1,006 913 732 103 91 83 108 71 104 161 67 104 87 652 1,086 174 87 57 58 72 39 64.43 57.44 53.72 79.85 59.89 88.87 139.06 61.07 95.84 86.83 70.53 73.68 64.61 44.81 51.44 57.70 78.47 53.24 -31.71 -21.22 -34.96 5.14 -19.66 26.80 43.32 -22.88 34.02 24.75 -1.12 -3.44 -1.62 -24.51 -22.16 -5.70 50.24 -12.71 -580,591 -298,449 -544,685 64,294 -211,585 266,856 593,093 -244,216 321,049 211,323 -89,976 -472,573 -34,813 -343,032 -196,323 -42,789 292,175 -69,173 -21.57 -9.52 -25.30 20.75 -7.73 45.63 64.60 -11.43 53.92 43.27 13.57 10.89 12.99 -13.30 -10.61 8.30 72.55 0.26 -343,881 -116,564 -343,212 226,097 -72,416 395,620 770,122 -106,216 443,062 321,742 950,334 1,301,571 243,337 -162,099 -81,797 54,281 367,368 1,223 1,409 1,370 1,450 6,227 2,110 1,934 1,495 77 54.80 63 45.85 66 45.74 483 77.62 134 63.27 139 72.09 114 76.50 -6.24 -17.81 -22.58 12.41 -0.13 3.41 24.06 -62,674 -166,000 -235,927 650,747 -2,183 56,099 270,511 7.68 -5.60 -11.09 29.10 14.70 18.77 42.48 67,187 -45,481 -100,841 1,328,785 208,632 268,697 415,886 1,053 6,458 5,270 12,412 2,968 1,683 1,573 1,348 1,289 1,004 1,001 1,845 6,155 2,340 1,374 2,982 3,411 5,564 2,978 1,033 73 69.06 430 66.52 374 70.90 612 49.29 128 43.26 89 52.86 84 53.10 87 64.66 101 78.23 27 26.88 41 41.37 53 28.89 232 37.62 171 72.90 102 74.14 250 83.97 336 98.52 372 66.84 144 48.32 59 57.10 -2.19 -3.10 -2.40 -19,894 -167,372 -112,126 -19.59 -1,817,238 -27.68 -21.17 -7.27 14.27 22.73 -52.20 -26.91 -44.05 -36.99 -599,354 -291,388 -79,866 132,767 227,806 -359,409 -185,956 -511,818 -1,658,178 -7.65 -2.75 18.87 22.16 -172,377 -35,180 484,752 743,351 -9.15 -14.55 -12.33 -456,772 -298,926 -101,188 12.33 11.29 12.09 -7.64 -16.94 -9.46 6.50 31.24 40.96 -45.10 -16.06 -35.74 -27.64 6.06 11.69 36.52 40.30 4.34 -1.87 0.69 97,361 531,699 491,587 -617,546 -319,374 -113,412 62,171 253,054 357,370 -270,380 -96,615 -361,585 -1,078,620 118,902 130,014 816,935 1,177,153 188,748 -33,364 4,909 Geographic Variation in Hospital Admission Rates in the Medicare Population 57 State Wisconsin Wisconsin Wyoming Wyoming CBSA Rural Wisconsin Aggregate small CBSAs Rural Wyoming Aggregate small CBSAs Count Benef 4,706 8,671 1,483 2,555 Count PPAs 228 509 83 155 PPAs per 1000 Benef 48.45 58.65 56.19 60.54 %(A-E)/E PPA Nat Norm -24.58 -12.36 -2.20 5.19 $(A-E) PPA Nat Norm -906,148 -875,051 -22,846 93,116 %(AE)/E PPA BP Norm -13.38 0.65 12.33 20.81 $(A-E) PPA BP Norm -429,399 40,047 111,521 324,997 Geographic Variation in Hospital Admission Rates in the Medicare Population 58 Health Information Systems 575 West Murray Boulevard Salt Lake City, UT 84123 U.S.A. 800 367 2447 www.3m.com/his 3M is a trademark of 3M Company. Please recycle. Printed in U.S.A. © 3M 2021. All rights reserved. Published 1/21Adobe Acrobat Pro DC 20.13.20074 Adobe Acrobat Pro DC 20.13.20074