Hospital Characteristics Associated With Postdischarge Hospital Readmission, Observation, and Emergency Department Utilization.

BACKGROUND: Whether types of hospitals with high readmission rates also have high overall postdischarge acute care utilization (including emergency department and observation care) is unknown. DESIGN: Cross-sectional analysis. SUBJECTS: Nonfederal United States acute care hospitals. MEASURES: Using methodology established by the Centers for Medicare & Medicaid Services, we calculated each hospital's "excess days in acute care" for fee-for-service (FFS) Medicare beneficiaries aged over 65 years discharged after hospitalization for acute myocardial infarction, heart failure (HF), or pneumonia, representing the mean difference between predicted and expected total days of acute care utilization in the 30 days following hospital discharge, per 100 discharges. We assessed the multivariable association of 8 hospital characteristics with excess days in acute care and the proportion of hospitals with each characteristic that were statistical outliers (95% credible interval estimate does not include 0). RESULTS: We included 2184 hospitals for acute myocardial infarction [228 (10.4%) better than expected, 549 (25.1%) worse than expected], 3720 hospitals for HF [484 (13.0%) better and 840 (22.6%) worse], and 4195 hospitals for pneumonia [673 (16.0%) better, 1005 (24.0%) worse]. Results for all conditions were similar. Worse than expected outliers for pneumonia included: 18.8% of safety net hospitals versus 26.1% of nonsafety net hospitals; 16.7% of public hospitals versus 33.1% of for-profit hospitals; 19.5% of nonteaching hospitals versus 52.2% of major teaching hospitals; 7.9% of rural hospitals versus 42.1% of large urban hospitals; 5.9% of hospitals with 24-<50 beds versus 58% of hospitals with >500 beds; and 29.0% of hospitals with nurse-to-bed ratios >1.0-1.5 versus 21.7% of hospitals with ratios >2.0. CONCLUSIONS: Including emergency department and observation stays in measures of postdischarge utilization produces similar results as measuring only readmissions in that major teaching, urban and for-profit hospitals still perform disproportionately poorly versus nonteaching or public hospitals. However, it enables identification of more outliers and a more granular assessment of the association of hospital factors and outcomes.

Hospital Characteristics Associated With Risk-standardized Readmission Rates.

BACKGROUND: Safety-net and teaching hospitals are somewhat more likely to be penalized for excess readmissions, but the association of other hospital characteristics with readmission rates is uncertain and may have relevance for hospital-centered interventions. OBJECTIVE: To examine the independent association of 8 hospital characteristics with hospital-wide 30-day risk-standardized readmission rate (RSRR). DESIGN: This is a retrospective cross-sectional multivariable analysis. SUBJECTS: US hospitals. MEASURES: Centers for Medicare and Medicaid Services specification of hospital-wide RSRR from July 1, 2013 through June 30, 2014 with race and Medicaid dual-eligibility added. RESULTS: We included 6,789,839 admissions to 4474 hospitals of Medicare fee-for-service beneficiaries aged over 64 years. In multivariable analyses, there was regional variation: hospitals in the mid-Atlantic region had the highest RSRRs [0.98 percentage points higher than hospitals in the Mountain region; 95% confidence interval (CI), 0.84-1.12]. For-profit hospitals had an average RSRR 0.38 percentage points (95% CI, 0.24-0.53) higher than public hospitals. Both urban and rural hospitals had higher RSRRs than those in medium metropolitan areas. Hospitals without advanced cardiac surgery capability had an average RSRR 0.27 percentage points (95% CI, 0.18-0.36) higher than those with. The ratio of registered nurses per hospital bed was not associated with RSRR. Variability in RSRRs among hospitals of similar type was much larger than aggregate differences between types of hospitals. CONCLUSIONS: Overall, larger, urban, academic facilities had modestly higher RSRRs than smaller, suburban, community hospitals, although there was a wide range of performance. The strong regional effect suggests that local practice patterns are an important influence. Disproportionately high readmission rates at for-profit hospitals may highlight the role of financial incentives favoring utilization.

Hospital Use of Observation Stays: Cross-sectional Study of the Impact on Readmission Rates.

BACKGROUND: The Centers for Medicare and Medicaid Services publicly reports hospital risk-standardized readmission rates (RSRRs) as a measure of quality and performance; mischaracterizations may occur because observation stays are not captured by current measures. OBJECTIVES: To describe variation in hospital use of observation stays, the relationship between hospitals observation stay use and RSRRs. MATERIALS AND METHODS: Cross-sectional analysis of Medicare fee-for-service beneficiaries discharged after acute myocardial infarction (AMI), heart failure, or pneumonia between July 2011 and June 2012. We calculated 3 hospital-specific 30-day outcomes: (1) observation rate, the proportion of all discharges followed by an observation stay without a readmission; (2) observation proportion, the proportion of observation stays among all patients with an observation stay or readmission; and (3) RSRR. RESULTS: For all 3 conditions, hospitals' observation rates were <2.5% and observation proportions were <12%, although there was variation across hospitals, including 28% of hospital with no observation stay use for AMI, 31% for heart failure, and 43% for pneumonia. There were statistically significant, but minimal, correlations between hospital observation rates and RSRRs: AMI (r=-0.02), heart failure (r=-0.11), and pneumonia (r=-0.02) (P<0.001). There were modest inverse correlations between hospital observation proportion and RSRR: AMI (r=-0.34), heart failure (r=-0.26), and pneumonia (r=-0.21) (P<0.001). If observation stays were included in readmission measures, <4% of top performing hospitals would be recategorized as having average performance. CONCLUSIONS: Hospitals' observation stay use in the postdischarge period is low, but varies widely. Despite modest correlation between the observation proportion and RSRR, counting observation stays in readmission measures would minimally impact public reporting of performance.

Development and use of an administrative claims measure for profiling hospital-wide performance on 30-day unplanned readmission.

BACKGROUND: Existing publicly reported readmission measures are condition-specific, representing less than 20% of adult hospitalizations. An all-condition measure may better measure quality and promote innovation. OBJECTIVE: To develop an all-condition, hospital-wide readmission measure. DESIGN: Measure development study. SETTING: 4821 U.S. hospitals. PATIENTS: Medicare fee-for-service beneficiaries aged 65 years or older. MEASUREMENTS: Hospital-level, risk-standardized unplanned readmissions within 30 days of discharge. The measure uses Medicare fee-for-service claims and is a composite of 5 specialty-based, risk-standardized rates for medicine, surgery/gynecology, cardiorespiratory, cardiovascular, and neurology cohorts. The 2007-2008 admissions were randomly split for development and validation. Models were adjusted for age, principal diagnosis, and comorbid conditions. Calibration in Medicare and all-payer data was examined, and hospital rankings in the development and validation samples were compared. RESULTS: The development data set contained 8 018 949 admissions associated with 1 276 165 unplanned readmissions (15.9%). The median hospital risk-standardized unplanned readmission rate was 15.8 (range, 11.6 to 21.9). The 5 specialty cohort models accurately predicted readmission risk in both Medicare and all-payer data sets for average-risk patients but slightly overestimated readmission risk at the extremes. Overall hospital risk-standardized readmission rates did not differ statistically in the split samples (P = 0.71 for difference in rank), and 76% of hospitals' validation-set rankings were within 2 deciles of the development rank (24% were more than 2 deciles). Of hospitals ranking in the top or bottom deciles, 90% remained within 2 deciles (10% were more than 2 deciles) and 82% remained within 1 decile (18% were more than 1 decile). LIMITATION: Risk adjustment was limited to that available in claims data. CONCLUSION: A claims-based, hospital-wide unplanned readmission measure for profiling hospitals produced reasonably consistent results in different data sets and was similarly calibrated in both Medicare and all-payer data. PRIMARY FUNDING SOURCE: Centers for Medicare & Medicaid Services.

National patterns of risk-standardized mortality and readmission after hospitalization for acute myocardial infarction, heart failure, and pneumonia: update on publicly reported outcomes measures based on the 2013 release.

BACKGROUND: The Centers for Medicare & Medicaid Services publicly reports risk-standardized mortality rates (RSMRs) within 30-days of admission and, in 2013, risk-standardized unplanned readmission rates (RSRRs) within 30-days of discharge for patients hospitalized with acute myocardial infarction (AMI), heart failure (HF), and pneumonia. Current publicly reported data do not focus on variation in national results or annual changes. OBJECTIVE: Describe U.S. hospital performance on AMI, HF, and pneumonia mortality and updated readmission measures to provide perspective on national performance variation. DESIGN: To identify recent changes and variation in national hospital-level mortality and readmission for AMI, HF, and pneumonia, we performed cross-sectional panel analyses of national hospital performance on publicly reported measures. PARTICIPANTS: Fee-for-service Medicare and Veterans Health Administration beneficiaries, 65 years or older, hospitalized with principal discharge diagnoses of AMI, HF, or pneumonia between July 2009 and June 2012. RSMRs/RSRRs were calculated using hierarchical logistic models risk-adjusted for age, sex, comorbidities, and patients' clustering among hospitals. RESULTS: Median (range) RSMRs for AMI, HF, and pneumonia were 15.1% (9.4-21.0%), 11.3% (6.4-17.9%), and 11.4% (6.5-24.5%), respectively. Median (range) RSRRs for AMI, HF, and pneumonia were 18.2% (14.4-24.3%), 22.9% (17.1-30.7%), and 17.5% (13.6-24.0%), respectively. Median RSMRs declined for AMI (15.5% in 2009-2010, 15.4% in 2010-2011, 14.7% in 2011-2012) and remained similar for HF (11.5% in 2009-2010, 11.9% in 2010-2011, 11.7% in 2011-2012) and pneumonia (11.8% in 2009-2010, 11.9% in 2010-2011, 11.6% in 2011-2012). Median hospital-level RSRRs declined: AMI (18.5% in 2009-2010, 18.5% in 2010-2011, 17.7% in 2011-2012), HF (23.3% in 2009-2010, 23.1% in 2010-2011, 22.5% in 2011-2012), and pneumonia (17.7% in 2009-2010, 17.6% in 2010-2011, 17.3% in 2011-2012). CONCLUSIONS: We report the first national unplanned readmission results demonstrating declining rates for all three conditions between 2009-2012. Simultaneously, AMI mortality continued to decline, pneumonia mortality was stable, and HF mortality experienced a small increase.

Pre-COVID-19 hospital quality and hospital response to COVID-19: examining associations between risk-adjusted mortality for patients hospitalised with COVID-19 and pre-COVID-19 hospital quality.

OBJECTIVES: The extent to which care quality influenced outcomes for patients hospitalised with COVID-19 is unknown. Our objective was to determine if prepandemic hospital quality is associated with mortality among Medicare patients hospitalised with COVID-19. DESIGN: This is a retrospective observational study. We calculated hospital-level risk-standardised in-hospital and 30-day mortality rates (risk-standardised mortality rates, RSMRs) for patients hospitalised with COVID-19, and correlation coefficients between RSMRs and pre-COVID-19 hospital quality, overall and stratified by hospital characteristics. SETTING: Short-term acute care hospitals and critical access hospitals in the USA. PARTICIPANTS: Hospitalised Medicare beneficiaries (Fee-For-Service and Medicare Advantage) age 65 and older hospitalised with COVID-19, discharged between 1 April 2020 and 30 September 2021. INTERVENTION/EXPOSURE: Pre-COVID-19 hospital quality. OUTCOMES: Risk-standardised COVID-19 in-hospital and 30-day mortality rates (RSMRs). RESULTS: In-hospital (n=4256) RSMRs for Medicare patients hospitalised with COVID-19 (April 2020-September 2021) ranged from 4.5% to 59.9% (median 18.2%; IQR 14.7%-23.7%); 30-day RSMRs ranged from 12.9% to 56.2% (IQR 24.6%-30.6%). COVID-19 RSMRs were negatively correlated with star rating summary scores (in-hospital correlation coefficient -0.41, p<0.0001; 30 days -0.38, p<0.0001). Correlations with in-hospital RSMRs were strongest for patient experience (-0.39, p<0.0001) and timely and effective care (-0.30, p<0.0001) group scores; 30-day RSMRs were strongest for patient experience (-0.34, p<0.0001) and mortality (-0.33, p<0.0001) groups. Patients admitted to 1-star hospitals had higher odds of mortality (in-hospital OR 1.87, 95% CI 1.83 to 1.91; 30-day OR 1.46, 95% CI 1.43 to 1.48) compared with 5-star hospitals. If all hospitals performed like an average 5-star hospital, we estimate 38 000 fewer COVID-19-related deaths would have occurred between April 2020 and September 2021. CONCLUSIONS: Hospitals with better prepandemic quality may have care structures and processes that allowed for better care delivery and outcomes during the COVID-19 pandemic. Understanding the relationship between pre-COVID-19 hospital quality and COVID-19 outcomes will allow policy-makers and hospitals better prepare for future public health emergencies.

Incorporating Present-on-Admission Indicators in Medicare Claims to Inform Hospital Quality Measure Risk Adjustment Models.

Importance: Present-on-admission (POA) indicators in administrative claims data allow researchers to distinguish between preexisting conditions and those acquired during a hospital stay. The impact of adding POA information to claims-based measures of hospital quality has not yet been investigated to better understand patient underlying risk factors in the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision setting. Objective: To assess POA indicator use on Medicare claims and to assess the hospital- and patient-level outcomes associated with incorporating POA indicators in identifying risk factors for publicly reported outcome measures used by the Centers for Medicare & Medicaid Services (CMS). Design, Setting, and Participants: This comparative effectiveness study used national CMS claims data between July 1, 2015, and June 30, 2018. Six hospital quality measures assessing readmission and mortality outcomes were modified to include POA indicators in risk adjustment models. The models using POA were then compared with models using the existing complications-of-care algorithm to evaluate changes in risk model performance. Patient claims data were included for all Medicare fee-for-service and Veterans Administration beneficiaries aged 65 years or older with inpatient hospitalizations for acute myocardial infarction, heart failure, or pneumonia within the measurement period. Data were analyzed between September 2019 and March 2020. Main Outcomes and Measures: Changes in patient-level (C statistics) and hospital-level (quintile shifts in risk-standardized outcome rates) model performance after including POA indicators in risk adjustment. Results: Data from a total of 6 027 988 index admissions were included for analysis, ranging from 491 366 admissions (269 209 [54.8%] men; mean [SD] age, 78.2 [8.3] years) for the acute myocardial infarction mortality outcome measure to 1 395 870 admissions (677 158 [48.5%] men; mean [SD] age, 80.3 [8.7] years) for the pneumonia readmission measure. Use of POA indicators was associated with improvements in risk adjustment model performance, particularly for mortality measures (eg, the C statistic increased from 0.728 [95% CI, 0.726-0.730] to 0.774 [95% CI, 0.773-0.776] when incorporating POA indicators into the acute myocardial infarction mortality measure). Conclusions and Relevance: The findings of this quality improvement study suggest that leveraging POA indicators in the risk adjustment methodology for hospital quality outcome measures may help to more fully capture patients' risk factors and improve overall model performance. Incorporating POA indicators does not require extra effort on the part of hospitals and would be easy to implement in publicly reported quality outcome measures.

Administrative Claims Measure for Profiling Hospital Performance Based on 90-Day All-Cause Mortality Following Coronary Artery Bypass Graft Surgery.

BACKGROUND: Coronary artery bypass graft (CABG) surgery is a focus of bundled and alternate payment models that capture outcomes up to 90 days postsurgery. While clinical registry risk models perform well, measures encompassing mortality beyond 30 days do not currently exist. We aimed to develop a risk-adjusted hospital-level 90-day all-cause mortality measure intended for assessing hospital performance in payment models of CABG surgery using administrative data. METHODS: Building upon Centers for Medicare and Medicaid Services hospital-level 30-day all-cause CABG mortality measure specifications, we extended the mortality timeframe to 90 days after surgery and developed a new hierarchical logistic regression model to calculate hospital risk-standardized 90-day all-cause mortality rates for patients hospitalized for isolated CABG. The model was derived from Medicare claims data for a 3-year cohort between July 2014 to June 2017. The data set was randomly split into 50:50 development and validation samples. The model performance was evaluated with C statistics, overfitting indices, and calibration plot. The empirical validity of the measure result at the hospital level was evaluated against the Society of Thoracic Surgeons composite star rating. RESULTS: Among 137 819 CABG procedures performed in 1183 hospitals, the unadjusted mortality rate within 30 and 90 days were 3.1% and 4.7%, respectively. The final model included 27 variables. Hospital-level 90-day risk-standardized mortality rates ranged between 2.04% and 11.26%, with a median of 4.67%. C statistics in the development and validation samples were 0.766 and 0.772, respectively. We identified a strong positive correlation between 30- and 90-day risk-standardized mortality rates, with a regression slope of 1.09. Risk-standardized mortality rates also showed a stepwise trend of lower 90-day mortality with higher Society of Thoracic Surgeons composite star ratings. CONCLUSIONS: We present a measure of hospital-level 90-day risk-standardized mortality rates following isolated CABG. This measure complements Centers for Medicare and Medicaid Services' existing 30-day CABG mortality measure by providing greater insight into the postacute recovery period. It offers a balancing measure to ensure efforts to reduce costs associated with CABG recovery and rehabilitation do not result in unintended consequences.

Impact of left ventricular assist devices and heart transplants on acute myocardial infarction and heart failure mortality and readmission measures.

BACKGROUND: Concern has been raised about consequences of including patients with left ventricular assist device (LVAD) or heart transplantation in readmission and mortality measures. METHODS: We calculated unadjusted and hospital-specific 30-day risk-standardized mortality (RSMR) and readmission (RSRR) rates for all Medicare fee-for-service beneficiaries with a primary diagnosis of AMI or HF discharged between July 2010 and June 2013. Hospitals were compared before and after excluding LVAD and heart transplantation patients. LVAD indication was measured. RESULTS: In the AMI mortality (n = 506,543) and readmission (n = 526,309) cohorts, 1,166 and 1,016 patients received an LVAD while 3 and 2 had a heart transplantation, respectively. In the HF mortality (n = 1,015,335) and readmission (n = 1,254,124) cohorts, 789 and 931 received an LVAD, while 212 and 202 received a heart transplantation, respectively. Less than 2% of hospitals had either ≥6 patients who received an LVAD or, independently, had ≥1 heart transplantation. The AMI mortality and readmission cohorts used 1.8% and 2.8% of LVADs for semi-permanent/permanent indications, versus 73.8% and 78.0% for HF patients, respectively. The rest were for temporary/external indications. In the AMI cohort, RSMR for hospitals without LVAD patients versus hospitals with ≥6 LVADs was 14.8% and 14.3%, and RSRR was 17.8% and 18.3%, respectively; the HF cohort RSMR was 11.9% and 9.7% and RSRR was 22.6% and 23.4%, respectively. In the AMI cohort, RSMR for hospitals without versus with heart transplantation patients was 14.7% and 13.9% and RSRR was 17.8% and 17.7%, respectively; in the HF cohort, RSMR was 11.9% and 11.0%, and RSRR was 22.6% and 22.6%, respectively. Estimations changed ≤0.1% after excluding LVAD or heart transplantation patients. CONCLUSION: Hospitals caring for ≥6 patients with LVAD or ≥1 heart transplantation typically had a trend toward lower RSMRs but higher RSRRs. Rates were insignificantly changed when these patients were excluded. LVADs were primarily for acute-care in the AMI cohort and chronic support in the HF cohort. LVAD and heart transplantation patients are a distinct group with differential care requirements and outcomes, thus should be considered separately from the rest of the HF cohort.

Quality Measure Public Reporting Is Associated with Improved Outcomes Following Hip and Knee Replacement.

BACKGROUND: Given the inclusion of orthopaedic quality measures in the Centers for Medicare & Medicaid Services national hospital payment programs, the present study sought to assess whether the public reporting of total hip arthroplasty (THA) and total knee arthroplasty (TKA) risk-standardized readmission rates (RSRRs) and complication rates (RSCRs) was temporally associated with a decrease in the rates of these outcomes among Medicare beneficiaries. METHODS: Annual trends in national observed and hospital-level RSRRs and RSCRs were evaluated for patients who underwent hospital-based inpatient hip and/or knee replacement procedures from fiscal year 2010 to fiscal year 2016. Hospital-level rates were calculated with use of the same measures and methodology that were utilized in public reporting. Annual trends in the distribution of hospital-level outcomes were then examined with use of density plots. RESULTS: Complication and readmission rates and variation declined steadily from fiscal year 2010 to fiscal year 2016. Reductions of 33% and 25% were noted in hospital-level RSCRs and RSRRs, respectively. The interquartile range decreased by 18% (relative reduction) for RSCRs and by 34% (relative reduction) for RSRRs. The frequency of risk variables in the complication and readmission models did not systematically change over time, suggesting no evidence of widespread bias or up-coding. CONCLUSIONS: This study showed that hospital-level complication and readmission rates following THA and TKA and the variation in hospital-level performance declined during a period coinciding with the start of public reporting and financial incentives associated with measurement. The consistently decreasing trend in rates of and variation in outcomes suggests steady improvements and greater consistency among hospitals in clinical outcomes for THA and TKA patients in the 2016 fiscal year compared with the 2010 fiscal year. The interactions between public reporting, payment, and hospital coding practices are complex and require further study. LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.

Trends in Hospital Readmission of Medicare-Covered Patients With Heart Failure.

BACKGROUND: The Medicare Hospital Readmissions Reduction Program has led to fewer readmissions following hospitalizations with a principal diagnosis of heart failure (HF). Patients with HF are frequently hospitalized for other causes. OBJECTIVES: This study sought to compare trends in Medicare risk-adjusted, 30-day readmissions following principal HF hospitalizations and other hospitalizations with HF. METHODS: This was a retrospective study of 12,973,853 Medicare hospitalizations with a principal or secondary diagnosis of HF between January 2008 and June 2015. Hospitalizations were categorized as follows: principal HF hospitalizations; principal acute myocardial infarction or pneumonia hospitalizations with secondary HF; and other hospitalizations with secondary HF. The study examined trends in risk-adjusted, 30-day, all-cause readmission rates for each cohort and trends in differences in readmission rates among cohorts by using linear spline regression models. RESULTS: Before passage of the Affordable Care Act in March 2010, risk-adjusted, 30-day readmission rates were stable for all 3 cohorts, with mean monthly rates of 26.1%, 24.9%, and 24.4%, respectively. Risk-adjusted readmission rates started declining after passage of the Affordable Care Act by 1.09% (95% confidence interval [CI]: 0.51% to 1.68%), 1.24% (95% CI: 0.92% to 1.57%), and 1.05% (95% CI: 0.52% to 1.58%) per year, respectively, until implementation of the Hospital Readmissions Reduction Program in October 2012 and then stabilized for all 3 cohorts. CONCLUSIONS: Patients with HF are often hospitalized for other causes, and these hospitalizations have high readmission rates. Policy changes led to decreases in readmission rates for both principal and secondary HF hospitalizations. Readmission rates in both groups remain high, suggesting that initiatives targeting all hospitalized patients with HF continue to be warranted.

Comparative Effectiveness of New Approaches to Improve Mortality Risk Models From Medicare Claims Data.

Importance: Risk adjustment models using claims-based data are central in evaluating health care performance. Although US Centers for Medicare & Medicaid Services (CMS) models apply well-vetted statistical approaches, recent changes in the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) coding system and advances in computational capabilities may provide an opportunity for enhancement. Objective: To examine whether changes using already available data would enhance risk models and yield greater discrimination in hospital-level performance measures. Design, Setting, and Participants: This comparative effectiveness study used ICD-9-CM codes from all Medicare fee-for-service beneficiary claims for hospitalizations for acute myocardial infarction (AMI), heart failure (HF), or pneumonia among patients 65 years and older from July 1, 2013, through September 30, 2015. Changes to current CMS mortality risk models were applied incrementally to patient-level models, and the best model was tested on hospital performance measures to model 30-day mortality. Analyses were conducted from April 19, 2018, to September 19, 2018. Main Outcomes and Measures: The main outcome was all-cause death within 30 days of hospitalization for AMI, HF, or pneumonia, examined using 3 changes to current CMS mortality risk models: (1) incorporating present on admission coding to better exclude potential complications of care, (2) separating index admission diagnoses from those of the 12-month history, and (3) using ungrouped ICD-9-CM codes. Results: There were 361 175 hospital admissions (mean [SD] age, 78.6 [8.4] years; 189 225 [52.4%] men) for AMI, 716 790 hospital admissions (mean [SD] age, 81.1 [8.4] years; 326 825 [45.6%] men) for HF, and 988 225 hospital admissions (mean [SD] age, 80.7 [8.6] years; 460 761 [46.6%] men) for pneumonia during the study; mean 30-day mortality rates were 13.8% for AMI, 12.1% for HF, and 16.1% for pneumonia. Each change to the models was associated with incremental gains in C statistics. The best model, incorporating all changes, was associated with significantly improved patient-level C statistics, from 0.720 to 0.826 for AMI, 0.685 to 0.776 for HF, and 0.715 to 0.804 for pneumonia. Compared with current CMS models, the best model produced wider predicted probabilities with better calibration and Brier scores. Hospital risk-standardized mortality rates had wider distributions, with more hospitals identified as good or bad performance outliers. Conclusions and Relevance: Incorporating present on admission coding and using ungrouped index and historical ICD-9-CM codes were associated with improved patient-level and hospital-level risk models for mortality compared with the current CMS models for all 3 conditions.

Development and Testing of Improved Models to Predict Payment Using Centers for Medicare & Medicaid Services Claims Data.

Importance: Predicting payments for particular conditions or populations is essential for research, benchmarking, public reporting, and calculations for population-based programs. Centers for Medicare & Medicaid Services (CMS) models often group codes into disease categories, but using single, rather than grouped, diagnostic codes and leveraging present on admission (POA) codes may enhance these models. Objective: To determine whether changes to the candidate variables in CMS models would improve risk models predicting patient total payment within 30 days of hospitalization for acute myocardial infarction (AMI), heart failure (HF), and pneumonia. Design, Setting, and Participants: This comparative effectiveness research study used data from Medicare fee-for-service hospitalizations for AMI, HF, and pneumonia at acute care hospitals from July 1, 2013, through September 30, 2015. Payments across multiple care settings, services, and supplies were included and adjusted for geographic and policy variations, corrected for inflation, and winsorized. The same data source was used but varied for the candidate variables and their selection, and the method used by CMS for public reporting that used grouped codes was compared with variations that used POA codes and single diagnostic codes. Combinations of use of POA codes, separation of index admission diagnoses from those in the previous 12 months, and use of individual International Classification of Diseases, Ninth Revision, Clinical Modification codes instead of grouped diagnostic categories were tested. Data analysis was performed from December 4, 2017, to June 10, 2019. Main Outcomes and Measures: The models' goodness of fit was compared using root mean square error (RMSE) and the McFadden pseudo R2. Results: Among the 1 943 049 total hospitalizations of the study participants, 343 116 admissions were for AMI (52.5% male; 37.4% aged ≤74 years), 677 044 for HF (45.5% male; 25.9% aged ≤74 years), and 922 889 for pneumonia (46.4% male; 28.2% aged ≤74 years). The mean (SD) 30-day payment was $23 103 ($18 221) for AMI, $16 365 ($12 527) for HF, and $17 097 ($12 087) for pneumonia. Each incremental model change improved the pseudo R2 and RMSE. Incorporating all 3 changes improved the pseudo R2 of the patient-level models from 0.077 to 0.129 for AMI, from 0.042 to 0.129 for HF, and from 0.114 to 0.237 for pneumonia. Parallel improvements in RMSE were found for all 3 conditions. Conclusions and Relevance: Leveraging POA codes, separating index from previous diagnoses, and using single diagnostic codes improved payment models. Better models can potentially improve research, benchmarking, public reporting, and calculations for population-based programs.

Effect of Hospital Readmission Reduction on Patients at Low, Medium, and High Risk of Readmission in the Medicare Population.

BACKGROUND: Hospitalization and readmission rates have decreased in recent years, with the possible consequence that hospitals are increasingly filled with high-risk patients. OBJECTIVE: We studied whether readmission reduction has affected the risk profile of hospitalized patients and whether readmission reduction was similarly realized among hospitalizations with low, medium, and high risk of readmissions. DESIGN: Retrospective study of hospitalizations between January 2009 and June 2015. PATIENTS: Hospitalized fee-for-service Medicare beneficiaries, categorized into 1 of 5 specialty cohorts used for the publicly reported hospital-wide readmission measure. MEASUREMENTS: Each hospitalization was assigned a predicted risk of 30-day, unplanned readmission using a risk-adjusted model similar to publicly reported measures. Trends in monthly mean predicted risk for each cohort and trends in monthly observed to expected readmission for hospitalizations in the lowest 20%, middle 60%, and highest 20% of risk of readmission were assessed using time series models. RESULTS: Of 47,288,961 hospitalizations, 16.2% (n = 7,642,161) were followed by an unplanned readmission within 30 days. We found that predicted risk of readmission increased by 0.24% (P = .03) and 0.13% (P = .004) per year for hospitalizations in the surgery/ gynecology and neurology cohorts, respectively. We found no significant increase in predicted risk for hospitalizations in the medicine (0.12%, P = .12), cardiovascular (0.32%, P = .07), or cardiorespiratory (0.03%, P = .55) cohorts. In each cohort, observed to expected readmission rates steadily declined, and at similar rates for patients at low, medium, and high risk of readmission. CONCLUSIONS: Hospitals have been effective at reducing readmissions across a range of patient risk strata and clinical conditions. The risk of readmission for hospitalized patients has increased for 2 of 5 clinical cohorts.

Variation in the Diagnosis of Aspiration Pneumonia and Association with Hospital Pneumonia Outcomes.

RATIONALE: National efforts to compare hospital outcomes for patients with pneumonia may be biased by hospital differences in diagnosis and coding of aspiration pneumonia, a condition that has traditionally been excluded from pneumonia outcome measures. OBJECTIVES: To evaluate the rationale and impact of including patients with aspiration pneumonia in hospital mortality and readmission measures. METHODS: Using Medicare fee-for-service claims for patients 65 years and older from July 2012 to June 2015, we characterized the proportion of hospitals' patients with pneumonia diagnosed with aspiration pneumonia, calculated hospital-specific risk-standardized rates of 30-day mortality and readmission for patients with pneumonia, analyzed the association between aspiration pneumonia coding frequency and these rates, and recalculated these rates including patients with aspiration pneumonia. RESULTS: A total of 1,101,892 patients from 4,263 hospitals were included in the mortality measure analysis, including 192,814 with aspiration pneumonia. The median proportion of hospitals' patients with pneumonia diagnosed with aspiration pneumonia was 13.6% (10th-90th percentile, 4.2-26%). Hospitals with a higher proportion of patients with aspiration pneumonia had lower risk-standardized mortality rates in the traditional pneumonia measure (12.0% in the lowest coding and 11.0% in the highest coding quintiles) and were far more likely to be categorized as performing better than the national mortality rate; expanding the measure to include patients with aspiration pneumonia attenuated the association between aspiration pneumonia coding rate and hospital mortality. These findings were less pronounced for hospital readmission rates. CONCLUSIONS: Expanding the pneumonia cohorts to include patients with a principal diagnosis of aspiration pneumonia can overcome bias related to variation in hospital coding.

Trends in readmission rates for safety net hospitals and non-safety net hospitals in the era of the US Hospital Readmission Reduction Program: a retrospective time series analysis using Medicare administrative claims data from 2008 to 2015.

OBJECTIVE: To compare trends in readmission rates among safety net and non-safety net hospitals under the US Hospital Readmission Reduction Program (HRRP). DESIGN: A retrospective time series analysis using Medicare administrative claims data from January 2008 to June 2015. SETTING: We examined 3254 US hospitals eligible for penalties under the HRRP, categorised as safety net or non-safety net hospitals based on the hospital's proportion of patients with low socioeconomic status. PARTICIPANTS: Admissions for Medicare fee-for-service patients, age ≥65 years, discharged alive, who had a valid five-digit zip code and did not have a principal discharge diagnosis of cancer or psychiatric illness were included, for a total of 52 516 213 index admissions. PRIMARY AND SECONDARY OUTCOME MEASURES: Mean hospital-level, all-condition, 30-day risk-adjusted standardised unplanned readmission rate, measured quarterly, along with quarterly rate of change, and an interrupted time series examining: April-June 2010, after HRRP was passed, and October-December 2012, after HRRP penalties were implemented. RESULTS: 58.0% (SD 15.3) of safety net hospitals and 17.1% (SD 10.4) of non-safety net hospitals' patients were in the lowest quartile of socioeconomic status. The mean safety net hospital standardised readmission rate declined from 17.0% (SD 3.7) to 13.6% (SD 3.6), whereas the mean non-safety net hospital declined from 15.4% (SD 3.0) to 12.7% (SD 2.5). The absolute difference in rates between safety net and non-safety net hospitals declined from 1.6% (95% CI 1.3 to 1.9) to 0.9% (0.7 to 1.2). The quarterly decline in standardised readmission rates was 0.03 percentage points (95% CI 0.03 to 0.02, p<0.001) greater among safety net hospitals over the entire study period, and no differential change among safety net and non-safety net hospitals was found after either HRRP was passed or penalties enacted. CONCLUSIONS: Since HRRP was passed and penalties implemented, readmission rates for safety net hospitals have decreased more rapidly than those for non-safety net hospitals.

Development and Validation of an Algorithm to Identify Planned Readmissions From Claims Data.

BACKGROUND: It is desirable not to include planned readmissions in readmission measures because they represent deliberate, scheduled care. OBJECTIVES: To develop an algorithm to identify planned readmissions, describe its performance characteristics, and identify improvements. DESIGN: Consensus-driven algorithm development and chart review validation study at 7 acute-care hospitals in 2 health systems. PATIENTS: For development, all discharges qualifying for the publicly reported hospital-wide readmission measure. For validation, all qualifying same-hospital readmissions that were characterized by the algorithm as planned, and a random sampling of same-hospital readmissions that were characterized as unplanned. MEASUREMENTS: We calculated weighted sensitivity and specificity, and positive and negative predictive values of the algorithm (version 2.1), compared to gold standard chart review. RESULTS: In consultation with 27 experts, we developed an algorithm that characterizes 7.8% of readmissions as planned. For validation we reviewed 634 readmissions. The weighted sensitivity of the algorithm was 45.1% overall, 50.9% in large teaching centers and 40.2% in smaller community hospitals. The weighted specificity was 95.9%, positive predictive value was 51.6%, and negative predictive value was 94.7%. We identified 4 minor changes to improve algorithm performance. The revised algorithm had a weighted sensitivity 49.8% (57.1% at large hospitals), weighted specificity 96.5%, positive predictive value 58.7%, and negative predictive value 94.5%. Positive predictive value was poor for the 2 most common potentially planned procedures: diagnostic cardiac catheterization (25%) and procedures involving cardiac devices (33%). CONCLUSIONS: An administrative claims-based algorithm to identify planned readmissions is feasible and can facilitate public reporting of primarily unplanned readmissions.

The performance of US hospitals as reflected in risk-standardized 30-day mortality and readmission rates for medicare beneficiaries with pneumonia.

BACKGROUND: Pneumonia is a leading cause of hospitalization and death in the elderly, and remains the subject of both local and national quality improvement efforts. OBJECTIVE: To describe patterns of hospital and regional performance in the outcomes of elderly patients with pneumonia. DESIGN: Cross-sectional study using hospital and outpatient Medicare claims between 2006 and 2009. SETTING: A total of 4,813 nonfederal acute care hospitals in the United States and its organized territories. PATIENTS: Hospitalized fee-for-service Medicare beneficiaries age 65 years and older who received a principal diagnosis of pneumonia. INTERVENTION: None. MEASUREMENTS: Hospital and regional level risk-standardized 30-day mortality and readmission rates. RESULTS: Of the 1,118,583 patients included in the mortality analysis 129,444 (11.6%) died within 30 days of hospital admission. The median (Q1, Q3) hospital 30-day risk-standardized mortality rate for patients with pneumonia was 11.1% (10.0%, 12.3%), and despite controlling for differences in case mix, ranged from 6.7% to 20.9%. Among the 1,161,817 patients included in the readmission analysis 212,638 (18.3%) were readmitted within 30 days of hospital discharge. The median (Q1, Q3) 30-day risk-standardized readmission rate was 18.2% (17.2%, 19.2%) and ranged from 13.6% to 26.7%. Risk-standardized mortality rates varied across hospital referral regions from a high of 14.9% to a low of 8.7%. Risk-standardized readmission rates varied across hospital referral regions from a high of 22.2% to a low of 15%. CONCLUSIONS: Risk-standardized 30-day mortality and, to a lesser extent, readmission rates for patients with pneumonia vary substantially across hospitals and regions and may present opportunities for quality improvement, especially at low performing institutions and areas.