Alternative Payment Models and Patient-Reported Quality of Preparation for Discharge: A Retrospective Longitudinal Study.

Preparing patients for posthospital care may improve readmission risk. Alternative payment models (APMs) incent hospitals to reduce readmissions by tying payment to outcomes. The impact of APMs on preparation for discharge is not well understood. We assessed whether patient-reported preparation for posthospital care was associated with reduced readmissions, and whether APM participation was associated with improved preparation for posthospital care. We used mixed-effects regression on retrospective (2013-2017) observational data for 2685 U.S. hospitals. We measured patient-reported preparation for posthospital care using the 3-Item Care Transition Measure and readmission using 30-day all-cause risk-adjusted readmissions from Hospital Compare. Participation in accountable care organizations (ACOs), Medical Homes, and Medicare's Bundled Payments for Care Improvement program was obtained from Medicare, the American Hospital Association's Annual Survey, and Leavitt Partner's ACO database. We found that APMs are not associated with improved preparation for posthospital care, even though it was associated with reduced readmissions (Marginal Effect: -0.012 percentage points). This may be because hospitals are not investing in patient engagement. This study has limited insight into causality and reduced generalizability among smaller, rural, and non-teaching hospitals.

The Health Information Technology special issue: making innovations impactful, ethical, and equitable.

A letter from the guest editor highlights how the findings in this special issue draw attention to critical questions that have arisen from health care's digital transformation.

Segregated Patterns of Hospital Care Delivery and Health Outcomes.

Importance: Residential segregation has been shown to be a root cause of racial inequities in health outcomes, yet little is known about current patterns of racial segregation in where patients receive hospital care or whether hospital segregation is associated with health outcomes. Filling this knowledge gap is critical to implementing policies that improve racial equity in health care. Objective: To characterize contemporary patterns of racial segregation in hospital care delivery, identify market-level correlates, and determine the association between hospital segregation and health outcomes. Design, Setting, and Participants: This cross-sectional study of US hospital referral regions (HRRs) used 2018 Medicare claims, American Community Survey, and Agency for Healthcare Research and Quality Social Determinants of Health data. Hospitalization patterns for all non-Hispanic Black or non-Hispanic White Medicare fee-for-service beneficiaries with at least 1 inpatient hospitalization in an eligible hospital were evaluated for hospital segregation and associated health outcomes at the HRR level. The data analysis was performed between August 10, 2022, and September 6, 2023. Exposures: Dissimilarity index and isolation index for HRRs. Main Outcomes and Measures: Health outcomes were measured using Prevention Quality Indicator (PQI) acute and chronic composites per 100 000 Medicare beneficiaries, and total deaths related to heart disease and stroke per 100 000 residents were calculated for individuals aged 74 years or younger. Correlation coefficients were used to compare residential and hospital dissimilarity and residential and hospital isolation. Linear regression was used to examine the association between hospital segregation and health outcomes. Results: This study included 280 HRRs containing data for 4386 short-term acute care and critical access hospitals. Black and White patients tended to receive care at different hospitals, with a mean (SD) dissimilarity index of 23 (11) and mean (SD) isolation index of 13 (13), indicating substantial variation in segregation across HRRs. Hospital segregation was correlated with residential segregation (correlation coefficients, 0.58 and 0.90 for dissimilarity and isolation, respectively). For Black patients, a 1-SD increase in the hospital isolation index was associated with 204 (95% CI, 154-254) more acute PQI hospitalizations per 100 000 Medicare beneficiaries (28% increase from the median), 684 (95% CI, 488-880) more chronic PQI hospitalizations per 100 000 Medicare beneficiaries (15% increase), and 6 (95% CI, 2-9) additional deaths per 100 000 residents (6% increase) compared with 68 (95% CI, 24-113; 6% increase), 202 (95% CI, 131-274; 8% increase), and 2 (95% CI, 0 to 4; 3% increase), respectively, for White patients. Conclusions and Relevance: This cross-sectional study found that higher segregation of hospital care was associated with poorer health outcomes for both Black and White Medicare beneficiaries, with significantly greater negative health outcomes for Black populations, supporting racial segregation as a root cause of health disparities. Policymakers and clinical leaders could address this important public health issue through payment reform efforts and expansion of health insurance coverage, in addition to supporting upstream efforts to reduce racial segregation in hospital care and residential settings.

The use of individual and multilevel data in the development of a risk prediction model to predict patients' likelihood of completing colorectal cancer screening.

Promotion of colorectal cancer (CRC) screening can be expensive and unnecessary for many patients. The use of predictive analytics promises to help health systems target the right services to the right patients at the right time while improving population health. Multilevel data at the interpersonal, organizational, community, and policy levels, is rarely considered in clinical decision making but may be used to improve CRC screening risk prediction. We compared the effectiveness of a CRC screening risk prediction model that uses multilevel data with a more conventional model that uses only individual patient data. We used a retrospective cohort to ascertain the one-year occurrence of CRC screening. The cohort was determined from a Health Maintenance Organization, in Oregon. Eligible patients were 50-75 years old, health plan members for at least one year before their birthday in 2018 and were due for screening. We created a risk model using logistic regression first with data available in the electronic health record (EHR), and then added multilevel data. In a cohort of 59,249 patients, 36.1% completed CRC screening. The individual level model included 14 demographic, clinical and encounter based characteristics, had a bootstrap-corrected C-statistic of 0.722 and sufficient calibration. The multilevel model added 9 variables from clinical setting and community characteristics, and the bootstrap-corrected C-statistic remained the same with continued sufficient calibration. The predictive power of the CRC screening model did not improve after adding multilevel data. Our findings suggest that multilevel data added no improvement to the prediction of the likelihood of CRC screening.

Exit Rates of Accountable Care Organizations That Serve High Proportions of Beneficiaries of Racial and Ethnic Minority Groups.

Importance: The Medicare Shared Savings Program provides financial incentives for accountable care organizations (ACOs) to reduce costs of care. The structure of the shared savings program may not adequately adjust for challenges associated with caring for patients with high medical complexity and social needs, a population disproportionately made up of racial and ethnic minority groups. If so, ACOs serving racial and ethnic minority groups may be more likely to exit the program, raising concerns about the equitable distribution of potential benefits from health care delivery reform efforts. Objective: To evaluate whether ACOs with a high proportion of beneficaries of racial and ethnic minority groups are more likely to exit the Medicare Shared Savings Program and identify characteristics associated with this disparity. Design, Setting, and Participants: This retrospective observational cohort study used secondary data on Medicare Shared Savings Program ACOs from January 2012 through December 2018. Bivariate and multivariate cross-sectional regression analyses were used to understand whether ACO racial and ethnic composition was associated with program exit, and how ACOs with a high proportion of beneficaries of racial and ethnic minority groups differed in characteristics associated with program exit. Exposures: Racial and ethnic composition of an ACO's beneficiaries. Main Outcomes and Measures: Shared savings program exit before 2018. Results: The study included 589 Medicare Shared Savings Program ACOs. The ACOs in the highest quartile of proportion of beneficaries of racial and ethnic minority groups were designated high-proportion ACOs (145 [25%]), and those in the lowest 3 quartiles were designated low-proportion ACOs (444 [75%]). In unadjusted analysis, a 10-percentage point increase in the proportion of beneficiaries of racial and ethnic minority groups was associated with a 1.12-fold increase in the odds of an ACO exit (95% CI, 1.00-1.25; P = .04). In adjusted analysis, there were significant associations among high-proportion ACOs between characteristics such as patient comorbidities, disability, and clinician composition and a higher likelihood of exit. Conclusions and Relevance: The study results suggest that ACOs that served a higher proportion of beneficaries of racial and ethnic minority groups were more likely to exit the Medicare Shared Savings Program, partially because of serving patients with greater disease severity and complexity. These findings raise concerns about how current payment reform efforts may differentially affect racial and ethnic minority groups.

Association Between Oregon's 12-Month Contraceptive Supply Policy and Quantity of Contraceptives Dispensed.

Importance: Eighteen states, including Oregon, have passed legislation requiring insurers to cover dispensation of a 12-month supply of short-acting, hormonal contraception. Objective: To determine whether Oregon's 2016 12-month supply law was associated with an increase in contraceptive supply received. Design Setting and Participants: This retrospective cohort study of hormonal contraceptive users using Oregon's All Payer All Claims database examined the quantity of contraceptive supply dispensed 3 years before and 3 years after the 2016 policy change. We also assessed changes among patients attributed to Title X clinics. Exposures: Legislation requiring insurers in Oregon to cover a 12-month supply of contraception to continuing users. Main Outcomes and Measures: Receipt of a 12-month supply of hormonal contraception. Results: This cohort study of insured users (mean [SD] age, 27.4 [2.1] years) of short-acting hormonal contraception included 639 053 contraceptive prescriptions. Results indicated that more than 80% of prescriptions for contraceptives cover 3 months or fewer. Most women in the study population used the oral contraceptive pill, lived in a metropolitan area, and were privately insured. We did not observe a significant association between the passage of the 12-month supply policy and receipt of a 12-month supply (aOR, 1.01; 95% CI, 0.74-1.38). Receipt of a 12-month supply was more common for Medicaid recipients than the privately insured (aOR, 9.40; 95% CI, 6.62-13.34). We did find a shift from 1 month to 2 to 3 months supply being dispensed. The policy change was associated with a small, overall increase in quantity dispensed (0.27 months supply; 95% CI, 0.15 to -0.38). Title X clinics prescribed 3 months more of contraceptive supply than non-Title X clinics (3.03 months supply; 95% CI, 2.64-3.41). However, the policy change was not associated with increased contraceptive supply dispensed at Title X clinics. Conclusions and Relevance: In this cohort study of insured users of short-acting hormonal contraception, the passage of a 12-month contraceptive supply policy was not associated with an increase in contraceptive supply dispensed.

Variation in State COVID-19 Disease Reporting Forms on Social Identity, Social Needs, and Vaccination Status.

To assess variation in the way state health departments collect data on social identity, social needs, and vaccination status for people testing positive for COVID-19, we reviewed all state health department Web sites for COVID-19 disease reporting forms. We located forms for 39 states and coded each one for the presence and format of fields on race, ethnicity, disability, language, sexual and gender identity, comorbidities, pregnancy status, and social needs such as housing stability, occupation, and prior COVID-19 vaccination status. We find considerable variation in both type and format of data collected. For example, 10% of state forms have fields for race and ethnicity that are more disaggregated than the 5 US Census categories, 18% have fields on mental or physical ability, 37% have nonbinary or other fields for gender identity, 28% have fields on housing stability, and 15% have fields on vaccination status. Our findings suggest that national efforts to address COVID-19 may be limited by lack of standardization of data collection methods.

Trends In Outpatient Mental Health Services Use Before And During The COVID-19 Pandemic.

The COVID-19 pandemic disrupted mental health services delivery across the US, but the extent and implications of these disruptions are unclear. This retrospective observational analysis used the claims clearinghouse Office Ally to compare outpatient mental health services use from March to December 2016-18 against use during the same period in 2020. We identified encounters for people ages twelve and older with primary diagnosis codes corresponding to mental health conditions and categorized encounters as in-person or telehealth, using Current Procedural Terminology and place-of-service codes. In-person mental health encounters were reduced by half in the early months of the pandemic, with rapid recovery of service delivery attributable to telehealth uptake (accounting for 47.9 percent of average monthly encounters). We found variation in the degree to which telehealth use increased across groups: People with schizophrenia made up a lower proportion of telehealth encounters relative to in-person visits (1.7 percent versus 2.7 percent), whereas those with anxiety and fear-related disorders accounted for a higher proportion (27.5 percent versus 25.5 percent). These findings highlight the importance of broadening access to services through new modalities without supplanting necessary in-person care for certain groups.

Rapid adoption of electronic health record and health information exchange among assisted living communities, 2010-2018.

Adoption of electronic health records (EHRs) and health information exchange (HIE) is a key tool to improving the quality of care in assisted living communities (ALC). We examined whether EHRs were being used in ALC to support HIE in 2010 and 2018. We found that adoption of EHR and HIE functions increased substantially over the study period. However, adoption of HIE functions lagged significantly behind EHR functions in both 2010 and 2018 and was accompanied by growing disparities in the adoption of EHR functions among smaller, nonchain, and for-profit communities. To improve the quality of care for this important and growing population, targeted policies are needed to support the adoption of both EHR and HIE functions in ALC.

Out-Of-Network Primary Care Is Associated With Higher Per Beneficiary Spending In Medicare ACOs.

Despite expectations that Medicare accountable care organizations (ACOs) would curb health care spending, their effect has been modest. One possible explanation is that ACOs' inability to prohibit out-of-network care limits their control over spending. To examine this possibility, we examined the association between out-of-network care and per beneficiary spending using national Medicare data for 2012-15. While there was no association between out-of-network specialty care and ACO spending, each percentage-point increase in receipt of out-of-network primary care was associated with an increase of $10.79 in quarterly total ACO spending per beneficiary. When we broke down total spending by place of service, we found that out-of-network primary care was associated with higher spending in outpatient, skilled nursing facility, and emergency department settings, but not inpatient settings. Our findings suggest an opportunity for the Medicare program to realize substantial savings, if policy makers developed explicit incentives for beneficiaries to seek more of their primary care within network.

Are Patients Electronically Accessing Their Medical Records? Evidence From National Hospital Data.

Substantial policy effort has been directed at improving patients' ability to access and use electronic health records. Using nationwide data from 2,410 hospitals for the period 2014-16, we examined associations between patient- and hospital-level characteristics and access to and use of electronic health record data among discharged patients. On average, hospitals gave 95 percent of discharged patients access to view, download, and transmit their information, but only about 10 percent of those with access used it-levels that were stagnant during the study period. Access rates were highest among system-member, teaching, and for-profit hospitals. In contrast, access rates were lower for hospitals in the highest quartile for disproportionate share hospital status and for hospitals located in counties with high proportions of residents who were dually eligible for Medicare and Medicaid; use rates were lower for hospitals in counties with a high proportion of residents who were dually eligible, lacked computer or internet access, or were Hispanic. Overall, our findings suggest that policy efforts have failed to engage a large proportion of patients in the electronic use of their data or to bridge the "digital divide" that accompanies health care disparities. Additional-possibly targeted-policy incentives, as well as higher thresholds for meeting the requirements of the Promoting Interoperability Program, merit policy makers' consideration.

Alternative payment models and hospital engagement in health information exchange.

OBJECTIVES: To assess whether hospital participation in alternative payment models (APMs) is associated with greater engagement in health information exchange (HIE) along 4 dimensions: volume of patients for whom information is exchanged, diversity of information types, breadth of partner types, and depth of technical approach. STUDY DESIGN: Pooled, cross-sectional analysis of data on US hospitals from 2014 to 2015. METHODS: APM participation came from Leavitt Partners data, Medicare public use files, and the American Hospital Association (AHA) Annual Survey. We used Medicare data to measure HIE volume for 798 hospitals attesting to stage 2 Meaningful Use and the AHA Information Technology Supplement to measure HIE diversity, breadth, and depth for 1730 hospitals. We used mixed-effects regression to estimate the association between participation in APMs and each dimension of HIE. RESULTS: Compared with nonparticipating hospitals, full-year APM participation was associated with lower HIE volume (data were sent for 11 percentage points fewer discharges; P = .003), greater HIE diversity (of 4 data types, 0.3 more were transmitted; P <.001), greater HIE breadth (of 3 partner types, data were sent to 0.3 more; P <.001), and greater HIE depth (the odds of using a push and pull approach were 1.68 times greater; P = .004). CONCLUSIONS: Our finding that APM participation was associated with greater HIE diversity, breadth, and depth suggests that value-based payment may be spurring improvements in HIE infrastructure. However, our finding that APM participation is associated with lower HIE volume suggests that there may be an incentive to focus HIE investments on a limited number of partners.

Coordination of Care Around Surgery for Colon Cancer: Insights From National Patterns of Physician Encounters With Medicare Beneficiaries.

PURPOSE: To improve care coordination for complex cancers, it is critical to establish a more nuanced understanding of the types of providers involved. As the number of provider types increases, strategies to support cancer care coordination must adapt to a greater variety of information needs, communication styles, and treatment strategies. METHODS: We categorized providers into 11 types, using National Provider Identifier specialties. Using Medicare claims, we counted the number of unique combinations of provider types billed during preoperative, operative, and postdischarge care for colon cancer surgery and assessed how this count varies across hospitals. The study included 70,567 beneficiaries in fee-for-service Medicare A and B for 6 months before and 60 days after an admission for colectomy for colon cancer between 2008 and 2011. RESULTS: We observed 1,554 preoperative provider-type combinations, 975 operative combinations, and 1,571 postdischarge combinations. The three most common combinations in the preoperative phase were general medicine only, other medical specialists only, and general medicine and other medical specialists. In the operative phase, the three most common combinations were primary surgery, anesthesiology, and pathology; general medicine, other medical specialists, radiology, primary surgery, anesthesiology, and pathology; and other medical specialists, radiology, primary surgery, anesthesiology, and pathology. In the postdischarge phase, the three most common combinations were general medicine, general medicine and other medical specialists, and general medicine and oncology. On average, each hospital had 15 preoperative, 11 operative, and 15 postoperative combinations. High-volume, larger, teaching, urban, and noncritical access hospitals had more combinations in all phases. CONCLUSION: Many provider-type combinations are involved in colon cancer surgery care. Substantial variation exists across hospitals types, suggesting that certain hospitals need additional resources and more flexible infrastructure to coordinate care.

Electronic Health Records Associated With Lower Hospital Mortality After Systems Have Time To Mature.

Evidence linking electronic health record (EHR) adoption to better care is mixed. More nuanced measures of adoption, particularly those that capture the common incremental approach of adding functions over time in US hospitals, could help elucidate the relationship between adoption and outcomes. We used data for the period 2008-13 to assess the relationship between EHR adoption and thirty-day mortality rates. We found that baseline adoption was associated with a 0.11-percentage-point higher rate per function. Over time, maturation of the baseline functions was associated with a 0.09-percentage-point reduction in mortality rate per year per function. Each new function adopted in the study period was associated with a 0.21-percentage-point reduction in mortality rate per year per function. We observed effect modification based on size and teaching status, with small and nonteaching hospitals realizing greater gains. These findings suggest that national investment in hospital EHRs should yield improvements in mortality rates, but achieving them will take time.

The relationship between hospital and ehr vendor market dynamics on health information organization presence and participation.

BACKGROUND: Health Information Organizations (HIOs) are third party organizations that facilitate electronic health information exchange (HIE) between providers in a geographic area. Despite benefits from HIE, HIOs have struggled to form and subsequently gain broad provider participation. We sought to assess whether market-level hospital and EHR vendor dynamics are associated with presence and level of hospital participation in HIOs. METHODS: 2014 data on 4523 hospitals and their EHR vendors were aggregated to the market level. We used multivariate OLS regression to analyze the relationship between hospital and vendor dynamics and (1) probability of HIO presence and (2) percent of hospitals participating in an HIO. RESULTS: 298 of 469 markets (64%) had HIO presence, and in those markets, 47% of hospitals participated in an HIO on average. In multivariate analysis, four characteristics were associated with HIO presence. Markets with more hospitals, markets with more EHR vendors, and markets with an EHR vendor-led HIE approach were more likely to have an HIO. Compared to markets with low hospital competition, markets with high hospital competition had a 25 percentage point lower probability of HIO presence. Two characteristics were associated with level of hospital HIO participation. Markets with more hospitals as well as markets with high vendor competition (compared to low competition) had lower participation. CONCLUSION: Both hospital and EHR vendor dynamics are associated with whether a market has an HIO as well as the level of hospital participation in HIOs.

Technology, Incentives, or Both? Factors Related to Level of Hospital Health Information Exchange.

OBJECTIVE: To assess whether the level of health information exchange (HIE) in U.S. hospitals is related to technology capabilities, incentives to exchange, or both. STUDY SETTING: A total of 1,812 hospitals attesting to stage 2 of Medicare's Meaningful Use Incentive Program through April 2016. STUDY DESIGN: Hospital-level, multivariate OLS regression with state fixed effects was used to analyze the relationship between technology capability and incentives measures, and percent of care transitions with summary of care records (SCRs) sent electronically to subsequent providers. PRINCIPAL FINDINGS: Stage 2 hospitals reported sending SCRs electronically for an average of 41 percent (median = 33 percent) of transitions. HIE level is related to four capability measures, one incentive measure, and one measure that is related to both capability and incentive. Percent of transitions with SCRs sent electronically was 3 percentage points higher (95 percent CI: 0.1-5.1) for hospitals with a third-party HIE vendor, 3 percentage points higher (95 percent CI: 0.5-5.4) for hospitals with an EHR vendor as their HIE vendor, and 3 percentage points higher (95 percent CI: 0.4-5.4) for hospitals that automatically alert primary care providers. The direction and statistical significance of the relationships between specific EHR vendor and electronic SCR transmission level varied by vendor. Nonprofits and government hospitals performed 5 percentage points higher (95 percent CI: 1.5-9.1) and 8 percentage points higher (95 percent CI: 3.4-12.3) than for-profits. Hospitals in systems performed 3 percentage points higher (95 percent CI: 0.8-6.1). CONCLUSION: The overall level of HIE is low, with hospitals sending an SCR electronically for less than half of patient transitions. Specific hospital characteristics related to both technology capabilities and incentives were associated with higher levels of HIE.

The Number Of Health Information Exchange Efforts Is Declining, Leaving The Viability Of Broad Clinical Data Exchange Uncertain.

The diffusion of health information exchange (HIE), in which clinical data are electronically linked to patients in many different care settings, is a top priority for policy makers. To drive HIE, community and state efforts were federally funded to broadly engage providers in exchanging data in ways that improved patient care. To assess the current landscape, we conducted a national survey of community and state HIE efforts soon after federal funding ended. We found 106 operational HIE efforts that, as a group, engaged more than one-third of all US providers in 2014. However, the number of operational HIE efforts is down from 119 in 2012, representing the first decline observed since the tracking of these efforts began in 2006. Only half of operational efforts reported being financially viable, and all efforts reported a variety of barriers to continuation. These findings raise important questions about whether the current vision for HIE efforts will allow for the broad exchange of clinical data, or whether alternative approaches would be more successful.

Assessing payer perspectives on health information exchange.

OBJECTIVE: To identify factors that impede payer engagement in a health information exchange (HIE), along with organizational and policy strategies that might effectively address the impediments. MATERIALS AND METHODS: Qualitative analysis of semi-structured interviews with leaders from 17 varied payer organizations from across the country (e.g., large, national payers; state Blues plans; local Medicaid managed care plans). RESULTS: We found a large gap between payers' vision of what optimal HIE should be and the current approach to HIE in the United States. Notably, payers sought to be active participants in HIE efforts--both providing claims data and accessing clinical data to support payer HIE use cases. Instead, payers were often asked by HIE efforts only to provide financial support without the option to participate in data exchange, or, when given the option, their data needs were secondary to those of providers. DISCUSSION: Efforts to engage payers in pursuit of more robust and sustainable HIE need to better align their value proposition with payer HIE use cases. This will require addressing provider concerns about payer access to clinical data. Policymakers should focus on creating the conditions for broader payer engagement by removing common obstacles, such as low provider engagement in HIE. CONCLUSION: Despite variation in the extent to which payers engaged with current HIE efforts, there was agreement on the vision of optimal HIE and the facilitators of greater payer engagement. Specific actions by those leading HIE efforts, complemented by policy efforts nationally, could greatly increase payer engagement and enhance HIE sustainability.