Medical Care最新文献

Background: The Department of Veterans Affairs (VA) offers a 1-year Post-Baccalaureate-Registered Nurse Residency (PB-RNR) Program. The impact of the PB-RNR program on local RN recruitment was unknown.

Objectives: We aimed to evaluate the effect of the size of the PB-RNR program at a VA facility on its time-to-fill RN vacancies.

Project design: We used an instrumental variable approach with a 2-stage residual inclusion specification.

Subjects: We included RN filled vacancies in the VA that were posted nationwide between 2020 and 2021.

Measures: Our independent variable was the facility-year level number of PB-RNR program allocations. The 3 binary outcomes were whether the RN vacancy was filled within 90, 60, or 30 days.

Results: An increase of one training allocation was significantly associated with a 5.60 percentage point (PP) (95% CI: 2.74-8.46) higher likelihood of filling a vacancy within 90 days, 7.34 PP (95% CI: 4.66-10.03) higher likelihood of filling a vacancy within 60 days, and 5.32 PP (95% CI: 3.18-7.46) higher likelihood of filling a vacancy within 30 days. The impact was significant in both 2020 and 2021 positions, and in facilities located in areas with lower social deprivation scores, higher-quality public schools, or with either no or partial primary care physician shortages.

Conclusions: We found favorable impacts of the size of the PB-RNR program at a VA facility on filling RN vacancies.

Background: Lack of health insurance is a public health crisis, leading to foregone care and financial strain. Hospital Presumptive Eligibility (HPE) is a hospital-based emergency Medicaid program that provides temporary (up to 60 d) coverage, with the goal that hospitals will assist patients in applying for ongoing Medicaid coverage. It is unclear whether HPE is associated with successful longer-term Medicaid enrollment.

Objective: To characterize Medicaid enrollment 6 months after initiation of HPE and determine sociodemographic, clinical, and geographic factors associated with Medicaid enrollment.

Design: This was a cohort study of all HPE approved inpatients in California, using claims data from the California Department of Healthcare Services.

Setting: The study was conducted across all HPE-participating hospitals within California between January 1, 2016 and December 31, 2017.

Participants: We studied California adult hospitalized inpatients, who were uninsured at the time of hospitalization and approved for HPE emergency Medicaid. Using multivariable logistic regression models, we compared HPE-approved patients who enrolled in Medicaid by 6 months versus those who did not.

Exposures: HPE emergency Medicaid approval at the time of hospitalization.

Main outcomes and measures: The primary outcome was full-scope Medicaid enrollment by 6 months after the hospital's presumptive eligibility approval.

Results: Among 71,335 inpatient HPE recipients, a total of 45,817 (64.2%) enrolled in Medicaid by 6 months. There was variability in Medicaid enrollment across counties in California (33%-100%). In adjusted analyses, Spanish-preferred-language patients were less likely to enroll in Medicaid (aOR 0.77, P <0.001). Surgical intervention (aOR 1.10, P <0.001) and discharge to another inpatient facility or a long-term care facility increased the odds of Medicaid enrollment (vs. routine discharge home: aOR 2.24 and aOR 1.96, P <0.001).

Conclusion: California patients who enroll in HPE often enroll in Medicaid coverage by 6 months, particularly among patients requiring surgical intervention, repeated health care visits, and ongoing access to care. Future opportunities include prospective evaluation of HPE recipients to understand the impact that Medicaid enrollment has on health care utilization and financial solvency.

Objective: The aim of this study was to determine response patterns to sexual orientation and gender identity (SOGI) questions in the Behavioral Risk Factor Surveillance System (BRFSS) over time and to assess nonresponse and indeterminate responses by demographic characteristics.

Methods: This is a secondary data analysis of the SOGI module of the BRFSS. We used data from 46 states and Guam that implemented SOGI questions between 2014 and 2022. We used weighted analyses that accounted for the sampling design, determined SOGI response patterns by year, and assessed nonresponse and indeterminate responses by demographic characteristics.

Results: Over time, increasing numbers self-reported as sexual and gender minority respondents, while heterosexual identity declined. Sexual orientation nonresponse and indeterminate responses increased with time, while respondents' reports of not knowing gender identity declined. Hispanic, older, respondents, those with lower education, and those who completed the questionnaire in Spanish had higher SOGI nonresponse and indeterminate responses.

Conclusions: The low amount of SOGI nonresponse and indeterminate responses in the BRFSS can be instructive for the implementation of SOGI questions in medical settings. SOGI data collection in all settings requires improving procedures for the groups that have been shown to have elevated nonresponse and indeterminate response.

Background: Social barriers to health care, such as food insecurity, financial distress, and housing instability, may impede effective clinical management for individuals with chronic illness. Systematic strategies are needed to more efficiently identify at-risk individuals who may benefit from proactive outreach by health care systems for screening and referral to available social resources.

Objective: To create a predictive model to identify a higher likelihood of food insecurity, financial distress, and/or housing instability among adults with multiple chronic medical conditions.

Research design and subjects: We developed and validated a predictive model in adults with 2 or more chronic conditions who were receiving care within Kaiser Permanente Northern California (KPNC) between January 2017 and February 2020. The model was developed to predict the likelihood of a "yes" response to any of 3 validated self-reported survey questions related to current concerns about food insecurity, financial distress, and/or housing instability. External model validation was conducted in a separate cohort of adult non-Medicaid KPNC members aged 35-85 who completed a survey administered to a random sample of health plan members between April and June 2021 (n = 2820).

Measures: We examined the performance of multiple model iterations by comparing areas under the receiver operating characteristic curves (AUCs). We also assessed algorithmic bias related to race/ethnicity and calculated model performance at defined risk thresholds for screening implementation.

Results: Patients in the primary modeling cohort (n = 11,999) had a mean age of 53.8 (±19.3) years, 64.7% were women, and 63.9% were of non-White race/ethnicity. The final, simplified model with 30 predictors (including utilization, diagnosis, behavior, insurance, neighborhood, and pharmacy-based variables) had an AUC of 0.68. The model remained robust within different race/ethnic strata.

Conclusions: Our results demonstrated that a predictive model developed using information gleaned from the medical record and from public census tract data can be used to identify patients who may benefit from proactive social needs assessment. Depending on the prevalence of social needs in the target population, different risk output thresholds could be set to optimize positive predictive value for successful outreach. This predictive model-based strategy provides a pathway for prioritizing more intensive social risk outreach and screening efforts to the patients who may be in greatest need.

Background: Collaborative care integrates mental health treatment into primary care and has been shown effective. Yet even in states where its use has been encouraged, take-up remains low and there are potential financial barriers to care.

Objective: Describe patient out-of-pocket costs and variations in referral patterns for collaborative care in Colorado.

Research design: Retrospective observational study using administrative medical claims data to identify outpatient visits with collaborative care. For individuals with ≥1 visit, we measure spending and visits at the month level. Among physicians with billings for collaborative care, we measure prevalence of eligible patients with collaborative care utilization.

Subjects: Patients with Medicare, Medicare Advantage, or commercial health insurance in Colorado, 2018-2019.

Outcomes: Out-of-pocket costs (enrollee payments to clinicians), total spending (insurer+enrollee payments to clinicians), percent of patients billed collaborative care.

Results: Median total spending (insurer+patient cost) was $48.32 (IQR: $41-$53). Median out-of-pocket cost per month in collaborative care was $8.35 per visit (IQR: $0-$10). Patients with commercial insurance paid the most per month (median: $15); patients with Medicare Advantage paid the least (median: $0). Among clinicians billing for collaborative care (n=193), a mean of 12 percent of eligible patients utilized collaborative care; family practice and advanced practice clinicians' patients utilized it most often.

Conclusions: Collaborative care remains underused with fewer than 1 in 6 potentially eligible patients receiving care in this setting. Out-of-pocket costs varied, though were generally low; uncertainty about costs may contribute to low uptake.

Background: Low response rates (RRs) can affect hospitals' data collection costs for patient experience surveys and value-based purchasing eligibility. Most hospitals use single-mode approaches, even though sequential mixed mode (MM) yields higher RRs and perhaps better patient representativeness. Some hospitals may be reluctant to incur MM's potential additional cost and complexity without knowing how much RRs would increase.

Objective: The aim of this study was to estimate the differences in RR and patient representation between MM and single-mode approaches and to identify hospital characteristics associated with the largest RR differences from MM of single-mode protocols (mail-only, phone-only).

Research design: Patients were randomized within hospitals to one of 3 modes (mail-only, phone-only, MM).

Subjects: A total of 17,415 patients from the 51 nationally representative US hospitals participating in a randomized HCAHPS mode experiment.

Results: Mail-only RRs were lowest for ages 18-24 (7%) and highest for ages 65+ (31%-35%). Phone-only RRs were 24% for ages 18-24, increasing to 37%-40% by ages 55+. MM RRs were 28% for ages 18-24, increasing to 50%-60% by ages 65-84. Lower hospital-level mail-only RRs strongly predicted greater gains from MM. For example, a hospital with a 15% mail-only RR has a predicted MM RR >40% (with >25% occurring in telephone follow-up).

Conclusion: MM increased representation of hard-to-reach (especially young adult) patients and hospital RRs in all mode experiment hospitals, especially in hospitals with low mail-only RRs.

Background: Advanced primary care models are key in moving primary care practices toward greater accountability for the quality and cost of a beneficiary's care. One critical but often overlooked detail in model design is the beneficiary attribution methodology. Attribution results are key inputs in calculating practice payments. Stable attribution yields predictable practice payments, fostering longer-term investments in advanced primary care.

Objective: We examine attribution stability for Medicare fee-for-service beneficiaries in Medicare's Comprehensive Primary Care Plus (CPC+) Model.

Design: To measure attribution stability, we calculate churn rates, which we define as the percentage of beneficiaries eligible for CPC+ who were not attributed to the same practice in a later period. Using 2017-2021 CPC+ program data and Medicare administrative data, we calculate churn rates for CPC+ overall and for beneficiary subgroups. To assess whether CPC+ attribution was responsive enough to changes in a beneficiary's practice, we calculate how long before attribution changes following a beneficiary's long-distance move.

Results: We find that for every 100 beneficiaries attributed to a CPC+ practice, 88 were still attributed to the same practice a year later (ie, churn rate of 12%), 79 were attributed 2 years later, 74 three years later, and 70 four years later. However, some vulnerable subgroups, such as disabled beneficiaries, had higher churn rates. Our analysis of long-distance movers reveals that only after 5 quarters did attribution change for more than half of these movers.

Conclusions: Overall, high attribution stability may have encouraged CPC+ practices to make longer-term investments in advanced primary care.

Background: Hospital inpatient data, coded using the International Classification of Diseases (ICD), is widely used to monitor diseases, allocate resources and funding, and evaluate patient outcomes. As such, hospital data quality should be measured before use; however, currently, there is no standard and international approach to assess ICD-coded data quality.

Objective: To develop a standardized method for assessing hospital ICD-coded data quality that could be applied across countries: Data quality indicators (DQIs).

Research design: To identify a set of candidate DQIs, we performed an environmental scan, reviewing gray and academic literature on data quality frameworks and existing methods to assess data quality. Indicators from the literature were then appraised and selected through a 3-round Delphi process. The first round involved face-to-face group and individual meetings for idea generation, while the second and third rounds were conducted remotely to collect online ratings. Final DQIs were selected based on the panelists' quantitative and qualitative feedback.

Subjects: Participants included international experts with expertise in administrative health data, data quality, and ICD coding.

Results: The resulting 24 DQIs encompass 5 dimensions of data quality: relevance, accuracy and reliability; comparability and coherence; timeliness; and Accessibility and clarity. These will help stakeholders (eg, World Health Organization) to assess hospital data quality using the same standard across countries and highlight areas in need of improvement.

Conclusions: This novel area of research will facilitate international comparisons of ICD-coded data quality and be valuable to future studies and initiatives aimed at improving hospital administrative data quality.

Introduction: Predictive models have proliferated in the health system in recent years and have been used to predict both health services utilization and medical outcomes. Less is known, however, on how these models function and how they might adapt to different contexts. The purpose of the current study is to shed light on the inner workings of a large-scale predictive model deployed in 2 distinct populations, with a particular emphasis on adaptability issues.

Methods: We compared the performance and functioning of a predictive model of avoidable hospitalization in 2 very different populations: Medicaid and Medicare enrollees in Maryland. Specifically, we assessed characteristics of the risk scores from March 2022 for the 2 populations, the predictive ability of the scores, and the driving risk factors behind the scores. In addition, we created and assessed the performance of an "unadapted" model by applying coefficients from the Medicare model to the Medicaid population.

Results: The model adapted to, and performed well in, both populations, despite demographic differences in these 2 groups. However, the most salient risk factors and their relative weightings differed, sometimes dramatically, across the 2 populations. The unadapted Medicaid model displayed poor performance relative to the adapted model.

Conclusions: Our findings speak to the need to "peek behind the curtain" of predictive models that may be applied to different populations, and we caution that risk prediction is not "one size fits all": for optimal performance, models should be adapted to, and trained on, the target population.

Background: The Elixhauser Comorbidity Index (ECI) is widely used, but its performance in homeless populations has not been evaluated.

Objectives: Using a national sample of inpatients, this study compared homeless and nonhomeless inpatients on common clinical diagnoses and evaluated ECI performance in predicting mortality among homeless inpatients.

Research design: A retrospective study was conducted using 2019 National Inpatient Sample (NIS) data, the largest publicly available all-payer inpatient health care database in the United States.

Subjects: Among 4,347,959 hospitalizations, 78,819 (weighted 1.8%) were identified as homeless.

Measures: The ECI consists of 38 medical conditions; homelessness was defined using the International Classification of Diseases Tenth Revision Clinical Modification (ICD-10-CM) diagnostic code, and clinical conditions were based on the Clinical Classifications Software Refined (CCSR) for ICD-10-CM.

Results: Leading clinical diagnoses for homeless inpatients included schizophrenia and other psychotic disorders (13.3%), depressive disorders (9.4%), and alcohol-related disorders (7.2%); leading diagnoses for nonhomeless inpatients were septicemia (10.2%), heart failure (5.2%), and acute myocardial infarction (3.0%). Metastatic cancer and liver disease were the most common ECI diagnoses for both homeless and nonhomeless inpatients. ECI indicators and summary scores were predictive of in-hospital mortality for homeless and nonhomeless inpatients, with all models yielding concordance statistics above 0.80, with better performance found among homeless inpatients.

Conclusions: These findings underlie the high rates of behavioral health conditions among homeless inpatients and the strong performance of the ECI in predicting in-hospital mortality among homeless inpatients, supporting its continued use as a case-mix control method and predictor of hospital readmissions.