Search

Camille Robichaux,1 Jordan Aron,2 Chris H Wendt,1,3 Jesse D Berman,4 Austin Rau,4 Ann Bangerter,5 R Adams Dudley,1,3,5 Arianne K Baldomero1,3,5 1Pulmonary, Allergy, Critical Care, and Sleep Medicine, University of Minnesota, Minneapolis, MN, USA; 2Division of Biostatistics, University of Minnesota School of Public Health, Minneapolis, MN, USA; 3Pulmonary, Allergy, Critical Care, and Sleep Medicine, Minneapolis VA Health Care System, Minneapolis, MN, USA; 4Division of Environmental Health Sciences, University of Minnesota School of Public Health, Minneapolis, MN, USA; 5Center for Care Delivery and Outcomes Research, Minneapolis VA Health Care System, Minneapolis, MN, USACorrespondence: Camille Robichaux, Division of Pulmonary, Allergy, Critical Care, and Sleep, University of Minnesota, 420 Delaware Street SE, MMC 276, Minneapolis, MN, 55455, USA, Tel +1 612 624-0999, Email robic020@umn.eduBackground: Chronic obstructive pulmonary disease (COPD) is a leading cause of mortality worldwide. Identifying both individual and community risk factors associated with higher mortality is essential to improve outcomes. Few population-based studies of mortality in COPD include both individual characteristics and community risk factors.Objective: We used geocoded, patient-level data to describe the associations between individual demographics, neighborhood socioeconomic status, and all-cause mortality.Methods: We performed a nationally representative retrospective cohort analysis of all patients enrolled in the Veteran Health Administration with at least one ICD-9 or ICD-10 code for COPD in 2016– 2019. We obtained demographic characteristics, comorbidities, and geocoded residential address. Area Deprivation Index and rurality were classified using individual geocoded residential addresses. We used logistic regression models to assess the association between these characteristics and age-adjusted all-cause mortality.Results: Of 1,106,163 COPD patients, 33.4% were deceased as of January 2021. In age-adjusted models, having more comorbidities, Black/African American race (OR 1.09 [95% CI: 1.08– 1.11]), and higher neighborhood disadvantage (OR 1.30 [95% CI: 1.28– 1.32]) were associated with all-cause mortality. Female sex (OR 0.67 [95% CI: 0.65– 0.69]), Asian race (OR 0.64, [95% CI: 0.59– 0.70]), and living in a more rural area were associated with lower odds of all-cause mortality. After adjusting for age, comorbidities, neighborhood socioeconomic status, and rurality, the association with Black/African American race reversed.Conclusion: All-cause mortality in COPD patients is disproportionately higher in patients living in poorer neighborhoods and urban areas, suggesting the impact of social determinants of health on COPD outcomes. Black race was associated with higher age-adjusted all-cause mortality, but this association was abrogated after adjusting for gender, socioeconomic status, comorbidities, and urbanicity. Future studies should focus on exploring mechanisms by which disparities arise and developing interventions to address these.Keywords: chronic obstructive pulmonary disease, social determinants, race, geography, socioeconomic status

Neal Yuan,1 Christy Boscardin,2 Nadra E Lisha,2 R Adams Dudley,3,4 Grace A Lin2 1Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA, USA; 2Division of General Internal Medicine, Department of Medicine, University of California, San Francisco, CA, USA; 3Division of Pulmonary, Allergy, Critical Care and Sleep Medicine, Department of Medicine, University of Minnesota, Minneapolis, MN, USA; 4School of Public Health, University of Minnesota, Minneapolis, MN, USACorrespondence: Neal YuanSmidt Heart Institute, Cedars-Sinai Medical Center, 8700 Beverly Blvd., Davis 1015, Los Angeles, CA 90048, USAEmail Neal.Yuan@cshs.orgBackground: In stable coronary artery disease (CAD), shared decision-making (SDM) is encouraged when deciding whether to pursue percutaneous coronary intervention (PCI) given similar cardiovascular outcomes between PCI and medical therapy. However, it remains unclear whether improving patient-provider communication and patient knowledge, the main tenets of SDM, changes patient preferences or the treatment chosen. We explored the relationships between patient-provider communication, patient knowledge, patient preferences, and the treatment received.Methods: We surveyed stable CAD patients referred for elective cardiac catheterization at seven hospitals from 6/2016 to 9/2018. Surveys assessed patient-provider communication, medical knowledge, and preferences for treatment and decision-making. We verified treatments received by chart review. We used linear and logistic regression to examine relationships between patient-provider communication and knowledge, knowledge and preference, and preference and treatment received.Results: Eighty-seven patients completed the survey. More discussion of the benefits and risks of both medical therapy and PCI associated with higher patient knowledge scores (β=0.28, p< 0.01). Patient knowledge level was not associated with preference for PCI (OR=0.78, 95% CI 0.57– 1.03, p=0.09). Black patients had more than four times the odds of preferring medical therapy to PCI (OR=4.49, 1.22– 18.45, p=0.03). Patients preferring medical therapy were not significantly less likely to receive PCI (OR=0.67, 0.16– 2.52, p=0.57).Conclusions: While communicating the risks of PCI may improve patient knowledge, this knowledge may not affect patient treatment preferences. Rather, other factors such as race may be significantly more influential on a patient’s treatment preferences. Furthermore, patient preferences are still not well reflected in the treatment received. Improving shared decision-making in stable CAD therefore may require not only increasing patient education but also better understanding and including a patient’s background and pre-existing beliefs.Keywords: stable coronary artery disease, stable angina, shared decision-making, percutaneous coronary intervention, stent, optimal medical therapy

© 2015 Wolters Kluwer Health, Inc. All rights reserved. BACKGROUND:: Hospital report cards and financial incentives linked to performance require clinical data that are reliable, appropriate, timely, and cost-effective to process. Pay-for-performance plans are transitioning to automated electronic health record (EHR) data as an efficient method to generate data needed for these programs. OBJECTIVE:: To determine how well data from automated processing of structured fields in the electronic health record (AP-EHR) reflect data from manual chart review and the impact of these data on performance rewards. RESEARCH DESIGN:: Cross-sectional analysis of performance measures used in a cluster randomized trial assessing the impact of financial incentives on guideline-recommended care for hypertension. SUBJECTS:: A total of 2840 patients with hypertension assigned to participating physicians at 12 Veterans Affairs hospital-based outpatient clinics. Fifty-two physicians and 33 primary care personnel received incentive payments. MEASURES:: Overall, positive and negative agreement indices and Cohen’s kappa were calculated for assessments of guideline-recommended antihypertensive medication use, blood pressure (BP) control, and appropriate response to uncontrolled BP. Pearson’s correlation coefficient was used to assess how similar participants’ calculated earnings were between the data sources. RESULTS:: By manual chart review data, 72.3% of patients were considered to have received guideline-recommended antihypertensive medications compared with 65.0% by AP-EHR review (κ=0.51). Manual review indicated 69.5% of patients had controlled BP compared with 66.8% by AP-EHR review (κ=0.87). Compared with 52.2% of patients per the manual review, 39.8% received an appropriate response by AP-EHR review (κ=0.28). Participants’ incentive payments calculated using the 2 methods were highly correlated (r≥0.98). Using the AP-EHR data to calculate earnings, participants’ payment changes ranged from a decrease of $91.00 (−30.3%) to an increase of $18.20 (+7.4%) for medication use (interquartile range, −14.4% to 0%) and a decrease of $100.10 (−31.4%) to an increase of $36.40 (+15.4%) for BP control or appropriate response to uncontrolled BP (interquartile range, −11.9% to −6.1%). CONCLUSIONS:: Pay-for-performance plans that use only EHR data should carefully consider the measures and the structure of the EHR before data collection and financial incentive disbursement. For this study, we feel that a 10% difference in the total amount of incentive earnings disbursed based on AP-EHR data compared with manual review is acceptable given the time and resources required to abstract data from medical records.

Abstract. Background: In December 2000, Medicare eliminated time limitations in immunosuppressant coverage after kidney transplant for beneficiaries age 65 and those who were disabled. This change did not apply to younger non-disabled beneficiaries who qualified for Medicare only because of their end-stage renal disease (ESRD). We sought to examine access to waitlisting for kidney transplantation in a cohort spanning this policy change. Methods. This was a retrospective cohort analysis of 241,150 Medicare beneficiaries in the United States Renal Data System who initiated chronic dialysis between 1/1/96 and 11/30/03. We fit interrupted time series Cox proportional hazard models to compare access to kidney transplant waitlist within 12months of initiating chronic dialysis by age/disability status, accounting for secular trends. Results: Beneficiaries age

Background: Systemic lupus erythematosus (SLE) affects 1 in 2500 Americans and is associated with significant morbidity and mortality. The recent development of SLE quality measures provides an opportunity to understand gaps in clinical care and to identify modifiable factors associated with variations in quality.Objective: To evaluate performance on SLE quality measures as well as differences in quality of care by demographic, socioeconomic, disease, and health system characteristics.Design and Patients: Cross-sectional analysis of data derived from the Lupus Outcomes Study, a prospective, longitudinal study of 814 individuals. Principal data collection was through annual structured telephone surveys between 2009-2010. Data on 13 SLE quality measures was collected. We used regression models to estimate demographic, socioeconomic, disease, and health system characteristics associated with performance on individual and overall quality measures.Outcome Measures: Performance on each quality measure and overall performance on all measures for which participants were eligible (pass rate).Results: Participants were eligible for a mean of five measures (range 2-12). Performance varied from 29 % (assessment of cardiovascular risk factors) to 90 % (sun avoidance counseling). The overall pass rate was 65 % (95 % CI 64 %, 65 %). In unadjusted analyses, younger age, minority race/ethnicity, poverty, shorter disease duration, fewer physician visits, and lack of health insurance, were associated with lower pass rates. Receiving care in public sector managed care organizations was associated with higher pass rates. After adjustment, younger age, having fewer physician visits and lacking health insurance remained significantly associated with lower performance; receiving care in public sector managed care organizations remained associated with higher performance.Conclusions: We identified a number of gaps in clinical care for SLE. Factors associated with the health care system, including presence and type of health insurance, were the primary determinants of performance on quality measures in SLE. [ABSTRACT FROM AUTHOR]

Background: To develop and compare ICU length-of-stay (LOS) risk-adjustment models using three commonly used mortality or LOS prediction models.Methods: Between 2001 and 2004, we performed a retrospective, observational study of 11,295 ICU patients from 35 hospitals in the California Intensive Care Outcomes Project. We compared the accuracy of the following three LOS models: a recalibrated acute physiology and chronic health evaluation (APACHE) IV-LOS model; and models developed using risk factors in the mortality probability model III at zero hours (MPM(0)) and the simplified acute physiology score (SAPS) II mortality prediction model. We evaluated models by calculating the following: (1) grouped coefficients of determination; (2) differences between observed and predicted LOS across subgroups; and (3) intraclass correlations of observed/expected LOS ratios between models.Results: The grouped coefficients of determination were APACHE IV with coefficients recalibrated to the LOS values of the study cohort (APACHE IVrecal) [R(2) = 0.422], mortality probability model III at zero hours (MPM(0) III) [R(2) = 0.279], and simplified acute physiology score (SAPS II) [R(2) = 0.008]. For each decile of predicted ICU LOS, the mean predicted LOS vs the observed LOS was significantly different (p Conclusions: APACHE IV and MPM(0) III were more accurate than SAPS II for the prediction of ICU LOS. APACHE IV is the most accurate and best calibrated model. Although it is less accurate, MPM(0) III may be a reasonable option if the data collection burden or the treatment effect bias is a consideration. [ABSTRACT FROM AUTHOR]

Background: Federal and state agencies are considering ICU performance assessment and public reporting; however, an accurate method for measuring performance must be selected. In this study, we determine whether a substantial variation in ICU mortality performance still exists in modern ICUs, and compare the predictive accuracy, reliability, and data burden of existing ICU risk-adjustment models.Methods: A retrospective chart review of 11,300 ICU patients from 35 California hospitals from 2001 to 2004 was performed. We calculated standardized mortality ratios (SMRs) for each hospital using the mortality probability model III (MPM(0) III), the simplified acute physiology score (SAPS) II, and the acute physiology and chronic health evaluation (APACHE) IV risk-adjustment models. We compared discrimination, calibration, data reliability, and abstraction time for the models.Results: Regardless of the model used, there was a large variation in SMRs among the ICUs studied. The discrimination and calibration were adequate for all risk-adjustment models. APACHE IV had the best discrimination (area under the receiver operating characteristic curve [AUC], 0.892) compared to MPM(0) III (AUC, 0.809), and SAPS II (AUC, 0.873; p < 0.001). The models differed substantially in data abstraction times, as follows: MPM(0)III, 11.1 min (95% confidence interval [CI], 8.7 to 13.4); SAPS II, 19.6 min (95% CI, 17.0 to 22.2); and APACHE IV, 37.3 min (95% CI, 28.0 to 46.6).Conclusions: We found substantial variation in the ICU risk-adjusted mortality rates that persisted regardless of the risk-adjustment model. With unlimited resources, the APACHE IV model offers the best predictive accuracy. If constrained by cost and manual data collection, the MPM(0) III model offers a viable alternative without a substantial loss in accuracy. [ABSTRACT FROM AUTHOR]

OBJECTIVE: To examine racial/ethnic and economic variation in cost-related medication underuse among insured adults with diabetes. RESEARCH DESIGN AND METHODS: We surveyed 5,086 participants from the multicenter Translating Research Into Action for Diabetes Study. Respondents reported whether they used less medication because of cost in the past 12 months. We examined unadjusted and adjusted rates of cost-related medication underuse, using hierarchical regression, to determine whether race/ethnicity differences still existed after accounting for economic, health, and other demographic variables. RESULTS: Participants were 48% white, 14% African American, 14% Latino, 15% Asian/Pacific Islander, and 8% other. Overall, 14% reported cost-related medication underuse. Unadjusted rates were highest for Latinos (23%) and African Americans (17%) compared with whites (13%), Asian/Pacific Islanders (11%), and others (15%). In multivariate analyses, race/ethnicity significantly predicted cost-related medication underuse (P = 0.048). However, adjusted rates were only slightly higher for Latinos (14%) than whites (10%) (P = 0.026) and were not significantly different for African Americans (11%), Asian/Pacific Islanders (7%), and others (11%). Income and out-of-pocket drug costs showed the greatest differences in adjusted rates of cost-related medication underuse (15 vs. 5% for participants with income $50,000 and 24 vs. 7% for participants with out-of-pocket costs >$150 per month vs.

Background: Maintaining independence in daily functioning is an important health outcome in older adults. A key measure of functional independence in elders is the ability to do activities of daily living (ADL) without the assistance of another person. However, few prognostic indices have been developed that stratify elders into groups at variable risk for developing ADL dependence.Objective: We sought to develop and validate a prognostic index that distinguishes between elders at different risk of ADL dependence.Research Design, Subjects, and Measures: We studied subjects enrolled in Asset and Health Dynamics Among the Oldest Old (AHEAD), a nationally representative cohort of elders older than the age of 70. We included 5239 subjects (mean age, 77) reporting that they could do each of 5 ADL (bathing, dressing, toileting, transferring, and eating) without the assistance of another person at baseline. Subjects were divided into development (n = 3245) and validation (n = 1994) samples based on region of the United States. Our primary outcome was the need for help (dependence) with at least one ADL at 2 years. We used logistic regression to select among predictor variables encompassing several domains: demographic characteristics, comorbid conditions, functional status, cognitive status, and general health indicators.Results: The 9 independent predictors of 2-year ADL dependence were age older than 80, diabetes, difficulty walking several blocks, difficulty bathing or dressing, need for help with personal finances, difficulty lifting 10 pounds, inability to name the Vice President, history of falling, and low body mass index. We created a risk score by assigning 1 point to each risk factor. In the development sample, rates of 2-year ADL dependence in subjects with 0, 1, 2, 3, 4, and 5 or more risk factors were 1.3%, 2.8%, 3.8%, 10%, 22%, and 33%, respectively (P < 0.001, roc area = 0.79). In the validation sample, the rates were 0.7%, 4.3%, 8.7%, 11%, 18%, and 40% (P < 0.001, roc area = 0.77). The risk score also discriminated between subjects at variable risk for a combined outcome of either ADL decline or death (4.3%, 7.6%, 15%, 21%, 30%, and 47%).Conclusion: Using data available from patient reports, we validated a simple risk index that distinguished between elders at variable risk of ADL dependence. This index may be useful for identifying elders at high risk of poor outcomes or for risk adjustment. [ABSTRACT FROM AUTHOR]

As drug costs rose in the 1990s, health maintenance organizations (HMOs) began transferring risk for prescription drug expenditures to physician groups. With principal-agent theory as a framework for understanding drug-risk transfer, we used a multiple case-study design to examine the relationship between the level of drug risk that a physician group accepts and the physician group's adoption of drug-use management strategies. The data demonstrated that adoption of drug-use management innovations was not related to level of risk for pharmacy costs and that factors other than drug-risk level (e.g., contracting and data issues, financial and market factors, and physician group assessments of the fairness and incentives of risk contracts) can influence the principal-agent relationship. The data also revealed a novel form of information asymmetry between physicians and HMOs and unexpected failures of HMOs to fully enable their physician-agents. We believe these observations reflect the complexity of relationships in the health care system and have implications for the use of incentives. Based on principal-agent theory and our findings, we offer an alternative approach to drug-risk contracting that reduces physicians responsibility for aspects of drug use that are beyond their control while maintaining the incentives to manage drug costs and use that were the original intent of drug-risk contracting. [ABSTRACT FROM AUTHOR]

OBJECTIVES: To determine whether passage of welfare and immigration policies was followed in California by changes in births to foreign-born women in California with respect to total numbers, payer sources, prenatal care use, or health outcomes. METHODS: Comparison of births to foreign-born and US-born women from 1990 to 1997 using adjusted odds ratios generated with multivariate logistic regression. RESULTS: Policies passed in 1994 and 1996 were followed by decreases in adjusted odds of births to foreign-born women with prenatal Medicaid coverage, without a corresponding increase in uninsured foreign-born women. There was no decline in the use of prenatal care by foreign-born women, and no worsening of birth outcomes after passage of the reforms. Foreign-born women, however, remained more likely to have inadequate prenatal care than US-born women, and the improvement in outcomes that occurred for US-born women from 1994 to 1997 did not occur for foreign-born women. CONCLUSIONS: In spite of the fact that pregnant immigrant women remained eligible for Medicaid after passage of welfare and immigration policies in California, the volume of births to foreign-born women using Medicaid declined. The lack of a corresponding increase in births to uninsured foreign-born women appears to have prevented deterioration in the use of prenatal care or birth outcomes. [ABSTRACT FROM AUTHOR]

Context: Evidence exists that high-volume hospitals (HVHs) have lower mortality rates than low-volume hospitals (LVHs) for certain conditions. However, few employers, health plans, or government programs have attempted to increase the number of patients referred to HVHs.Objectives: To determine the difference in hospital mortality between HVHs and LVHs for conditions for which good quality data exist and to estimate how many deaths potentially would be avoided in California by referral to HVHs.Design, Setting, and Patients: Literature in MEDLINE, Current Contents, and First-Search Social Abstracts databases from January 1, 1983, to December 31, 1998, was searched using the key words hospital, outcome, mortality, volume, risk, and quality. The highest-quality study assessing the mortality-volume relationship for each given condition was identified and used to calculate odds ratios (ORs) for in-hospital mortality for LVHs vs HVHs. These ORs were then applied to the 1997 California database of hospital discharges maintained by the California Office of Statewide Health Planning and Development to estimate potentially avoidable deaths.Main Outcome Measures: Deaths that potentially could be avoided if patients with conditions for which a mortality-volume relationship had been treated at an HVH vs LVH.Results: The articles identified in the literature search were grouped by condition, and predetermined criteria were applied to choose the best article for each condition. Mortality was significantly lower at HVHs for elective abdominal aortic aneurysm repair, carotid endarterectomy, lower extremity arterial bypass surgery, coronary artery bypass surgery, coronary angioplasty, heart transplantation, pediatric cardiac surgery, pancreatic cancer surgery, esophageal cancer surgery, cerebral aneurysm surgery, and treatment of human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS). A total of 58,306 of 121,609 patients with these conditions were admitted to LVHs in California in 1997. After applying the calculated ORs to these patient populations, we estimated that 602 deaths (95% confidence interval, 304-830) at LVHs could be attributed to their low volume. Additional analyses were performed to take into account emergent admissions and distance traveled, but the impact of loss of continuity of care for some patients and reduction in the availability of specialists for patients remaining at LVHs could not be assessed.Conclusions: Initiatives to facilitate referral of patients to HVHs have the potential to reduce overall hospital mortality in California for the conditions identified. Additional study is needed to determine the extent to which selective referral is feasible and to examine the potential consequences of such initiatives. [ABSTRACT FROM AUTHOR]

In the past two decades, relationships among health plans, medical groups, and providers have grown more complex and the number of clinical management strategies has increased. In this context, determining the independent effect of a particular organizational strategy on quality of care has become more difficult. The authors review some of the issues a researcher must address when studying the relationship between organizational characteristics and quality of care. They offer criteria for selecting a research question, list organizational characteristics that may influence quality, and suggest sampling and study design techniques to reduce confounding. Since this type of research often requires a health care organization as collaborator, the authors discuss strategies for developing research partnerships and collecting data from the partner organization. Finally, they offer suggestions for translating research into policy. [ABSTRACT FROM AUTHOR]

Medicare beneficiaries are now facing advertising from an unprecedented number of health plans that are offering prescription drug coverage. Previous Medicare managed care efforts have been undermined by risk selection, the practice of enrolling healthier and therefore less costly patients. In this study we explore how the content of health plan advertising is related to the competitiveness of the health plan market. We find that increased competition is associated with greater use of advertising that targets healthier patients. [ABSTRACT FROM AUTHOR]

Objective: Access to firearms is associated with increased suicide risk. Our aim was to develop a natural language processing approach to characterizing firearm access in clinical records., Materials and Methods: We used clinical notes from 36 685 Veterans Health Administration (VHA) patients between April 10, 2023 and April 10, 2024. We expanded preexisting firearm term sets using subject matter experts and generated 250-character snippets around each firearm term appearing in notes. Annotators labeled 3000 snippets into three classes. Using these annotated snippets, we compared four nonneural machine learning models (random forest, bagging, gradient boosting, logistic regression with ridge penalization) and two versions of Bidirectional Encoder Representations from Transformers, or BERT (specifically, BioBERT and Bio-ClinicalBERT) for classifying firearm access as "definite access", "definitely no access", or "other"., Results: Firearm terms were identified in 36 685 patient records (41.3%), 33.7% of snippets were categorized as definite access, 9.0% as definitely no access, and 57.2% as "other". Among models classifying firearm access, five of six had acceptable performance, with BioBERT and Bio-ClinicalBERT performing best, with F1s of 0.876 (95% confidence interval, 0.874-0.879) and 0.896 (95% confidence interval, 0.894-0.899), respectively., Discussion and Conclusion: Firearm-related terminology is common in the clinical records of VHA patients. The ability to use text to identify and characterize patients' firearm access could enhance suicide prevention efforts, and five of our six models could be used to identify patients for clinical interventions., (Published by Oxford University Press on behalf of the American Medical Informatics Association 2024.)

Background: Many patients receive guideline-discordant inhaler regimens after chronic obstructive pulmonary disease (COPD) hospitalization. Geography and fragmented care across multiple providers likely influence prescription of guideline-discordant inhaler regimens, but these have not been comprehensively studied. We assessed patient-level differences in guideline-discordant inhaler regimens by rurality, drive time to pulmonary specialty care, and fragmented care., Methods: Retrospective cohort analysis using national Veterans Health Administration (VA) data among patients who received primary care and prescriptions from the VA. Patients hospitalized for COPD exacerbation between 2017 and 2020 were assessed for guideline-discordant inhaler regimens in the subsequent 3 months. Guideline-discordant inhaler regimens were defined as short-acting inhaler/s only, inhaled corticosteroid (ICS) monotherapy, long-acting beta-agonist (LABA) monotherapy, ICS + LABA, long-acting muscarinic antagonist (LAMA) monotherapy, or LAMA + ICS. Rural residence and drive time to the closest pulmonary specialty care were obtained from geocoded addresses. Fragmented care was defined as hospitalization outside the VA. We used multivariable logistic regression models to assess associations between rurality, drive time, fragmentated care, and guideline-discordant inhaler regimens. Models were adjusted for age, sex, race/ethnicity, Charlson Comorbidity Index, Area Deprivation Index, and region., Findings: Of 33,785 patients, 16,398 (48.6%) received guideline-discordant inhaler regimens 3 months after hospitalization. Rural residents had higher odds of guideline-discordant inhalers regimens compared to their urban counterparts (adjusted odds ratio [aOR] 1.18 [95% CI: 1.12-1.23]). The odds of receiving guideline-discordant inhaler regimens increased with longer drive time to pulmonary specialty care (aOR 1.38 [95% CI: 1.30-1.46] for drive time >90 min compared to <30 min). Fragmented care was also associated with higher odds of guideline-discordant inhaler regimens (aOR 1.56 [95% CI: 1.48-1.63])., Interpretation: Rurality, long drive time to care, and fragmented care were associated with greater prescription of guideline-discordant inhaler regimens after COPD hospitalization. These findings highlight the need to understand challenges in delivering evidence-based care., Funding: NIHNCATS grants KL2TR002492 and UL1TR002494., Competing Interests: AKB reported grants from the National Institutes of Health. CHS reported grants from the Health Resources and Services Administration, Federal Office of Rural Health Policy, the National Institutes of Health and the CDC; payment or honoraria from Iowa Primary Care Association, Southern Gerontological Society, and the University of Michigan; and leadership or fiduciary role in the National Rural Health Association Board of Trustees. KMK reported personal fees from Nuvaira and Organicell (Data and Safety Monitoring Boards) outside of this work; consulting fees from Allergan/AbbVie and leadership or fiduciary role in the American Thoracic Society Proposal Review Committee, the American Thoracic Society Clinical Problems, and the Assembly Planning Committee.