RESUMEN
OBJECTIVE: Evaluate whether cost-sharing decreases led high-deductible health plans (HDHP) enrollees to increase their use of healthcare. DATA SOURCES, STUDY SETTING: National sample of chronically-ill patients age 18-64 from 2018 to 2020 (n = 1,318,178). STUDY DESIGN: Difference-in-differences analyses using entropy-balancing weights were used to evaluate the effect of a policy shift to $0 cost-sharing for telehealth on utilization for HDHP compared with non-HDHP enrollees. Due to this shock, HDHP enrollees experienced substantial declines in cost-sharing for telehealth, while non-HDHP enrollees experienced small declines. Event study models were also used to evaluate changes over time. DATA COLLECTION/EXTRACTION METHODS: Outcomes included use of any outpatient care; use of $0 telehealth; use of $0 telehealth as a proportion of all outpatient care; and use of any telehealth. To test whether any differences were due to preferences for care modality versus cost-sharing, we further evaluated use of non-$0 telehealth as a placebo test. PRINCIPAL FINDINGS: There was no difference in change in overall outpatient visits (p = 0.84), with chronicall-ill HDHP enrollees using less care both before and after the policy shift. However, compared with non-HDHP enrollees, HDHP enrollees increased their use of $0 telehealth by 0.08 visits over a 9-month period, a 27% increase (95% CI 0.07-0.09, p < 0.001) and shifted 1.2 percentage points more of their care to $0 telehealth, a 15% increase (ß = 0.01, 95% CI 0.01, 0.01, p < 0.001). However, HDHP enrollees had lower uptake of non-$0 telehealth than non-HDHP enrollees (ß = -0.01, 95%CI -0.02, 0.00, p = 0.04). CONCLUSIONS: Recent-but-expiring federal legislation exempts telehealth from HDHP deductibles for care provided in 2023 and 2024. Our results indicate that extending the protections provided by this legislation could help reduce the gap in access to care for chronically-ill persons enrolled in HDHPs.
RESUMEN
BACKGROUND: Many healthcare systems have implemented intensive outpatient primary care programs with the hopes of reducing healthcare costs. OBJECTIVE: The Veterans Health Administration (VHA) piloted primary care intensive management (PIM) for patients at high risk for hospitalization or death, or "high-risk." We evaluated whether a referral model would decrease high-risk patient costs. DESIGN: Retrospective cohort study using a quasi-experimental design comparing 456 high-risk patients referred to PIM from October 2017 to September 2018 to 415 high-risk patients matched on propensity score. PARTICIPANTS: Veterans in the top 10th percentile of risk for 90-day hospitalization or death and recent hospitalization or emergency department (ED) visit. INTERVENTION: PIM consisted of interdisciplinary teams that performed comprehensive assessments, intensive case management, and care coordination services. MAIN OUTCOMES AND MEASURES: Change in VHA and non-VHA outpatient utilization, inpatient admissions, and costs 12 months pre- and post-index date. KEY RESULTS: Of the 456 patients referred to PIM, 301 (66%) enrolled. High-risk patients referred to PIM had a marginal reduction in ED visits (- 0.7; [95% CI - 1.50 to 0.08]; p = 0.08) compared to propensity-matched high-risk patients; overall outpatient costs were similar. High-risk patients referred to PIM had similar number of medical/surgical hospitalizations (- 0.2; [95% CI, - 0.6 to 0.16]; p = 0.2), significant increases in length of stay (6.36; [CI, - 0.01 to 12.72]; p = 0.05), and higher inpatient costs ($22,628, [CI, $3587 to $41,669]; p = 0.02) than those not referred to PIM. CONCLUSIONS AND RELEVANCE: VHA intensive outpatient primary care was associated with higher costs. Referral to intensive case management programs targets the most complex patients and may lead to increased utilization and costs, particularly in an integrated healthcare setting with robust patient-centered medical homes. TRIAL REGISTRATION: PIM 2.0: Patient Aligned Care Team (PACT) Intensive Management (PIM) Project (PIM2). NCT04521816. https://clinicaltrials.gov/study/NCT04521816.
RESUMEN
Postsurgical falls have significant patient and societal implications but remain challenging to identify and track. Detecting postsurgical falls is crucial to improve patient care for older adults and reduce healthcare costs. Large language models (LLMs) offer a promising solution for reliable and automated fall detection using unstructured data in clinical notes. We tested several LLM prompting approaches to postsurgical fall detection in two different healthcare systems with three open-source LLMs. The Mixtral-8×7B zero-shot had the best performance at Stanford Health Care (PPV = 0.81, recall = 0.67) and the Veterans Health Administration (PPV = 0.93, recall = 0.94). These results demonstrate that LLMs can detect falls with little to no guidance and lay groundwork for applications of LLMs in fall prediction and prevention across many different settings.
RESUMEN
Importance: Inpatient clinical deterioration is associated with substantial morbidity and mortality but may be easily missed by clinicians. Early warning scores have been developed to alert clinicians to patients at high risk of clinical deterioration, but there is limited evidence for their effectiveness. Objective: To evaluate the effectiveness of an artificial intelligence deterioration model-enabled intervention to reduce the risk of escalations in care among hospitalized patients using a study design that facilitates stronger causal inference. Design, Setting, and Participants: This cohort study used a regression discontinuity design that controlled for confounding and was based on Epic Deterioration Index (EDI; Epic Systems Corporation) prediction model scores. Compared with other observational research, the regression discontinuity design facilitates causal analysis. Hospitalized adults were included from 4 general internal medicine units in 1 academic hospital from January 17, 2021, through November 16, 2022. Exposure: An artificial intelligence deterioration model-enabled intervention, consisting of alerts based on an EDI score threshold with an associated collaborative workflow among nurses and physicians. Main Outcomes and Measures: The primary outcome was escalations in care, including rapid response team activation, transfer to the intensive care unit, or cardiopulmonary arrest during hospitalization. Results: During the study, 9938 patients were admitted to 1 of the 4 units, with 963 patients (median [IQR] age, 76.1 [64.2-86.2] years; 498 males [52.3%]) included within the primary regression discontinuity analysis. The median (IQR) Elixhauser Comorbidity Index score in the primary analysis cohort was 10 (0-24). The intervention was associated with a -10.4-percentage point (95% CI, -20.1 to -0.8 percentage points; P = .03) absolute risk reduction in the primary outcome for patients at the EDI score threshold. There was no evidence of a discontinuity in measured confounders at the EDI score threshold. Conclusions and Relevance: Using a regression discontinuity design, this cohort study found that the implementation of an artificial intelligence deterioration model-enabled intervention was associated with a significantly decreased risk of escalations in care among inpatients. These results provide evidence for the effectiveness of this intervention and support its further expansion and testing in other care settings.
Asunto(s)
Inteligencia Artificial , Deterioro Clínico , Humanos , Masculino , Femenino , Anciano , Persona de Mediana Edad , Estudios de Cohortes , Puntuación de Alerta Temprana , Hospitalización/estadística & datos numéricos , Equipo Hospitalario de Respuesta Rápida , Unidades de Cuidados IntensivosRESUMEN
BACKGROUND: Although methamphetamine use associated heart failure (MU-HF) is increasing, data on its clinical course are limited due to a preponderance of single center studies and significant heterogeneity in the definition of MU-HF in the published literature. Our objective was to evaluate left ventricular ejection fraction (LVEF) distribution, methamphetamine use treatment engagement and postdischarge healthcare utilization among Veterans with heart failure hospitalization in the department of Veterans Affairs (VA) medical centers for MU-HF versus HF not associated with methamphetamine use (other-HF). METHODS: Observational study including a cohort of Veterans with a first heart failure hospitalization during 2007 - 2020 using data in the VA Corporate Data Warehouse. MU-HF was identified based on the presence of an ICD-code for methamphetmaine use or positive toxicology results within 1-year of heart failure hospitalization. LVEF values entered in the medical record were identified using a validated natural language processing algorithm. Healthcare utilization data was obtained using clinic stop-codes and hosptilaization records. RESULTS: Of 203,005 first-time heart failure hospitlaizations, 4080 were categorized as MU-HF. Median (interquartile range) of LVEF was 30 (20-45) % for MU-HF versus 40 (25-55)% for other-HF (P < .0001). Eighteen percent of MU-HF had LVEF ≥ 50% compared to 28% in other-HF. Discharge against medical advice was higher in MU-HF (8% vs 2%). Among Veterans with MU-HF, post hospital discharge methamphetamine use treatment engagement was low (18% at 30 days post discharge), with higher follow-up in primary care (76% at 30 days). Post discharge emergency department visits (33% versus 22% at 30 days) and rehospitalizations (24% versus 18% at 30 days) were higher in MU-HF compared to other-HF. CONCLUSIONS: While the majority of MU-HF hospitalizations are HFrEF, a sizeable minority have HFpEF. This finding has implications for accurate MU-HF classification, treatment, and prognosis. Patients with MU-HF have low addiction treatment receipt and high postdischarge unplanned healthcare utilization. Increasing substance use disorder treatment in this population must be a priority to improve health outcomes. Care-coordination and linkage interventions are urgently needed to increase post-hospitalization addiction treatment and follow-up in an effort to increase evidence-base care and mitigate unplanned healthcare utilization.
Asunto(s)
Insuficiencia Cardíaca , Función Ventricular Izquierda , Humanos , Volumen Sistólico , Insuficiencia Cardíaca/terapia , Insuficiencia Cardíaca/tratamiento farmacológico , Cuidados Posteriores , Alta del Paciente , Hospitalización , Pronóstico , Aceptación de la Atención de SaludRESUMEN
BACKGROUND: Predictive models show promise in healthcare, but their successful deployment is challenging due to limited generalizability. Current external validation often focuses on model performance with restricted feature use from the original training data, lacking insights into their suitability at external sites. Our study introduces an innovative methodology for evaluating features during both the development phase and the validation, focusing on creating and validating predictive models for post-surgery patient outcomes with improved generalizability. METHODS: Electronic health records (EHRs) from 4 countries (United States, United Kingdom, Finland, and Korea) were mapped to the OMOP Common Data Model (CDM), 2008-2019. Machine learning (ML) models were developed to predict post-surgery prolonged opioid use (POU) risks using data collected 6 months before surgery. Both local and cross-site feature selection methods were applied in the development and external validation datasets. Models were developed using Observational Health Data Sciences and Informatics (OHDSI) tools and validated on separate patient cohorts. RESULTS: Model development included 41 929 patients, 14.6% with POU. The external validation included 31 932 (UK), 23 100 (US), 7295 (Korea), and 3934 (Finland) patients with POU of 44.2%, 22.0%, 15.8%, and 21.8%, respectively. The top-performing model, Lasso logistic regression, achieved an area under the receiver operating characteristic curve (AUROC) of 0.75 during local validation and 0.69 (SD = 0.02) (averaged) in external validation. Models trained with cross-site feature selection significantly outperformed those using only features from the development site through external validation (P < .05). CONCLUSIONS: Using EHRs across four countries mapped to the OMOP CDM, we developed generalizable predictive models for POU. Our approach demonstrates the significant impact of cross-site feature selection in improving model performance, underscoring the importance of incorporating diverse feature sets from various clinical settings to enhance the generalizability and utility of predictive healthcare models.
Asunto(s)
Ciencia de los Datos , Informática Médica , Humanos , Modelos Logísticos , Reino Unido , FinlandiaRESUMEN
BACKGROUND: Patient portals play an increasingly critical role in engaging patients in their health care. They have the potential to significantly impact the health of those living with chronic diseases, such as HIV, for whom consistent care engagement is both critical and complex. OBJECTIVE: The primary aim was to examine the longitudinal relationships between individual portal tool use and health-related outcomes in patients living with HIV. DESIGN: Retrospective cohort study using electronic health record data to examine the relationship between patient portal tool use and key HIV-specific, health-related outcomes in patients engaged in care in the Veterans Health Administration (VA) through the application of marginal structural models. PARTICIPANTS: A national sample of patients living with HIV (PLWH) active in VA care who were registered to use the VA's patient portal, My HealtheVet (MHV; n = 18,390) between 10/1/2012 and 4/1/2017. MAIN MEASURES: The MHV tools examined were prescription refill (including prescription refill of an antiretroviral (ART) medication and any medication), secure messaging, view appointments, and view labs. Primary outcomes were viral load test receipt, viral load suppression, and ART medication adherence (measured as proportion of days covered). KEY RESULTS: The use of prescription refill for any medication or for ART was positively associated with ART adherence. Secure messaging was positively associated with ART adherence but not with viral load test receipt or viral load suppression. The use of view appointments was positively associated with ART adherence and viral load test receipt but not viral load suppression. The use of view labs was positively associated with viral load suppression but not ART adherence or viral load test receipt. CONCLUSIONS: These findings highlight the valuable role patient portals may play in improving health-related outcomes among PLWH and have implications for patients living with other types of chronic disease.
Asunto(s)
Fármacos Anti-VIH , Infecciones por VIH , Portales del Paciente , Comportamiento del Uso de la Herramienta , Humanos , Fármacos Anti-VIH/uso terapéutico , Estudios Retrospectivos , Carga Viral , Infecciones por VIH/tratamiento farmacológico , Cumplimiento de la MedicaciónRESUMEN
DESIGN: Retrospective cohort study. OBJECTIVE: We sought to examine whether disruptions in follow-up intervals contributed to hypertension control. BACKGROUND: Disruptions in health care were widespread during the coronavirus disease 2019 pandemic. PATIENTS AND METHODS: We identified a cohort of individuals with hypertension in both prepandemic (March 2019-February 2020) and pandemic periods (March 2020-February 2022) in the Veterans Health Administration. First, we calculated follow-up intervals between the last prepandemic and first pandemic blood pressure measurement during a primary care clinic visit, and between measurements in the prepandemic period. Next, we estimated the association between the maintenance of (or achieving) hypertension control and the period using generalized estimating equations. We assessed associations between follow-up interval and control separately for periods. Finally, we evaluated the interaction between period and follow-up length. RESULTS: A total of 1,648,424 individuals met the study inclusion criteria. Among individuals with controlled hypertension, the likelihood of maintaining control was lower during the pandemic versus the prepandemic (relative risk: 0.93; 95% CI: 0.93, 0.93). Longer follow-up intervals were associated with a decreasing likelihood of maintaining controlled hypertension in both periods. Accounting for follow-up intervals, the likelihood of maintaining control was 2% lower during the pandemic versus the prepandemic. For uncontrolled hypertension, the likelihood of gaining control was modestly higher during the pandemic versus the prepandemic (relative risk: 1.01; 95% CI: 1.01, 1.01). The likelihood of gaining control decreased with follow-up length during the prepandemic but not pandemic. CONCLUSIONS: During the pandemic, longer follow-up between measurements contributed to the lower likelihood of maintaining control. Those with uncontrolled hypertension were modestly more likely to gain control in the pandemic.
Asunto(s)
COVID-19 , Hipertensión , Veteranos , Humanos , Estudios de Cohortes , Pandemias , Estudios Retrospectivos , COVID-19/epidemiología , Hipertensión/epidemiologíaRESUMEN
BACKGROUND: Normalization Process Theory (NPT) is an implementation theory that can be used to explain how and why implementation strategies work or not in particular circumstances. We used it to understand the mechanisms that lead to the adoption and routinization of palliative care within hemodialysis centers. METHODS: We employed a longitudinal, mixed methods approach to comprehensively evaluate the implementation of palliative care practices among ten hemodialysis centers participating in an Institute for Healthcare Improvement Breakthrough- Series learning collaborative. Qualitative methods included longitudinal observations of collaborative activities, and interviews with implementers at the end of the study. We used an inductive and deductive approach to thematic analysis informed by NPT constructs (coherence, cognitive participation, collective action, reflexive monitoring) and implementation outcomes. The NoMAD survey, which measures NPT constructs, was completed by implementers at each hemodialysis center during early and late implementation. RESULTS: The four mechanisms posited in NPT had a dynamic and layered relationship during the implementation process. Collaborative participants participated because they believed in the value and legitimacy of palliative care for patients receiving hemodialysis and thus had high levels of cognitive participation at the start. Didactic Learning Sessions were important for building practice coherence, and sense-making was solidified through testing new skills in practice and first-hand observation during coaching visits by an expert. Collective action was hampered by limited time among team members and practical issues such as arranging meetings with patients. Reflexive monitoring of the positive benefit to patient and family experiences was key in shifting mindsets from disease-centric towards a patient-centered model of care. NoMAD survey scores showed modest improvement over time, with collective action having the lowest scores. CONCLUSIONS: NPT was a useful framework for understanding the implementation of palliative care practices within hemodialysis centers. We found a nonlinear relationship among the mechanisms which is reflected in our model of implementation of palliative care practices through a learning collaborative. These findings suggest that the implementation of complex practices such as palliative care may be more successful through iterative learning and practice opportunities as the mechanisms for change are layered and mutually reinforcing. TRIAL REGISTRATION: ClinicalTrials.gov, NCT04125537 . Registered 14 October 2019 - Retrospectively registered.
Asunto(s)
Buceo , Cuidados Paliativos , Humanos , Natación , Atención a la Salud , Encuestas y Cuestionarios , Investigación CualitativaRESUMEN
BACKGROUND: Peripheral arterial disease (PAD) is underdiagnosed, partially due to a high prevalence of atypical symptoms and a lack of physician and patient awareness. Implementing clinical decision support tools powered by machine learning algorithms may help physicians identify high-risk patients for diagnostic workup. OBJECTIVE: This study aims to evaluate barriers and facilitators to the implementation of a novel machine learning-based screening tool for PAD among physician and patient stakeholders using the Consolidated Framework for Implementation Research (CFIR). METHODS: We performed semistructured interviews with physicians and patients from the Stanford University Department of Primary Care and Population Health, Division of Cardiology, and Division of Vascular Medicine. Participants answered questions regarding their perceptions toward machine learning and clinical decision support for PAD detection. Rapid thematic analysis was performed using templates incorporating codes from CFIR constructs. RESULTS: A total of 12 physicians (6 primary care physicians and 6 cardiovascular specialists) and 14 patients were interviewed. Barriers to implementation arose from 6 CFIR constructs: complexity, evidence strength and quality, relative priority, external policies and incentives, knowledge and beliefs about intervention, and individual identification with the organization. Facilitators arose from 5 CFIR constructs: intervention source, relative advantage, learning climate, patient needs and resources, and knowledge and beliefs about intervention. Physicians felt that a machine learning-powered diagnostic tool for PAD would improve patient care but cited limited time and authority in asking patients to undergo additional screening procedures. Patients were interested in having their physicians use this tool but raised concerns about such technologies replacing human decision-making. CONCLUSIONS: Patient- and physician-reported barriers toward the implementation of a machine learning-powered PAD diagnostic tool followed four interdependent themes: (1) low familiarity or urgency in detecting PAD; (2) concerns regarding the reliability of machine learning; (3) differential perceptions of responsibility for PAD care among primary care versus specialty physicians; and (4) patient preference for physicians to remain primary interpreters of health care data. Facilitators followed two interdependent themes: (1) enthusiasm for clinical use of the predictive model and (2) willingness to incorporate machine learning into clinical care. Implementation of machine learning-powered diagnostic tools for PAD should leverage provider support while simultaneously educating stakeholders on the importance of early PAD diagnosis. High predictive validity is necessary for machine learning models but not sufficient for implementation.
RESUMEN
This cohort study uses a deidentified national administrative claims database to assess the association of eligibility expansion with abdominal aortic aneurysm screening and diagnosis.
Asunto(s)
Aneurisma de la Aorta Abdominal , Tamizaje Masivo , Humanos , Aneurisma de la Aorta Abdominal/diagnóstico por imagen , UltrasonografíaRESUMEN
BACKGROUND: Opioids are commonly prescribed for postoperative pain, but may lead to prolonged use and addiction. Diabetes impairs nerve function, complicates pain management, and makes opioid prescribing particularly challenging. METHODS: This retrospective observational study included a cohort of postoperative patients from a multisite academic health system to assess the relationship between diabetes, pain, and prolonged opioid use (POU), 2008-2019. POU was defined as a new opioid prescription 3-6 months after discharge. The odds that a patient had POU was assessed using multivariate logistic regression controlling for patient factors (e.g., demographic and clinical factors, as well as prior pain and opiate use). FINDINGS: A total of 43,654 patients were included, 12.4% with diabetes. Patients with diabetes had higher preoperative pain scores (2.1 vs 1.9, p<0.001) and lower opioid naïve rates (58.7% vs 68.6%, p<0.001). Following surgery, patients with diabetes had higher rates of POU (17.7% vs 12.7%, p<0.001) despite receiving similar opioid prescriptions at discharge. Patients with Type I diabetes were more likely to have POU compared to other patients (Odds Ratio [OR]: 2.22; 95% Confidence Interval [CI]:1.69-2.90 and OR:1.44, CI: 1.33-1.56, respectively). INTERPRETATION: In conclusion, surgical patients with diabetes are at increased risk for POU even after controlling for likely covariates, yet they receive similar postoperative opiate therapy. The results suggest a more tailored approach to diabetic postoperative pain management is warranted.
Asunto(s)
Diabetes Mellitus , Alcaloides Opiáceos , Trastornos Relacionados con Opioides , Humanos , Analgésicos Opioides/efectos adversos , Manejo del Dolor , Pautas de la Práctica en Medicina , Dolor Postoperatorio/tratamiento farmacológico , Diabetes Mellitus/tratamiento farmacológicoRESUMEN
BACKGROUND: Many healthcare systems seek to improve care for complex high-risk patients, but engaging such patients to actively participate in their healthcare can be challenging. OBJECTIVE: To identify and describe types of patient engagement strategies reported as successfully deployed by providers/teams and experienced by patients in a Veterans Health Administration (VA) intensive primary care (IPC) pilot program. METHODS: We conducted semi-structured qualitative telephone interviews with 29 VA IPC staff (e.g., physicians, nurses, psychologists) and 51 patients who had at least four IPC team encounters. Interviews were recorded, transcribed, and analyzed thematically using a combination a priori/inductive approach. RESULTS: The engagement strategies successfully deployed by the IPC providers/teams could be considered either more "facilitative," i.e., facilitated by and dependent on staff actions, or more "self-sustaining," i.e., taught to patients, thus cultivating their ongoing patient self-care. Facilitative strategies revolved around enhancing patient access and coordination of care, trust-building, and addressing social determinants of health. Self-sustaining strategies were oriented around patient empowerment and education, caregiver and/or community support, and boundaries and responsibilities. When patients described their experiences with the "facilitative" strategies, many discussed positive proximal outcomes (e.g., increased access to healthcare providers). Self-sustaining strategies led to positive (self-reported) longer-term clinical outcomes, such as behavior change. CONCLUSION: We identified two categories of strategies for successfully engaging complex, high-risk patients: facilitative and self-sustaining. Intensive primary care program leaders may consider thoughtfully building "self-sustaining" engagement strategies into program development. Future research can confirm their effectiveness in improving health outcomes.
Asunto(s)
Médicos , Humanos , Personal de Salud , Pacientes , Participación del PacienteRESUMEN
BACKGROUND: Treatment of surgical pain is a common reason for opioid prescriptions. Being able to predict which patients are at risk for opioid abuse, dependence, and overdose (opioid-related adverse outcomes [OR-AE]) could help physicians make safer prescription decisions. We aimed to develop a machine-learning algorithm to predict the risk of OR-AE following surgery using Medicaid data with external validation across states. METHODS: Five machine learning models were developed and validated across seven US states (90-10 data split). The model output was the risk of OR-AE 6-months following surgery. The models were evaluated using standard metrics and area under the receiver operating characteristic curve (AUC) was used for model comparison. We assessed calibration for the top performing model and generated bootstrap estimations for standard deviations. Decision curves were generated for the top-performing model and logistic regression. RESULTS: We evaluated 96,974 surgical patients aged 15 and 64. During the 6-month period following surgery, 10,464 (10.8%) patients had an OR-AE. Outcome rates were significantly higher for patients with depression (17.5%), diabetes (13.1%) or obesity (11.1%). The random forest model achieved the best predictive performance (AUC: 0.877; F1-score: 0.57; recall: 0.69; precision:0.48). An opioid disorder diagnosis prior to surgery was the most important feature for the model, which was well calibrated and had good discrimination. CONCLUSIONS: A machine learning models to predict risk of OR-AE following surgery performed well in external validation. This work could be used to assist pain management following surgery for Medicaid beneficiaries and supports a precision medicine approach to opioid prescribing.
Asunto(s)
Analgésicos Opioides , Alcaloides Opiáceos , Humanos , Analgésicos Opioides/uso terapéutico , Medicaid , Pautas de la Práctica en Medicina , Manejo del Dolor , Estudios RetrospectivosRESUMEN
INTRODUCTION: There are limited longitudinal data on the cost of treating patients with cirrhosis, which hampers value-based improvement initiatives. METHODS: We conducted a retrospective cohort study of patients with cirrhosis seen in the Veterans Affairs health care system from 2011 to 2015. Patients were followed up through 2019. We identified a sex-matched and age-matched control cohort without cirrhosis. We estimated incremental annual health care costs attributable to cirrhosis for 4 years overall and in subgroups based on severity (compensated, decompensated), cirrhosis complications (ascites, encephalopathy, varices, hepatocellular cancer, acute kidney injury), and comorbidity (Deyo index). RESULTS: We compared 39,361 patients with cirrhosis with 138,964 controls. The incremental adjusted costs for caring of patients with cirrhosis were $35,029 (95% confidence interval $32,473-$37,585) during the first year and ranged from $14,216 to $17,629 in the subsequent 3 years. Cirrhosis complications accounted for most of these costs. Costs of managing patients with hepatic encephalopathy (year 1 cost, $50,080) or ascites ($50,364) were higher than the costs of managing patients with varices ($20,488) or hepatocellular cancer ($37,639) in the first year. Patients with acute kidney injury or those who had multimorbidity were the most costly at $64,413 and $66,653 in the first year, respectively. DISCUSSION: Patients with cirrhosis had substantially higher health care costs than matched controls and multimorbid patients had even higher costs. Cirrhosis complications accounted for most of the excess cost, so preventing complications has the largest potential for cost saving and could serve as targets for improvement.
RESUMEN
AIMS: To compare healthcare costs and use between United States (US) Veterans Health Administration (VHA) patients with opioid use disorder (OUD) who experienced an opioid overdose (OD cohort) and patients with OUD who did not experience an opioid overdose (non-OD cohort). DESIGN: This is a retrospective cohort study of administrative and clinical data. SETTING: The largest integrated national health-care system is the US Veterans Health Administration's healthcare systems. PARTICIPANTS: We included VHA patients diagnosed with OUD from October 1, 2017 through September 30, 2018. We identified the index date of overdose for patients who had an overdose. Our control group, which included patients with OUD who did not have an overdose, was randomly assigned an index date. A total of 66 513 patients with OUD were included for analysis (OD cohort: n = 1413; non-OD cohort: n = 65 100). MEASUREMENTS: Monthly adjusted healthcare-related costs and use in the year before and after the index date. We used generalized estimating equation models to compare patients with an opioid overdose and controls in a difference-in-differences framework. FINDINGS: Compared with the non-OD cohort, an opioid overdose was associated with an increase of $16 890 [95% confidence interval (CI) = $15 611-18 169; P < 0.001] in healthcare costs for an estimated $23.9 million in direct costs to VHA (95% CI = $22.1 million, $25.7 million) within the 30 days following overdose after adjusting for baseline characteristics. Inpatient costs ($13 515; 95% CI = $12 378-14 652; P < 0.001) reflected most of this increase. Inpatient days (+6.15 days; 95% CI, = 5.33-6.97; P < 0.001), inpatient admissions (+1.01 admissions; 95% CI = 0.93-1.10; P < 0.001) and outpatient visits (+1.59 visits; 95% CI = 1.34-1.84; P < 0.001) also increased in the month after opioid overdose. Within the overdose cohort, healthcare costs and use remained higher in the year after overdose compared with pre-overdose trends. CONCLUSIONS: The US Veterans Health Administration patients with opioid use disorder (OUD) who have experienced an opioid overdose have increased healthcare costs and use that remain significantly higher in the month and continuing through the year after overdose than OUD patients who have not experienced an overdose.