Clinical Outcome and Utilization Profiles Among Latent Groups of High-Risk Patients: Moving from Segmentation Towards Intervention.

BACKGROUND: The ability of latent class models to identify clinically distinct groups among high-risk patients has been demonstrated, but it is unclear how healthcare data can inform group-specific intervention design. OBJECTIVE: Examine how utilization patterns across latent groups of high-risk patients provide actionable information to guide group-specific intervention design. DESIGN: Cohort study using data from 2012 to 2015. PATIENTS: Participants were 934,787 patients receiving primary care in the Veterans Health Administration, with predicted probability of 12-month hospitalization in the top 10th percentile during 2014. MAIN MEASURES: Patients were assigned to latent groups via mixture-item response theory models based on 28 chronic conditions. We modeled odds of all-cause mortality, hospitalizations, and 30-day re-hospitalizations by group membership. Detailed outpatient and inpatient utilization patterns were compared between groups. KEY RESULTS: A total of 764,257 (81.8%) of patients were matched with a comorbidity group. Groups were characterized by substance use disorders (14.0% of patients assigned), cardiometabolic conditions (25.7%), mental health conditions (17.6%), pain/arthritis (19.1%), cancer (15.3%), and liver disease (8.3%). One-year mortality ranged from 2.7% in the Mental Health group to 14.9% in the Cancer group, compared to 8.5% overall. In adjusted models, group assignment predicted significantly different odds of each outcome. Groups differed in their utilization of multiple types of care. For example, patients in the Pain group had the highest utilization of in-person primary care, with a mean (SD) of 5.3 (5.0) visits in the year of follow-up, while the Substance Use Disorder group had the lowest, with 3.9 (4.1) visits. The Substance Use Disorder group also had the highest rates of using services for housing instability (25.1%), followed by the Liver group (10.1%). CONCLUSIONS: Latent groups of high-risk patients had distinct hospitalization and utilization profiles, despite having comparable levels of predicted baseline risk. Utilization profiles pointed towards system-specific care needs that could inform tailored interventions.

Two-year change in latent classes of comorbidity among high-risk Veterans in primary care: a brief report.

BACKGROUND: Segmentation models such as latent class analysis are an increasingly popular approach to inform group-tailored interventions for high-risk complex patients. Multiple studies have identified clinically meaningful high-risk segments, but few have evaluated change in groupings over time. OBJECTIVES: To describe population-level and individual change over time in latent comorbidity groups among Veterans at high-risk of hospitalization in the Veterans Health Administration (VA). RESEARCH DESIGN: Using a repeated cross-sectional design, we conducted a latent class analysis of chronic condition diagnoses. We compared latent class composition, patient high-risk status, and patient class assignment in 2018 to 2020. SUBJECTS: Two cohorts of eligible patients were selected: those active in VA primary care and in the top decile of predicted one-year hospitalization risk in 2018 (n = 951,771) or 2020 (n = 978,771). MEASURES: Medical record data were observed from January 2016-December 2020. Latent classes were modeled using indicators for 26 chronic health conditions measured with a 2-year lookback period from study entry. RESULTS: Five groups were identified in both years, labeled based on high prevalence conditions: Cardiometabolic (23% in 2018), Mental Health (18%), Substance Use Disorders (16%), Low Diagnosis (25%), and High Complexity (10%). The remaining 8% of 2018 patients were not assigned to a group due to low predicted probability. Condition prevalence overall and within groups was stable between years. However, among the 563,725 patients identified as high risk in both years, 40.8% (n = 230,185) had a different group assignment in 2018 versus 2020. CONCLUSIONS: In a repeated latent class analysis of nearly 1 million Veterans at high-risk for hospitalization, population-level groups were stable over two years, but individuals often moved between groups. Interventions tailored to latent groups need to account for change in patient status and group assignment over time.

Dual Receipt of Prescription Opioids From the Department of Veterans Affairs and Medicare Part D and Prescription Opioid Overdose Death Among Veterans: A Nested Case-Control Study.

Background: More than half of enrollees in the U.S. Department of Veterans Affairs (VA) are also covered by Medicare and can choose to receive their prescriptions from VA or from Medicare-participating providers. Such dual-system care may lead to unsafe opioid use if providers in these 2 systems do not coordinate care or if prescription use is not tracked between systems. Objective: To evaluate the association between dual-system opioid prescribing and death from prescription opioid overdose. Design: Nested case-control study. Setting: VA and Medicare Part D. Participants: Case and control patients were identified from all veterans enrolled in both VA and Part D who filled at least 1 opioid prescription from either system. The 215 case patients who died of a prescription opioid overdose in 2012 or 2013 were matched (up to 1:4) with 833 living control patients on the basis of date of death (that is, index date), using age, sex, race/ethnicity, disability, enrollment in Medicaid or low-income subsidies, managed care enrollment, region and rurality of residence, and a medication-based measure of comorbid conditions. Measurements: The exposure was the source of opioid prescriptions within 6 months of the index date, categorized as VA only, Part D only, or VA and Part D (that is, dual use). The outcome was unintentional or undetermined-intent death from prescription opioid overdose, identified from the National Death Index. The association between this outcome and source of opioid prescriptions was estimated using conditional logistic regression with adjustment for age, marital status, prescription drug monitoring programs, and use of other medications. Results: Among case patients, the mean age was 57.3 years (SD, 9.1), 194 (90%) were male, and 181 (84%) were non-Hispanic white. Overall, 60 case patients (28%) and 117 control patients (14%) received dual opioid prescriptions. Dual users had significantly higher odds of death from prescription opioid overdose than those who received opioids from VA only (odds ratio [OR], 3.53 [95% CI, 2.17 to 5.75]; P < 0.001) or Part D only (OR, 1.83 [CI, 1.20 to 2.77]; P = 0.005). Limitation: Data are from 2012 to 2013 and cannot capture prescriptions obtained outside the VA or Medicare Part D systems. Conclusion: Among veterans enrolled in VA and Part D, dual use of opioid prescriptions was independently associated with death from prescription opioid overdose. This risk factor for fatal overdose among veterans underscores the importance of care coordination across health care systems to improve opioid prescribing safety. Primary Funding Source: U.S. Department of Veterans Affairs.

Impact of deprescribing AChEIs on aggressive behaviors and antipsychotic prescribing.

INTRODUCTION: We evaluated the impact of deprescribing acetylcholinesterase inhibitors (AChEIs) on aggressive behaviors and incident antipsychotic use in nursing home (NH) residents with severe dementia. METHODS: We conducted a retrospective study of Medicare claims, Part D, Minimum Data Set for NH residents aged 65+ with severe dementia receiving AChEIs in 2016. Aggressive behaviors were measured using the aggressive behavior scale (ABS; n = 30,788). Incident antipsychotic prescriptions were evaluated among antipsychotic non-users (n = 25,188). Marginal structural models and inverse probability of treatment weights were used to evaluate associations of AChEI deprescribing and outcomes. RESULTS: The severity of aggressive behaviors was low at baseline (mean ABS = 0.5) and was not associated with deprescribing AChEIs (0.002 increase in ABS, P = .90). Incident antipsychotic prescribing occurred in 5.1% of residents and was less likely with AChEI deprescribing (adjusted odds ratio = 0.52 [0.40-0.68], P <.001]). DISCUSSION: Deprescribing AChEIs was not associated with a worsening of aggressive behaviors or incident antipsychotic prescriptions.

Receipt of Overlapping Opioid and Benzodiazepine Prescriptions Among Veterans Dually Enrolled in Medicare Part D and the Department of Veterans Affairs: A Cross-sectional Study.

Background: Overlapping use of opioids and benzodiazepines is associated with increased risk for overdose. Veterans receiving medications concurrently from the U.S. Department of Veterans Affairs (VA) and Medicare may be at higher risk for such overlap. Objective: To assess the association between dual use of VA and Medicare drug benefits and receipt of overlapping opioid and benzodiazepine prescriptions. Design: Cross-sectional. Setting: VA and Medicare. Participants: All veterans enrolled in VA and Medicare Part D who filled at least 2 opioid prescriptions in 2013 (n = 368 891). Measurements: Outcomes were the proportion of patients with a Pharmacy Quality Alliance (PQA) measure of opioid-benzodiazepine overlap (≥2 filled prescriptions for benzodiazepines with ≥30 days of overlap with opioids) and the proportion of patients with high-dose opioid-benzodiazepine overlap (≥30 days of overlap with a daily opioid dose >120 morphine milligram equivalents). Augmented inverse probability weighting regression was used to compare these measures by prescription drug source: VA only, Medicare only, or VA and Medicare (dual use). Results: Of 368 891 eligible veterans, 18.3% received prescriptions from the VA only, 30.3% from Medicare only, and 51.4% from both VA and Medicare. The proportion with PQA opioid-benzodiazepine overlap was larger for the dual-use group than the VA-only group (23.1% vs. 17.3%; adjusted risk ratio [aRR], 1.27 [95% CI, 1.24 to 1.30]) and Medicare-only group (23.1% vs. 16.5%; aRR, 1.12 [CI, 1.10 to 1.14]). The proportion with high-dose overlap was also larger for the dual-use group than the VA-only group (4.7% vs. 2.3%; aRR, 2.23 [CI, 2.10 to 2.36]) and Medicare-only group (4.7% vs. 2.9%; aRR, 1.06 [CI, 1.02 to 1.11]). Limitation: Data are from 2013 and cannot capture medications purchased without insurance; unmeasured confounding may remain in this cross-sectional study. Conclusion: Among a national cohort of veterans dually enrolled in VA and Medicare, receiving prescriptions from both sources was associated with greater risk for receiving potentially unsafe overlapping prescriptions for opioids and benzodiazepines. Primary Funding Source: U.S. Department of Veterans Affairs.

Physicians' Perspectives Regarding Prescription Drug Monitoring Program Use Within the Department of Veterans Affairs: a Multi-State Qualitative Study.

BACKGROUND: The Department of Veterans Affairs (VA) has implemented robust strategies to monitor prescription opioid dispensing, but these strategies have not accounted for opioids prescribed by non-VA providers. State-based prescription drug monitoring programs (PDMPs) are a potential tool to identify VA patients' receipt of opioids from non-VA prescribers, and recent legislation requires their use within VA. OBJECTIVE: To evaluate VA physicians' perspectives and experiences regarding use of PDMPs to monitor Veterans' receipt of opioids from non-VA prescribers. DESIGN: Qualitative study using semi-structured interviews. PARTICIPANTS: Forty-two VA primary care physicians who prescribed opioids to 15 or more Veterans in 2015. We sampled physicians from two states with PDMPs (Massachusetts and Illinois) and one without prescriber access to a PDMP at the time of the interviews (Pennsylvania). APPROACH: From February to August 2016, we conducted semi-structured telephone interviews that addressed the following topics regarding PDMPs: overall experiences, barriers to optimal use, and facilitators to improve use. KEY RESULTS: VA physicians broadly supported use of PDMPs or desired access to one, while exhibiting varying patterns of PDMP use dictated by state laws and their clinical judgment. Physicians noted administrative burdens and incomplete or unavailable prescribing data as key barriers to PDMP use. To facilitate use, physicians endorsed (1) linking PDMPs with the VA electronic health record, (2) using templated notes to document PDMP use, and (3) delegating routine PDMP queries to ancillary staff. CONCLUSIONS: Despite the time and administrative burdens associated with their use, VA physicians in our study broadly supported PDMPs. The application of our findings to ongoing PDMP implementation efforts may strengthen PDMP use both within and outside VA and improve the safe prescribing of opioids.

Impact of Dual Use of Department of Veterans Affairs and Medicare Part D Drug Benefits on Potentially Unsafe Opioid Use.

OBJECTIVES: To estimate the prevalence and consequences of receiving prescription opioids from both the Department of Veterans Affairs (VA) and Medicare Part D. METHODS: Among US veterans enrolled in both VA and Part D filling 1 or more opioid prescriptions in 2012 (n = 539 473), we calculated 3 opioid safety measures using morphine milligram equivalents (MME): (1) proportion receiving greater than 100 MME for 1 or more days, (2) mean days receiving greater than 100 MME, and (3) proportion receiving greater than 120 MME for 90 consecutive days. We compared these measures by opioid source. RESULTS: Overall, 135 643 (25.1%) veterans received opioids from VA only, 332 630 (61.7%) from Part D only, and 71 200 (13.2%) from both. The dual-use group was more likely than the VA-only group to receive greater than 100 MME for 1 or more days (34.3% vs 10.9%; adjusted risk ratio [ARR] = 3.0; 95% confidence interval [CI] = 2.9, 3.1), have more days with greater than 100 MME (42.5 vs 16.9 days; adjusted difference = 16.4 days; 95% CI = 15.7, 17.2), and to receive greater than 120 MME for 90 consecutive days (7.8% vs 3.1%; ARR = 2.2; 95% CI = 2.1, 2.3). CONCLUSIONS: Among veterans dually enrolled in VA and Medicare Part D, dual use of opioids was associated with more than 2 to 3 times the risk of high-dose opioid exposure.

Dual Health Care System Use and High-Risk Prescribing in Patients With Dementia: A National Cohort Study.

BACKGROUND: Recent federal policy changes attempt to expand veterans' access to providers outside the Department of Veterans Affairs (VA). Receipt of prescription medications across unconnected systems of care may increase the risk for unsafe prescribing, particularly in persons with dementia. OBJECTIVE: To investigate the association between dual health care system use and potentially unsafe medication (PUM) prescribing. DESIGN: Retrospective cohort study. SETTING: National VA outpatient care facilities in 2010. PARTICIPANTS: 75 829 veterans with dementia who were continuously enrolled in Medicare from 2007 to 2010; 80% were VA-only users, and 20% were VA-Medicare Part D (dual) users. MEASUREMENTS: Augmented inverse propensity weighting was used to estimate the effect of dual-system versus VA-only prescribing on 4 indicators of PUM prescribing in 2010: any exposure to Healthcare Effectiveness Data and Information Set (HEDIS) high-risk medication in older adults (PUM-HEDIS), any daily exposure to prescriptions with a cumulative Anticholinergic Cognitive Burden (ACB) score of 3 or higher (PUM-ACB), any antipsychotic prescription (PUM-antipsychotic), and any PUM exposure (any-PUM). The annual number of days of each PUM exposure was also examined. RESULTS: Compared with VA-only users, dual users had more than double the odds of exposure to any-PUM (odds ratio [OR], 2.2 [95% CI, 2.2 to 2.3]), PUM-HEDIS (OR, 2.4 [CI, 2.2 to 2.8]), and PUM-ACB (OR, 2.1 [CI, 2.0 to 2.2]). The odds of PUM-antipsychotic exposure were also greater in dual users (OR, 1.5 [CI, 1.4 to 1.6]). Dual users had an adjusted average of 44.1 additional days of any-PUM exposure (CI, 37.2 to 45.0 days). LIMITATION: Observational study design of veteran outpatients only. CONCLUSION: Among veterans with dementia, rates of PUM prescribing are significantly higher among dual-system users than with VA-only users. PRIMARY FUNDING SOURCE: U.S. Department of Veterans Affairs.

Association between aggressive care and bereaved families' evaluation of end-of-life care for veterans with non-small cell lung cancer who died in Veterans Affairs facilities.

BACKGROUND: To the authors' knowledge, little is known regarding the relationship between patients' and families' satisfaction with aggressive end-of-life care. Herein, the authors examined the associations between episodes of aggressive care (ie, chemotherapy, mechanical ventilation, acute hospitalizations, and intensive care unit admissions) within the last 30 days of life and families' evaluations of end-of-life care among patients with non-small cell lung cancer (NSCLC). METHODS: A total of 847 patients with NSCLC (34% of whom were aged <65 years) who died in a nursing home or intensive care, acute care, or hospice/palliative care (HPC) unit at 1 of 128 Veterans Affairs Medical Centers between 2010 and 2012 were examined. Data sources included Veterans Affairs administrative and clinical data, Medicare claims, and the Bereaved Family Survey. The response rate for the Bereaved Family Survey was 62%. RESULTS: Greater than 72% of veterans with advanced lung cancer who died in an inpatient setting had at least 1 episode of aggressive care and 31% received chemotherapy within the last 30 days of life. For all units except for HPC, when patients experienced at least 1 episode of aggressive care, bereaved families rated care lower compared with when patients did not receive any aggressive care. For patients dying in an HPC unit, the associations between overall ratings of care and ≥2 inpatient admissions or any episode of aggressive care were not found to be statistically significant. Rates of aggressive care were not associated with age, and family ratings of care were similar for younger and older patients. CONCLUSIONS: Aggressive care within the last month of life is common among patients with NSCLC and is associated with lower family evaluations of end-of-life care. Specialized care provided within an HPC unit may mitigate the negative effects of aggressive care on these outcomes. Cancer 2017;123:3186-94. © 2017 American Cancer Society.

Race/Ethnicity and End-of-Life Care Among Veterans.

BACKGROUND: Few studies have examined comprehensively racial/ethnic variations in quality of end-of-life care. OBJECTIVE: Examine end-of-life care quality received by Veterans and their families, comparing racial/ethnic minorities to nonminorities. RESEARCH DESIGN: This is a retrospective, cross-sectional analysis of chart review and survey data. SUBJECTS: Nearly all deaths in 145 Veterans Affairs Medical Centers nationally (n=94,697) in addition to Bereaved Family Survey (BFS) data (n=51,859) from October 2009 to September 2014. MEASURES: Outcomes included 15 BFS items and 4 indicators of high-quality end-of-life care, including receipt of a palliative care consult, chaplain visit, bereavement contact, and death in hospice/palliative care unit. Veteran race/ethnicity was measured via chart review and defined as non-Hispanic white, non-Hispanic black, Hispanic, or other. RESULTS: In adjusted models, no differences were observed by race/ethnicity in receipt of a palliative care consult or death in a hospice unit. Although black Veterans were less likely than white Veterans to receive a chaplain visit, Hispanic Veterans were more likely than white Veterans to receive a chaplain visit and to receive a bereavement contact. Less favorable outcomes for racial/ethnic minorities were noted on several BFS items. In comparison with family members of white Veterans, families of minority Veterans were less likely to report excellent overall care, and this difference was largest for black Veterans (48% vs. 62%). CONCLUSIONS: Bereaved family members of minority Veterans generally rate the quality of end-of-life care less favorably than those of white Veterans. Family perceptions are critical to the evaluation of equity and quality of end-of-life care.

Pharmacy Use in the First Year of the Veterans Choice Program: A Mixed-methods Evaluation.

BACKGROUND: The Veterans Choice Program (VCP) was created to ensure timely access to health care in the Department of Veterans Affairs (VA). Under this program, medications may be ordered by select non-VA clinicians to be dispensed by VA pharmacies, creating new challenges in ensuring medication safety. OBJECTIVES: To examine pharmaceutical use during the first year of the VCP and to understand barriers and facilitators for VA pharmacists to dispensing medications under the VCP. STUDY DESIGN: Mixed-methods evaluation. METHODS: We captured all prescriptions dispensed through the VCP and described the demographics of VCP users and their medications. We also conducted semistructured interviews of VA pharmacists, focusing on VA formulary management and experiences dispensing opioid and hepatitis C (HCV) medications. Codebook development and coding followed iterative qualitative methods. RESULTS: Overall, 17,346 Veterans received 56,426 VCP prescriptions from November 7, 2014 through November 7, 2015. The total medication cost was $27 million, 90% of which was for only 2772 HCV prescriptions. Topical eye drops and opioids represented the most commonly dispensed prescriptions (15.6% and 9.2% of all prescriptions, respectively). Pharmacists reported numerous challenges to dispensing VCP medications, including time required to contact non-VA clinicians about formulary issues, requiring controlled substance prescriptions to be hand delivered to VA pharmacies, and lack of access to laboratory data required to safely dispense medications. CONCLUSIONS: HCV-related medication costs predominated the first year of VCP, but this is likely to change going forward. The safe use of opioids, efficient management of nonformulary medications, and unintended new barriers to access created by the VCP must be addressed.

The Impact of Medication-Based Risk Adjustment on the Association Between Veteran Health Outcomes and Dual Health System Use.

BACKGROUND: Veterans commonly receive care from both Veterans Health Administration (VA) and non-VA sources (i.e., dual use). A major challenge in comparing health outcomes between dual users and VA-predominant users is applying an accurate method of risk adjustment. OBJECTIVE: To determine how different comorbidity indices affect the association between patterns of dual use and health outcomes. DESIGN: Retrospective cohort. PARTICIPANTS: A total of 316,775 community-dwelling Veterans (≥65 years) with type 2 diabetes who were enrolled in VA and fee-for-service Medicare from 2008 to 2010. METHODS: We determined the associations between dual use and death or diabetes-related hospitalization in FY 2010 using multivariable models incorporating claims-based (Elixhauser) or medication-based (RxRisk-V) risk adjustment. Dual use was classified using four previously identified groups of health services users: 1) VA-predominant, 2) VA + Medicare visits and labs, 3) VA + Medicare test strips, and 4) VA + Medicare medications. KEY RESULTS: Controlling for Elixhauser comorbidities, dual-use groups 2-4 had significantly decreased odds of death or hospitalization compared to VA-predominant users. Controlling for RxRisk-V comorbidities, groups 2-4 had increased odds of death compared to VA-predominant users, but variable odds of hospitalization, with group 2 having increased odds (OR 1.06, CI 1.04-1.09), while groups 3 (OR 0.96, CI 0.94-0.99) and 4 (OR 0.93, CI 0.89-0.97) had decreased odds. CONCLUSIONS: The method of risk adjustment drastically influences the direction of effect in health outcomes among dual users of VA and Medicare. These findings underscore the need for standardized and reliable risk adjustment methods that are not susceptible to measurement differences across different health systems.

Overlapping buprenorphine, opioid, and benzodiazepine prescriptions among veterans dually enrolled in Department of Veterans Affairs and Medicare Part D.

BACKGROUND: Buprenorphine is a key tool in the management of opioid use disorder, but there are growing concerns about abuse, diversion, and safety. These concerns are amplified for the Department of Veterans Affairs (VA), whose patients may receive care concurrently from multiple prescribers within and outside VA. To illustrate the extent of this challenge, we examined overlapping prescriptions for buprenorphine, opioids, and benzodiazepines among veterans dually enrolled in VA and Medicare Part D. METHODS: We constructed a cohort of all veterans dually enrolled in VA and Part D who filled an opioid prescription in 2012. We identified patients who received tablet or film buprenorphine products from either source. We calculated the proportion of buprenorphine recipients with any overlapping prescription (based on days supply) for a nonbuprenorphine opioid or benzodiazepine, focusing on veterans who received overlapping prescriptions from a different system than their buprenorphine prescription (Part D buprenorphine recipients receiving overlapping opioids or benzodiazepines from VA and vice versa). RESULTS: There were 1790 dually enrolled veterans with buprenorphine prescriptions, including 760 (43%) from VA and 1091 (61%) from Part D (61 veterans with buprenorphine from both systems were included in each group). Among VA buprenorphine recipients, 199 (26%) received an overlapping opioid prescription and 11 (1%) received an overlapping benzodiazepine prescription from Part D. Among Part D buprenorphine recipients, 208 (19%) received an overlapping opioid prescription and 178 (16%) received an overlapping benzodiazepine prescription from VA. Among VA and Part D buprenorphine recipients with cross-system opioid overlap, 25% (49/199) and 35% (72/208), respectively, had >90 days of overlap. CONCLUSIONS: Many buprenorphine recipients receive overlapping prescriptions for opioids and benzodiazepines from a different health care system than the one in which their buprenorphine was filled. These findings highlight a previously undocumented safety risk for veterans dually enrolled in VA and Medicare.

VA and Medicare Utilization Among Dually Enrolled Veterans with Type 2 Diabetes: A Latent Class Analysis.

BACKGROUND: Many Veterans treated within the VA Healthcare System (VA) are also enrolled in fee-for-service (FFS) Medicare and receive treatment outside the VA. Prior research has not accounted for the multiple ways that Veterans receive services across healthcare systems. OBJECTIVE: We aimed to establish a typology of VA and Medicare utilization among dually enrolled Veterans with type 2 diabetes. DESIGN: This was a retrospective cohort. PARTICIPANTS: 316,775 community-dwelling Veterans age ≥ 65 years with type 2 diabetes who were dually enrolled in the VA and FFS Medicare in 2008-2009. METHODS: Using latent class analysis, we identified classes of Veterans based upon their probability of using VA and Medicare diabetes care services, including patient visits, laboratory tests, glucose test strips, and medications. We compared the amount of healthcare use between classes and identified factors associated with class membership using multinomial regression. KEY RESULTS: We identified four distinct latent classes: class 1 (53.9%) had high probabilities of VA use and low probabilities of Medicare use; classes 2 (17.2%), 3 (21.8%), and 4 (7.0%) had high probabilities of VA and Medicare use, but differed in their Medicare services used. For example, Veterans in class 3 received test strips exclusively through Medicare, while Veterans in class 4 were reliant on Medicare for medications. Living ≥ 40 miles from a VA predicted membership in classes 3 (OR 1.1, CI 1.06-1.15) and 4 (OR 1.11, CI 1.04-1.18), while Medicaid eligibility predicted membership in class 4 (OR 4.30, CI 4.10-4.51). CONCLUSIONS: Veterans with diabetes can be grouped into four distinct classes of dual health system use, representing a novel way to characterize how patients use multiple services across healthcare systems. This classification has applications for identifying patients facing differential risk from care fragmentation.

Trajectories of Diabetes Medication Adherence and Hospitalization Risk: A Retrospective Cohort Study in a Large State Medicaid Program.

BACKGROUND: Numerous interventions are available to boost medication adherence, but the targeting of these interventions often relies on crude measures of poor adherence. Group-based trajectory models identify individuals with similar longitudinal prescription filling patterns. Identifying distinct adherence trajectories may be more useful for targeting interventions, although the association between adherence trajectories and clinical outcomes is unknown. OBJECTIVE: To examine the association between adherence trajectories for oral hypoglycemics and subsequent hospitalizations among diabetes patients. DESIGN: Retrospective cohort study. PATIENTS: A total of 16,256 Pennsylvania Medicaid enrollees, non-dually eligible for Medicare, initiating oral hypoglycemics between 2007 and 2009. MAIN MEASURES: We used group-based trajectory models to identify trajectories of oral hypoglycemics in the 12 months post-treatment initiation, using monthly proportion of days covered (PDC) as the adherence measure. Multivariable Cox proportional hazard models were used to examine the association between trajectories and time to first diabetes-related hospitalization/emergency department (ED) visits in the following year. We used the C-index to compare prediction performance between adherence trajectories and dichotomous cutpoints (annual PDC <80 vs. ≥80 %). RESULTS: The mean annual PDC was 0.58 (SD 0.32). Seven trajectories were identified: perfect adherers (9 % of the cohort), nearly perfect adherers (31.4 %), moderate adherers (21.0 %), low adherers (11.0 %), late discontinuers (6.8 %), early discontinuers (9.7 %), and non-adherers with only one fill (11.1 %). Compared to perfect adherers, trajectories of moderate adherers (HR = 1.48, 95 % CI 1.25, 1.75), low adherers (HR = 1.51, 95 % CI 1.25, 1.83), and non-adherers with only one fill (HR = 1.35, 95 % CI 1.09, 1.67) had greater risk of diabetes-related hospitalizations/ED visits. Predictive accuracy was improved using trajectories compared to dichotomized cutpoints (C-index = 0.714 vs. 0.652). CONCLUSIONS: Oral hypoglycemic treatment trajectories were highly variable in this large Medicaid cohort. Low and moderate adherers and those filling only one prescription had a modestly higher risk of hospitalizations/ED visits compared to perfect adherers. Trajectory models may be valuable in identifying specific non-adherence patterns for targeting interventions.

Cost-Utility Analysis of Long-Acting Beta Agonists versus Leukotriene Receptor Antagonists in Older Adults with Persistent Asthma Receiving Concomitant Inhaled Corticosteroid Therapy.

BACKGROUND: Long-acting beta agonists (LABA) and leukotriene receptor antagonists (LTRA) are the major add-on treatments in older adults with persistent asthma when inhaled corticosteroids (ICS) fail to achieve adequate asthma control. OBJECTIVES: To evaluate the cost-utility of ICS + LABA treatment compared with ICS + LTRA treatment in older adults with asthma. METHODS: A Markov model was used to estimate the incremental costs and quality-adjusted life expectancy associated with ICS + LABA treatment versus ICS + LTRA treatment in older adults with asthma in the United States from the health system perspective. The HCUPnet 2010 national statistics were used to extract the costs associated with asthma and cardiovascular hospitalizations, and inpatient mortality associated with these events. Event probabilities were predicted using Medicare 2009-2010 claims for older adults with asthma. Treatment costs were estimated on the basis of average wholesale drug price listings, and utility estimates were extracted from the literature. To account for uncertainty, one-way sensitivity analysis and probabilistic sensitivity analysis were performed. RESULTS: The model predicted that, compared with ICS + LTRA treatment, ICS + LABA treatment costs $5,823 more while gaining 0.03 quality-adjusted life-years (QALYs), resulting in an incremental cost-effectiveness ratio of $209,090 per QALY. Hospitalization probabilities and posthospitalization utilities were the most influential parameters in the one-way sensitivity analysis. Probabilistic uncertainty analysis using Monte-Carlo simulations showed that the probabilities that ICS + LTRA treatment is cost-effective compared with ICS + LABA treatment are 77% and 62% at $50,000 and $100,000 per QALY gained willingness-to-pay thresholds, respectively. CONCLUSIONS: The cost-effectiveness of ICS + LABA treatment is economically unfavorable in older adults when compared with LTRA as add-on treatment.

Antidepressant Use and Recurrent Falls in Community-Dwelling Older Adults: Findings From the Health ABC Study.

BACKGROUND: Few studies have compared the risk of recurrent falls across various antidepressant agents-using detailed dosage and duration data-among community-dwelling older adults, including those who have a history of a fall/fracture. OBJECTIVE: To examine the association of antidepressant use with recurrent falls, including among those with a history of falls/fractures, in community-dwelling elders. METHODS: This was a longitudinal analysis of 2948 participants with data collected via interview at year 1 from the Health, Aging and Body Composition study and followed through year 7 (1997-2004). Any antidepressant medication use was self-reported at years 1, 2, 3, 5, and 6 and further categorized as (1) selective serotonin reuptake inhibitors (SSRIs), (2) tricyclic antidepressants, and (3) others. Dosage and duration were examined. The outcome was recurrent falls (≥2) in the ensuing 12-month period following each medication data collection. RESULTS: Using multivariable generalized estimating equations models, we observed a 48% greater likelihood of recurrent falls in antidepressant users compared with nonusers (adjusted odds ratio [AOR] = 1.48; 95% CI = 1.12-1.96). Increased likelihood was also found among those taking SSRIs (AOR = 1.62; 95% CI = 1.15-2.28), with short duration of use (AOR = 1.47; 95% CI = 1.04-2.00), and taking moderate dosages (AOR = 1.59; 95% CI = 1.15-2.18), all compared with no antidepressant use. Stratified analysis revealed an increased likelihood among users with a baseline history of falls/fractures compared with nonusers (AOR = 1.83; 95% CI = 1.28-2.63). CONCLUSION: Antidepressant use overall, SSRI use, short duration of use, and moderate dosage were associated with recurrent falls. Those with a history of falls/fractures also had an increased likelihood of recurrent falls.

Asthma medication use among late midlife U.S. Adults.

OBJECTIVE: Despite substantial prevalence of asthma, little is known about asthma in late midlife adults (50-64 years). The objective of this study was to examine the factors associated with the use of asthma medications among late midlife adults. METHODS: Pooled data were obtained from the 2006 to 2010 Medical Expenditure Panel Survey. Medication use outcome variables include: (a) daily use of a preventive asthma medication and (b) use of more than three canisters of rescue inhalers in last 3 months. The Andersen Behavioral Model of Health Services Utilization was used to guide the selection of independent variables. Descriptive, unadjusted and adjusted logistic regression analyses were performed. Point estimates were weighted to the US civilian population and variance estimates were adjusted to obtain appropriate standard errors. All analyses were conducted using STATA (version 12). RESULTS: A total of 1414 (weighted sample of 15,030,364) self-reported late midlife asthmatics were identified. About 31% of late midlife adults with asthma were using a preventive medication on a daily basis while 11% reported overusing acute medications. Adjusted analyses found that race, rurality and smoking were related to poor use of asthma medications among late midlife adults. CONCLUSION: Results suggest that asthma medication use is far from optimal among vulnerable groups of late midlife US adults.

Using machine learning to examine medication adherence thresholds and risk of hospitalization.

BACKGROUND: Quality improvement efforts are frequently tied to patients achieving ≥80% medication adherence. However, there is little empirical evidence that this threshold optimally predicts important health outcomes. OBJECTIVE: To apply machine learning to examine how adherence to oral hypoglycemic medications is associated with avoidance of hospitalizations, and to identify adherence thresholds for optimal discrimination of hospitalization risk. METHODS: A retrospective cohort study of 33,130 non-dual-eligible Medicaid enrollees with type 2 diabetes. We randomly selected 90% of the cohort (training sample) to develop the prediction algorithm and used the remaining (testing sample) for validation. We applied random survival forests to identify predictors for hospitalization and fit survival trees to empirically derive adherence thresholds that best discriminate hospitalization risk, using the proportion of days covered (PDC). OUTCOMES: Time to first all-cause and diabetes-related hospitalization. RESULTS: The training and testing samples had similar characteristics (mean age, 48 y; 67% female; mean PDC=0.65). We identified 8 important predictors of all-cause hospitalizations (rank in order): prior hospitalizations/emergency department visit, number of prescriptions, diabetes complications, insulin use, PDC, number of prescribers, Elixhauser index, and eligibility category. The adherence thresholds most discriminating for risk of all-cause hospitalization varied from 46% to 94% according to patient health and medication complexity. PDC was not predictive of hospitalizations in the healthiest or most complex patient subgroups. CONCLUSIONS: Adherence thresholds most discriminating of hospitalization risk were not uniformly 80%. Machine-learning approaches may be valuable to identify appropriate patient-specific adherence thresholds for measuring quality of care and targeting nonadherent patients for intervention.

Anticholinergic Use and Recurrent Falls in Community-Dwelling Older Adults: Findings From the Health ABC Study.

BACKGROUND: Although it is generally accepted that anticholinergic use may lead to a fall, results from studies assessing the association between anticholinergic use and falls are mixed. In addition, direct evidence of an association between use of anticholinergic medications and recurrent falls among community-dwelling elders is not available. OBJECTIVE: To assess the association between anticholinergic use across multiple anticholinergic subclasses, including over-the-counter medications, and recurrent falls. METHODS: This was a longitudinal analysis of 2948 participants, with data collected via interview at year 1 from the Health, Aging and Body Composition study and followed through year 7 (1997-2004). Self-reported use of anticholinergic medication was identified at years 1, 2, 3, 5, and 6 as defined by the list from the 2015 American Geriatrics Society Beers Criteria. Dosage and duration were also examined. The main outcome was recurrent falls (≥2) in an ensuing 12-month period from each medication data collection. RESULTS: Using multivariable generalized estimating equation models, controlling for demographic, health status/behaviors, and access-to-care factors, a 34% increase in likelihood of recurrent falls in anticholinergic users (adjusted odds ratio = 1.34; 95% CI = 0.93-1.93) was observed, but the results were not statistically significant; similar results were found with higher doses and longer duration of use. CONCLUSION: Increased point estimates suggest an association of anticholinergic use with recurrent falls, but the associations did not reach statistical significance. Future studies are needed for more definitive evidence and to examine other measures of anticholinergic burden and associations with more intermediate adverse effects such as cognitive function.