BACKGROUND:The COVID-19 pandemic affected sexually transmitted infection (STI) testing and diagnosis rates in the United States, but these patterns have not been well characterized among people with HIV. METHODS:We analyzed medical records data of HIV Outpatient Study (HOPS) participants seen for HIV care from January 2019 to March 2021, with ≥1 CD4+ cell count and viral load test results recorded. We used Poisson regression models to estimate rate ratio (RR) and 95% confidence interval (CI) to compare STI testing and diagnoses rates on/after versus before March 1, 2020 (early COVID-19 pandemic [pandemic] vs. prepandemic). RESULTS:Of 2311 eligible patients, STI tests (STI cases, primarily defined as test results that were positive) were as follows during the analysis time frame: 4991 gonorrhea (157), 4978 chlamydia (135), and 4216 syphilis (114). Comparing pandemic versus prepandemic periods, STI testing RRs were 0.78 for both gonorrhea (95% CI, 0.73-0.82) and chlamydia (95% CI, 0.73-0.83), and 0.93 for syphilis (95% CI, 0.88-0.99); diagnosis rates were not statistically different. Multivariable models showed reduced testing for gonorrhea (adjusted RR, 0.79; 95% CI, 0.72-0.87) and chlamydia (adjusted RR, 0.78; 95% CI, 0.71-0.86) for men who have sex with men, but not for other HIV transmission groups. CONCLUSIONS:The fallout of the COVID-19 pandemic on sexual health may not be seen for some time. Despite reduced STI testing, rates of STI diagnoses did not decrease. It will take a return to more routine screening and improved access to sexual health care to uncover the true impact of undetected or untreated STIs.

This article aimed at analyzing the acute impact and the longer-term recovery of COVID-19 pandemic effects on clinical encounter types, HIV viral load (VL) testing, and suppression (HIV VL < 200 copies/mL). This study was a longitudinal cohort study of participants seen during 2019-2022 at nine HIV Outpatient Study (HOPS) sites. Generalized linear mixed models (GLMMs) estimated monthly rates of all encounters, office and telemedicine visits, and HIV VL tests using 2010-2022 data. We examined factors associated with nonsuppressed VL (VL >= 200 copies/mL) and not having ambulatory care visits during the pandemic using GLMM for logistic regression with 2017-2022 and 2019-2022 data, respectively. Of 2351 active participants, 76.0% were male, 57.6% aged >= 50 years, 40.7% non-Hispanic White, 38.2% non-Hispanic Black, 17.3% Hispanic/Latino, and 51.0% publicly insured. The monthly rates of in-person and telemedicine visits varied during 2020 through mid-year 2022. Multivariable logistic regression showed that persons with no encounters were more likely to be male or have VL >= 200 copies/mL. For participants with >= 1 VL test, the prevalence rate of HIV VL >= 200 copies/mL during 2020 was close to the rates from 2014 to 2019. The change in probability of viral suppression was not associated with participant's age, sex, race/ethnicity, or insurance type. In the HOPS, overall patient encounters declined over 2 years during the pandemic with variations in telemedicine and in-person events, with relative maintenance of viral suppression. Ongoing recovery from the impact of COVID-19 on ambulatory care will require continued efforts to improve retention and patient access to medical services.

Current U.S. guidelines recommend integrase strand transfer inhibitor (INSTI)-based antiretroviral therapy (ART) as initial treatment for people with HIV (PWH). We assessed long-term effects of INSTI use on lipid profiles in routine HIV care. We analyzed medical record data from the HIV Outpatient Study’s participants in care from 2007 to 2021. Hyperlipidemia was defined based on clinical diagnoses, treatments, and laboratory results. We calculated hyperlipidemia incidence rates and rate ratios (RRs) during initial ART and assessed predictors of incident hyperlipidemia by using Poisson regression. Among 349 eligible ART-naïve PWH, 168 were prescribed INSTI-based ART (36 raltegravir (RAL), 51 dolutegravir (DTG), and 81 INSTI-others (elvitegravir and bictegravir)) and 181 non-INSTI-based ART, including 68 protease inhibitor (PI)-based ART. During a median follow-up of 1.4 years, hyperlipidemia rates were 12.8, 22.3, 22.7, 17.4, and 12.6 per 100 person years for RAL-, DTG-, INSTI-others-, non-INSTI-PI-, and non-INSTI-non-PI-based ART, respectively. In multivariable analysis, compared with the RAL group, hyperlipidemia rates were higher in INSTI-others (RR = 2.25; 95% confidence interval (CI): 1.29–3.93) and non-INSTI-PI groups (RR = 1.89; CI: 1.12–3.19) but not statistically higher for the DTG (RR = 1.73; CI: 0.95–3.17) and non-INSTI-non-PI groups (RR = 1.55; CI: 0.92–2.62). Other factors independently associated with hyperlipidemia included older age, non-Hispanic White race/ethnicity, and ART without tenofovir disoproxil fumarate. PWH using RAL-based regimens had lower rates of incident hyperlipidemia than PWH receiving non-INSTI-PI-based ART but had similar rates as those receiving DTG-based ART, supporting federal recommendations for using DTG-based regimens as the initial therapy for ART-naïve PWH.

Background:The timing and magnitude of antiretroviral therapy-associated weight change attributions are unclear.Setting:HIV Outpatient Study participants.Methods:We analyzed 2007-2018 records of virally suppressed (VS) persons without integrase inhibitor (INSTI) experience who switched to either INSTI-based or another non-INSTI-based ART, and remained VS. We analyzed BMI changes using linear mixed models, INSTI- and tenofovir alafenamide (TAF) contributions to BMI change by linear mixed models-estimated slopes, and BMI inflection points.Results:Among 736 participants (5316 person-years), 441 (60%) switched to INSTI-based ART; the remainder to non-INSTI-based ART. The mean follow-up was 7.15 years for INSTI recipients and 7.35 years for non-INSTI. Preswitch, INSTI and non-INSTI groups had similar median BMI (26.3 versus 25.9 kg/m(2), P = 0.41). INSTI regimens included raltegravir (178), elvitegravir (112), and dolutegravir (143). Monthly BMI increases postswitch were greater with INSTI than non-INSTI (0.0525 versus 0.006, P < 0.001). A BMI inflection point occurred 8 months after switch among INSTI users; slopes were similar regardless of TAF use immediately postswitch. Among INSTI + TAF users, during 8 months postswitch, 87% of BMI slope change was associated with INSTI use, 13% with TAF use; after 8 months, estimated contributions were 27% and 73%, respectively. For non-INSTI+TAF, 84% of BMI gain was TAF-associated consistently postswitch. Persons switching from TDF to TAF had greater BMI increases than others (P < 0.001).Conclusion:Among VS persons who switched ART, INSTI and TAF use were independently associated with BMI increases. During 8 months postswitch, BMI changes were greatest and most associated with INSTI use; afterward, gradual BMI gain was largely TAF-associated.

Introduction At the population level, Black and Hispanic adults in the United States have increased risk of dying from COVID-19, yet whether race and ethnicity impact on risk of mortality among those hospitalized for COVID-19 is unclear. Methods Retrospective cohort study using data on adults hospitalized with COVID-19 from the electronic health record from 52 health systems across the United States contributing data to Cerner Real World Data TM . In-hospital mortality was evaluated by race first in unadjusted analysis then sequentially adjusting for demographics and clinical characteristics using logistic regression. Results Through August 2020, 19,584 patients with median age 52 years were hospitalized with COVID-19, including n = 4,215 (21.5%) Black and n = 5,761 (29.4%) Hispanic patients. Relative to white patients, crude mortality was slightly higher in Black adults [22.7% vs 20.8%, unadjusted OR 1.12 (95% CI 1.02–1.22)]. Mortality remained higher among Black adults after adjusting for demographic factors including age, sex, date, region, and insurance status (OR 1.13, 95% CI 1.01–1.27), but not after including comorbidities and body mass index (OR 1.07, 95% CI 0.93–1.23). Compared with non-Hispanic patients, Hispanic patients had lower mortality both in unadjusted and adjusted models [mortality 12.7 vs 25.0%, unadjusted OR 0.44(95% CI 0.40–0.48), fully adjusted OR 0.71 (95% CI 0.59–0.86)]. Discussion In this large, multicenter, EHR-based analysis, Black adults hospitalized with COVID-19 had higher observed mortality than white patients due to a higher burden of comorbidities in Black adults. In contrast, Hispanic ethnicity was associated with lower mortality, even in fully adjusted models.

Background: COVID-19 pandemic effects on ambulatory care services for persons living with HIV in the United States, including on frequency of office visits and HIV viral load (VL) testing, have not been well described. Methods: We analyzed longitudinal data of active patients (any encounter after January 1, 2020) at 8 HIV Outpatient Study (HOPS) sites. Monthly rates of all-inclusive encounters (office, lab, pharmacy, hospital, telemedicine [TM], phone, other), office and TM (O&T) visits, and HIV VL tests were derived from generalized linear mixed models, using data available from January 2010 to June 2020. We then assessed temporal trends and the adjusted effects of sociodemographic and clinical factors on the rates of O&T visits during the COVID-19 pandemic, in multivariable Poisson regression of data from January to June 2020. Results: Of 1251 active patients, 71% were male, 57% aged ≥ 50 years, 36% non-Hispanic white, 42% non-Hispanic black, 19% Hispanic/Latino, and 49% publicly insured. Median CD4 count was 680 cells/mm3 and 93% had suppressed ( 0.20). The increase in TM visits (2%) did not offset the decline in office visits (26%). The HIV VL testing rate fell by 50% in the first 6 months of 2020 among patients who had VL test done in 2019 (Figure). Conclusion: In the HOPS, the rates of office visits and HIV VL tests dropped precipitously after March 2020. The long-term implications for clinical outcomes and HIV viral suppression may not be evident at this time in the COVID-19 pandemic but HIV care sites need strategies to ensure patients maintain engagement in care and HIV laboratory monitoring.

Human immunodeficiency virus (HIV) and hepatitis C virus (HCV) coinfection carries substantial risk for all-cause mortality and liver-related morbidity and mortality, yet many persons coinfected with HIV/HCV remain untreated for HCV. We explored demographic, clinical, and sociodemographic factors among participants in routine HIV care associated with prescription of direct-acting antivirals (DAAs). The HIV Outpatient Study (HOPS) is an ongoing longitudinal cohort study of persons with HIV in care at participating clinics since 1993. There are currently eight study sites in six US cities. We analyzed medical records data of HOPS participants diagnosed with HCV since June 2010. Sustained virological response (SVR) was documented with first undetectable HCV viral load (VL). We assessed factors associated with being prescribed DAAs by multi-variable logistic regression and described the cumulative rate of SVR. Among 306 eligible participants, 131 (43%) were prescribed DAA therapy. Factors associated with greater odds of being prescribed DAA were older age, private health insurance, higher CD4 cell count, being a person who injects drugs, and receiving care at publicly funded sites (p<0.05). Of 127 (97%) participants with at least 1 follow-up HCV VL, 110 (87%) achieved SVR at 12 weeks. Of the total 131 participants, 123 (94%) eventually achieved SVR. Less than half of HIV/HCV coinfected patients in HOPS have been prescribed DAAs. Interventions are needed to address deficits in DAA prescription, including among patients with public or no health insurance, younger age, and lower CD4 cell count.

BACKGROUND:Direct oral anticoagulants (DOACs) have become first-line treatment for venous thrombotic events. DOAC prescribing trends among people living with human immunodeficiency virus (PWH) are not well described. The coadministration of DOACs with the antiretroviral (ARV) pharmacokinetic boosters ritonavir (RTV) or cobicistat (COBI) may be complicated by pharmacokinetic interactions.METHODS:A longitudinal cohort study was conducted using the D.C. Cohort Database in Washington, D.C., from January 2011 to March 2017, to describe oral anticoagulant prescribing among PWH ≥ 18 years old and the prevalence of DOAC use with RTV or COBI. Data collection included demographic and clinical characteristics, ARV and anticoagulant prescriptions, and International Classification of Diseases Ninth and Tenth Edition diagnosis codes.RESULTS:Among 8315 PWH, there were 236 anticoagulant prescriptions (96 DOAC, 140 warfarin) for 206 persons. PWH prescribed anticoagulants were predominantly Black (82%) and male (82%), with a mean age at anticoagulant initiation of 56 years. DOAC use increased from 3% of total anticoagulant prescribing in 2011 to 43% in 2016, accounting for 64% of all newly recorded anticoagulant prescriptions by 2016. There were 19 bleeding events recorded among 16 individuals. Despite the Food and Drug Administration label recommendation to avoid rivaroxaban with boosted ARVs, 41% remained on boosted ARVs after rivaroxaban initiation.CONCLUSIONS:DOAC use increased substantially in PWH by 2016. Although rivaroxaban is not recommended with RTV or COBI, concomitant use was recorded in 41% of rivaroxaban recipients in this cohort. As DOAC usage increases, clinicians need to be aware of potential DOAC/ARV interactions in order to select the most appropriate oral anticoagulant and monitoring plan for PWH.

Background: Despite widely available access to HIV care in Washington, DC, inequities in HIV outcomes persist. We hypothesized that laboratory monitoring and virologic outcomes would not differ significantly based on insurance type. Methods: We compared HIV monitoring with outcomes among people with HIV (PWH) with private (commercial payer) versus public (Medicare, Medicaid) insurance receiving care at community and hospital clinics. The DC Cohort follows over 8,000 PWH from 14 clinics. We included those ≥18 years old enrolled between 2011-2015 with stable insurance. Outcomes included frequency of CD4 count and HIV RNA monitoring ( > 2 lab measures/year, > 30 days apart) and durable viral suppression (VS; HIV RNA <50 copies/mL at last visit and receiving antiretroviral therapy (ART) for ≥12 months). Multivariable logistic regression models examined impact of demographic and clinical factors. Results: Among 3,908 PWH, 67.9% were publicly-insured and 58.9% attended community clinics. Compared with privately insured participants, a higher proportion of publicly insured participants had the following characteristics: female sex, Black race, heterosexual, unemployed, and attending community clinics. Despite less lab monitoring, privately-insured PWH had greater durable VS than publicly-insured PWH (ART-naïve: private 70.0% vs public 53.1%, p=0.03; ART-experienced: private 80.2% vs public 69.4%, p<0.0001). Privately-insured PWH had greater durable VS than publicly-insured PWH at hospital clinics (AOR=1.59, 95% CI: 1.20, 2.12; p=0.001). Conclusions: Paradoxical differences between HIV monitoring and durable VS exist among publicly and privately-insured PWH in Washington, DC. Programs serving PWH must improve efforts to address barriers creating inequity in HIV outcomes.

INTRODUCTION:Prior studies found renal disease was common among HIV-infected outpatients. We updated incident renal disease estimates in this population, comparing those with and without tenofovir exposure. METHODS:We conducted a retrospective analysis of the DC Cohort, a longitudinal study of HIV patients in Washington, DC, from 2011 to 2015. We included adults prescribed antiretroviral therapy (ART) with baseline glomerular filtration rate (GFR) ≥15 ml/min per 1.73 m2. We defined renal disease as 50% decrease in GFR or doubled serum creatinine (Cr) within 3 months. We defined cumulative viral load as area under the curve (AUC) of log10 transformed longitudinal HIV RNA viral load (VL). Correlates of time to incident renal disease were identified using Cox proportional hazard regression models, adjusted for demographics and known risk factors for kidney disease. RESULTS:Among 6068 adults, 77% were Black and median age was 48 years. Incident renal disease rate was 0.77 per 100 person-years (95% confidence interval [CI]: 0.65-0.9). Factors associated with renal disease were age (adjusted hazard ratio [aHR]: 1.4; CI 1.1-1.7 per 10 years), public non-Medicaid, non-Medicare insurance (aHR: 3.4; CI: 1.9-6.4), AUC VL (aHR: 1.1; CI: 1.1-1.2), diabetes mellitus (aHR: 1.6; CI: 1.0-2.4), and mildly reduced GFR (60-89 ml/min per 1.73 m2) (aHR: 1.5; CI: 1.0-2.3); recent tenofovir exposure was not associated with renal disease (aHR: 0.7; CI: 0.5-1.1). CONCLUSION:Our study revealed a substantial burden of renal disease among HIV patients. Cumulative VL was associated with renal disease, suggesting that early VL suppression may decrease its incidence.

Objective: To evaluate associations of mood, anxiety, stress-/trauma-related, and psychotic disorders, both treated and untreated, with duration of unsuppressed HIV viral load (VL) among persons living with HIV (PLWH). Setting: The DC Cohort, an observational clinical cohort of PLWH followed from 2011 to 2018 at 14 sites in Washington, DC. Methods: Among PLWH ≥18 years old who received primary care at their HIV clinic, we determined in a time-updated manner whether participants had diagnoses and pharmacologic prescriptions for mood, anxiety, stress-/trauma-related, and/or psychotic disorders. Associations between psychiatric disorders/treatments and the proportion of subsequent days with VL ≥200 copies/mL were assessed using multivariable Poisson regression with generalized estimating equations. Results: Among 5904 participants (median age 51; 70% men; 82% Black), 45% had ≥1 psychiatric disorder, including 38% with mood disorders (50% treated), 18% with anxiety or stress-/trauma-related disorders (64% treated), and 4% with psychotic disorders (52% treated). Untreated major depressive disorder (adjusted rate ratio = 1.17; 95% confidence interval: 1.00 to 1.37), untreated other/unspecified depressive disorder (1.23; 1.01 to 1.49), untreated bipolar disorder (1.39; 1.15 to 1.69), and treated bipolar disorder (1.25; 1.02 to 1.53) (vs. no mood disorder) predicted more time with VL ≥200 copies/mL. Treated anxiety disorders (vs. no anxiety disorder) predicted less time (0.78; 0.62 to 0.99). Associations were weaker and nonsignificant for treated depressive disorders (vs. no mood disorder) and untreated anxiety disorders (vs. no anxiety disorder). Conclusions: PLWH with depressive and bipolar disorders, particularly when untreated, spent more time with unsuppressed VL than PLWH without a mood disorder. Treatment of mood disorders may be important for promoting sustained viral suppression.

Hepatitis B (HBV) co-infection increases the risk for liver-related morbidity among HIV-infected patients. Department of Health and Human Services (DHHS) guidelines for the management of HIV/HBV co-infection recommend six-monthly monitoring of blood tests. We assessed longitudinal adherence to these guidelines among patients enrolled in the DC Cohort Study, a city-wide clinical cohort in Washington, DC. Patients ≥18 years old who were enrolled between January 1, 2011 and March 31, 2016 and had ≥6 months of follow-up were included. Advanced liver fibrosis was defined as having a FIB-4 score >3.25 calculated from same-day platelet, AST and ALT results. Viral suppression (VS) was defined as having undetectable HIV VL (<200 copies/mL). Chronic HBV status was determined using ICD 9 and 10 diagnosis codes. Clinical targets were defined as the proportion tested for HBV viral load (VL), platelet count and AST (markers of HBV care) along with HIV VL and CD4 count tests (markers of HIV care) every six months following enrollment. Among 7,631 HIV-infected patients, 354 (4.6%) had chronic HBV among whom 22 (6.2%) had advanced fibrosis. Compared with HIV-mono-infected patients, HIV/HBV patients were more likely to be male (86% vs 72%, P < 0.001) and had a history of AIDS (49% vs 37%, P < 0.001). Although HIV VS was high in both groups, co-infected patients were less likely to have undetectable HIV VL (88% in HIV only vs 84% in HIV/HBV, P = 0.026). HIV/HBV patients were nearly four times less likely to be tested for HBV than for HIV VL in the first six months (21% vs 79%, P <0.0001) and eight times less likely in the last six months of observation (6% vs 49%, P <0.0001). Comparing the first six months to the last six months of observation, the proportion of patients tested for AST (77% and 50% P < 0.0001) and platelets (76% and 57% P < 0.0001) declined. Adherence to the DHHS management guidelines for monitoring HBV VL among HIV/HBV co-infected patients was low. AST and platelet counts were monitored at a similar frequency to HIV VL, suggesting that these markers may have been checked as part of routine HIV care than for HBV monitoring. Focusing on adherence to guidelines may ensure early detection of complications and provide patients with timely and appropriate care. All authors: No reported disclosures.

Objective: To investigate the epidemiology, treatment, length of stay (LOS) and costs for neonatal and pediatric inpatients with invasive candidiasis (IC). Methods: The Cerner Health Facts Database was used to assess inpatients (2005-2014) identified by positive blood or cerebrospinal fluid (CSF) Candida cultures. Log-transformed LOS and cost were examined in candidemia-only patients (n = 191) using multivariable linear regression. Results: A total of 202 patients had a positive culture (blood: n = 192; CSF: n = 10; both: n = 2). The most prevalent species were C. parapsilosis (n = 70, 34.7%), and C. albicans (n = 66, 32.7%). Mean (SD) age was 5 (5.5) years; 30 (14.9%) patients were <4 months. Comorbidities included sepsis (n = 85, 42.1%), coagulation disorders (n = 57, 28.2%), cancer (n = 64, 31.7%), and low birthweight (n = 26, 12.9%). Antifungals (AFs) included azoles (57.4%), polyenes (28.7%), and echinocandins (35.1%); 20.8% of patients received no AF during their encounter. The mean (SD) cost per encounter was $ 97,392 ($ 149,253), with a mean (SD) LOS of 45.6 (59.5) days and 9.9% mortality at discharge. Results were similar across Candida species. In regression analysis, intensive care unit (ICU) exposure, central catheter, sepsis, AF >48 hours prior to index culture, and age <4 months were associated with increased LOS; treatment at a non-teaching hospital was associated with reduced LOS (p<. 05). AF use >48 hours before index, in-hospital mortality, Midwest region and ventricular shunt were associated with increased cost (p <=.05). Conclusions: This analysis confirms the association between pediatric candidemia and increased resource utilization and LOS. Given high observed rates of potential under-treatment, an opportunity may exist to improve AF therapy in this population.

Background.This retrospective observational study was conducted to gain insight into epidemiology, treatment, length of stay (LOS) and costs for hospitalized neonate and pediatric patients with invasive candidiasis (IC).Methods.The Cerner Health Facts Database, a multicenter US electronic health record database, was used to assess the clinical and economic impact of pediatric IC in inpatients (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014).Patient encounters were identified by positive blood/cerebrospinal fluid (CSF) cultures for Candida, and encounter characteristics were evaluated.The impact of multiple factors on (log-transformed) LOS and cost were examined in patients with candidemia only (n = 191) using multivariable linear regression.Model parameter uncertainty was evaluated with bootstrap analysis.Results.From 2005 to 2014, 202 patients had a positive culture (blood: n = 192; CSF: n = 12).The most prevalent species at index culture was C. parapsilosis (n = 70, 34.7%), followed by C. albicans (n = 66, 32.7%).Mean (SD) age was 5 (5.5) years, with 30 patients (14.9%) <4 months of age.Common comorbidities included sepsis (n = 85, 42.1%), coagulation disorders (n = 57, 28.2%), cancer (n = 64, 31.7%), and low birthweight (n = 26, 12.9%).Antifungal exposure included azoles (57.4%), polyenes (28.7%), and echinocandins (35.1%); 20.8% of patients had no record of receiving an antifungal during their index encounter.The mean cost per encounter was $97,392 ($149,253), with a mean LOS of 45.6 (59.5) days and 9.9% mortality at discharge.Results did not differ greatly across Candida species.In regression analysis, ICU exposure, central catheter, sepsis, receipt of an antifungal >48 hours prior to index culture, and age <4 months were significantly associated with increased LOS, while treatment at a non-teaching hospital was associated with reduced LOS ( p<0.05).Antifungal use >48 hours before index, alive at discharge, Midwest/ west region and ventricular shunt were significantly associated with increased cost (p ≤ 0.05).Conclusion.While limited by small sample size, this analysis confirms the association between neonatal and pediatric candidemia and increased resource utilization.However, given high observed rates of potential undertreatment, an opportunity may exist to improve antifungal therapy in this population.

Medication adherence broadly describes the extent to which individuals comply with prescribed treatment. Many approaches are used to assess medication adherence, including formulas to measure dosing/frequency compliance and persistence over time. To provide insight into their appropriateness for population health research, this study compared several measures of medication adherence through a retrospective claims analysis of onsite pharmacies. To cross validate and increase statistical power, multiple calculations were used to assess medication possession ratio (MPR) and percentage of days covered (PDC) among health plan members who used onsite or community pharmacies to fill prescriptions for certain chronic conditions. Additionally, the percentage of members without 30-day gaps in supplied medication and average days until discontinuation were calculated. The linear mixed model approach was used to gauge the onsite pharmacy's impact while adjusting for covariate effects. Across medication types, onsite pharmacy users demonstrated significantly higher adherence rates than community pharmacy users based on the calculations for MPR over 365 days (P < 0.0001), MPR over 730 days (P < 0.001), average PDC over 365 days (P < 0.0001), percentage of members without 30-day coverage gaps over 180 days (P < 0.05) and average number of days until discontinuation (P < 0.0001). This trend was supported by the mixed model analysis that grouped all medication types and estimated that the covariate adjusted odds of greater adherence was statistically significant (OR = 3.44; CI95%= 2.84, 4.16; P < 0.0001). The formula for the percentage of members without 30-day coverage gaps over 1 year yielded few differences between groups. Characteristics of the study population, methodology and data availability will influence estimates of medication adherence. To best assess the full scope of medication adherence within a population, it is therefore recommended that a variety of measures be considered and reported in research studies.

BACKGROUND:Poor medication adherence is associated with worsened health outcomes and higher health care expenditures. An increasing number of employers are sponsoring wellness initiatives designed to support healthy lifestyles, improve productivity, and offer a return on investment. Onsite pharmacies may facilitate higher medication adherence rates by providing employees a convenient, low-cost option for filling prescriptions that is integrated with other onsite health services.OBJECTIVES:To (a) assess the impact of an employer's onsite pharmacy on health plan members' medication adherence using multiple measures of medication adherence and persistence, including medication possession ratio (MPR), average number of days until discontinuation (60-day gap in coverage), and percentage of members without a 30-day gap in coverage, and (b) evaluate these outcomes between those members who participated in condition management programs and those who did not.METHODS:A retrospective analysis of a self-insured employer's claims data was undertaken. Medication adherence was assessed among the self-insured employer's health plan members, which included subscribers and their dependents who filled an asthma, depression, diabetes, hypertension, or hyperlipidemia medication at an onsite pharmacy, compared with those who used a community pharmacy. Multiple standard measures of medication adherence were considered. These measures included MPR, which was assessed for 1- and 2-year time periods. MPR was chosen because it is one of the most commonly referenced formulas in the literature and represents adherence over a fixed period of time. In addition, medication persistence was estimated by 30-day gaps in coverage and discontinuation of treatment. To assess the impact of onsite pharmacy use and account for covariate effects, the linear mixed model approach was applied with the logit transformed MPR as the response variable. An analysis of MPR among condition management participants was also performed.RESULTS:In total, 2,498 subscribers and their dependents were included in the analysis. The average MPR at 365 days was significantly higher (P  <  0.0001) among onsite pharmacy users for all medication types, ranging from 13% higher for depression medications to 20% higher for hypertension medications. This trend persisted at 730 days (P  <  0.001), with average MPRs ranging from 6% higher for hyperlipidemia medications to 11% higher for hypertension medications. A mixed model analysis indicated that members who used the onsite pharmacy were 3.44 times more likely to demonstrate medication adherence (95% CI = 2.84-4.16; P  <  0.0001) at 365 days. Likewise, at 180 and 365 days, onsite pharmacy users were less likely to have 30-day gaps in treatment. The average number of days until discontinuation (defined as a 60-day gap) was also significantly longer (P  <  0.0001) among onsite pharmacy users, ranging from an average of 56 additional days for depression medications to 105 additional days for hypertension medications. While the average MPR tended to be higher among those subscribers and their dependents who participated in condition management programs, this trend was not statistically significant for all medication types.CONCLUSIONS:Based on multiple measures, onsite pharmacy use was associated with higher medication adherence, while the results were inconclusive for condition management participation.

Musculoskeletal conditions are the primary cause of physical disability in the United States, and have implications for workplace productivity and employer health costs. This study determined the influence of employer-sponsored, onsite chiropractic care on health care utilization. A retrospective claims analysis of members of an employer-sponsored health plan receiving chiropractic care exclusively onsite or offsite from 2010-2012. Data were obtained from the employer’s health benefits administration program. Eligible participants had continuous enrollment in the health plan and ≥1 billing code for chiropractic services during the 36-month study period. Utilization was assessed in 2 categories: radiological procedures and clinical care. Utilization differences were evaluated by having ≥1 health care event in any category. Continuous variables were summarized as means, and binary variables as counts and proportions. Differences were assessed via chi-square test for categorical variables and t-test and Kruskal-Wallis non-parametric test for continuous variables. Comparisons were considered significant at alpha = 0.05. The analysis included 876 onsite and 759 offsite participants. The populations were similar in gender; the onsite group was slightly younger. The offsite group received more radiology services overall (55.5% vs 38.2%, P<0.001) including x-ray (46.0% vs 26.6%; P < 0.0001); ultrasound (15.8% vs 10.7%, P<0.0001), and magnetic resonance imaging (14.6% vs 12.4%, P<0.0001); had higher outpatient (47.3% vs 30.2%, P<0.0001) and emergency department (19.0% vs 13.1%; P = 0.022) utilization; and demonstrated greater use of chiropractic care (mean 15.2 vs 9.16 visits; P<0.0001) and physical therapy (mean 9.6 vs 1.5 visits; P<0.0001). In patients needing chiropractic care, those utilizing onsite services had lower health care utilization, including radiology services , clinical care, and counts of chiropractic and physical therapy services. The study results support the value of chiropractic services offered at onsite health centers, where more tightly managed and evidence-based approaches to musculoskeletal conditions can be facilitated.