Relationship between medication adherence for opioid use disorder and health care costs and health care events in a claims dataset.

INTRODUCTION: Medication for opioid use disorder (MOUD) has well-documented benefits for treating OUD, though its efficacy depends on patient adherence. We know little about outcomes of MOUD nonadherence compared to treatment regimens without MOUD, and this article aims to address the gap. This analysis focused on office-based MOUD treatment (buprenorphine and naltrexone) to evaluate the long-term impact of adherence on subsequent health care costs and health care events. METHODS: With claims data from 2017 to 2019, we used propensity score (PS) weighting to create three comparable cohorts of patients: 1) Adherent: filled MOUD prescription & ≧80 % of days covered by MOUD (N = 1045); 2) Nonadherent: filled MOUD & < 80 % of days covered (N = 1116), 3) did not fill MOUD (N = 16,784). The study defined three time intervals based on a patient's most recent MOUD episode: A 6-month baseline period before initiation of MOUD or random index date for those with MOUD; a 6-month treatment period, during which adherence or nonadherence was established; and a 12-month follow-up period to evaluate outcome measures. The study used generalized PS methodology to examine the effect of proportion of days covered (PDC) as a continuous measure of adherence. RESULTS: Among patients who filled MOUD, adherence to MOUD was significantly predicted by having less severe OUD, being older, having fewer inpatient visits and lower outpatient costs before the start of treatment. Adherent patients displayed significantly lower health care costs in the follow-up period compared to nonadherent MOUD patients, and lower odds of experiencing health care events. The nonadherent MOUD group displayed significantly higher odds of health care events compared to patients who had no evidence of receiving MOUD in claims data (NO-MOUD). Among patients prescribed MOUD, each 10 % increase in PDC was associated with a significant decrease in inpatient/outpatient costs and in odds of health care events. CONCLUSIONS: This analysis aligns with previous findings about the importance of maintaining long-term adherence to MOUD in supporting patient outcomes. The results also suggest a novel finding that despite confounder control via PS methods, nonadherent patients display poorer outcomes compared to similar NO-MOUD patients.

Characterizing the 'feel-good experience' in multiple sclerosis patients treated with natalizumab or other therapies.

Aim: Patients with relapsing-remitting multiple sclerosis (RRMS) treated with natalizumab have anecdotally reported a 'feel-good experience' (FGE). The authors characterized the FGE using survey data from patients with RRMS treated with natalizumab or other disease-modifying therapies (other-DMT). Methods: Questionnaire data from RRMS patients who use MyMSTeam, an online patient social network, were analyzed. Results: The survey included 347 patients (95 natalizumab; 252 other-DMT). More natalizumab than other-DMT patients self-reported having an FGE (62.1 vs 44.8%; p = 0.001) as well as other physical, emotional and cognitive benefits. Conclusion: This study demonstrates that physical, emotional and cognitive benefits were more commonly reported by patients treated with natalizumab than those treated with other disease-modifying therapies and helps characterize patient-reported factors associated with the FGE.

No difference in radiologic outcomes for natalizumab patients treated with extended interval dosing compared with standard interval dosing: Real-world evidence from MS PATHS.

BACKGROUND: Extended interval dosing (EID; average dosing interval approximately every 6 weeks) of natalizumab is associated with significantly lower risk of progressive multifocal leukoencephalopathy than standard interval dosing (SID; every 4 weeks) in patients with relapsing-remitting multiple sclerosis (MS). Real-world studies, though limited, suggest that natalizumab effectiveness is generally maintained in patients who switch to EID after initiation of stable treatment with SID. MS PATHS (Multiple Sclerosis Partners Advancing Technology and Health Solutions) is a collaborative, multicenter learning health system that generates real-world clinical and MRI data using highly standardized acquisition protocols. We compared MRI outcomes in MS PATHS patients treated with natalizumab EID versus SID. We also compared MRI outcomes in patients treated with natalizumab (EID and/or SID) versus injectable MS platform therapy. METHODS: Natalizumab infusion data from the TOUCH Prescribing Program database and MS PATHS MRI assessment data from seven US sites as of July 23, 2020, were used to identify patients with relapsing-remitting MS who had received natalizumab EID or SID in the interval between two MRI scans (an MRI segment). Patients who received injectable platform MS therapy between two MRI scans were also identified. MRI data were used to determine the incidence rate and odds of developing new or enlarging T2 lesions, annualized percentage change in T2 lesion volume (T2LV), and annualized percentage change in brain parenchymal fraction (BPF). MRI outcomes were compared for 1) natalizumab EID treatment versus natalizumab SID treatment, 2) natalizumab treatment (EID + SID) versus platform therapy, and 3) natalizumab EID versus platform therapy. Propensity score-based weighting or matching were used to balance covariates at the start of MRI segments for all comparisons. RESULTS: The MRI outcomes observed with natalizumab EID treatment did not differ significantly from those observed with natalizumab SID treatment. The odds ratio for any new or enlarging T2 lesion was 1.07 (95% confidence interval [CI]: 0.93, 1.24; p = 0.355), and the rate ratio (95% CI) for new or enlarging T2 lesions was 1.62 (0.93, 2.82; p = 0.090). Differences (95% CI) between EID and SID patients in mean annualized percentage change in T2LV and BPF were 1.56% (-3.77%, 6.90%; p = 0.566) and -0.11% (-0.25%, -0.10%; p = 0.096), respectively. Conversely, when MRI outcomes in natalizumab and platform therapy patients were compared, there were significant differences favoring natalizumab in all assessments: the odds of any new or enlarging T2 lesion (odds ratio: 0.69 [95% CI: 0.64, 0.75]; p<0.001), the incidence rate of new or enlarging T2 lesions (rate ratio: 0.47 [95% CI: 0.37, 0.61]; p<0.001), annualized percentage change (decrease) in T2LV (difference: -3.68% [95% CI: -7.06%, -0.30%]; p = 0.033), and annualized percentage change (increase) in BPF (difference: 0.22% [95% CI: 0.16%, 0.29%]; p<0.001). Results of the subgroup comparison of natalizumab EID patients with platform therapy patients were similar to those of the overall-natalizumab-group-versus-platform-therapy comparison. CONCLUSIONS: The results indicate that natalizumab EID and SID provide comparable real-world effectiveness on quantitative MRI metrics. These data further demonstrate that natalizumab EID can provide superior real-world effectiveness to injectable platform therapy on quantitative MRI metrics.

No evidence for loss of natalizumab effectiveness with every-6-week dosing: a propensity score-matched comparison with every-4-week dosing in patients enrolled in the Tysabri Observational Program (TOP).

BACKGROUND: Extended interval dosing of natalizumab is associated with significantly lower progressive multifocal leukoencephalopathy risk compared with every-4-week (Q4W) dosing in patients with relapsing-remitting multiple sclerosis. Previous studies have suggested that natalizumab effectiveness is maintained in patients who switch from Q4W to extended interval dosing but have been limited by a lack of well-matched patient cohorts. METHODS: Tysabri Observational Program (TOP) data as of November 2019 were used to identify patients with relapsing-remitting multiple sclerosis treated with natalizumab Q4W and those with a single physician-indicated dosing change from Q4W to every-6-week (Q6W) dosing after ⩾1 year of Q4W treatment. Patients were propensity score matched at the time of the switch from Q4W to Q6W dosing. Clinical outcomes (annualized relapse rate and probability of remaining relapse free or free of 24-week confirmed disability worsening) and safety outcomes were assessed for the two cohorts. RESULTS: This study included 219 pairs of propensity score-matched Q6W and Q4W patients. Annualized relapse rates were similar for Q6W (0.150) and Q4W (0.157) patients. The probability of remaining relapse free [hazard ratio = 1.243 (95% confidence interval = 0.819-1.888); p = 0.307] and of remaining free of 24-week confirmed disability worsening [hazard ratio = 0.786 (95% confidence interval = 0.284-2.176); p = 0.644] did not differ significantly between Q6W and Q4W patients. Summarized safety results for the matched Q6W and Q4W patients are also presented. CONCLUSION: These real-world findings in well-matched patient cohorts from TOP demonstrate that natalizumab effectiveness is maintained in patients who switch to Q6W dosing after ⩾1 year of Q4W dosing. CLINICALTRIALSGOV IDENTIFIER: NCT00493298.

Uncertainty in the design stage of two-stage Bayesian propensity score analysis.

The two-stage process of propensity score analysis (PSA) includes a design stage where propensity scores (PSs) are estimated and implemented to approximate a randomized experiment and an analysis stage where treatment effects are estimated conditional on the design. This article considers how uncertainty associated with the design stage impacts estimation of causal effects in the analysis stage. Such design uncertainty can derive from the fact that the PS itself is an estimated quantity, but also from other features of the design stage tied to choice of PS implementation. This article offers a procedure for obtaining the posterior distribution of causal effects after marginalizing over a distribution of design-stage outputs, lending a degree of formality to Bayesian methods for PSA that have gained attention in recent literature. Formulation of a probability distribution for the design-stage output depends on how the PS is implemented in the design stage, and propagation of uncertainty into causal estimates depends on how the treatment effect is estimated in the analysis stage. We explore these differences within a sample of commonly used PS implementations (quantile stratification, nearest-neighbor matching, caliper matching, inverse probability of treatment weighting, and doubly robust estimation) and investigate in a simulation study the impact of statistician choice in PS model and implementation on the degree of between- and within-design variability in the estimated treatment effect. The methods are then deployed in an investigation of the association between levels of fine particulate air pollution and elevated exposure to emissions from coal-fired power plants.

Sequential screening for depression in humanitarian emergencies: a validation study of the Patient Health Questionnaire among Syrian refugees.

BACKGROUND: Despite the need for mental health surveillance in humanitarian emergencies, there is a lack of validated instruments. This study evaluated a sequential screening process for major depressive disorder (MDD) using the two- and eight-item Patient Health Questionnaires (PHQ-2 and PHQ-8, respectively). METHODS: This study analyzed data collected during a cross-sectional survey in a Syrian refugee camp in Greece (n = 135). The response rate for each instrument was assessed, and response burden was calculated as the number of items completed. The sequential screening process was simulated to replicate the MDD classifications captured if the PHQ-2 was used to narrow the population receiving the full PHQ-8 assessment. All respondents were screened using the PHQ-2. Only respondents scoring ≥ 2 are considered at risk for symptoms of MDD and complete the remaining six items. The positive and negative percent agreement of this sequential screening process were evaluated. RESULTS: The PHQ-2, PHQ-2/8 sequential screening process, and PHQ-8 were completed by 91%, 87%, and 84% of respondents, respectively. The sequential screening process had a positive percent agreement of 89% and a negative percent agreement of 100%, and eliminated the need to complete the full PHQ-8 scale for 34 (25%) respondents. CONCLUSIONS: The benefits of the sequential screening approach for the classification of MDD presented here are twofold: preserving classification accuracy relative to the PHQ-2 alone while reducing the response burden of the PHQ-8. This sequential screening approach is a pragmatic strategy for streamlining MDD surveillance in humanitarian emergencies.

Major depressive disorder prevalence and risk factors among Syrian asylum seekers in Greece.

BACKGROUND: Over one million Syrian asylum seekers have travelled to Greece with the ultimate aim of reaching other countries in western Europe. Depression prevalence and associated sociodemographic and displacement characteristics have been reported for resettled migrants. However, the prevalence of major depressive disorder (MDD) and its risk factors have not been described among migrants engaged in the asylum process ensuing from the Syrian crisis. This study provides new data about the mental health status of migrants in transition in the context of protracted asylum procedures. METHODS: We conducted a cross-sectional survey in a Syrian refugee camp in the Attica region of Greece from January 16-31, 2017. Individuals ≥18 years of age with verbal Arabic or English language skills were eligible to participate. The Patient Health Questionnaire-8 (PHQ-8), an eight-item validated diagnostic and severity measure, was used to screen for MDD. We analysed the relationships between MDD and sociodemographic and displacement characteristics using multivariable logistic regression. RESULTS: A total of 135 surveys were completed, representing 40% of the adult population in the refugee camp. The mean age of the participants was 30 years (18-61 years); women comprised 41% of the sample; 74% of the participants had ever married; 67% had children; and 33% of participants had not attended secondary school, including 11% who had never attended school. Median time since departing the country of origin was 12 months (< 1-74 months). Median time spent in the asylum process in Greece was 10 months (< 1-49 months). MDD was detected in 44% (95% CI: 37-50) of participants. Being a woman (Adjusted Odds Ratio [AOR]: 3.23, p = 0.019), each additional child (AOR: 1.61, p = 0.006), and increased time in the asylum process in Greece (AOR: 1.15, p = 0.043) were significant risk factors for MDD. Ever being married was associated with reduced odds of MDD (AOR: 0.23, p = 0.042). CONCLUSIONS: Syrian migrants face an extraordinarily high burden of MDD as they seek asylum. Incorporation of screening and treatment into service provision within refugee camps is urgently needed, particularly as migrants spend extended periods of time in transition due to protracted asylum procedures.