A Retrospective Claims Analysis of Medication Adherence and Persistence Among Patients Taking Antidepressants for the Treatment of Major Depressive Disorder (MDD)

Thesis (Master's)--University of Washington, 2015

Background: Pharmacotherapy to treat Major Depressive Disorder (MDD) has proven to be effective, yet improving adherence and persistence (A&P) to antidepressant (AD) therapy continues to be challenging. Guidelines recommend initiating AD therapy through initiation phase (3 months of therapy), and continuation phase (6 months beyond initiation phase), with a recommended total duration of 9 months of therapy. Even at the minimal recommended duration of therapy (6 months), it is estimated that only 12-34% of patients remain adherent to AD therapy, increasing the chance of relapse back into the depressive episode and resource utilization. Further research on A&P is warranted due to lack of estimates for new AD therapy and incomparability among existing findings. Methods: To date, this was the largest US insurance claims analysis conducted of A&P to initial AD therapy. Truven Marketscan® Commercial, Medicare, and Medicaid databases were queried for MDD patients between the years 2003-2014. Qualifying MDD diagnoses for inclusion included either one inpatient, or one outpatient service claim with a second confirmatory claim (either inpatient or outpatient), after ensuring 6 months of negative MDD diagnosis and AD prescription history. Patients with a diagnosis of other mental disorders (Schizophrenia, Bipolar, Alzheimer’s, Dementia, depressive type psychosis, or other mood disorders) were excluded if diagnosed from 6 months before to 12 months after the IDD. 527,907 patients were identified who initiated therapy with one AD within 60 days of a qualifying MDD diagnosis and had continuous insurance coverage from 6 months before to 12 months after this index AD prescription date (IPD). A&P was calculated to initial AD medication, to initial therapeutic class, and overall, to any AD therapy, over the first 3, 6, 9, and 12 months from the IPD, using Medication Possession Ratio (MPR)/Proportion of Days Covered (PDC) and days to discontinuation, respectively. Adherence to continuation phase, or adherence between 4-9 months, was calculated similarly. The proportion adherent was defined as the proportion whose MPR/PDC was greater than or equal to 0.80, where the proportion persistent was defined as the proportion remaining persistent over the time frame. Chi-squared testing was used to test differences in the proportion adherent or persistent, when grouped by initial AD or comparing proportions over time. Odds ratios of adherence comparing initial ADs to sertraline were estimated at 3, 6, 9, and 12 months after the IPD through multivariable logistic regression, adjusting for age, gender, insurance source, region, insurance plan type, MDD diagnosis code, Charlson Comorbidity Index, and comorbid anxiety or chronic pain disorders. In a separate logistic model adjusted for the same covariates, the odds ratio of adherence to continuation phase among those adherent/not adherent to initiation phase was estimated. Results: The proportion adherent to initial AD medication over time frames was 0.43/0.41 (MPR/PDC) at 3 months, and 0.23/0.21 at 12 months (p value<0.0001). The proportion persistent to initial AD medication was 0.44 at 3 months and 0.17 at 12 months (p value<0.0001). A&P to initial therapeutic class and overall, to any AD therapy was slightly higher than to initial AD medication, yet each measure similarly decreased over time (p value<0.0001). For each logistic regression for the primary model, association of initial AD and adherence to the time frame was found to be significant (p value <0.0001). Compared to sertraline, patients initiating with desvenlafaxine, duloxetine, and venlafaxine XR had greater odds of adherence at 6 months (adjusted odds ratios of 1.30, 1.07, and 1.19, respectively; p values all <0.0001). In a separate logistic model, adherence to the first 3 months of treatment was associated with adherence to continuation phase, or months 4-9 (p value<0.0001). Patients who were adherent to the first 3 months of therapy had 13.7 times greater odds of adherence to continuation phase, when adjusted for covariates (95% Confidence Interval 13.4-14.0). Conclusion: Results support that the choice of initial AD is an important clinical decision, associated with A&P, and thus, has the potential to improve clinical outcomes and decrease resource utilization.