Prescribing Patterns Associated With Biologic Therapies for Psoriasis from a United States Medical Records Database

August 2019 | Volume 18 | Issue 8 | Original Article | 745 | Copyright © August 2019


Megan H. Noe MD MPH MSCE, Daniel B. Shin PhD, Jalpa A. Doshi PhD, David J. Margolis MD PhD, Joel M. Gelfand MD MSCE

University of Pennsylvania Perelman School of Medicine, Philadelphia, PA

METHODS

Data Source
We used data from OptumInSight’s electronic health records database (OEHR), a de-identified dataset containing information from over 81 million individuals in the United States. The OEHR database includes information from healthcare encounters, claims, and prescriptions. Inclusion in the database is based on receiving care within the OEHR network, not a specific insurance provider. Patients can seek medical care from sources outside of the network, but approximately 70% of individuals are from integrated delivery networks and thus all their healthcare encounters are captured. The EHR dataset available for this research was a psoriasis cohort derived from the full EHR dataset including patients with a diagnosis of psoriasis from January 1, 2007 to June 30, 2017 and a 10% random sample of all patients during the same period. This study was deemed exempt from review by the Institutional Review Board of the University of Pennsylvania.

Study Population
Patients were classified as having psoriasis if they had 2 diagnosis codes for psoriasis, on two separate days or 1 diagnosis for psoriasis and a subsequent prescription for a systemic psoriasis therapy or phototherapy on a separate day. Follow-up time began (index date) at the second qualifying event and was continued until the patient left the database or died. Only patients with at least 1 prescription for a biologic medication (adalimumab, etanercept, infliximab, ixekizumab, secukinumab, and ustekinumab) were included in this analysis.

Statistical Analysis
Descriptive statistics were used to examine demographics, medical comorbidities, and psychiatric comorbidities for all patients who received at least 1 prescription for a biologic medication. Medical and psychiatric comorbidities were determined by the presence of at least 1 ICD-9/ICD-10 diagnosis code prior to the index date. A prescription episode was defined as all prescriptions written for a single medication to a single individual. Individuals may have multiple prescription episodes if they received prescriptions for more than one biologic medication; however only the first episode for each drug was included in the analysis. Time between first and last prescription was used as a measure of drug survival in this EHR dataset, because information about drug supply and refills were missing from the majority of prescriptions. At least 1 prescription every 180 days was required to be considered continuous therapy. A time period of greater than 180 days between prescriptions was considered to be a lapse in therapy. Sensitivity analyses including only prescriptions written in the last 3 years of the dataset (2014 – 2017) to look for changes in prescribing practices over time and varying the definition of continuous therapy in the calculation of time between first and last prescription were performed. To evaluate concomitant use of topical medications, phototherapy, and other systemic agents in patients receiving prescriptions for biologics we identified a subgroup of patients with prescription episodes of at least 12 months duration. Individuals were included in this subgroup if they had two prescriptions for the same biologic within a 12-month period and a third prescription within the next 6 months. All prescriptions written for concomitant psoriasis therapy during this prescription episode were collected. Duplicate prescriptions written for the same medication on the same day were excluded. Prescriptions were analyzed by category of topical medications (any topical, topical steroid, vitamin D analogs, and calcineurin inhibitors) and oral systemic therapies (acitretin, apremilast, cyclosporine, and methotrexate). Additionally, CPT codes for any type of phototherapy during the biologic prescription episode were counted. Sensitivity analyses were performed excluding prescriptions written for topical therapies in the first 6 months of the biologic prescription episode and excluding oral therapies written in the first 3 months to exclude concomitant prescriptions only written during possible disease flares. Sensitivity analyses were also performed excluding patients with a history of psoriatic arthritis.

RESULTS

There were 34,714 patients who met the inclusion criteria for psoriasis and also were prescribed at least one biologic medication (Table 1). Of these patients, 17,888 (51.5%) were female with a mean age of 47.8 years (standard deviation (SD): 14.2) and a median follow-up time in the database of 3.1 years (interquartile range (IQR): 1.5 – 5.1). The majority of patients were Caucasian (84.7%) and non-Hispanic (85.7%). A history of psoriatic arthritis was present in 11,469 (33.0%). Consistent with what has been previously reported in the literature, medical comorbidities including hyperlipidemia (26.7%), hypertension (27.9%), and obesity (15.5%) were common in the biologics cohort. Psychiatric comorbidities, including anxiety (12.4%) and depression (16.5%) were also common. The rates of self-harm, suicidal ideation, and suicide attempt were low.

Overall, there were 19,890 individuals whom received a prescription for adalimumab (43.0%), 14,108 (30.5%) for etanercept, 6561 (14.2%) for ustekinumab, 2787 (6.0%) for infliximab, 2356 (5.1%) for secukinumab, and 539 (1.2%) for ixekizumab. Between 27.7% and 53.1% of patients only received 1 prescription for the drug. Patients prescribed ixekizumab were the least likely to receive a second ixekizumab prescription (Table 2). Patients prescribed infliximab had the longest median time between first and last prescription (7.0 months, IQR: 2.9 – 20.8) and patients prescribed IL-17 inhibitors had the shortest median time with 3.3 months (IQR: 1.4 – 5.1) for ixekizumab and 3.4 months for secukinumab (IQR: 1.3 – 7.3; Table 3). Sensitivity analyses that changed the definition of continuous treatment to < 365 days between each prescription increased the median time between first and last prescription for all drugs (data not shown).