Layton JB, Ritchey ME, Huang Z, Anderson-Smits C, Hill G. Identifying immunoglobulin use in claims data with procedure coding and pharmacy dispensing data. Poster presented at the 2020 36th ICPE International Virtual Conference on Pharmacoepidemiology & Therapeutic Risk Management; September 16, 2020.

BACKGROUND: Immunoglobulins (Ig) may be administered intravenously (IVIg) or subcutaneously (SCIg), inpatient, at an infusion center or at home, and some Ig may be self-administered. Differences in administration routes, treatment guidelines, and reimbursement policies may result in Ig use recorded as dispensed pharmacy products or procedures in US claims data. Correctly identifying users of Ig treatments is critical for valid generation of real-world evidence for these products.

OBJECTIVES: To describe the relative distribution and characteristics of Ig initiators identified via procedure and pharmacy claims.

METHODS: Patients with diagnoses of primary immunodeficiency (PID) or chronic inflammatory demyelinating polyneuropathy (CIDP) were identified in US commercial, Medicare, or Medicaid claims data. Patients with PID initiating IVIg or SCIG and patients with CIDP initiating IVIg were identified at the first recorded claim for an Ig product in inpatient or outpatient procedure coding (i.e., HCPCS) or pharmacy dispensing drug claims (i.e., NDC). Characteristics of Ig initiators were described, stratified by the source of the first Ig claim.

RESULTS: For patients with PID initiating Ig (N=17,931), 9% were first identified through pharmacy claims; those identified in pharmacy claims were more likely to have Medicaid insurance than those identified in procedure claims (30% vs. 17%). The characteristics of pharmacy-identified patients differed from procedure-identified patients, with some characteristics higher among those identified in pharmacy claims (e.g. asthma, 55% vs. 40%), but others were higher among procedure-identified patients (e.g. leukemia, 22% vs. 4%). Among those with CIDP (N=3,975), 11% were first identified through pharmacy claims, and they were more likely to have Medicaid (19% vs. 8%) or Medicare (38% vs. 23%) than those identified through procedure claims. Pharmacy-identified patients generally had similar or higher levels of most comorbidities compared to procedure-identified patients. Procedure-identified Ig were more likely to have some markers of health care screening and utilization, such as blood laboratory tests, in both diseases.

CONCLUSIONS: While most Ig initiators were identified in procedure claims, pharmacy dispensing claims are an important source of information about Ig use and identifying initiation of Ig products. There may be key differences in patient populations identified by procedure or pharmacy coding reflective of the payer policies or treatment guidelines across disease areas.

Share on: