Inconclusive Evidence: Can Pharmacoeconomics Tilt The Balance?

Abstract

To examine the role of pharmacoeconomics in ascertaining the choice of appropriate therapeutics. This is a retrospective analysis, conducted using the clinical records from a tertiary referral hospital in western Maharashtra, India. Data on patients of ankylosing spondylitis and rheumatoid arthritis was collected. Biological response modifiers (BRMs) are indicated for NSAID failure and DMARD failure cases, respectively. We compared the input costs and relative benefit of various BRMs Out of 35 patients of ankylosing spondylitis who had NSAID failure, 15 were given etanercept (24 injections/patient) and 20 were given infliximab (8 injections/patient). Mean reduction in Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score in etanercept and infliximab group was 1.79 and 2.41 respectively (difference not statistically significant). The total input cost per patient of etanercept and infliximab was INR 1,56,984 and INR 2,55,360 respectively. The average cost incurred for per unit reduction in mean BASDAI score was INR 87,700 (1,56,984/1.79) for etanercept and 105958 for infliximab. Though the gross input cost of infliximab was 1.62 (2,55,360/1,56,984) times higher, its cost per unit benefit offered was only 1.2 (1,05,958/87,700) times higher than etanercept. Similarly, in 28 cases of DMARD failure RA cases, 15 were given infliximab and 13 were given etanercept. The average cost incurred for per unit reduction in mean disease activity score (DAS-28) was INR 1,22,643 for etanercept and INR 1,78,573 for infliximab. Cost of infliximab per unit benefit offered was 1.4 times higher than etanercept In view of its lower cost for comparable clinical effect, etanercept appears to be a more appropriate BRM. The case study demonstrates that pharmacoeconomic analysis of competing therapies can single out the appropriate treatment choice. However, for drawing generalizable conclusions, prospective studies with indirect costs and longer time horizon are required.