Using pharmacoeconomic analysis to make drug insurance coverage decisions

Our objective was to institute a cost-effectiveness-based reimbursement eligibility and coverage scheme for drugs in the Canadian province of British Columbia. All applications from drug manufacturers requesting Pharmacare (British Columbia government-funded drug insurance plan) coverage were evaluated by the Pharmacoeconomic Initiative (PI) of British Columbia. PI recommendations are according to a majority decision reached by a multidisciplinary volunteer expert committee and are based on a critical evaluation of pharmacoeconomic studies submitted by manufacturers seeking reimbursement eligibility. Coverage for drugs is universal and completely free for the financially indigent. Others are charged a small copayment and/or a deductible. PI assessments are evidence-based. Published guidelines from the Canadian Coordinating Office of Health Technology Assessment (CCOHTA) and/or Ontario Ministry of Health guidelines for the economic evaluation of pharmaceuticals are recommended for preparing submissions to the PI. Between January 1996 and December 1996, the PI made recommendations on 21 submissions; 4 of these were cost-effectiveness or cost-utility analyses; 3 were cost-minimisation analyses; 6 were cost comparisons or cost-consequence analyses; and 8 were provincial formulary budget impact studies. Of the 21 PI recommendations, 18 were accepted by Pharmacare and decisions are pending for 2 others, thus providing a concordance rate of 95% (18/19; kappa = 0.89). A total of 7 of the 21 products were recommended for formulary inclusion by the PI; 4 were as per drug company requests (i.e. full-benefit status) and 3 were recommended under restricted use. Only 5 of 21 submissions, of which 4 had favourable reviews, complied with either the CCOHTA or the Ontario Ministry of Health guidelines. Most studies were conducted, not from a societal perspective, but from the perspective of the provincial healthcare system. Most of the analysis were short term and therefore discounting was not applied. Sensitivity analysis was not performed in more than half (52%) of the submissions, and 48% of applications used inappropriate comparators. Ontario is the only other Canadian province with a similar process, with cost-effectiveness criteria for reimbursement eligibility. However, analysis in that province during the same approximate time period demonstrated a low concordance between Ontario Drug Benefit and PI decisions (kappa = 0.07). Currently, the mandated or suggested use of technology assessments of pharmaceuticals with cost effectiveness as the primary end-point is a reality in several countries worldwide. Our results, based on actual experience from implementing such a programme, suggest that while industry is slow to adapt to the new reporting requirements it may also be sceptical about the importance of cost effectiveness and guideline compliance in decision-making.