Addressing the challenges to immunization practice with an economic algorithm for vaccine selection.

The biotechnology revolution is producing a growing bounty of new vaccines which pose difficult choices in selecting among many products. Some major public and private purchasers of vaccine may offer individual physicians and clinics their choice in assembling vaccine inventories. Others might purchase only a limited stock of products that would satisfactorily immunize a typical child. In either case, current vaccine selection decisions are based principally on purchase price alone without systematic consideration of other factors of fiscal consequence. As a potential tool for decision making, we developed an economic algorithm for vaccine selection that would minimize the overall costs of disease control through immunization by considering: (1) purchase price, (2) number of doses needed, (3) preparation time, (4) route of administration, (5) cold storage needs, (6) shelf life, (7) earliest age of full immunity, (8) adverse events frequency, and (9) efficacy of protection. To demonstrate the algorithm, variables (1) to (4) above were incorporated into a pilot binary-integer linear programming model that satisfied the recommended immunization schedule for diphtheria, tetanus, pertussis, Haemophilus influenzae b, and hepatitis B, using eleven vaccines (DTaP, DTaP-Hib, Hib, HepB and Hib-HepB) from four manufacturers. Five (or six) opportunities to vaccinate were modeled at (1), 2, 4, 6, 12-18, and 60 months of life, assuming US$40 per clinic visit, $15 per injection, and $0.50 per minute of nurse preparation time. Vaccine costs were varied using actual March and September 1997 US Federal vaccine prices, as well as estimates for unpriced new vaccines. Over 16,000 distinct vaccine stocking lists by vaccine type and brand were possible. Including a 1-month visit, the lowest-cost 'solution' of the algorithm was $529.41 per child in the March cost-assumption case, and $490.32 in the September one (both included four doses of DTaP-Hib, three HepB, and one DTaP). Without a 1-month visit, the lowest-cost solution in the March case cost $486.67 (four DTaP, two Hib-HepB, one DTaP-Hib, and one HepB), while the September case cost $450.32 (four DTaP-Hib, three HepB, and one DTaP). Ensuring at least one product was selected from each of the four manufacturers increased costs about $13.00, and the needed injections rose from eight to nine. The most economical selection of vaccines to use cannot be intuitively predicted, as permutations are large and solutions are sensitive to minor changes in costs and constraints. A transparent, objective selection method that weighs the economic value of distinguishing features among competing vaccines might offer the 'best value' to vaccine purchasers, while also creating strong market incentives for continuing innovation and competition in the vaccine industry.