Using benefit harm tradeoffs to estimate sufficiently important difference: the case of the common cold.

摘要

CONTEXT: The term "sufficiently important difference" (SID) refers to the overall amount of benefit that people consider sufficient to justify the costs and risks of treatment. Little is known about patient preferences regarding benefits and harms of common cold treatments. OBJECTIVES: To develop methods to assess SID and to estimate SID for common cold. DESIGN: The authors conducted in-person and telephone interviews with people with colds, using benefit harm tradeoff methods. The hypothetical benefit of reduction in length of illness was traded off against best estimates of costs and risks. First, the authors briefly described costs, risks, and possible symptomatic benefits of 4 treatment scenarios, based on evidence regarding vitamin C, echinacea, zinc, and pleconaril, an antiviral. Hypothetical benefit (reduction of illness duration) was then varied until the cold sufferer indicated that the treatment was minimally desirable. PARTICIPANTS: Interviews were conducted in person with 149 community-recruited adult participants, once at the beginning of their colds, and then again within a few days after symptoms had resolved. Additionally, 162 adult callers with self-identified colds completed interviews via telephone. RESULTS: A total of 460 benefit harm tradeoff interviews (1840 treatment scenarios) estimated overall mean SID as 52.6 h (95% CI, 50.6 to 54.6). For the scenario based on vitamin C, mean SID was estimated as 26.1 h (95% CI, 23.2 to 29.3), with 142 of 460 (31%) saying they would take it regardless of duration benefit, and 22 of 460 (5%) saying they would not take it, regardless of duration benefit. For the echinacea-based scenario, mean SID was estimated at 36.8 h (33.4 to 40.2), with 105 (23%) favoring and 41 (9%) rejecting treatment, regardless of duration benefit. For the zinc lozenge-based scenario, mean SID was estimated as 64.8 h (61.0 to 67.9), with 42 (9%) favoring and 109 (24%) rejecting treatment. For the prescription antiviral-based scenario, mean SID was estimated as 82.6 h (78.7 to 86.7), with 29 (6%) favoring and 223 (48%) rejecting. Severity of illness at the time of interview did not appear to significantly influence responses. Possible side effects, treatment type (tablet v. lozenge v. liquid), monetary costs, and opportunity costs (e.g., getting to the doctor or pharmacy, dosing frequency) did appear to be important in influencing these preference patterns. CONCLUSIONS: Our study suggests that, on average, people want the duration of their colds to be reduced by between 26 and 65 h to justify potential harms of popular cold treatments. A prescription antiviral would require a greater benefit (83 h) to justify larger perceived risks.