Clinical decision support tools: focus on dietary supplement databases.

Objective: To assess the content of dietary supplement databases used for clinical decision support.

Methods: Four online dietary supplement databases were assessed according to scope, completeness, and ease of use. Additionally, a composite score integrating all 3 criteria was determined. One hundred two question-and-answer pairs were developed and distributed among 10 weighted categories dealing with the most clinically relevant aspects of dietary supplement therapy. Descriptive and inferential statistics were used to summarize the evaluation components and to compare databases. Chi-square was used to investigate differences in scope scores. Scheffe's multiple comparison procedure was used to determine statistical differences in completeness and ease of use.

Results: The percentage of questions that each database answered successfully were as follows: Natural Medicines Comprehensive Database, 84.3%; Natural Standard, 80.4%; AltMedDex, 74.5%; and Lexi-Natural Products, 53.9%, indicating 2 statistical groupings (P < .05) in which the first 3 databases performed significantly better than Lexi-Natural. Completeness scores were similarly stratified. Ease-of-use scores were Natural Standard, 1.96; Natural Medicines Comprehensive Database, 2.00; Lexi-Natural, 2.02; and AltMedDex, 2.50. Composite scores indicating overall performance were Natural Medicines Comprehensive Database, 81.5%; Natural Standard, 76.9%; AltMedDex, 71.6%; and Lexi-Natural, 50.7%.

Conclusion: Many clinicians may be unprepared to deal with patient-related dietary supplement questions; therefore, clinical decision support tools that address this knowledge gap are needed. There was significant heterogeneity in the content of dietary supplement-focused online databases, with clustering in the top tier.