Article type
Year
Abstract
Introduction/Objective: A key component of the usefulness of a health care intervention is the balance between effectiveness and potential harm. This paper aims to measure this important indicator of usefulness of the Cochrane Collaboration's reviews by analysing how adverse effects are considered.
Methods: Each of the 159 completed reviews in Issue 1, 1997 of The Cochrane Library was analysed against seven questions relating to adequacy of information about adverse effects. Reviews were also analysed for discussion of treatment alternatives and costs.
Results: Initial analysis showed that 23% of Cochrane reviews were solely concerned with effectiveness of interventions, rather than their effects; 19% of reviews made no mention of adverse effects or safety whatsoever. This was also true for reviews dealing with drug treatments: 24% were solely concerned with effectiveness, and 18% did not mention adverse effects at all. Almost one in three reviews did not specify the nature of potential or experienced adverse effects, and information on incidence and severity of adverse effects is even more limited. In 40% of reviews, no trial data on adverse effects were available, but 58% of these reviews did not specify that this should be considered in future trials. Only 14% of reviews mention costs of interventions, and often only in very vague terms.
Discussion: Cochrane reviews purport to help people make well-informed decisions about the "tradeoff between expected benefits, harms and costs." At present, a significant number of reviews do not meet this goal. At least one in five Cochrane reviews makes no mention of adverse effects of the intervention even if it is a drug treatment. When mentioned, the range of adverse effects is often very limited and far from comprehensive. Insufficient trial data hampers adequacy of data on adverse effects, but when trial data are available they are often not fully reported or commented on by reviewers. Where trial data are adequate to consider harmful effects, many reviewers do not provide such information from other sources, and more than half do not specify this as important for furture trials. If Cochrane reviews are to be more useful, the potential adverse effects of interventions need to be considered and addressed more specifically and systematically. This suggests important roles for reviewers.
Methods: Each of the 159 completed reviews in Issue 1, 1997 of The Cochrane Library was analysed against seven questions relating to adequacy of information about adverse effects. Reviews were also analysed for discussion of treatment alternatives and costs.
Results: Initial analysis showed that 23% of Cochrane reviews were solely concerned with effectiveness of interventions, rather than their effects; 19% of reviews made no mention of adverse effects or safety whatsoever. This was also true for reviews dealing with drug treatments: 24% were solely concerned with effectiveness, and 18% did not mention adverse effects at all. Almost one in three reviews did not specify the nature of potential or experienced adverse effects, and information on incidence and severity of adverse effects is even more limited. In 40% of reviews, no trial data on adverse effects were available, but 58% of these reviews did not specify that this should be considered in future trials. Only 14% of reviews mention costs of interventions, and often only in very vague terms.
Discussion: Cochrane reviews purport to help people make well-informed decisions about the "tradeoff between expected benefits, harms and costs." At present, a significant number of reviews do not meet this goal. At least one in five Cochrane reviews makes no mention of adverse effects of the intervention even if it is a drug treatment. When mentioned, the range of adverse effects is often very limited and far from comprehensive. Insufficient trial data hampers adequacy of data on adverse effects, but when trial data are available they are often not fully reported or commented on by reviewers. Where trial data are adequate to consider harmful effects, many reviewers do not provide such information from other sources, and more than half do not specify this as important for furture trials. If Cochrane reviews are to be more useful, the potential adverse effects of interventions need to be considered and addressed more specifically and systematically. This suggests important roles for reviewers.