Article type
Year
Abstract
Background: Assessment of benefit and harm are an important part of decision-making with drug therapy for both prescriber and patient. Reviewers should therefore aim to provide a balanced summary of both the benefits and harm, which may follow any particular treatment. We undertook a survey of completed Cochrane reviewers with a view to describing how Cochrane reviewers do this.
Method: We listed the titles of all the Cochrane reviewers in The Cochrane Library, 2000 Issue 2 to identify therapeutic interventions of particular interest from an adverse reaction perspective. We chose three pharmacotherapeutic areas for detailed study: (i) corticosteroid therapy for chronic conditions (ii) rheumatoid arthritis and (iii) schizophrenia. The reviewers in these areas were printed and read and sections dealing with adverse effects were highlighted. All reviewers were read by at least two authors.
Results: In several reviewers the evaluation of potential adverse outcomes was the primary objective (e.g. Beclomethasone for asthma in children: effect on linear growth). Interestingly, several others considered therapy aimed at countering adverse effects of pharmacotherapy (e.g. Bisphosphonates for steroid-induced osteoporosis). Tables 1-3 summarise how the Cochrane reviewers assessed the adverse effects of the pharmacotherapies considered. Cochrane reviewers generally restrict their discussions of the adverse consequences of drug therapy to those reported in the trials meeting their eligibility criteria. As most trials are short-term, even for treatments intended for repeated or continuous use in chronic diseases, there is a risk that the adverse consequences of drug therapy are not given their appropriate weighting in such systematic reviews. Some reviewers (e.g. Wahlberg, Cheine and Essali, 1999 and the reviews of of psychotropic drugs in general) discuss adverse effects of drug therapy extensively. This may be because adverse effects of antipsychotic drugs are often debilitating and lead to withdrawal. Indeed in some of the reviews, withdrawal from the study was used as a surrogate for intolerable adverse effect. In other reviews the same outcome is also used as evidence of lack of efficacy, attesting to its poor specificity. Reviews that discussed adverse effects rather sparsely, seen to do so mainly because the trials assessed also report adverse effects poorly.
Conclusion: Cochrane reviewers place insufficient emphasis on adverse drug effects. We would recommend that all Cochrane reviews of drug therapy should discuss adverse effects not captured by current randomised controlled trials. Possible ways of addressing this will be discussed in the presentation.
Method: We listed the titles of all the Cochrane reviewers in The Cochrane Library, 2000 Issue 2 to identify therapeutic interventions of particular interest from an adverse reaction perspective. We chose three pharmacotherapeutic areas for detailed study: (i) corticosteroid therapy for chronic conditions (ii) rheumatoid arthritis and (iii) schizophrenia. The reviewers in these areas were printed and read and sections dealing with adverse effects were highlighted. All reviewers were read by at least two authors.
Results: In several reviewers the evaluation of potential adverse outcomes was the primary objective (e.g. Beclomethasone for asthma in children: effect on linear growth). Interestingly, several others considered therapy aimed at countering adverse effects of pharmacotherapy (e.g. Bisphosphonates for steroid-induced osteoporosis). Tables 1-3 summarise how the Cochrane reviewers assessed the adverse effects of the pharmacotherapies considered. Cochrane reviewers generally restrict their discussions of the adverse consequences of drug therapy to those reported in the trials meeting their eligibility criteria. As most trials are short-term, even for treatments intended for repeated or continuous use in chronic diseases, there is a risk that the adverse consequences of drug therapy are not given their appropriate weighting in such systematic reviews. Some reviewers (e.g. Wahlberg, Cheine and Essali, 1999 and the reviews of of psychotropic drugs in general) discuss adverse effects of drug therapy extensively. This may be because adverse effects of antipsychotic drugs are often debilitating and lead to withdrawal. Indeed in some of the reviews, withdrawal from the study was used as a surrogate for intolerable adverse effect. In other reviews the same outcome is also used as evidence of lack of efficacy, attesting to its poor specificity. Reviews that discussed adverse effects rather sparsely, seen to do so mainly because the trials assessed also report adverse effects poorly.
Conclusion: Cochrane reviewers place insufficient emphasis on adverse drug effects. We would recommend that all Cochrane reviews of drug therapy should discuss adverse effects not captured by current randomised controlled trials. Possible ways of addressing this will be discussed in the presentation.