Article type
Year
Abstract
Objective: To review approaches to applicability used by authors of recently published reports of RCTs and systematic reviews.
Methods: RCTs and systematic reviews published in six key general medical journals during 1997 and new systematic reviews in the 1998 Issue 1 edition of the Cochrane Library are being identified. We are using the previously described five-step approach as the framework for this review. Two people are independently reviewing each report and summarising the information provided and analyses undertaken.
Results: To date, 35 systematic reviews published during 1997 in the British Medical Journal (n=12), Lancet (n=3), Journal of the American Medical Association (n=l 0), New England Journal of Medicine (n=0), Annals of Internal Medicine (n=3) and Archives of Internal Medicine (n=7) have been assessed. The first three steps require information to be provided for making a judgment about applicability. The numbers of papers considering (at least in pant) each step are (n) - Step 1: Report all likely outcomes, both beneficial and adverse (11); Step 2: Look for treatment effect modifiers (14); Step 3: Examine the effects of baseline risk on treatment effect (5). One systematic review provided data for all three steps and another two had data for two steps. The final two steps are - Step 4: Estimate the benefits for an individual using their baseline risk (1); Step 5: Weigh up the benefits and harms to make a decision about treatment (1). Screening of titles and abstracts of 510 citations from Medline has identified 315 potential RCTs published in the six key journals during 1997. These are being reviewed further (BMJ - 43, Lancet - 79, JAMA - 36, NEJM - 96, Annals -31, Archives - 30). The results of the review of the RCTs and new Cochrane reviews will be included in the presentation.
Discussion: Applicability of the results of RCTs and systematic reviews is a neglected issue in reports and publications. While some authors are starting to include some data and analyses (mainly relating to the investigation of effect modification) there is a general failure to consider or report all patient relevant outcomes, to explore the relationship between (relative) treatment effect and baseline risk or other effect modifiers. These components are essential to assist clinicians and individual patients weigh up the potential benefits and harms of new medical interventions.
Methods: RCTs and systematic reviews published in six key general medical journals during 1997 and new systematic reviews in the 1998 Issue 1 edition of the Cochrane Library are being identified. We are using the previously described five-step approach as the framework for this review. Two people are independently reviewing each report and summarising the information provided and analyses undertaken.
Results: To date, 35 systematic reviews published during 1997 in the British Medical Journal (n=12), Lancet (n=3), Journal of the American Medical Association (n=l 0), New England Journal of Medicine (n=0), Annals of Internal Medicine (n=3) and Archives of Internal Medicine (n=7) have been assessed. The first three steps require information to be provided for making a judgment about applicability. The numbers of papers considering (at least in pant) each step are (n) - Step 1: Report all likely outcomes, both beneficial and adverse (11); Step 2: Look for treatment effect modifiers (14); Step 3: Examine the effects of baseline risk on treatment effect (5). One systematic review provided data for all three steps and another two had data for two steps. The final two steps are - Step 4: Estimate the benefits for an individual using their baseline risk (1); Step 5: Weigh up the benefits and harms to make a decision about treatment (1). Screening of titles and abstracts of 510 citations from Medline has identified 315 potential RCTs published in the six key journals during 1997. These are being reviewed further (BMJ - 43, Lancet - 79, JAMA - 36, NEJM - 96, Annals -31, Archives - 30). The results of the review of the RCTs and new Cochrane reviews will be included in the presentation.
Discussion: Applicability of the results of RCTs and systematic reviews is a neglected issue in reports and publications. While some authors are starting to include some data and analyses (mainly relating to the investigation of effect modification) there is a general failure to consider or report all patient relevant outcomes, to explore the relationship between (relative) treatment effect and baseline risk or other effect modifiers. These components are essential to assist clinicians and individual patients weigh up the potential benefits and harms of new medical interventions.