Background: In several countries, clinical practice guidelines for lung cancer recommend that multidisciplinary (MD) teams should be used to plan the management of all lung cancer patients. We conducted a systematic review to evaluate and critically appraise the effectiveness of multidisciplinary teams for lung cancer.
Materials and methods: Medline searches were carried out for the period 1984 to July 2007. We included any study that mentioned team working among specialists with diagnostic and curative therapeutic intent, where members of the team met at a specified time, either in person or by video or teleconferencing, to discuss the diagnosis and management of patients with suspected lung cancer. All study designs were included. We were particularly interested in whether multidisciplinary working improved survival but also considered other outcomes such as practice patterns and waiting times.
Results: Sixteen studies met the criteria for inclusion. Statistical pooling was not possible due to clinical heterogeneity. Only two of the primary studies reported an improvement in survival. Both were before-and-after designs, providing weak evidence of a causal association. Evidence of the effect of MD teams was stronger for changing patient management than for affecting survival. Six of the studies reported an increase in the percentage of patients undergoing surgical resection or an increase in the percentage of patients undergoing chemotherapy or radiotherapy with curative intent.
Conclusion: This systematic review shows limited evidence linking MD teams with improved lung cancer survival. This does not mean that MD teams do not improve survival, merely that currently available evidence of this is limited. It seems intuitively obvious that MD teams should improve outcomes for lung cancer patients, but there are difficulties in conducting randomised trials to show this. The best way forward would be prospective evaluation of the effectiveness of MD teams as they are implemented, paying particular attention to collecting data on potential confounders.