Meta-analysis of data on costs from trials of counselling in primary care: using individual patient data to overcome sample size limitations in economic analyses
British Medical Journal 326 1247-1250
Available online: 5 June 2003
Abstract
Objective To assess the feasibility of overcoming sample size limitations in economic analyses of clinical trials through meta-analysis of data on individual patients from multiple trials. Design Meta-analysis of individual patient data from trials of counselling in primary care compared with usual care by a general practitioner. Setting Primary care. Patients People with mental health problems. Main outcome measures Direct treatment costs, depressive symptoms, and cost effectiveness. Results Meta-analysis of individual patient data proved feasible. The results showed that the previous analyses of individual trials were underpowered to provide useful conclusions about the cost comparisons. The results are sensitive to assumptions made about the costs of sessions with a counsellor and the management of patients by a general practitioner. Conclusions Meta-analysis of individual patient data may assist in overcoming sample size limitations in economic analyses. Although feasible, such analysis has shortcomings that may limit the validity of the results. The relative costs and benefits of this method, as opposed to further collection of primary data, are as yet unclear.
Objective To assess the feasibility of overcoming sample size limitations in economic analyses of clinical trials through meta-analysis of data on individual patients from multiple trials. Design Meta-analysis of individual patient data from trials of counselling in primary care compared with usual care by a general practitioner. Setting Primary care. Patients People with mental health problems. Main outcome measures Direct treatment costs, depressive symptoms, and cost effectiveness. Results Meta-analysis of individual patient data proved feasible. The results showed that the previous analyses of individual trials were underpowered to provide useful conclusions about the cost comparisons. The results are sensitive to assumptions made about the costs of sessions with a counsellor and the management of patients by a general practitioner. Conclusions Meta-analysis of individual patient data may assist in overcoming sample size limitations in economic analyses. Although feasible, such analysis has shortcomings that may limit the validity of the results. The relative costs and benefits of this method, as opposed to further collection of primary data, are as yet unclear.