The economic search terms used for this systematic review are based on validated search algorithms published previously (Wilczynski et al. 2004). For the Medline database, the search terms “cost-benefit analysis.sh. OR costs.tw. OR cost effective.tw.” produces the best combination of sensitivity (95.7%) and specificity (97.2%) of locating methodologically sound economic publications (Wilczynski et al. 2004). The search terms “cost.tw. OR costs.tw.” used in the EMBASE database produces the optimal combination of sensitivity (96.8%) and specificity (97.6%) for economic articles (McKinlay et al. 2006).
Once all full-text articles were screened, the remaining articles were reviewed by two reviewers (AMc and BC) and critically appraised using the Quality of Health Economic Studies (QHES) checklist developed by Chiou and colleagues (2003). This information was used to quickly identify potential method limitations in the individual studies, as well as in the area of comparative economic studies in SCI as a whole.
The QHES is a 16 item checklist that was developed in 2003. The checklist items were selected by consensus from a broader pool of items by a panel of 8 Health Economists. Conjoint analysis methods were used to produce utility values for each of the checklist items in relation to the questionnaire. A total of 98 researchers in the field of Health Economics participated in this analysis. The questionnaire was then validated by comparing the checklist scores of a scientific article to global scores reported by experts (Chiou et al. 2003). In a recent quality assessment of comparative economic study critical appraisal tools, Langer (2012) observed that the QHES and the economic evaluation checklist produced by the British Medical Journal were the most comprehensive checklists. Among all the checklists reviewed, only the QHES has been validated for construct validity and pre-tested among a group of experts. Further, the QHES is also the only checklist where the items are weighted, allowing for a quality rating.
A brief summary of each study will be presented, describing the outcomes used, the primary results, and the methodological strengths and limitations. Due to the variability in methods, disease population, cohort country of origin and interventions, only descriptive results are reported.