Page 50 - Effective healthcare cost containment policies Using the Netherlands as a case study - Niek W. Stadhouders

Chapter 3
articles were screened in full text independently by two researchers. References of the excluded reviews were checked for relevant articles and screened using a similar process as the original articles (see Appendix figure 3.2 for the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram).
We extracted relevant information of the included articles and assessed the quality. To this aim, we reviewed ten quality assessment tools (Bero et al.; Evers et al., 2005; Glasgow et al., 1999; Group, 2004; Harbour and Miller, 2001; Olsen, 2002; Peersman et al., 1997; Sterne et al., 2014; Walker et al., 2012; West et al., 2002). Based on these validated frameworks, we designed a quality assessment tool specifically for policy evaluations (Appendix 3.1). We use five quality domains: content validity, selection bias, confounding bias, measurement bias and reliability. Literature reviews were assessed separately using A MeaSurement Tool to Assess systematic Reviews (AMSTAR) (Shea et al., 2007). The results of the included systematic reviews were incorporated directly in the narrative synthesis without a separate assessment of the papers included in that review (Whitlock et al., 2008). Two reviewers independently performed quality assessments and review assessments.
To structure our results, we use an existing framework of mutually exclusive groups, covering both macro-policies and micro-interventions (Stadhouders et al., 2016). Comparing our results with this framework allows detection of knowledge gaps, which may inspire future policy evaluations. Although other categorisations of cost-containment policies exist (Ellis and McGuire, 1993; Mongan et al., 2008; Moreno-Serra, 2014; Pane and Taliaferro, 1994; Vogler et al., 2016; White, 1999), this categorisation was deemed most suitable to detect knowledge gaps.
In total, 72 policies from 43 empirical papers and 18 systematic reviews were included. The 18 systematic reviews covered a total of 341 studies. Overall, the included papers were of mixed quality. Of the 43 empirical papers, 29 articles scored low on at least one of the five quality domains. Systematic review ratings ranged from 3 out of 12 to 12 out of 12 points on the AMSTAR scale. We were unable to perform a meta-analysis due to the methodological heterogeneity and context dependency of the studies (Schünemann et al., 2008). Therefore, results for each intervention are discussed separately. First, we summarise the literature. Next, we identify knowledge gaps. Last, we synthesise the results.
3.3 Results
Appendix table 3.4 summarises our results, structured according to each of the four primary policy groups: budgets, price controls, volume controls and market-oriented policies, where
42
3.3.1. Literature overview