Gilbody Simon M., Petticrew Mark
NHS Centre for Reviews and Dissemination, University of York, YO10 5DD, UK.
J Ment Health Policy Econ. 1999 Sep 1;2(3):99-106. doi: 10.1002/(sici)1099-176x(199909)2:3<99::aid-mhp51>3.0.co;2-c.
"Systematic reviews" have come to be recognized as the most rigorous method of summarizing confusing and often contradictory primary research in a transparent and reproducible manner. Their greatest impact has been in the summarization of epidemiological literature - particularly that relating to clinical effectiveness. Systematic reviews also have a potential to inform rational decision-making in healthcare policy and to form a component of economic evaluation. AIMS OF THE STUDY: This article aims to introduce the rationale behind systematic reviews and, using examples from mental health, to introduce the strengths and limitations of systematic reviews, particularly in informing mental health policy and economic evaluation. METHODS: Examples are selected from recent controversies surrounding the introduction of new psychiatric drugs (anti-depressants and anti-schizophrenia drugs) and methods of delivering psychiatric care in the community (case management and assertive community treatment). The potential for systematic reviews to (i) produce best estimates of clinical efficacy and effectiveness, (ii) aid economic evaluation and policy decision-making and (iii) highlight gaps in the primary research knowledge base are discussed. Lastly examples are selected from outside mental health to show how systematic reviews have a potential to be explicitly used in economic and health policy evaluation. RESULTS: Systematic reviews produce the best estimates of clinical efficacy, which can form an important component of economic evaluation. Importantly, serious methodological flaws and areas of uncertainty in the primary research literature are identified within an explicit framework. Summary indices of clinical effectiveness can be produced, but it is difficult to produce such summary indices of cost effectiveness by pooling economic data from primary studies. Modelling is commonly used in economic and policy evaluation. Here, systematic reviews can provide the best estimates of effectiveness and, importantly, highlight areas of uncertainty that can be used in "sensitivity analysis". DISCUSSION: Systematic reviews are an important recent methodological advance, the potential for which has only begun to be realized in mental health. This use of systematic reviews is probably most advanced in producing critical summaries of clinical effectiveness data. Systematic reviews cannot produce valid and believable conclusions when the primary research literature is of poor quality. An important function of systematic reviews will be in highlighting this poor quality research which is of little use in mental health decision making. IMPLICATIONS FOR HEALTH PROVISION: Health care provision should be both clinically and cost effective. Systematic reviews are a key component in ensuring that this goal is achieved. IMPLICATIONS FOR HEALTH POLICIES: Systematic reviews have potential to inform health policy. Examples presented show that health policy is often made without due consideration of the research evidence. Systematic reviews can provide robust and believable answers, which can help inform rational decision-making. Importantly, systematic reviews can highlight the need for important primary research and can inform the design of this research such that it provides answers that will help in forming healthcare policy. IMPLICATIONS FOR FURTHER RESEARCH: Systematic reviews should precede costly (and often unnecessary) primary research. Many areas of health policy and practice have yet to be evaluated using systematic review methodology. Methods for the summarization of economic data are methodologically complex and deserve further research
“系统评价”已被公认为是以透明且可重复的方式总结令人困惑且往往相互矛盾的原始研究的最严谨方法。其最大影响在于对流行病学文献的总结，尤其是与临床疗效相关的文献。系统评价还有助于为医疗保健政策中的合理决策提供信息，并成为经济评估的一个组成部分。
本文旨在介绍系统评价背后的基本原理，并通过心理健康领域的实例，介绍系统评价的优势和局限性，特别是在为心理健康政策和经济评估提供信息方面。
实例选自围绕新型精神科药物（抗抑郁药和抗精神分裂症药物）的引入以及社区精神科护理提供方式（病例管理和积极社区治疗）的近期争议。讨论了系统评价在以下方面的潜力：（i）得出临床疗效和有效性的最佳估计值；（ii）辅助经济评估和政策决策；（iii）突出原始研究知识库中的差距。最后，实例选自心理健康领域之外，以展示系统评价如何有可能被明确用于经济和卫生政策评估。
系统评价能得出临床疗效的最佳估计值，这可构成经济评估的重要组成部分。重要的是，能在一个明确的框架内识别原始研究文献中严重的方法学缺陷和不确定性领域。可以生成临床有效性的汇总指标，但通过汇总原始研究的经济数据来生成成本效益的此类汇总指标则很困难。建模常用于经济和政策评估。在此，系统评价可提供有效性的最佳估计值，重要的是，突出可用于“敏感性分析”的不确定性领域。
系统评价是近期一项重要的方法学进展，其潜力在心理健康领域才刚刚开始得到认识。系统评价在生成临床有效性数据的批判性总结方面的应用可能最为先进。当原始研究文献质量较差时，系统评价无法得出有效且可信的结论。系统评价的一项重要功能将是突出这类对心理健康决策几乎没有用处的质量较差的研究。
医疗保健提供应在临床和成本效益方面都具有成效。系统评价是确保实现这一目标的关键组成部分。
系统评价有潜力为卫生政策提供信息。所举实例表明，卫生政策的制定往往未充分考虑研究证据。系统评价可提供有力且可信的答案，有助于为合理决策提供信息。重要的是，系统评价可突出进行重要原始研究的必要性，并可为该研究的设计提供信息，使其提供有助于形成医疗保健政策的答案。
在进行成本高昂（且往往不必要）的原始研究之前应先进行系统评价。许多卫生政策和实践领域尚未使用系统评价方法进行评估。经济数据汇总方法在方法学上很复杂，值得进一步研究。
