Garrett Zoe, Thomas James
Social Science Research Unit, Institute of Education, London, UK.
Int J Lang Commun Disord. 2006 Jan-Feb;41(1):95-105. doi: 10.1080/13682820500071542.
The advent of evidence-based healthcare has seen a rise in the use of systematic reviews to bring together the findings from research studies. The use of systematic reviews in speech and language therapy (SLT) was criticized in this journal in 2004 by T. R. Pring. It was claimed that their findings are misleading due to the potential inclusion of biased data, and uninformative due to a lack of detail in the reporting of interventions. It is argued that outcome research should be carried out in a series of phases in which small-scale research precedes large-scale research. This, it is argued, is most likely to demonstrate statistically significant effects and also help to ensure that therapies become sufficiently defined so that clinicians can apply them in practice.
This paper argues that the above criticism of systematic reviews is based on a narrow conception of their capabilities: on the popular misapprehension that all systematic reviews answer effectiveness questions using only experimental studies and contain meta-analyses. Different methods for systematic reviews are described and their application within clinical outcome research is discussed with reference to a phased structure for empirical enquiry.
Systematic reviews seek to identify and synthesize information within a given topic area. They are used to answer a wide range of research questions and the studies they include are not limited exclusively to experimental designs. Methods of synthesis can include both statistical approaches, such as meta-analysis, and 'qualitative' approaches, such as meta-ethnography and thematic analysis. Knowledge of the current state of research is essential for a sequentially phased approach within outcome research to operate. Since systematic reviews are summaries of research activity, they can provide this knowledge and should therefore be considered a valuable tool within outcome research.
A systematic review using 'qualitative' and/or statistical methods for combining studies can be carried out within or across any of the phases within outcome research. Far from being uninformative, this can help bring together what is, and what is not, known and indicate the kinds of therapies that may be beneficial in the clinical setting and therapies which would benefit from further research and development.
循证医疗保健的出现使得系统评价的应用有所增加，以整合来自各项研究的结果。2004年，T. R. 普林在本期刊上对言语和语言治疗（SLT）中系统评价的应用提出了批评。据称，由于可能纳入有偏差的数据，其结果具有误导性；又因干预措施报告缺乏细节，所以没有参考价值。有人认为，结果研究应分一系列阶段进行，其中小规模研究先于大规模研究。据称，这样最有可能证明具有统计学意义的效果，也有助于确保疗法得到充分界定，以便临床医生能够在实践中应用。
本文认为，上述对系统评价的批评基于对其能力的狭隘理解：基于一种普遍的误解，即所有系统评价仅使用实验研究来回答有效性问题且包含荟萃分析。本文描述了系统评价的不同方法，并参照实证研究的分阶段结构，讨论了它们在临床结果研究中的应用。
系统评价旨在识别和综合给定主题领域内的信息。它们用于回答广泛的研究问题，所纳入的研究并不局限于实验设计。综合方法可以包括统计方法，如荟萃分析，以及“定性”方法，如元民族志和主题分析。了解研究现状对于结果研究中按顺序分阶段进行的方法的运作至关重要。由于系统评价是研究活动的总结，它们可以提供这种知识，因此应被视为结果研究中的一种有价值的工具。
使用“定性”和/或统计方法对研究进行综合的系统评价可以在结果研究的任何阶段内或跨阶段进行。这远非没有参考价值，反而有助于整合已知和未知的内容，并指出在临床环境中可能有益的疗法类型以及哪些疗法将受益于进一步的研究和开发。
