In this chapter, I will discuss the methods of data retrieval and storage that help you to subsequently extract and analyse outcomes, bias and confounding factors, with particular reference to the systematic review of experimental studies.

There has been very little empirical research on how different methods of data retrieval and storage affect the results of systematic reviews. Most research has focussed on variables in the early part of the process, such as blinding data extractors to the authors, institute and publishing journal of each trial.

Because of the paucity of evidence I have written a pragmatic chapter based upon my own experience as an author and editor of Cochrane systematic reviews. Therefore you should not accord my conclusions with the same weight you would give to conclusions in other chapters that are based upon more evidence.

Introduction

Your aim is to find out what results your patient can expect from an intervention and how reliable are the effects. To do this you have to retrieve data from studies accurately without introducing bias. You determined the participants, interventions and outcomes for which you want to retrieve data when you planned your protocol (Chapter 2). Your search strategy determined the studies that you found (Chapter 3). In this chapter I will explain how best to retrieve data from those studies.