When a meta-analysis includes studies reporting only a single PRO, presented as a continuous variable, a pooled result will generate a mean difference. The problem with this mean difference is that clinicians may have difficulty with its interpretation. For example, if told that the mean difference between rehabilitation and standard care in a series of randomized trials using the Chronic Respiratory Questionnaire was 1.0 (95% CI 0.6 – 1.5), many readers would have no idea if this represents a trivial, small but important, moderate, or large effect.
The systematic review author can aid interpretation by reporting the range of possible results and the range of mean results in treatment and control groups in the studies. Most useful, however – if it is available – is an estimate of the smallest difference that patients are likely to consider important (the minimally important difference or MID). There are a variety of methods for generating estimates of the MID, including use of global ratings of change (Guyatt 2002). Ideally, review authors will present estimates of the MID in the abstract. For example, investigators examining the impact of respiratory rehabilitation in patients with chronic lung disease on health-related quality of life reported, in their abstract, that “for two important features of HRQL, dyspnea and mastery, the overall effect was larger than the MCID: 1.0 (95% CI 0.6-1.5) and 0.8 (0.5-1.2), respectively, compared with an MCID of 0.5.” (Lacasse 1996).
While this is very helpful, it potentially tempts clinicians to make inappropriate inferences. If the MID is 0.5 and the mean difference between treatments is 0.4, clinicians may infer that nobody benefits from the intervention. If the mean difference is 0.6, they may conclude that everyone benefits. Both inferences may be misguided. First, they ignore the uncertainty (confidence intervals) around the point estimate. More importantly, they ignore the variation (standard deviation) in responses across individuals.
It is also possible for investigators to provide a ‘responder’ definition to help interpret outcomes (see Chapter 12, Section 12.6.1). It is useful to know the definition that characterizes an individual patient as a responder to treatment. Such a responder definition is based upon pre-specified criteria backed by empirically derived evidence supporting the responder definition as a measure of benefit. Methods for defining a responder include: (1) a pre-specified change from baseline on one or more scales; (2) a change in score of a certain size or greater (e.g. a 2-point change on an 8-point scale); and (3) a percentage change from baseline.