Personalised pulmonary rehabilitation with multidimensional response profiling could aid efficiency and effectiveness http://ow.ly/TouTt
Clinical practice should be based on the best scientific evidence, which does not always correspond to evidence-based medicine, involving guidelines, meta-analyses and protocols as a basis for clinical approach [1, 2]. There is no need of further randomised controlled trials (RCTs) for evidence that pulmonary rehabilitation (PR) improves daily symptoms, exercise performance and health status in patients with chronic obstructive pulmonary disease (COPD), independently of disease stage and complexity . Although the direct demonstration of a clear benefit in survival is lacking (as is also true for most therapies used in COPD), PR is well recognised as a fundamental part of the integrated care of these patients and it has been incorporated in most guidelines for their management .
Due to the current restraints in healthcare resources and the different opportunities in healthcare policies all over the world , we need solid outcomes to decide how to optimally invest money in this field. Clinical observations have attempted to define predictors of PR success, evaluating outcome measures able to discriminate responders from nonresponders to treatment [6–8]. Nevertheless, the improvement of a single outcome measure like exercise tolerance in an RCT may not be enough to catch the real and/or perceived benefit that the individual patient may achieve, just like the glycaemia control might not be a perfect or unique index of control of diabetes and related complications in different environments and stages along the natural history of this disease .
As reported in this issue of the European Respiratory Journal, SPRUIT et al.  looked for a profile of multidimensional response to PR in >2000 COPD patients based on eight different outcome measures, including symptoms, exercise performance, health and mood status, and activities of daily life. Analysis led to clustering of patients into four groups with different response profiles and different prevalence: “very good responder” (18.3%), “good responder” (35.9%), “moderate responder” (35.4%) and “poor responder” (10.5%). The “very good responder” had more severe dyspnoea, more hospitalisations per year, worse exercise tolerance, worse performance and satisfaction scores for activities of daily life, more severe anxiety and depression, worse health status, and was more likely to follow an inpatient PR programme compared with the other three categories. Interestingly, even the “poor responder” may show that PR is a therapy of clinical value in those COPD patients who are symptomatic despite optimal medical therapy.
Overall, it seems that the best responders are the most severe patients, as probably expected.