Article Text

Download PDFPDF

44 Opening the black box of ‘usual care’ and finding a black hole: a numerical systematic review on ‘usual care’ control groups in stroke rehabilitation RCTs
  1. Chiara Arienti1,
  2. Chiara Arienti2,
  3. Riccardo Buraschi2,
  4. Joel Pollet2,
  5. Massimiliano Gobbo2
  1. 1University of Brescia, Senago, Italy
  2. 2IRCCS Fondazione Don Carlo Gnocchi, Milan, Italy


Objectives Ethically, in every experimental trial new methods should be tested against the ‘best’ or, if not defined, the ‘usual’ care according to current Guidelines. To understand the latter, we focused on stroke rehabilitation as a case in point because stroke is a leading cause of disability and rehabilitation is a complex intervention whose need is increasing due to growth of chronicity and disability (World Health Organization). Our aim was to appraise stroke rehabilitation interventions provided as ‘usual care’ and the terminology used, comparing high to low quality randomised-controlled trials (RCTs).

Method Systematic review of RCTs (2006–16). Inclusion criteria were as follows: stroke survivor patients, intervention, rehabilitation, control: rehabilitative ‘usual care’, outcome: lower limb function. We used the Cochrane ‘risk of bias’ tool, rating studies as low or high quality. We identified the terminology used to describe the Control Group Rehabilitation Program (CGP), performed a knowledge synthesis process and conducted a frequency analysis to sort the heterogeneity through the itemised identification of the CGP contents. Two quality groups of studies (high-low risk of bias) have been compared.

Results We included 86 publications (23% low-risk of bias). Nine per cent of articles did not describe the CGP. In the remaining 78, we identified 64 interventions: 53 were proposed once, 8 were proposed twice and 3 were proposed in 3 papers. Two interventions (gait and balance) were proposed in 52% and 51% of papers, respectively. Results did not differ in the two RCT quality groups.

Conclusions This case study on stroke rehabilitation has shown that the term ‘usual care’ in CGP is not adequately used: a good terminology to define interventions is missing and respecting CONSORT checklists does not avoid these problems. These results are clear, and involve authors, reviewers and both field-specific and generalist journals. Nevertheless, they should be verified by future studies in other fields, with this or other methodologies. Reporting guidelines should probably give better expert guidance on this issue.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.