Evaluating the impact of using an Option Grid for osteoarthritis of the knee in an interface musculoskeletal clinic: a stepped wedge trial.

Talk Code: 

The problem

Implementing shared decision making (SDM) is a difficult due to a multitude of practical and attitudinal barriers. There was hope that SDM could be achieved by introducing information to patients using decision aids (PDAs). There is evidence that PDAs given to patients ahead of their appointment lead to improved knowledge but there is little evidence that this approach has a significant effect on the process of SDM. However, tools that have been designed for use in consultations have more promise. The aim of this trial was to investigate the effect of introducing a short decision aid, called an Option Grid, into clinical practice.

The approach

We planned a trial where six physiotherapists working in an interface clinic for musculoskeletal problems in Oldham were introduced to an Option Grid that described treatments for osteoarthritis of the knee: they were also instructed in how to use it. A stepped wedge design was used to introduce the intervention sequentially to the physiotherapists; a design useful in situations where concurrent intervention delivery is not practical. Each physiotherapist consulted with 6 patients in the pre-intervention phase, and a further 6 in the post-intervention phase, giving a total patient sample of 72. This sample provided 90% power to detect a cluster-altered difference of t = 3.22. Consultations were audio-recorded and the extent of SDM was measured using the Observer OPTION measure, and Decision Quality was evaluated using an adapted version of a published measure. Demographic (including age, gender, deprivation quintile, ethnicity, educational level, and employment status), and health literacy (REALM-R) data were also collected.


The mean OPTION score was 29.40 (SD 12.9) in the pre-intervention phase and increased to 37.83 (SD 8.4) in the post-intervention phase, and thus the 95% confidence interval (CI) for the difference (8.34) was 4.42 to 12.27, indicating statistical significance. The ICC was 0.101 which implies notable clustering of physiotherapists on their OPTION scores. Health literacy (as measured by REALM-R scores) and patients' Adjusted Index of Multiple Deprivation Quintile had no effect on the levels of SDM achieved. On average, there was a 0.9 point difference (score range 0-5), 95% CI 0.3 to 1.5, on the knowledge component of the Decision Quality measure; baseline patient education level had no effect on these scores.Consequence After brief training in how to use the osteoarthritis knee Option Grid, the extent of SDM as measured by an observer-based measure increased in the post-intervention phase of this stepped wedge trial. This significant increase in SDM was accompanied by a similar increase in the patient knowledge component of the Decision Quality measure. Neither levels of deprivation nor health literacy had any impact on the amount of SDM measured in a consultation.


  • Tim Pickles, Pennine MSK Partnership Ltd, Oldham, UK
  • Glyn Elwyn
  • Adrian Edwards, South East Wales Trials Unit, Cardiff, UK
  • Katharine Kinsey, The Dartmouth Institute for Health Policy and Clinical Practice, Hanover, USA
  • Katherine Brain, South East Wales Trials Unit, Cardiff, UK
  • Robert G. Newcombe, South East Wales Trials Unit, Cardiff, UK
  • Jill Firth, The Dartmouth Institute for Health Policy and Clinical Practice, Hanover, USA
  • Alan Nye, The Dartmouth Institute for Health Policy and Clinical Practice, Hanover, USA
  • Fiona Wood, South East Wales Trials Unit, Cardiff, UK