How do White British and Pakistani people rate communication within simulated GP-patient consultations? A national experimental vignette study

Talk Code: 

The problem

In many countries, minority ethnic groups report lower patient experience scores than the majority population. In the UK, South Asians report particularly low scores, for example in the General Practice Patient Survey (GPPS). Around half of the difference in scores is explained by the concentration of South Asian patients in low-scoring practices. The remaining difference may arise because South Asian patients rate the same care lower, or because South Asian patients receive worse care.The aim of this study was to examine whether people from a Pakistani background rate simulated GP consultations differently than White British people. If these groups rated simulated consultations the same when viewing identical videos, this would increase the likelihood that low scores among minority ethnic groups in real surveys reflect real differences in quality of care.

The approach

We conducted an experimental vignette study, in which we showed video-recordings of simulated GP-patient consultations to members of the public. Participants were recruited using an in home face to face approach, coordinated by a market research agency (Ipsos MORI). Trained fieldworkers recruited 1,128 participants (564 White British and 564 Pakistani adults: in both groups half of respondents were aged 18-54 and half aged over 55), working within pre-defined Output Areas (around 130 addresses each), selected according to their ethnic and age composition.Fieldworkers completed Computer-Assisted Personal Interviews with participants, using a standardised script. Participants rated three simulated consultations in turn, using the five GP-patient communication items taken from GPPS. From these a 0-100 communication score was derived. Consultations were shown in a random order, selected from a pool of 16. We also asked participants their gender, age and self-rated health. Mean differences in communication score between White British and South Asian patients were estimated from linear regression.


Pakistani participants, on average, scored consultations 9.8 points higher than White British participants (95%CI 8.0, 11.7, p<0.001). When adjusted for age, gender, deprivation, self-rated health, and video, the difference increased to 11.0 points (95%CI 8.5, 13.6, p<0.001). There was evidence that the effect of ethnicity was moderated by age (p<0.001) and whether the video was scripted to be good or poor for communication (p<0.001); the largest differences were seen when participants were older (>55) and where communication was scripted to be poor.


This vignette study found substantial differences in ratings of simulated consultations between Pakistani and White British participants, which are in the opposite direction to those observed in national surveys. If we take these findings at face value, suggesting that Pakistani patients rate the same care higher, then we would conclude that the lower scores they report in national surveys represent genuinely worse care.


  • Jenni Burt, University of Exeter, Exeter, UK
  • Gary Abel, University of Exeter, Exeter, UK
  • Natasha Elmore, University of Exeter, Exeter, UK
  • Cathy Lloyd, Ipsos MORI, London, UK
  • John Benson, University of Exeter, Exeter, UK
  • Lara Sarson
  • Anna Carluccio
  • John Campbell, RAND, Santa Monica, USA
  • Marc Elliott, Open University, Milton Keynes, UK