Simulated patients and their reality: An inquiry into theory and method

Tuberculosis diagnosis and management in the public versus private sector: a standardised patients study in Mumbai, India

“…Because these data are scarce and because different sectors even in the same location have different client mixes and different data collection protocols, there are nearly no ‘apples to apples’ comparisons of the quality of care between sectors that use valid counterfactuals. 15 These statistical issues are unavoidable in administrative data, 25 leading to literatures that produce extremely heterogeneous results in meta-analysis 54 or that rely on intermediate indicators rather than quality outcomes. 55 …”

“…First, that single-interaction SPs like those done here are typically predictive of the same provider’s behaviour until the patient initiates a change in course. 25 Second, referral chains are highly complex in terms of ultimate outcomes and require careful sample selection that we cannot track in this sample. 61 For this study, our presumption was that any referral is best practice from the perspective of provider behaviour when the provider believes they cannot handle the case themselves.…”

“…Because these data are scarce and because different sectors even in the same location have different client mixes and different data collection protocols, there are nearly no 'apples to apples' comparisons of the quality of care between sectors that use valid counterfactuals. 15 These statistical issues are unavoidable in administrative data, 25 leading to literatures that produce extremely heterogeneous results in meta-analysis 54 or that rely on intermediate indicators rather than quality outcomes. 55 Only very recent studies-such as one conducted with standardised patients in Xi'an, China, among 212 public providers and 27 private providers-have begun to directly compare equivalent quality outcomes across sectors with gold standard measurement methods.…”

“…They are recruited from local communities, speak the local language and are extensively trained to portray a predetermined and scripted medical condition to those healthcare providers, as well as to recall accurately and in great detail the actions of the provider 13. Previous publications provide more details on the SP methodology,12 25 29–31 and prior studies using the method for TB in India, China, Kenya and South Africa have demonstrated the validity of the method for TB presentation with no significant risks to providers, SPs or other patients 12 17 26 28 30 32–34…”

“…To measure the quality of care in the public and private sector, we used SPs—trained members of the community who present the same case to multiple providers posing as patients. This method has gained rapid acceptance in quality measurement in countries around the world 13 25 26. The use of SPs is informed by the fact that the most easily observable characteristics of facilities—including facility equipment, provider education levels, medication supplies and patient access—are poor predictors of key quality metrics 27.…”