Quantity inequalities may be dwarfed by quality inequalities
In my last post on UC I argued that UC is best thought of as a means to achieving lower inequalities and improved financial protection in the health sector, but that in practice UC is unlikely to be sufficient – and may not even be necessary – for us to achieve these goals.
In this post, I argue that our focus on narrowing inequalities in the quantity of care is leading us to ignore another and potentially more important type of inequality in the health sector: inequality in the quality of care.
Measuring the quality of ambulatory care
My colleague Jishnu Das – whose work I’m going to draw on in this post – distinguishes between the competence of a doctor and the effort he or she makes during a consultation. What the doctor does in a real-life consultation depends on both – how much the doctor knows, and how much effort he makes to put his knowledge to use in diagnosing and treating the patient.
To measure competence, Das and his coauthors – like a growing number of other researchers working in the field – use “vignettes”. These are hypothetical standardized cases that can be implemented through a questionnaire or acted out by the doctor and the interviewer who plays the role of the patient. The doctor knows the interaction is an exercise and the patient isn’t really ill, but the doctor can ask questions. The “patient” is often accompanied by a second person who can answer the questions a patient wouldn’t know the answer to, e.g. body temperature, test results, etc. Through vignettes researchers are able to capture (a) things that doctors say they’d do that they should do, as well as (b) things that they say they’d do that they shouldn’t do.
To assess effort, Das and others measure the time doctors spend with the patient in real-life consultations, the number of questions doctors ask the patient, the number of exams they perform, and so on. These items are recorded by an observer, or by the patient during an exit interview.
Finally, to get at practice – what doctors actually do – Das and others watch doctors during their consultations with patients. There’s a risk of a Hawthorne effect here, of course. But any Hawthorne effect will lead the researcher to overestimate the quality of care; in the event, research by Ken Leonard suggests that the effect wears off after about 90 minutes. Sometimes observers sit in on consultations with real patients. Sometimes Das & Co send fake patients – actors who have learnt the part of being, say, an anxious mother with a young child at home suffering from diarrhea. Doctors know some of their patients will be fake, but they rarely guess after the fact which were fake and which were real!
Performance on vignettes throws up large within-country variations in competence, measured in terms of the intent to ask the right questions and do the right things. Some doctors display alarmingly low competence, but even the best fall a long way short of the ideal. For example, in urban Delhi the doctors in the top 20 percent of the overall competence index asked or did only about 30 percent of the essential questions / tasks in the diarrhea vignette.
Watching what doctors do throws up large inter-country variations in effort. Take the amount of time a doctor spends with a patient. In Delhi, it’s very low. In Paraguay, it’s more like the length of consultation seen in an OECD country. Das and coauthors also see a variation too in the number of questions asked, the number of physical exams, and the number of medicines given; doctors in Delhi ask 3 questions, do 1 exam, and give 3 medicines; in Paraguay, they ask 8 questions, do 3 exams, and give 1.5 medicines.
Das & Co find that competence and effort are positively correlated. Furthermore, they find that more competent doctors also ask more of the essential questions and do more of the essential tasks. Intriguingly, though, for most doctors there’s a gap between what doctors know and what they do. This ‘know-do’ gap varies by provider type. In Delhi, unqualified doctors do all the essential tasks they know about, but they know only 20 percent of the essential tasks. Qualified private doctors know 40 percent of essential tasks but do only 25 percent of them. By contrast, public-sector doctors know 30 percent of essential tasks but do only 8 percent! In related research by Ken Leonard in Tanzania, public-sector and NGO doctors performed similarly on vignettes, and did less well in real consultations—but the ‘know-do’ gap was smaller for NGO doctors.
Inequalities in quality
How does all this relate to the UC debate? The answer is simply that if we’re focusing on narrowing gaps in utilization and ignoring quality, we could be missing a big part of the inequality story.
In fact, we are. It turns out that when it comes to the quality of health care – at least in India – the poor are trebly disadvantaged.
In each of the neighborhoods they were working in, Das & Co were able to get an estimate of the average household income in the neighborhood from the national household expenditure survey. It turned out that for each type of doctor – whether an unqualified “quack” or a super-qualified hospital doctor – competence was lower in poorer neighborhoods.Disadvantage number 1.
Less competent doctors in all neighborhoods make less effort – they ask fewer questions, and ask fewer of the essential questions. So, doctors in poor neighborhoods make less effort. Disadvantage number 2.
Finally, within a neighborhood when the poor go to a public facility they are more likely to go to a primary health center (where competence is low) than to a hospital (where competence is higher). Disadvantage number 3.
There are a couple of potential further disadvantages that – fortunately for the poor – don’t seem to occur in practice. In the private sector (which is used more frequently by everyone in Delhi than the public sector), there isn’t any competence difference between doctors seen by the poor and doctors seen by the better off – after controlling for neighborhood. In addition, at least in Paraguay, doctors don’t seem to make more effort (ask more questions, etc.) with richer patients. They don’t discriminate against the poor, in other words.
Quality and UC
What all this suggests to me is that we ought to be recasting our equity goal in the health sector to factor in the quality of care. Equalizing utilization – in terms of visit rates, admission rates, etc. – simply won’t do. It could be consistent with large inequalities in health improvements through the poor’s treble quality disadvantage.
But it also suggest to me that we need to start finding a way to map quality – measured in terms of the percentage of essential tasks and questions done and asked – into health improvements. At the moment, we don’t know whether the ambulatory visit of a poor person – in health terms – is worth three quarters, a half, or a quarter of a rich person’s visit. What we do know is it’s not worth the same. And that’s worrying.