Commentary

David Ogilvie takes you through the paper and discusses the implications of the results

Should we tell patients in advance that they might be seen by a medical student? Attitudes have moved towards getting better informed consent in all aspects of medical practice. But on the other hand, we do not want to discourage patients from co-operating and so miss the chance to learn from them. This paper sheds some light on the dilemma.

What did they do?

Randomised trials are most familiar in the assessment of drug treatments, but the approach can also be used to study other types of intervention. It is essential to start with a well defined hypothesis that can be tested, and the authors make theirs clear: does advance written information lead patients to decline to take part in teaching? A randomised trial is a type of experiment in which the participants—in this case, people waiting for outpatient appointments—are randomly divided into two groups. One group gets the intervention— advance written information—and the other (control) group gets whatever is standard procedure. The idea is that there should be no difference between the groups except for the provision (or not) of the advance information. That way, we can reasonably assume that any difference which arises between the groups afterwards is due to the intervention and not to anything else.
To see why this is important, imagine an alternative study in which patients attending hospital A got the information, and patients attending hospital B did not.
If patients at hospital A were less keen on taking part in teaching that might be due to the information they were given. But it could also reflect other differences, such as consulting styles between the hospitals, or some other factor.
The classic way of doing a randomised trial is the double blind method. This means that neither the patient nor the researcher knows whether the patient has received the intervention or not and is supposed to minimise the risk of bias in assessing the effect of the intervention. This study is described as a double blind trial, but of course the patients must have known what information they were given. It turns out that the double blind effect was achieved by the interesting method of not telling the patients that they were in a trial until afterwards. You could not do this in a trial of a new drug as it would be unethical.

What did they find?

The main finding was that no patient in either group declined to take part in teaching, which suggests that giving advance information did not put anyone off. This seems straightforward, although it is worth asking yourself how many of these patients you would expect to refuse. It is always possible that a larger study might have found a difference which was missed in this small trial. What about the rest of the data? Fisher's exact test and the chi square (X²) test are ways of assessing whether two variables are associated or not. The results show, for example, that only one out of 32 informed patients wanted to be alone with the doctor and not have medical students present, compared with eight out of 39 control patients. The P value obtained from the Fisher's exact test indicates that this is a significant association—the likelihood of wanting to be alone with the doctor was related to whether or not advance information was given. The same applies to the likelihood of feeling free to say no to students. (Remember from my commentary in the July issue, that a P value of less than 0.05 is usually considered as statistically significant.¹)
The table shows how patients rated the information they were given. It is interesting to note that the majority of patients in both groups thought that the information was good or rather good. It is well known that patients tend to give good ratings in satisfaction surveys. Perhaps people do not like to seem critical or perhaps the standard procedures in Swedish hospitals are already very good.

So what?

We can conclude that giving patients advance information about medical students did not put them off. We can also conclude that the patients who got advance information were less likely to want to consult alone without medical students present, and more likely to feel able to say no. So should we just get on and give out advance information to everyone? It certainly does not seem to have caused any problems in this study.
But there is a question mark over generalisability. This study was conducted on a small number of gynaecological outpatients in Sweden. We do not know whether patients in other countries have a different attitude to medical students, or whether the results obtained in gynaecology would apply equally to other specialities. Sounds like an opportunity for further research.

avid Ogilvie, specialist registrar in public health medicine, Hamilton, Lanarkshire
Email: david.ogilvie@lanarkshirehb.scot.nhs.uk


studentBMJ 2001;09:357-398 October ISSN 0966-6494

1. Ogilvie D. Commentary on: Increasing obesity in primary school children. studentBMJ 2001;9:242. (July.)