Fifteen years after retirement, Geoffrey Marsh is offering advice on How to be Happy in General Practice: ‘Overall I tried to diagnose and treat patients without extensive investigations …’ (page 295). How wise. We use investigations all the time, and they provide endless information. The skill is in being sure what the answers mean. On page 243 a team of Dutch researchers have reported their conclusions from a study looking into the use of investigations for patients presenting with fatigue. Among all the patients included in the study, 8% were found to have somatic illness detected by blood tests; much larger numbers had test results found to be false positives, and naturally enough, the more tests done the higher were the number of false positives. Patients in the study were also randomised to having tests immediately or after a delay. Of those randomised to the delay group, only a quarter returned for tests after 4 weeks. Based on this the authors advise that doctors should delay before offering tests, but Harvey and Wessely in the leader on page 237 are less sure, pointing out that the proportions of true and false positives were similar in the immediate and delayed groups. Another Dutch group on page 250, trying to predict poor outcome at presentation, tried a different approach. Readers might guess that baseline severity would predict a longer course, but the finding that patients' expectations also predicted worse outcomes was a surprise to me. The theme of false positive results is taken up on page 297, following the recent controversy onmammography screening set off by Peter Gøtzsche's piece in the BMJ. Mike Fitzpatrick ends with a plea for statistical literacy as a requirement for an educated citizenship (echoing the crusade that Ben Goldacre is engaged on – see the review of his book in March's BJGP).
The opposite problem is illustrated on page 256 in an examination of a diagnostic rule to help identify patients at high risk of osteoporosis. Here the rule did identify many who turned out not to have the problem, as one would expect; the surprise was the low sensitivity of the rule, so that many with osteoporosis would have been classified as low risk, and would not have been considered for DEXA scanning. Some false positives, of course, but too many false negatives for the rule to be of much use. The accompanying leader on page 239 reminds us of the importance of osteoporosis. At the same time it makes an interesting comparison with hypertension: neither really a disease entity in itself, but contributing to a long-term risk. Just as we are succeeding in incorporating blood pressure into an overall assessment of heart disease risk we should stop separating osteoporosis from other risk factors, and start to use it as a component of a risk score for fragility fractures. But it remains to be seen whether this general approach will do much better. Part of statistical literacy (for all of us) is accepting that such an approach will prevent some but not all of the problem: even with a low risk score patients can suffer from cardiovascular disease – or a fractured neck of femur.
The first element for a happy life Geoffrey Marsh identified was looking after a personal list. Continuity of care is, however, a slippery concept that the paper on page 276 sets out to pin down. The paper concludes that research must be explicit about what kind of continuity is being measured, and why. It matters because it matters to patients; because ‘having positive consultations with the same doctor over time builds depth in the patient–doctor relationship which, in turn, may promote further longitudinal care’ (page 268), and the review discusses the main elements of such depth. The leader on page 236 points out that the more assertive of our patients manage to navigate whatever systems are put in place, but we have a duty to organise services to make sure that the more vulnerable are equally able to get what they want. Here in the UK our system of primary care has always been defended on the grounds of providing a personal service. The charge of hypocrisy has been made against us for defending vociferously something we seem willing to negotiate away (for instance giving up 24-hour responsibility), although the figures being able to consult the doctor quoted in the discussion paper look creditably high (page 276). It still matters, and such papers give us the tools with which to support this core value in the face of scepticism from others, and to ensure that it is not ignored and then swept away as the landscape around changes beyond recognition.
- © British Journal of General Practice, 2009.
