My Cumbrian GP friend, who has the perspective of a doctor with experience of medicolegal work, thinks that the medical textbooks need to be rewritten. Taking subarachnoid haemorrhage and pulmonary embolism as prime examples, he points out that the textbooks list the clinical signs, without always adding that both problems can exist with no signs at all. (Favourite joke about embolism: The News Chronicle, a highly respected liberal-minded newspaper, ceased production in 1960. At the time it had an astonishing readership of something over a million, but an incompetent management. The peerless journalist James Cameron, who had worked on the News Chronicle, wrote that it died of an embolism — a good circulation impeded by clots).
The textbooks are still mostly written by hospital doctors. As Paul Hodgkin points out on page 130, hospital doctors' approach patients with the expectation that there is a high probability of serious disease. In contrast, one of the primary concerns in general practice is to exclude serious illness with a degree of confidence, so the task is subtly different. Or at least it was. When it comes to coronary heart disease, modern opportunities for early effective intervention, together with the increased sensitivity of diagnostic tests, have generated more of an impetus not to miss a positive diagnosis. The review on page 105 sets out to shed some additional light on this area. In their downbeat way, the authors report that ‘ … it was not possible to define an important role for signs and symptoms in the diagnosis of acute myocardial infarction … ’, but the presence of chest-wall tenderness was found to rule out acute myocardial infarction. Which rather proves my Cumbrian friend's point.
There is the suggestion of another legacy of the textbook problem on page 102, with the authors suggesting that GPs are not as precise as they might be when diagnosing headache. In the accompanying editorial on page 77, David Watson wonders if we don't take headaches seriously enough — they just don't kill. At the same time, the only people who can teach about headache are experienced GPs, and perhaps we have collectively failed to build up an independent body of knowledge. The guidelines offered by websites cited in the editorial may help, but patients with headache need to be treated as individuals just as much as anyone else. The paper on page 98 reports on a survey which asked patients about their suitability and acceptance of migraine prophylaxis. Again the suggestion was that GPs weren't offering prophylaxis as often as they could have done. Just as interesting was the finding of discordance among patients: those with little migraine who wanted to try regular prophylaxis and those stoical souls with high rates of migraine but who didn't want to consider it.
Paul Hodgkin's piece (page 130) offers a vision in which the clinical symptoms for very large numbers of patients, and the final diagnoses are pooled on big databases. The Holy Grail at the end of this quest is perfect knowledge (or at least knowledge much better than we have at present) of the predictive value for individual symptoms and symptom clusters. Such a system would, he suggests, obviate the need to maintain two separate groups of doctors created around probability of serious illness. Such talk is, however, a reminder of the difficulties presented by large databases, of which the losses of data reported a few weeks ago is only one. On page 131 Gordon Baird reports his questioning of his patients supposed consent to their records being loaded onto a central database. In the editorial on page 75 Ross Anderson has concerns beyond consent that should worry patients and doctors alike. His fear is that without robust arrangements for consent and opting out, anyone receiving NHS care will have data transferred to a number of different databases. Crucially he suggests alternative models to the single centralised database that are already being developed in mainland Europe.
There are numerous advantages to systems that enable us to work more efficiently, not least further elimination of paper records; and there are lots of well-meaning people working hard to make the system work well and securely. But some of us fear we are colluding in the creation of a monster. Anyone going to join the opt-out campaign?
- © British Journal of General Practice, 2008.