BMJ 25 Aug 2007 Vol 335

Every week or two we detect an irregular pulse and send the patient off for an ECG, but we’ve never audited how many with ?AF written on the form actually have it. Some researchers in the Birmingham (UK) Department of Primary Care, however, have made something of a specialty of atrial fibrillation and the results of their labours are to be found in two papers here and in the BAFTA study published a fortnight back in The Lancet. The ECGs for these studies were read by (a) software, (b) GPs, (c) practice nurses and (d) by Mick Davies or Greg Lip, who are Birmingham cardiologists, and therefore gold standard. GPs and PNs are equally good (bad) at spotting atrial fibrillation on ECG tracings; software is very specific but misses a few (sensitivity 83%). If you combine the two you will miss less than 10% of AF as defined by a Birmingham cardiologist.

Atrial fibrillation gets commoner the older you are and is a potent risk factor for stroke and heart failure, so how might we best screen for it in primary care? Should we do ECGs on everybody over a certain age (65 in this study) or should we just take the pulse opportunistically? Although at first glance you might think this study addresses this question, in fact it doesn’t. Both the ECG group and the pulse-taking group were called in for examination, so this was not “opportunistic” in the sense we usually use the word, i.e. done if the patient happens to turn up. Oddly enough, one third of patients who were found to have an irregular pulse refused to have an ECG – something I have never known happen in 30 years. But whatever the quirks of this study, it does appear to show that pulse-taking followed by ECG is as good as mass ECG screening for detecting AF, and much cheaper.

Children should not have needles stuck into them without good reason. Does acute pyelonephritis count as one? Not as far as antibiotic treatment goes: oral co-amoxiclav proved as good as parenteral ceftriaxone in this large Italian study. They did however have to have a needle for the dimercaptosuccinic acid (DMSA) scans they all had.

DMSA scanning at 4-6 months after an acute UTI in children under 3 years old is perhaps the most important innovation in the NICE guideline for urinary tract infection in children, which does away with the barbaric practice of testing for vesico-ureteric reflux.