Hip-fracture and stroke care: parallel problems in evidence

Although Gary Naglie and colleagues' study of postoperative care for geriatric patients with hip fracture1 produced a neutral result, it may well have been underpowered, as the authors note in their interpretation. In looking for an absolute risk reduction of 17%, they may have missed a clinically important difference of 5%. Here I see a parallel with the development of evidence in favour of stroke unit care.

The benefit of stroke unit care was convincingly shown only in a meta-analysis of 19 trials.2 The absolute benefit in reduction in mortality or dependency is about 6%, a figure similar to the absolute (nonsignificant) benefit in reduction in mortality and ambulatory deterioration of 5.6% seen in Naglie and colleagues' trial. Equally, among stroke patients, roving stroke units are probably less effective than geographically focused units.3 Perhaps the physical centralization of geriatric hip- fracture patients is similarly important.

It is still unknown in definitive terms why stroke units are effective. Common sense gives us reasons but, broadly speaking, perhaps focused multidisciplinary care could improve outcomes for relatively homogeneous patient populations in a wide range of disciplines. It would be worth while to pursue a larger multicentre study of interdisciplinary hip-fracture care with sufficient power to detect small benefits. A 5% absolute benefit would be clinically important in Canada, with obvious relevance as the population ages.

Michael D. Hill Stroke Neurologist University of Calgary Calgary, Alta.