Hip-fracture and stroke care: parallel problems in evidence

We thank Michael Hill for his comments and agree that our study¹ may have missed a clinically important difference because of a lack of statistical power. As we stated in our interpretation, the 95% confidence interval for the primary outcome measure (–5.6% to 17.0%) allowed for the possibility of a clinically important effect. We strongly support the need for a larger multicentre trial to study the effectiveness of interdisciplinary care for elderly people with hip fracture. However, as we stated in our paper, we recommend that the intervention be targeted to a subgroup of patients that may be more likely to benefit than the heterogeneous population included in our study.

Hill writes of the potential importance of physically centralizing geriatric hip-fracture patients, as is done with stroke patients. In our study, the intervention patients were located together in the hospital.

Gary Naglie Barry Goldlist University Health Network Ed Etchells Sunnybrook and Women's College Health Sciences Centre George Tomlinson University Health Network University of Toronto Toronto, Ont.