Upper limb regional anaesthesia and altered coagulation function.

Editor—We thank Dr Dolenska for her letter and completely agree that the finding with respect to the predictive value of patient age is highly counterintuitive. Age is generally accepted as a powerful predictor of survival in many contexts. However, our results would suggest that it is the decline in fitness with advancing age which is the real predictor, rather than age per se. We do not believe that the finding could be erroneous. Specifically, when the trillions of models potentially explaining our data were distilled down to the few hundred very best models via the Bayesian model averaging process, 66% of those models predicted survival without requiring the input of patient age. The general linkage between age and fitness makes age a reasonable surrogate for fitness in the absence of fitness data. However, where one has access to such data, it renders age largely irrelevant. The day may be fast approaching where life insurers require a cardiopulmonary exercise testing before quoting a premium.