Sir,

I read the above paper with interest. Unfortunately, because of relatively small numbers and a heterogenous group of patients, despite this representing a large body of work on which the authors should be commended, I feel that important messages are being lost. This may lead the inexperienced surgeon to persist with mistakes in management that have dogged the improvement of trauma outcomes for three decades.

The grade of trauma is significant. It is well recognised that the energy transfer at the point of trauma is a major factor in determining management choices and success. We should not draw conclusions from a group that lumps together all Gustillo grades. Gustillo, in his various writing, points out the importance of the grading in determining management choices and outcomes.

I was saddened by the conclusion that timing is unimportant and feel that this, too, has been drawn because of the flawed nature of the study. [1] showed, quite clearly, in a large, pure series of Gustillo III fractures that there is a linear relationship between time of repair and the important complications of delayed union, nonunion and infection. That is a hard-learned lesson and could almost be described as an "inconvenient truth" when it comes to designing trauma services and organisation and its impact on surgeons' lifestyles.

I feel that osteomyelitis is a separate and difficult problem in its own right. As such, observations about its successful management, which is largely about adequate debridement and manipulation of blood supply [2], should not be used to inform management decisions in the acute trauma setting.