Failures teach more than successes. Therefore it is important to examine in what cases the method fails to sterilise wounds, and what are the causes of these failures.

A. Wounds Of The Soft Parts Not Accompanied By Bone Injuries

If the surgical sterilisation of a wound be considered as the object of the method, it might be asserted that no failure has been observed since the month of December, 1915. But if the role of the method is to prepare for the secondary closure of wounds, the proportion of failures rises to about five per cent.

These failures are due to the following causes:-

(a) Errors in the bacteriological examination. In spite of the absence of microbes from the smears, suture of the wound was followed by infection. This accident was very rare and always without serious results. It was met with twice in the course of 333 cases of wound-closing. Wounds, the seat of infection, were reopened and sterilised in a few days. This accident was the consequence of specimens for the smears being badly taken; and can be avoided by taking multiple specimens, especially from the most obscure parts of the irregularities of the wound.

{b) Loss of tissue-substance. In some cases, the loss of integumental substance was very extensive, and union impracticable. In other cases, union became possible, if traction more or less great by sutures were employed. But these cut the skin and union remained imperfect.

(c) Closure without bacteriological examination. It sometimes happened that, seeing a wound of good red colour, without secretion, and with margins perfectly supple, the surgeon did not wait for a laboratory report, and sutured. Under these conditions, the operation sometimes resulted in failure. This mistake has not been committed in our hospital since the month of December, 1915. Before that date, it happened several times.

When the treatment was commenced after a period of suppuration more or less long, numerous failures of the method might have been expected. Nevertheless, all the suppurating wounds arrived at the stage of surgical sterilisation, and no failure was registered. The closure of these wounds was practised at a later date than when dealing with fresh wounds, but nearly all the cases were sutured.

B. Wounds Of The Soft Parts Accompanied By Injuries To Bone

In compound fractures, and especially in those which had suppurated before the commencement of sterilisation, we did not always achieve surgical sterilisation. From this point of view, the results of treatment of compound fractures can be clearly separated from those of wounds of the soft parts. Failures were more frequent. We look upon as failures cases in which some microbes persisted in the secretions up to the spontaneous closure of the seat of fracture, and those in which a sinus or fistula persisted.

We have observed no failure in the treatment of compound fractures of the small bones, short bones, radius, and ulna. But some compound fractures of the humerus, tibia, and femur did not respond completely to treatment. The statistics of these cases will not afford any indication of interest, because the methods have been progressively modified, and the results are improving more and more. In the last fifteen cases of fracture of the humerus which have been under our care, several of which were freely suppurating at the commencement of treatment, only four were not sutured. In two cases, suture was not practised because of loss of substance. In only two cases surgical sterilisation of the seat of fracture was not attained. One case was a "smash-up" of the end of the diaphysis and the adjoining head of the humerus,, and the other case was a highly comminuted fracture of the shaft. In both cases the secretions contained some microbes up to complete cicatrisation. Recovery took place without a sinus remaining.

In compound fractures of the tibia, the loss of substance is often too great to allow of the soft tissues being completely brought together. In similar fractures of the femur, approximation of the tissues was always possible, but microbes often remained in the secretions, and prevented suture being carried out.