Laparotomy 1

In a consecutive series of 1207 major laparotomies, the incidence of wound infections, wound failures and chest complications was prospectively recorded; this allowed certain predisposing factors to be identified, both avoidable and unavoidable. Among the former were operative bacterial contamination, haemorrhagic and septic shock and the use of nasogastric tubes; the most important of the latter were male sex and pre-existing bronchitis.


Introduction
The surgical journals publish a vast number of papers designed to tell us of the success (or, more rarely, failure) of particular techniques applied to individual diseases. These are valuable in that they advance our knowledge in defined areas. Nevertheless, studies of this kind neglect the fact that the work of a surgeon is more than the scholarly investigation of specific diseases. He must see consecutive patients of all ages with a wide variety of problems. In treating abdominal conditions he must often make a laparotomy to diagnose and treat. In doing so he should aim to expose the patient to the smallest possible risk, irrespective of the hazards of the disease for which the operation is undertaken.
In an effort to minimize the dangers of laparotomy and to find out which features of management are important, all patients having major laparotomies under my care (not including grid-iron appendicectomies and hernia repairs) have been studied in random controlled clinical trials which seek to answer specific questions about postoperative complications. These trials have allowed us to accumulate a large amount of information about the outcome of abdominal surgery and what the laparotomy itself contributes to the result.
Patients and methods One thousand two hundred and seven consecutive patients who required abdominal exploration in the five years 1974-1979 are the basis of this report. All patients were prospectively documented during their inpatient stay; survivors were assessed at one month after operation to see if they had developed a wound infection after leaving hospital and again five months later to diagnose incisional hernia.
Postoperative physiotherapy was standarddeep breathing exercises, postural drainage and percussion. Antibiotic prophylaxis and wound closure techniques varied, as will be indicated in the paper.
There were 577 men and 630 women. The age range was 9-96 years, half the patients were over 65, and 270 (22%) presented as emergencies.

Mortality
There are five reasons why people die after abdominal operations: lethal disease; preventable complications; bad surgery (euphemistically called 'technical failure'); unforseeable complications; and the wrong choice of operation.
'Based on Presidential Address to Section of Surgery, 31 October 1980. Accepted 31 March 1981

Morbidity
Complications which could be regarded as being due to the abdominal incision were: (1) Wound infectiondefined stringently as any wound which did not remain completely dry throughout the healing process. It is appropriate to distinguish between primary infection sustained at the time of laparotomy, and secondary infection which takes place subsequently. Also a practical distinction should be made between major and minor infections: they are classed as major if they make the patient ill, delay discharge from hospital and need several dressings a day; and minor if they lack these characteristics.
(2) Respiratorythe patients' status was scored using a system previously described (Leaper et al. 1977), which distinguishes between minor and clinically important complications.
(3) Quality of abdominal wound healinga patient was said to have burst his abdomen if peritoneal fluid or abdominal contents were extruded. Incisional hernia was diagnosed at six months if there was a bulge visible on straining together with a palpable defect in the abdominal wall.

Deaths
One hundred and twenty-two patients died (either in hospital or at home) within 30 days of their operations, and necropsies were performed in 73%o. The adverse influences of male sex, old age and emergency operations are shown in Table 1. Forty-three patients were classified as having died of their disease, though there will always be doubt whether different management could have averted the fatal outcome. The processes responsible are indicated in Table 2. Theoretically preventable complications caused the deaths of 42 patients (Table 3)  All differences are statistically significant except those marked with *, *, A  and included organ failure in 27, pulmonary embolism in 8 and cholangitis or pancreatitis after combined supraduodenal and transduodenal choledochotomy in 6.
Technical inadequacy, with sepsis as its result, contributed to the deaths of 18 patients. Twelve were the consequence of anastomotic breakdown and the other causes are indicated in Table 4.
Thirteen patients died from unforseeable complications: 10 -while convalescing uneventfully from the surgical point of viewof cardiac infarction proved at necropsy, one of a stroke and two of necrotizing enteritis.
Five patients could be said to have succumbed because of the wrong operation. All were in a group of 19 with perforated peptic ulcers treated by simple suture. In contrast, 30 other patients with perforated ulcer were treated by vagotomy and pyloroplasty; there was only one death and that from pulmonary embolism.

Complications
Only 635 patients (53%O of the entire series) recovered without any complication whatsoever.
Wound infection: Fifty-three patients who died within a week of their operation without infection have been excluded and the percentages are, therefore, out of 1154.
Wound infection was the most common postoperative complication, occurring in 312 patients (27%). Eighty-one (7%O) suffered major primary wound infections and in 3 -all emergency colon resectionsit was a contributory cause of death. One hundred and eighty (16%) had minor infections. Heavy parietal contamination and inefficient antibiotic prophylaxis during operations were the most important causes of primary wound infection (Table 5). abdominal operations * Cephalosporin prophylaxis was by a single dose of I g cephaloridine intravenously or intra-incisionally, or a single dose of 1.5 g cefuroxime intravenously * Prepared indicates that patients were given an antiaerobic, together with an antianaerobic, drug by mouth for at least 24 hours before operation; ill-prepared indicates that only a single agent was used The incidence of secondary wound infection did not differ significantly between low-risk (upper abdominal) and high-risk (ileocolorectal) operations, nor did adequate mechanical and antimicrobial bowel preparation nor the use of a single dose of cephalosporin seem to have any effect.
Chest complications: Two hundred and ninety-two patients (25%) had a postoperative chest score of four or more (Leaper et al. 1977) and 16 of them died of pulmonary failure or infection within a month of operation. The most important single determinant of chest complications proved to be the sex of the patient -200 men (35%O) had high scores, compared with 92 women (15%). This was not due solely to the fact that more men smoked or had preexisting bronchitis (Table 6); all differences in this table are statistically significant (P < 0.001). In addition to sex and preoperative pulmonary status, strong relationships appeared between chest complications and excessive blood loss, serious sepsis and aspiration of gastric contents; there was a weaker association with depressed postoperative ventilation. A detailed analysis of these associations will be given elsewhere.
Disruptive complications. Five patients burst their abdomens, an incidence of 0.400. They were all men and the dehiscences could be blamed on technical faults in four and overwhelming infection in the fifth. The four due to faulty technique all occurred in the first 600 patients when a comparison of closure of the musculo-aponeurotic layers by deep-bite, far-and-near (Smead-Jones) interrupted mass sutures of either monofilament steel or polyglycolic acid with layered monofilament nylon closure was being made. The fifth patient had an ultimately fatal disruption in association with abdominal wall gangrene.
Nine hundred and sixty-one surviving patients were subsequently examined and 96 incisional hernias (10%) discovered. (2%) and 15 out of 229 (70%). Other significant determining factors were wound infection and a postoperative chest score of four or more (Table 7) and these seemed to be additive; when neither complication arose, 6% of men and 1% of women developed incisional hernias compared with 440% of men and 15% of women when both complications occurred (x2 = 54.44 and 20.21, P<0.001).

Discussion
Assessment of the results of treatment of intra-abdominal disease must take into account the influence of the means of treatment. In a surgeon's practice this usually includes an incision in the abdominal wall which carries its own risks and after effects. What I have attempted to do is identify how the laparotomy may influence results. The important conclusions which emerge from the studyand it should be emphasized that we are looking at consecutive patients in one man's practiceare as follows. Women tolerate laparotomy for comparable conditions better than men, perhaps because they smoke less, have a more highly developed survival instinct, and are intrinsically fitter in later life.
Wound infection is still a cause of mortality (rare) and morbidity (common). Many trials including our own (Pollock 1979) show that it is possible to reduce morbidity, and surgeons have a duty to look critically at their practice in this regard. Though burst abdomen and major sepsis can be largely prevented, incisional hernia remains a problem.
Respiratory complications of laparotomy remain common and can be lethal. Furthermore, they make for more wound complications. Improvement will come from a variety of sources, not least of which may be the better control of pain (Rutter et al. 1980).
Finally, intensive prospective study has revealed two things. First, that there are identifiable causes of death (such as in perforated ulcer) which may lead to better choice of operation irrespective of how the incision is managed and, secondly, that advances in the craft of surgery come only from analysis of results and identification of problems.