Jejuno-Ileal perforationsVG Mehendale, AB Samsi
Department of Surgery, K.E.M. Hospital and Seth G.S. Medical College, Parel, Bombay 400 012, India
Thirty-two consecutive cases of small bowel perforations have been analysed.
Peritonitis resulting from jejuno-ileal perforations is an extremely common surgical problem met with in this country. New trends in the diagnostic measures like laparoscopies, along with medical termination of pregnancy with vaginal sterilisation and tubal ligation done through laparoscopy have become quite common and have contributed to the aetiological factors. Inspite of better understanding of the pathology and consequent improvement in the management of these cases, the mortality and morbidity still remain high, both here and in other countries reporting such cases.
Thirty-two cases of jejuno-ileal perforations, admitted to King Edward VII Memorial Hospital, Bombay, over a period of one year were analysed. The method of study followed was as follows:
A complete detailed history and clinical examination of every patient was done. Clinical examination was repeated as often as necessary till a definite diagnosis was arrived at. A vigilant check was kept on urine output, central venous pressure, blood pressure and other parameters to ensure good peripheral tissue perfusion.
(1) Routine haemogram and urine examination.
(2) Biochemistry: (a) B.U.N., (b) serum creatinine and (c) liver function tests.
(3) Serology: Widal test was positive only in one case.
(4) Radiology: X-ray of the abdomen in an erect posture in every case. Twentyone cases showed pneumoperitoneum. In the remaining, probably, early sealing of the perforation or a slow leak prevented the formation of pneumoperitoneum.
(5) Peritoneal tap: Four quadrant paracentesis was necessary in 15 cases, while one quadrant tap was sufficient in 8 cases.
(6) Bacteriological examination of the peritoneal exudate.
(7) Histopathology of the edge of the perforation in each case.
Operative findings and procedures, post-operative management and complications, and post-mortem findings in fatal cases were completely analysed.
There was a predominance of males over females giving a ratio of three males to one female. This may be an apparent contradiction considering that incidence of iatrogenic perforations in females is on an increase.
Most of the patients were in the active phase of life. 84% of the cases were in the age group of 11-40 years [Table 1].
The diagnosis was not always easy. Important symptoms were pain in the abdomen, vomiting, fever and retention of urine. (1) If a patient who has been running moderate fever starts complaining of pain in the abdomen, the symptom should be considered seriously. (2) Retention of urine following laparoscopy, medical termination of pregnancy, vaginal sterilisation should be considered as a danger signal. (3) All the standard signs of perforative peritonitis were not convincingly present in all the cases. (4) Restriction of abdominal movement is suggestive of peritonitis and the sign was positive in every case. (5) Absence of peristalsis is an ominous sign and suggests severe degree of peritonitis. Those patients admitted in a state of peripheral circulatory failure were found to have gross contamination of peritoneal cavity and the prognosis was grave. Diagnosis of peritonitis was delayed in two of these cases due to non-recognition of abdominal signs [Table 2].
Amongst the investigations, X-ray of the abdomen in an erect posture and peritoneal paracentesis were most helpful, Widal test was positive only in one case with a titre of 1:250.This patient died due to septicemia following the perforation. Peritoneal paracentesis in more than one quadrant was more beneficial to reach the diagnosis.
Definite diagnosis of typhoid or nonspecific ulcer could be made only after the histopathological findings were available because the clinical picture and the operative findings were almost identical in two types of cases, and Widal test was not helpful. The etiology was determined by biopsy.
After pre-operative preparation with gastrointestinal decompression, intravenous fluids and electrolytes, patients were taken up for surgery. Pre-operative blood transfusion was necessary in two patients. Parenteral broad spectrum antibiotic therapy was instituted in all the cases.
[Table 3] outlines the operative procedures employed.
The amount of peritoneal contamination determined the drainage. When the peritoneal exudate was more than 1000 ml drainage was used. Peritoneal lavage was done in all the cases.
Extraperitonealisation was done when it was felt that the tissues were oedematous and friable and therefore not likely to hold the sutures well.
Exteriorisation was done when patient's condition was rather poor and the local tissues also were very friable. Resection was done when patient's condition was satisfactory but his bowel looked very markedly inflamed and congested.
In all except two cases the perforations were situated in the terminal ileum. Among these 12 were within the last 15 to 20 cms of ileum from ileocaecal junction. In 22 cases distal ileum looked inflamed or congested. Later on, these cases turned out to be of either nonspecific or of typhoid ulcers.
Only in two cases the perforations were situated in the proximal jejunum-one following intestinal obstruction due to tuberculous strictures and the other was due to blunt trauma to the abdominal wall. The size of the perforations varied from 2 mm to 15 mm. None of the perforations were localised or covered with greater omentum.
In all but 3 patients who expired, the duration of perforation was more than 72 hours. In two cases, the duration was 24 hours and in one case 48 hours; however, in these patients there was severe peritoneal contamination, peritoneal exudate being more than 1000 ml and they were admitted with established shock.
These are shown in [Table 4]. The most common complication was septicemia occurring in 13 cases out of which 11' cases succumbed and only 2 could recover.
In this series, 12 patients died giving an overall mortality rate of 37.5% , individual number of deaths being tuberculosis-1 out of 2 (50%), enteric fever-4 out of 9 (44.5%), non-specific inflammation-5 out of 13 (38.4%) , intestinal obstruction-1 out of 3 (33.3%) and trauma-1 out of 5 (20%).
Inspite of the adyent of new broad spectrum antibiotics, better understanding of the pathophysiology of shock and better diagnostic measures, the mortality of perforative peritonitis from small bowel lesions still remains quite high. In this series of 32 cases, 12 succumbed giving an overall mortality rate of 37.5% amongst the others, morbidity was high (40%).
Since the early symptoms and signs are rather vague and are not conducive to a definite diagnosis, most of the cases come late or are detected late. Early diagnosis is of profound importance, because peritonitis spreads early in these cases. There is no localisation of inflammation as the perforation is not covered with the omentum. , Peritonitis sets in early, possibly, because from stomach onwards acidity of the contents falls and the bacterial flora increases.  Once septic shock sets in the mortality is prohibitive in spite of a very aggressive treatment.
We feel that a high index of suspicion is essential to detect a perforation early. The important factors to look out for are: (1) persistent pain, (2) restriction of abdominal movements with respiration (3) absence of peristalsis, (4) retention of urine and (5) persistent tachycardia. Other classical signs may be totally absent in these cases. Tenderness may be minimal and vague, rigidity may only appear late and distension may be a terminal event. Diagnosis of peritonitis is liable to be missed in typhoid or tuberculosis, as pain and discomfort due to distension and toxaemia may be present much prior to the perforation and may be neglected by the physician. We have found this to be true also in cases of nonspecific perforations.
In any case where perforation is suspected, reliance has been placed in the past, on a plain skiagram for the presence of free gas under the diaphragm. In our opinion this sign, when present, is certainly diagnostic; but one would like to emphasize that absence of gas does not rule out perforative peritonitis. In our series this was absent in 11 cases (34.41%) . We feel that a four quadrant abdominal paracentesis is a much more conclusive test and is often thoroughly reliable. Occasionally it may so happen, that the paracentesis shows clear straw coloured fluid. This may be due to settling down of thick pus in the pelvis, the reactionary fluid on the surface being aspirated. Therefore when one aspirates clear fluid easily in suspicious cases, even if it is clear and straw coloured, one should not hesitate to diagnose perforative peritonitis.
Once the diagnosis is established the treatment should be aggressive. Huckstep  reports good results from conservative treatment in cases of typhoid ulcer perforations. He treats his patients on Oschner-Sherren regime. We strongly advocate operative treatment for small bowl perforations. Huckstep's series included only typhoid ulcer perforations, while we have found in our series that the number of non-specific perforations exceeds that of typhoid. Also peritoneal toilet lessens toxaemia and the closure of perforation eliminates continuous contamination of peritoneal cavity. 
The aetiological type of perforation is often not apparent on the table; and in point of fact may not really alter the course of the patient. The type of operative procedure too seems to have little bearing on the ultimate outcome. No single operative procedure can be said to be ideal. The more conservative the surgery, the better is the prognosis. The best policy seems to be, to do minimum possible surgical intervention and hope for the best. Though some of the recent reporters , give good results with resection and anastomosis, such an extreme procedure increases duration of anaesthesia that would be hazardous in these cases, at least in our country.
The incidence of proved typhoid perforation is much lower than before. The most frequent aetiological factor seems to be nonspecific ulcer, the exact cause of which is unknown.
It was our experience that the patients with non-specific and typhoid ulcer perforations had similar clinical features. Patients with non-specific perforations also had history of fever, vomiting, distension and gastrointestinal haemorrhage. Toxaemia and shock were equally common in typhoid and nonspecific inflammation. Since Widal test was negative in all except one patient we had to rely on histopathological diagnosis. In typhoid as well as in nonspecific ulcers the perforations were within the last 3 feet of ileum; the terminal ileum was inflamed in both type of cases. Seven out of 13 cases of nonspecific perforations did not have fever.
Histology: in the edges of ulcers,
(i) (i) Typhoid: Marked proliferation of reticuloendothelial cells of the lymphoid follicles was seen. The reticulum cells, reticuloendothelial cells and imigrant macrophages in the resultant aggregation contained large cells having an abundant pink cytoplasm often containing bacteria, cellular debris and red cells. Erythrophagocytosis was prominent and virtually marked the lesion as being due to enteric fever.
(ii) Non-specific ulcers: Large number of polymorphonuclear cells with cellular debris were seen. There was no evidence of large macrophages.
Evert and Black  and Goehrs et al  from Mayo Clinic gave a theory of submucosal vascular embolisation as the cause of nonspecific perforations. Many workers proved, clinically and experimentally, that enteric coated tablets of potassium chloride lead to nonspecific ulcers in the distal ileum, which perforated easily. ,
We had no evidence to support the above views. Clinical features, operative findings or histology did not reveal any evidence suggestive of vascular embolic phenomena, nor was there history of potassium chloride therapy in any case. Whether other drugs like salicylates, antibiotics or steroids cause non-specific ulcers in the small bowel should be studied. As yet, definite aetiological factors of small bowel perforations due to nonspecific inflammation remain to be defined. Whatever the aetiological factor, the best policy seems to be early diagnosis and quick, minimum surgery.
We thank the Dean, K.E.M. Hospital, Bombay, for permitting us to report the hospital data.
[Table 1], [Table 2], [Table 3], [Table 4]