Predicting Mortality and Morbidity of Patients Operated on for Perforated Peptic Ulcers FREE

IN THE ERA OF open surgery, 3 prognostic factors (preoperative shock, perforation for more than 24 hours, and associated medical diseases) have been identified in patients with perforated peptic ulcer.¹ With the introduction of laparoscopic repair techniques in the treatment of perforated peptic ulcer in 1990, many retrospective^2- 5 and prospective nonrandomized trials have been published.^6- 9 It is important to stratify patients into various risk groups in comparing the laparoscopic and conventional open techniques. There is no doubt that the Boey scoring system accurately predicts mortality after open surgery¹⁰; however, it is not known if the Boey scoring system is equally good for predicting morbidity. This is important because decrease in morbidity is a major outcome measure in minimally invasive surgery.

The aim of this study was to evaluate the usefulness of 2 scoring systems, the Boey scoring system and the APACHE II (Acute Physiology and Chronic Health Evaluation II) score as a potential tool to stratify patients with perforated peptic ulcer.

This retrospective review involved patients operated on for perforated peptic ulcers from January 1989 to December 1998. A total of 436 patients were recruited. Case notes were reviewed with particular attention to the preoperative risk factors. The Boey score and the APACHE II score were calculated based on data recorded at hospital admission. These patients were operated on by open and laparoscopic means, and reasons for conversion were sought from the operative record. Two types of laparoscopic repair were performed, the suture repair and the sutureless repair (fibrin glue technique). The latter procedure has previously been reported by our group.¹¹ Ninety-three patients were randomized into 4 different operative approaches from August 1992 to December 1994: open suture, open sutureless, laparoscopic suture, and laparoscopic sutureless repair. Results of that randomized trial have been previously published.¹² Because of the promising outcome of laparoscopic repair, patients with suspected perforated peptic ulcers were subjected to laparoscopic repair from 1995 onward unless the surgeon or the anesthetist considered the patient not suitable for laparoscopic repair. However, the choice between laparoscopic suture and laparoscopic sutureless repair depended on the individual surgeon's expertise and preference. Outcome measures evaluated in this study included mortality and postoperative complications (morbidity). The following complications were of particular interest: wound infection, prolonged ileus, intra-abdominal abscess, and leakage. Exact causes of death were individually examined.

The Pearson χ² test, the Mann-Whitney U test, and logistic regression were used for statistical analysis; P<.05 was considered statistically significant.

The mean ± SD age for the 436 patients was 51.5 ± 18.3 years (range, 14-92 years), with male predominance (365 male vs 71 female). Seventy-seven patients (17.7%) had an underlying medical illness, and 14 patients (3.2%) had malignant neoplasms.

The duodenum was the most common site of perforation, accounting for 344 (78.9%) of 436 cases. One hundred ninety-six patients underwent open suture repair, and 46 received laparoscopic suture repair. Sutureless repair was performed in 132 patients: laparoscopic fibrin glue repair was performed in 109 patients, and open glue repair was performed in 23 patients. Definitive procedure was performed in 59 patients (Table 1). The initial planned procedure for 209 patients was laparoscopic glue repair (153 patients) and laparoscopic suture repair (56 patients). Conversion to open suture repair or definitive procedure was required in 46 patients and 10 patients, respectively. The conversion rate of laparoscopic glue repair and laparoscopic suture repair showed no statistically significant difference (28.8% vs 21.4%; P = .3, χ² test). The 2 most common reasons for conversion were difficulty in identifying the perforation site (34.8%) and huge ulcer considered not safe for laparoscopic repair (47.2%).

The mortality rate was 34 (7.8%) per 436 cases, with sepsis associated with multiorgan failure being the most common cause of death (15 of 34 cases). Leakage after simple closure accounted for only 3 of 15 deaths. The other causes of death are as follows:

The median hospital stay among the survivors was 6 days (range, 2-77 days), and the median hospital stay for the nonsurvivors was 6.5 days (range, 0-64 days). One survivor stayed exceedingly long (77 days) because of necrotizing fasciitis, which required multiple operations for debridement and skin grafting.

Among the survivors, 89 patients (22.1%) had postoperative morbidities, including wound infection, prolonged ileus, leakage, intra-abdominal abscess, and other complications. Wound infection was the most common morbidity (Table 1). The leakage rate (including radiologically and clinically detected leak) after simple closure (including suture and sutureless repair) was 23 (6.1%) per 374 cases.

Thirty patients (6.9%) had a Boey score of 2 or more. The mortality rate increased progressively, with increasing numbers of Boey risk factors: 1.5%, 14.4%, 32.1%, and 100% for 0, 1, 2, and 3 factors, respectively (P<.001, Pearson χ² test). The morbidity rates for 0, 1, and 2 Boey risk factors were 17.4%, 30.1%, and 42.1%, respectively (P = .002, Pearson χ² test). The median APACHE II score was 5 (range, 0-24). The APACHE II score was higher among the nonsurvivors than among the survivors (P<.001, Mann-Whitney U test) and among patients with postoperative complications (P<.001, Mann-Whitney U test). The median size of the ulcer was 5 mm. When compared with that of the survivors, the median ulcer size of the nonsurvivors was significantly larger (P<.001, Mann-Whitney U test). Similarly, the ulcer was significantly larger in patients with morbidity (P<.001, Mann-Whitney U test). It appeared that the Boey score, the APACHE II score, and the ulcer size independently predicted mortality and morbidity of patients with perforated peptic ulcer. Using multivariate analysis (logistic regression, forward stepwise), patients' likelihood of death could be predicted by the Boey score (P = .02) and the APACHE II score (P<.001) but not the ulcer size (P = .88). In terms of morbidity, only the APACHE II score (P<.001) could predict the risk of complications (logistic regression, forward stepwise); the Boey score and the ulcer size were nonsignificant (P = .88 and P = .47, respectively).

The overall conversion rate for laparoscopic attempted repair was 26.8%. The conversion rates for 0, 1, and 2 Boey risk factors were 21.4%, 30.2%, and 81.8%, respectively. No patients with 3 risk factors were subjected to laparoscopic repair. A significantly higher conversion rate was noted in the high-risk group (P<.001, Pearson χ² test). On the other hand, using univariate analysis, the conversion rate for patients with worse APACHE II score was also significantly higher (P = .005, Mann-Whitney U test). However, multivariate analysis confirmed that only the Boey score predicted the risk of conversion (logistic regression, forward stepwise, P<.001). With regard to outcome measures, conversion to open surgery was not associated with higher mortality (P = .33, χ² test) or morbidity (P = .37, χ² test).

Using the median APACHE II score as the dividing line, patients could be divided into 2 groups, a low-risk group with APACHE II score less than or equal to 5 (256 patients) and a high-risk group with a score greater than 5 (180 patients). After stratification, the mortality rate of different procedures showed no statistically significant difference (P = .22, Pearson χ² test) in the high-risk group. Zero mortality was recorded in the low-risk group. Likewise, the morbidity rate of different procedures showed no difference in both the low- and high-risk group. Before stratification, the mortality rate of open suture repair (13.3%) was apparently higher than the other techniques. This could be explained by the fact that a greater proportion of high-risk patients (104/180) were treated by open suture repair. The technique per se sustained no impact on survival outcome after risk stratification.

A previous study by Boey et al¹⁰ in the era of open surgery validated a set of risk factors to stratify patients with perforated duodenal ulcers. Patients with 0, 1, 2, and all 3 risk factors were noted to have mortality rates of 0%, 10%, 45.5%, and 100%, respectively.¹⁰ Similar findings were observed in this study. Although the Boey scoring system accurately predicted the chances of survival in patients with perforated peptic ulcer, it failed to estimate the likelihood of postoperative complications. When evaluating clinical outcome of perforated peptic ulcer, the procedure-related morbidity rate is equally important. The APACHE II scoring system is commonly used in the modern-day management of surgical intensive care patients.^13- 18 This scoring system gives a detailed documentation of the acute physiological disturbance. In contrast, the Boey scoring system takes into consideration only one of the physiological parameters, ie, preoperative hypotension in classifying patients into different risk categories. In the present study, worse APACHE II score was predictive of high mortality and morbidity rate. Therefore, the APACHE II score should be considered an appropriate prognostic marker in managing patients with perforated peptic ulcers. More important, when comparing treatment outcome (mortality and morbidity) of different procedures, patients can be accurately stratified into various risk groups according to their APACHE II score before comparison.

Recent studies have demonstrated that laparoscopic repair of perforated peptic ulcer is feasible and as safe as conventional open surgery.^{3- 5,7- 9,12} The overall conversion rate in this study was 26.8%. Although conversion to open surgery apparently did not affect the clinical outcome of perforated peptic ulcer, the total operative time was prolonged and the work load of theater staff was increased. Only the Boey scoring system could predict the risk of conversion as confirmed by multivariate analysis. The conversion rate for patients with 2 points was more than 80%. Laparoscopic attempt in these patients did not appear to be beneficial at all.

In conclusion, the APACHE II score, which could predict the mortality and the morbidity rate, should be used for patient stratification in clinical research setting. On the other hand, one may contemplate the use of the Boey scoring system preoperatively to select patients for laparoscopic repair. Thus, both scoring systems serve as valuable predictors in the modern-day management of perforated peptic ulcers.

Corresponding author and reprints: Wan Yee Lau, MD, Department of Surgery, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong, China (e-mail: josephlau@cuhk.edu.hk).