The timing of elective colectomy in diverticulitis: a decision analysis

Commentary

Although diverticulitis is a common disease, there are relatively few data from high-quality studies on the natural history of diverticular disease, and no published randomized controlled trials have assessed the value of elective colectomy. Nevertheless, it is common practice to recommend elective colectomy after 2 episodes of diverticulitis in older individuals and after 1 episode in younger individuals. This decision analysis is timely, because it is reasonable to reconsider this dictum. Whereas acute diverticulitis, especially perforated diverticulitis, was once associated with extremely high morbidity and mortality rates, with the advent of new imaging modalities and the availability of percutaneous drainage and antibiotics, the risk of serious complications and death is rare. On the other hand, the disutility of elective surgery has decreased with the introduction of minimally invasive techniques.

Decision analysis is often useful in situations where there is more than one treatment alternative and no data from randomized controlled trials to assist in decision-making. The process of decision analysis explicitly and quantitatively considers both the risks and benefits of all treatment alternatives. In this particular model, the base case considered were patients aged 35 years or 50 years who had recovered from a nonsurgically treated episode of diverticulitis. Four different strategies were considered: elective colectomy after the first, second, third or fourth episodes of diverticulitis. Seventeen different health states were considered: recurrent diverticulitis treated medically or surgically, and if the latter, performed either electively or emergently and with or without a stoma. With each strategy, individuals could recover or die. To determine the probability of each outcome, the authors state that they performed a comprehensive review of the literature and used population-based hospital discharge data from Washington state. One of the poorer aspects of this study is that utilities were not derived directly from individuals with diverticular disease but were derived from limited studies of other similar cohorts of patients. Thus, an assumption was made that individuals with recurrent episodes of diverticulitis are in perfect health between episodes, but this may not be correct. They may have chronic symptoms attributable to the diverticular disease. The utility for postoperative status at 6 months was assumed to be 0.7, but that may not be true if individuals have elective surgery, especially if surgery is performed laparoscopically. Because sensitivity analyses were not performed by varying the utilities, one cannot be certain that the results of the decision analysis would not be altered by different utilities.

On the basis of these assumptions, the authors found that a strategy of surgery after the fourth episode of diverticulitis is preferred. This would result in 0.5% fewer deaths and 0.7% fewer colostomies in individuals over 50 years and 0.1% fewer deaths and 2% fewer colostomies in younger individuals.

Although decision analysis is often used to assist in decision-making when evidence from randomized controlled trials is not available, a decision analysis is limited by the validity of the data used in the analysis. This decision analysis is limited by the fact that most of the data came from retrospective case series. Data concerning the frequency and severity of outpatient-treated episodes of diverticulitis are not available, nor is it known whether quality of life was impaired between episodes of diverticulitis. Because many patients may be treated as outpatients, this may represent a high proportion of patients with diverticulitis whose data are not captured in the data sources used by the authors. Furthermore, some of these episodes may have been severe or complicated even though they were treated outside the hospital.

The authors also failed to consider the impact of laparoscopy, which has become the preferred method for resection of diverticulitis. Furthermore, they limited the surgical options to resection with anastomosis or resection with the Hartmann procedure, but there are other surgical possibilities.

The generalizability of this decision analysis may also be limited, because the results are not applicable to patients with diverticulitis complicated by stricture, fistula, chronic abdominal pain and bleeding. In addition, it may not be pertinent to patients who were treated successfully using percutaneous drainage without surgery for diverticular abscess and localized perforation. Similarly, the results are not applicable to immunocompromised patients with recurring episodes of diverticulitis.

Although this study has limitations, it should cause surgeons to reconsider current guidelines recommending elective surgery after 1 or 2 episodes. Furthermore, the surgical community should be encouraged to perform clinical trials to provide better evidence upon which to base treatment decisions.