Article 1
Sigmoid resection with primary anastomosis versus the Hartmann’s procedure for perforated diverticulitis with purulent or fecal peritonitis: a systematic review and meta-analysis. Lambrichts DPV, Edomskis PP, van der Bogt RD, Kleinrensink GJ, Bemelman WA, Lange JF. Int J Colorectal Dis. 2020 Jun 5.
Summary:
This was a systematic review of four scientific databases comparing sigmoid resection with primary anastomosis (+/- diverting ileostomy) to Hartmann’s procedure in the setting of Hinchey grade III or IV disease. Embase, MEDLINE, Cochrane, and Web of Science were queried with 1560 articles identified. After deduplication and exclusion criteria application, there were 4 RCTs and 10 observational studies eligible for inclusion.
Primary anastomosis (PA) was found to be better than Hartmann’s procedure (HP) in terms of stoma reversal rates and related morbidity, but there were no differences in mortality, morbidity, or reintervention rates between PA and HP. The authors also attempted to include cost-utility analysis and patient reported outcomes, but the data was insufficient to perform a meaningful analysis. There was also limited data on Hinchey III vs. IV disease.
Based on these findings, the authors conclude that PA seems to be preferred in selected patients with Hinchey III or IV diverticulitis.
Critique/assessment
The systematic review is done appropriately, with attempts to analyze RCTs and Observational studies separately, but still struggles to gain sufficient information to make a strong conclusion. They appropriately note that between-study heterogeneity is significant, not only among the observational studies, but among the RCTs. For instance, 2 of the 4 RCTs specifically excluded hemodynamically unstable patients, which introduces selection bias and makes the results, and randomization happened at different points of the clinical course. The random effects model is appropriate in this circumstance, but cannot overcome the inherent bias that is included in all of the studies. Additionally, almost all of the patients included in the review were Hinchey III, with very little data on Hinchey IV, and many of the observational studies were not clear on which patient with PA may or may not have had a divert ileostomy. Because of recruitment difficulties in the RCTs, they are likely underpowered, even in aggregate as a meta-analysis.
Predictably, PA had more reversals and fewer overall stomas, since the stomas were ileostomies rather than colostomies, and indication for reversal/readiness is different between the two. There is insufficient data to sort out initial selection bias and confounding by indication for both the primary and subsequent procedures.
Other endpoints of mortality and morbidity do not show a significant difference – of note, they state that mortality is lower for PA in the observational studies, but their pooled statistic is not valid, as the point estimates for the studies should not be pooled when they fall on both sides of the null hypothesis. Thus none of the Forest plots in Figures 2 or 3 should be pooled.
Take-home point
PA may be preferable for selected patients, but we do not know who these selected patients might be. A well-designed prospective multi-center study needs to be done to minimize clinical and methodological heterogeneity.
Article 2
Sex differences in the treatment and outcome of emergency general surgery. Rucker D, Warkentin LM,
Huynh H, Kadaroo RG. PLos One. 2019 Nov. 2019 Nov 4;14(11):e0224278.
Summary:
This was a post-hoc analysis of patients from the Elder-Friendly Approaches to the Surgical Environment (EASE) study, reviewing 512 patients admitted to the emergency general surgery services at two tertiary care hospitals in Canada (322 patients over 65, and 190 younger patients from the EASE excluded cohort). Inclusion criteria was having an emergency abdominal operation, but transfers and palliative operations were excluded. Objective was to investigate differences in pre-, intra, and post-operative care and outcomes between men and women.
When examining the patients excluded from surgical intervention, older women were significantly more likely to undergo non-operative management than men, in a system with presumably equal access to care. Once the decision was made for surgery, however, there were no significant differences in time to surgery, time in the OR, hospital LOS, discharge disposition, or mortality. The only difference was in regard to respiratory complications, which were higher in men.
They concluded that differences in outcomes of surgical disease are less related to biologic factors, and more likely related to decisions regarding seeking care or being referred for surgical care prior to emergent presentation.
Critique/assessment
As a post-hoc analysis of a particular cohort (those aged greater than 65), they are potentially limited since studies showing sex differences reveal that this is age-related to some extent. Because they had extensive information on all patients otherwise eligible for EASE but excluded by age, they were able to create a reasonably balanced dataset.
The strengths of this study are that it was done in a universal healthcare system, with presumed equal access to care, which limits confounding by insurance status. Additionally, there was extensive, uniform data collection in the context of the original study, making the data robust for analysis.
The authors acknowledge, however, that once a decision is finally made for surgery, there are minimal outcome sex differences. This means that everything that happens prior to the urgent presentation impacts outcomes, as does the discussion surrounding whether surgery should be offered or not, and neither of these points are addressed with this study. Thus we are left with minimal additional information as to where we can actually intervene, and the continuing charge to look beyond insurance and biology to explain differences in how we care for people.
Take-home point
Surgical disparities in patients presenting with emergent disease are unlikely to be addressed by biology or interventions to simply increase insurance coverage. Social determinants of health beyond insurance status must be explored and addressed proximate to urgent presentation of surgical disease.
