Should I exclude postoperative adhesion outcomes from my planned network meta‑analysis and focus only on anastomotic leak and length of hospital stay?

Should You Exclude Postoperative Adhesion Outcomes from Your Network Meta-Analysis?

Yes, you should exclude postoperative adhesion outcomes and focus exclusively on anastomotic leak and length of hospital stay for your network meta-analysis. These outcomes are more clinically meaningful, better reported in the literature, and directly impact morbidity, mortality, and quality of life in ways that are measurable and actionable.

Rationale for Exclusion

Adhesions Are Poorly Suited for Network Meta-Analysis

Postoperative adhesions present significant methodological challenges that undermine their utility in network meta-analysis:

• Inconsistent detection and reporting: Adhesions are often clinically silent and only discovered incidentally during reoperation or imaging for other indications. The 2018 WSES guidelines on adhesive small bowel obstruction 1 note that while adhesions cause 60% of small bowel obstructions and affect a substantial proportion of patients, their detection is highly variable and depends on whether patients undergo subsequent surgery.

• Long latency period: Adhesions may not manifest clinically for months to years after surgery, making short-term perioperative studies inadequate for capturing this outcome. Most surgical trials focus on 30-90 day outcomes, missing the true incidence of clinically significant adhesions.

• Lack of standardized measurement: Unlike anastomotic leak (which has clear International Study Group criteria) or length of stay (objective measurement), adhesion assessment lacks uniform grading systems across studies, making cross-study comparisons unreliable.

Anastomotic Leak and Length of Stay Are Superior Outcomes

Anastomotic leak directly impacts mortality, morbidity, and quality of life:

• The 2024 ECCO guidelines on Crohn's disease surgery 2 demonstrate that anastomotic leak rates are consistently reported across surgical techniques, with stapled side-to-side anastomoses showing OR 0.45 (95% CI: 0.20-1.00) compared to hand-sewn techniques.

• Recent evidence 3 confirms AL is associated with increased mortality (19% in one cohort), prolonged hospitalization, and impaired oncologic outcomes in colorectal surgery.

• The 2023 ERAS guidelines for emergency laparotomy 4 emphasize that AL is "a life-threatening major complication" with well-documented risk factors and scoring systems, making it ideal for comparative effectiveness research.

Length of hospital stay serves as a composite measure of recovery:

• The 2022 ERAS guidelines for low-middle income countries 5 consistently cite length of stay as a key outcome across multiple interventions, with clear associations to complications and recovery.

• The 2024 ECCO guidelines 2 note that while anastomotic technique affects overall complications, "the choice of anastomosis technique did not seem to affect...length of hospital stay" in some analyses, making it a useful discriminator between interventions.

• Multiple network meta-analyses 6, 7, 8 successfully use length of stay as a primary outcome with consistent reporting across studies.

Practical Considerations for Your Network Meta-Analysis

Data Availability and Quality

The PRISMA-NMA extension 9 emphasizes that network geometry and connectivity depend on consistent outcome reporting. Adhesion data will likely:

• Create sparse networks with missing connections between treatment nodes
• Introduce heterogeneity from inconsistent follow-up periods and detection methods
• Reduce statistical power due to underreporting

In contrast, anastomotic leak and length of stay are:

• Reported in virtually all surgical RCTs
• Measured during the index hospitalization with standardized definitions
• Suitable for both direct and indirect comparisons

Clinical Relevance

From a clinical decision-making perspective:

• Anastomotic leak informs immediate surgical technique choices (stapled vs. hand-sewn, anastomotic configuration) and perioperative management (diversion strategies, drain placement)

• Length of stay reflects overall recovery, complication burden, and healthcare resource utilization—all critical for guideline development and implementation

• Adhesions, while important long-term, are influenced by factors beyond the index operation and are better addressed through separate long-term cohort studies rather than perioperative intervention trials

Alignment with Current Guidelines

The 2017 ESPEN guidelines on clinical nutrition in surgery 10 and 2022 ERAS guidelines 5 consistently prioritize anastomotic leak and length of stay as primary outcomes when evaluating surgical interventions. The 2021 guidelines on elderly rectal cancer patients 11 similarly focus on anastomotic leak rates, complications, and hospital stay rather than adhesions when comparing surgical approaches.

Common Pitfalls to Avoid

• Do not attempt to include adhesions as a secondary outcome "for completeness"—this will introduce bias if only selectively reported studies contribute data
• Do not conflate postoperative ileus (short-term, well-reported) with adhesive complications (long-term, poorly reported)
• Do ensure your protocol clearly defines anastomotic leak using International Study Group criteria to maintain consistency across included studies
• Do consider including surgical site infection as an additional outcome if data permit, as it's well-reported and clinically relevant 11, 5

By focusing on anastomotic leak and length of hospital stay, your network meta-analysis will have greater statistical power, clinical applicability, and methodological rigor than if diluted with poorly reported adhesion outcomes that require different study designs to adequately assess.