EP576 - Developing Consensus for Colorectal Quality Improvement: A Delphi Study

Image or Table (Do Not add text here):

Purpose/Background: Data are essential to guide meaningful quality improvement (QI) in colorectal surgery. Enhanced recovery programs have been developed to analyze process measures and subsequent outcomes; however, guidelines are heterogeneous and there is a lack of consensus on which variables have the highest impact on QI efforts. Data collection is time- and resource-intensive. The goal of this Delphi study was to determine which data (i.e. process measures such as compliance with preoperative bowel preparation, and outcome measures such as anastomotic leak rate) are most important to colorectal surgeons to guide QI efforts.

Methods/Interventions: A two-round modified Delphi study was conducted using a panel of 34 colorectal surgeon experts with at least 5 years of research experience or years in practice. At least 1/3rd of experts had significant involvement in QI-related research. A total of 63 process and outcome measures were identified through a literature review of prior QI research, and were divided into preoperative, intraoperative, and postoperative process measures as well as outcome measures. In the first round, respondents ranked each measure or outcome on a 9-point Likert scale with “1” representing extremely unimportant (i.e., a measure you have never and would never utilize) and “9” representing extremely important (i.e., a measure you always utilize). In the subsequent round, each expert received summarized group results, was reminded of their prior item scores, and asked to complete an updated questionnaire with new measures identified by the experts as well as any measures that had not previously met consensus or inappropriateness. Group median scores were calculated for each item. Qualitatively, the suggestions and comments of the experts were summarized. Consensus was defined a priori as a median score of ≥7, and a median score of ≤3 was deemed inappropriate. All other scores were deemed indeterminate.

Results/Outcomes: The response rates after the first and second round were 61.8% (21/34) and 90.5% (19/21), respectively. Of the 63 outcome and process measures analyzed in round 1, 43 (68.3%) items reached consensus, and an additional 8 new measures were recommended for inclusion in round 2. At the end of Round 2, a consensus was reached on a total of 52 (73.2%) process and outcome measures, and 3 (4.2%) were deemed inappropriate. The remaining 15 (21.2%) measures, despite re-review, remained indeterminate. Process measures achieving consensus included pre- and postoperative stoma teaching, preoperative bowel preparation, surgical approach, postoperative early mobilization, and several others (Table 1). Outcome measures including anastomotic leak, return to the OR, mortality, and overall morbidity achieved the strongest consensus. Additional outcomes that reached consensus included SSI, DVT, and readmission.

Conclusion/Discussion: This Delphi study identified several key data points deemed important by experts to perform QI research in colorectal surgery. The measures most meaningful to surgeons were outcome measures; however, several process measures (i.e., stoma teaching, surgical approach, and early mobilization) also achieved high levels of consensus. Expert consensus on meaningful data to guide QI efforts in colorectal surgery will help hospitals and providers deploy data collection resources more effectively.