Adherence to 20 Emergency General Surgery Best Practices

Results of a National Survey

Angela M. Ingraham, MD, MS; M. Didem Ayturk, MS; Catarina I. Kiefe, MD, PhD; Heena P. Santry, MD, MS


Annals of Surgery. 2019;270(2):270-280. 

In This Article


Data Source

We conducted a national survey to determine variations in the delivery of EGS care across all US hospitals where an adult with a general surgery emergency might seek care (selection criteria below). The questionnaire was created using an iterative process informed by prior qualitative research, a pilot survey, and the expert panel convened by American Association for the Surgery of Trauma in 2014 to develop EGS quality indicators.[14–16] Early versions of the questionnaire were tested by surgeons (N = 26) who perform EGS at academic and community-based hospitals but who would not be targeted for final survey implementation. Questionnaire items included quality indicators as determined by our expert panel[14] (Appendix 1, or questions introduced into preliminary versions based on the final query "Is there anything else we should include in this survey?" that tested well in subsequent iterations as an important measure of EGS delivery. The questionnaire also queried the hospital's approach to EGS care by asking respondents to choose from "A dedicated clinical team whose scope encompasses EGS (±trauma, ±elective general surgery, ±burns)," "A traditional approach with an ad hoc 'general surgeon on call' schedule," or "Other (please specify)."

In this manuscript, we present a select combination (N = 20) of quality indicators and best practices that are broadly applicable to EGS care. A copy of the questionnaire can be found in Appendix 2, Hereafter, we utilize the term "best practices" to refer to all 20 measures detailed in this manuscript. Specific practices related to the structure of EGS teams, ancillary hospital services, and human resources are addressed in other manuscripts.

Hospitals to target were identified using 2 criteria. First, the 2013 American Hospital Association (AHA) Annual Survey of US hospitals was accessed to identify acute care general hospitals (nonfederal, short-term general, and other special hospitals including academic medical centers or other teaching hospitals) accessible to the general public (ie, excludes prison hospitals and college infirmaries) (N = 6356).[17] Second, these hospitals were assessed for the presence of an emergency room (ER) and at least 1 operating room (OR). The 3331 acute care general hospitals identified were assessed for eligibility using a grassroots approach that included Internet searches and direct outreach to hospital chief medical officers (CMOs) and ER staff. Through this process, we excluded 469 hospitals (14.1%) that did not provide EGS services (24-hour ER access and at least 1 OR) and 9 hospitals in Puerto Rico, leaving us with a final sample of 2844. After questionnaire implementation, an additional 33 hospitals responded that they did not have 24-hour ER access and/or at least 1 OR (contrary to what was indicated in the AHA database); thus, a total of 2811 hospitals remained in the final sample.

Through our grassroots approach, we simultaneously identified a single respondent at each hospital who was responsible for oversight of EGS care at that hospital using the algorithm shown in Figure 1 and created a database including mailing addresses and email addresses if accessible. Owing to the historically low response rate among surgeons and our target of a minimum 50% response, we used the same algorithm (excluding the position of Chief of ACS) to generate a list of secondary (back-up) respondents and their contact information. At hospitals in which we could only identify a single surgeon, we listed the hospital CMO as the secondary respondent. Overall, 97.8% of potential respondents had valid postal mailing addresses; 60.5% of potential respondents had valid email addresses.

Figure 1.

Algorithm utilized to identify a single respondent who was responsible for oversight of emergency general surgery coverage at hospitals in the United States where an adult with a general surgery emergency might seek care. CMO indicates Chief Medical Officer.

We then employed a hybrid approach to survey implementation. We first targeted primary respondents simultaneously with a paper questionnaire in an expense paid envelope as well as an electronic solicitation. An upfront incentive was included in the first questionnaire mailing. If a response was not received after the first mailing/electronic solicitation, 2 electronic reminders were sent to all those with a valid email address reminding them to fill out the paper questionnaire they had already received or to respond electronically. Those lacking a valid email address were sent a second paper questionnaire (without incentive) to encourage participation in the survey. The first survey implementation occurred from August 13 to October 26, 2015. Our overall response rate was 41.8%; therefore, we used the same survey implementation method (beginning with a paper questionnaire with upfront incentive and electronic solicitation followed by either 2 electronic reminders or a follow-up paper questionnaire by postal mail) for secondary surgeons or CMOs for all hospitals that did not have a questionnaire response by November 1, 2015. The second survey implementation occurred from November 2 to December 22, 2015. By the end of 2 unique survey implementations, there were 1690 responses from the 2811 eligible acute care hospitals, representing a total response rate of 60.1%. More than 95% of the hospitals (N = 1610) in our sample had data from surgeons who answered our questionnaire items, whereas 4.7% (N = 80) had data from CMOs who answered our questionnaire items.

Statistical Analysis

Questionnaire responses on adherence to EGS best practices were tabulated and compared by hospital characteristics (geographic region, ownership type, hospital location, teaching status, inpatient bed capacity, medical school affiliation, and trauma certification as reported by the AHA in 2015 and self-reported ACS model adoption) using univariate comparisons (Chi2). The independent effect of any given hospital characteristic on adherence to a specific EGS best practice was determined using multivariable logistic regression. Hospital characteristics found to have P < 0.20 in univariate comparisons were included in the model.

Finally, subanalyses were conducted using the same multivariable methods described above to determine what effect formally adopting an ACS model encompassing EGS care would have on adherence to EGS best practices. For these analyses, we compared hospitals that reported an ACS model to hospitals that reported a traditional general surgeon on call model during the year before the questionnaire completion. Ninety-nine hospitals that did not report a full year of an ACS model or had a response other than general surgeon on call model and 94 additional hospitals that did not report a specific model adoption were excluded from these analyses.

All analyses were performed using SAS 9.4 (SAS Institute, Cary, NC, 2015). This study was reviewed and deemed exempt by the University of Massachusetts Institutional Review Board.