Implementation of an enhanced recovery after surgery (ERAS) protocol for total abdominal hysterectomies in the division of gynecologic oncology: a network-wide quality improvement initiative

Methods

Our QI project was considered exempt through the St. Luke’s University Health Network’s Institutional Review Board (Local ID SLIR 2021-73). The Departments of Gynecologic Oncology and Anesthesia developed a network-specific ERAS protocol derived from previously reported literature (Figure 1). The protocol focused on measures that would benefit the patient during the pre-, intra-, and postoperative periods. The first component of the QI initiative consisted of interdisciplinary meetings to discuss the implementation of the ERAS protocol with preoperative nurses, postanesthesia care unit (PACU) nurses, and floor nurses. Standardization of the ERAS protocol was achieved through creation of an order set within the electronic medical record (EMR) system. The order set was built with input from the Departments of Pharmacy, Information Technology (IT), and the QI committee within the gynecologic oncology department. An opioid conversion scale was utilized to convert intravenous and oral opioid medications into OMEs (Figure 2). A chart review was performed to assess compliance with the ERAS protocol.

Figure 2:

Opioid conversion chart utilized in our ERAS protocol for TAHs within the Division of Gynecologic Oncology. ERAS, Enhanced Recovery After Surgery; IV, intravenous administration; mcg, micrograms; mg, milligrams; OME, oral morphine equivalent; PO, oral administration; TAH, total abdominal hysterectomy.

The primary outcome of interest was postoperative opioid consumption. OMEs were utilized as the measure for which the pre- and postopioid consumption analysis was conducted. Secondary outcomes of interest included the ERAS protocol compliance, length of stay, and cost. Patient data were collected retrospectively. All patients who underwent a total abdominal hysterectomy (TAH) from 2017 to 2019 by a gynecologic oncologist (IZ, NT, RB, SP, AG) in our hospital network were included. The same clinical team cared for the patients in both the pre- and post-ERAS groups. Statistical analysis was performed utilizing SPSS version 26 (IBM Corp, Armonk, NY). Unadjusted comparisons between treatment groups utilizing chi-square tests or Fisher’s exact tests were utilized, as appropriate. Mann Whitney U tests were utilized to assess length of stay, postoperative opioid requirements, and total cost variables between groups. For all comparisons, p<0.05 denoted statistical significance, with no adjustment for multiple testing. If the p value was >0.05 and <0.09, this was considered a trend toward significant.

Results

After conducting our QI initiative, a total of 183 patients were included in the final analysis, with 59 in the non-ERAS cohort and 124 in the ERAS cohort. The non-ERAS cohort consisted of patients who had a TAH performed in 2017, whereas the patients who underwent a TAH during 2018 or 2019 were considered part of the ERAS cohort after implementation. Demographically, the mean age of ERAS patient was 59 (37–84) years old, whereas the mean age of the non-ERAS patients was 57.9 (26–83) years old. Among the ERAS patients, 83.1 % are white and 54.8 % are married, compared to 86.4 and 50.4 %, respectively, of non-ERAS patients (Table 1). Of note, the participants in the study were classified into their racial categories by the predetermined options on hospital admission. They were self-identified and not designated by the research team. We recognize that racial and ethnic distinctions are not considered absolute, and this is a limitation of our study.

The median OME consumption among the non-ERAS cohort was 56 OMEs (range, 8–137) compared to 40 OMEs (range, 0–129) among the ERAS cohort. This was statistically significant (p=0.03). Figure 3 stratifies the changes in opioid consumption based upon the OME of the opioids consumed. Prior to implementing the ERAS protocol, there was an equal distribution in the proportion of patients receiving 0 to 40 OMEs (34 %), 41 to 60 OMEs (15 %), 61 to 90 OMEs (34 %), or >90 OMEs (17 %) postoperatively. In comparison, the majority of patients received less than 60 OMEs after ERAS protocol implementation (69 %).

Figure 3:

OME requirements during the postoperative period for the non-ERAS and ERAS cohorts. 0–40=minimum opioid usage; 41–60=moderate opioid usage; 61–90=severe opioid usage; >90=excessive opioid usage. ERAS, Enhanced Recovery After Surgery; OME, oral morphine equivalent.

The median length of stay decreased from 5.18 days (range 2–21 days) to 4.17 days (range 1–16 days) after ERAS implementation. Although this measure was not statistically significant (p=0.07), there is a shorter trend for the ERAS group. No statistically significant difference in cost was noted between the groups, with a median cost of $13,342.00 in the non-ERAS cohort and $13,703.00 in the ERAS cohort (p=0.8).

Of the 19 endpoints that assessed ERAS protocol compliance, eight showed statistically significant differences between the two cohorts (Table 2). Two categories could not be analyzed secondary to insufficient data. There was a significant increase in preoperative carb drink (0–82, p<0.0001), preoperative nonnarcotic pain medication utilization (5–55, p<0.0001), usage of surgical site infection (SSI) measures (35–113, p<0.001), preoperative deep vein thrombosis (DVT) prophylaxis (59–114, p=0.03), epidural or tap block utilization (29–104, p<0.001), proportion of patients receiving postoperative Tylenol (acetaminophen, 36 to 122, p<0.001), intravenous fluids (IVFs) stopped on postoperative day 1 (12–52, p=0.01), and diet advanced by postoperative day 1 (38–112, p<0.001). A compliance rate of >90 % was seen in 7 of the 19 endpoints.

Discussion

ERAS protocols for TAHs within the Gynecologic Oncology division have previously been shown to be highly efficacious in reducing length of stay, postoperative opioid consumption, and overall cost in single-center academic and community hospital settings. Nelson et al. [9] demonstrated that early feeding, maintaining euvolemic status, and providing multimodal analgesia have decreased the rate of ileus in women with complex open gynecologic cancer surgeries. Jimenez et al. [7] implemented an ERAS protocol at a large, intercity community hospital and utilized gabapentin and multimodal analgesia to reduce postoperative opioid usage. Results demonstrated significant reduction in opioid usage with a mild elevation in postoperative pain scores and no change in length of stay [7]. Mendivil et al. [4] conducted a similar initiative, but unlike the insignificant difference in hospital stay seen in the initiative conducted by Jimenez et al. [7], they were able to demonstrate a 3-day mean reduction in hospital length of stay with an approximately $2,500 per patient reduction in cost and a reduction in readmission rates by 2 % [4]. Our work differs from the previously mentioned studies in that an ERAS protocol was implemented for gynecologic oncology TAHs within our community-based, multi-hospital network that consisted of both urban and rural settings. Our results were as successful as previously published studies regarding reducing postoperative usage of opioid medications from 48 to 34 %. Unfortunately, we were unable to demonstrate significant change among length of stay or cost. The most significant contribution that our work provides is that an ERAS protocol can be created and standardized through a large community-based hospital network. This was shown in our work, with an overall high compliance rate of >90 % compliance in 7 of the 19 endpoints. Overall, this lends to the ability to implement higher standards of care within a large network of community-based hospitals, an aspect in healthcare improvement that has not been robustly depicted in the previous literature.

One aspect of our success in initiating this protocol can be attributed to the multidisciplinary team approach, which consisted of contributions from Pharmacy, IT, QI, the Department of Gynecology, and the Department of Anesthesia. It has been shown that the simple practice of involving others in the design and implementation of a QI initiative can improve adherence to clinical practice guidelines. Jun et al. [10] utilized this concept to demonstrate that nurses were more willing to following clinical practice guidelines when they were included in the implementation of such guidelines. The collaboration in the design and build of the order set led to each respective department taking a personal level of ownership in the project, which can therefore increase compliance rates. Along with incorporating a multidisciplinary approach, another potential strength of our study involved the use of an EMR to partially automate the QI initiative. Electronic order sets have been shown to be a significant driver of clinician decision making. By changing the default setting from “optional” to “preselected,” Olson et al. [11] was able to significantly increase the rate of posttransfusion platelet counts ordered from 7 % of platelet transfusions to 59.4 % (p<0.0001). Another example seen was a computer-based algorithm in an EMR that provided a sepsis-specific workup and resuscitation plan for patients at risk of sepsis in the emergency department; it was capable of increasing bundle compliance by 154 %, increasing compliance from 28 to 71 % [12]. Change should continue to be driven through EMR order sets to make adherence to new protocols as easy as possible.

Even in the presence of several important strengths, several limitations of our study must be acknowledged. The first major limitation includes the small sample size. Increasing the sample size by including data from previous years, as well as including further data points from postimplementation, may have revealed more significant secondary outcomes. Our study was only conducted the year following implementation and may have failed to capture statistical trends as compliance rates further increased. Another limitation is the unplanned nature of some TAHs. Conversion from a minimally invasive surgery (MIS) approach to TAH usually involves radical surgical intervention from complication or cancer extent. These patients would likely require a longer length of stay and require higher opioids secondary to longer operative times and more extensive dissection, and thus may have skewed the data.

One future direction could be to stratify the unplanned TAH cohort in order to establish the effect a conversion from an MIS approach to a TAH that may play on the same endpoints. Another future direction could include a more targeted approach to further reduce length of stay, which could subsequently reduce overall cost [13]. Lastly, more in-depth multidisciplinary approaches could increase compliance and secondary endpoints further, such as utilizing physical therapy services to help improve postoperative ambulation. After demonstrating the success of the ERAS protocol for TAHs done within the Gynecologic Oncology division, we aim to implement the protocol to other surgical procedures within the Department of Gynecology and surgical departments.

The osteopathic philosophy highlights the body as a whole, with each component being independently important in healing. By utilizing ERAS protocols to optimize each organ system to work at its fullest capacity, such as minimizing nasogastric tubes, early ambulation, and early feeding, the ERAS protocol aligns with the tenet of the body as self-healing as we optimize function as a whole to heal from surgery. Two papers (one by Goldstein et al. [14] and the other by Martingano et al. [15]) showed that osteopathic manipulative treatment (OMT) is safe in the postoperative period as well as in cases of gynecologic-related cancer. Although these papers did not report a significant drop in opioid use postoperatively, there was a trend toward lower opioid use. Another future direction would be to incorporate OMT into the ERAS protocol for further study in opioid reduction.

Overall, the previous literature has shown feasibility in the implementation of ERAS protocols among single institutions [5]. However, unlike previously established studies, we were able to successfully implement this protocol across an entire hospital network. At our multicenter institution, we utilized an ERAS protocol within an EMR order set to increase the ease of use across the network while simultaneously increasing compliance. By implementing this order set, we had a significant reduction in opioid use, and a nonsignificant but clinically relevant reduction in length of stay and hospital cost.

Conclusions

Gynecological oncology ERAS protocols for TAHs are effective in reducing postoperative opioid consumption, length of stay, and cost. Most of the published work on TAH ERAS initiatives on QI has only been demonstrated in single centers, and given that upwards of 84 % of hospitals across the United States are community based, it is critical to demonstrate that such an initiative can be implemented in a large-scale community setting with similar levels of excellence in postoperative recovery, opioid use reduction, and decreased costs. By utilizing a multidisciplinary approach and an EMR order set to implement an ERAS protocol for TAHs, we were able to demonstrate that the same levels of excellent postoperative recovery seen at single medical centers can be upheld within a large-scale multi-hospital network consisting of both urban and rural settings.