Using Nurse-Driven Protocols to Eliminate Routine Gastric Residual Volume Measurements: A Retrospective Study

At Atrium Health Navicent Medical Center, a 650-bed teaching hospital in Macon, Georgia, routine assessment of GRV has remained a mainstay of determining enteral nutrition tolerance. Nutrition protocols in place at this institution called for routine GRV measurements every 4 hours and withholding enteral nutrition for a minimum of 4 hours if volumes exceeded 400 mL. This practice was thought to be one of several (including withholding feedings for patient turns and procedures) leading to inadequate enteral nutrition delivery. An internal survey conducted at the institution in 2016 confirmed this hypothesis, concluding that on average patients received only 55% of estimated necessary intake of enteral nutrition while in the intensive care unit (ICU; Atrium Health Navicent, unpublished data, 2018). Despite updated practice guidelines and verification of inadequate enteral nutrition provision, practices remained unchanged. Administrative and bedside staff attributed this lack of change to deeply ingrained work habits as well as lack of understanding of current evidence-based guidelines. Therefore, the institution sought to adopt the elimination of GRV assessments through implementation of nursing-driven protocols and algorithms. Developing, adopting, and sustaining these protocols are key to ensuring permanent practice change.¹⁰ 

The purpose of this project was to determine the overall impact of implementation of enteral feeding guidelines that eliminate routine GRV assessments on patient care outcomes in an adult ICU. Variables reviewed included quantity of enteral nutrition delivered over the course of the ICU stay as well as incidences of emesis and aspiration pneumonia. Additionally, the study was intended to assess nursing staff compliance with the new guidelines after their implementation.

This evidence-based practice project was conducted in a 14-bed adult medical ICU at a 630-bed, level I trauma, 4-time Magnet-designated facility. The study involved a retrospective pre-post review of medical records over a 60-day period.

Before the intervention, the ICU required assessment of GRV every 4 hours in patients receiving enteral nutrition, with nutrition withheld for a minimum of 4 hours if GRV exceeded 400 mL at 2 consecutive measurements. Aspirated contents were returned to the patient. This task was performed by the nurse assigned to provide care for the patient during a given shift. Patients were fed according to ASPEN/SCCM critical care guidelines, with nutrition care goals of introducing the standard enteral nutrition formula within 24 to 48 hours, starting at a rate of 20 mL/h and advancing as tolerated over the first week to a goal rate according to a weight-based equation. Equations used included the Penn State equation for patients who were receiving mechanical ventilation and the Mifflin St Jeor equation for those who were breathing spontaneously.¹¹  Morbidly obese patients (body mass index [BMI], calculated as weight in kilograms divided by height in meters squared, > 40) were fed according to recommendations as described in ASPEN/SCCM guidelines.²  Nutrition initiation and advancement practices were not altered during the course of this study. All patients were screened for malnutrition within 24 hours of admission using the Malnutrition Screening Tool (MST).¹² 

The new protocol (Figure 1) called for the elimination of routine GRV monitoring but also provided nursing staff guidelines for appropriate measurement of GRV. The newly adopted protocol instructed nursing staff members to monitor for alternate signs of intolerance every 12 hours rather than assessing GRV. Although the definition of gastrointestinal (GI) intolerance varies in the literature,¹³  for the purposes of this study, signs of GI intolerance included abdominal distention, nausea and/or vomiting, and abdominal guarding or discomfort. If any of these signs were present, nursing staff members were instructed to contact the medical staff provider for further direction and next steps. The protocol instructed staff members to measure GRV only for certain patient populations (including general surgery and trauma patients classified as critically ill by the attending physician, those with traumatic brain injury, abdominal surgery patients, and those receiving vasopressors) when a separate order was placed by the medical provider. Although GRV measurement has traditionally been a nursing decision, the requirement for a separate order placed by the medical provider was incorporated into the new protocol to facilitate practice change. In the situation in which GRV monitoring was indicated, it was assessed every 4 hours. If no signs of intolerance were noted, GRV monitoring was stopped after 48 hours. If GRV was found to be elevated (> 400 mL) at 2 consecutive measurements, enteral nutrition was withheld.

The approach to implementing evidence-based practice guidelines of Melnyk et al¹⁴  was used to effect nursing change. This approach consists of 7 steps in which the most effective therapeutic practices as established by the results of well-designed clinical trials are incorporated into current practice. Using this approach was helpful in establishing the new protocol. Cultivating a spirit of inquiry and collecting and analyzing current evidence allowed nursing staff members to integrate current evidence-based practice into their routine nursing care. The final step in the process was evaluating the outcomes of the practice change and disseminating the results.

The process began with provision of informational materials to the ICU staff. The principal investigator (PI; M.S.) educated ICU staff members about the practice change via mandatory in-service training in the designated workroom of the ICU. All in-service training sessions met the hospital’s educational standards. Content for these training sessions included a handout of the guidelines and an algorithm detailing evidence supporting the practice change, a copy of a stop sign (Figure 2) that would be placed above the bed in every patient’s room, and an educational flyer explaining and summarizing the change in practice (Figure 3). Each educational session lasted approximately 5 to 10 minutes. Staff members were then given the opportunity to ask any questions or address any concerns related to the practice change. The PI’s contact information was posted in the unit in case any clinical staff members wanted to ask questions or raise concerns about the practice change during the course of the project.

All nursing staff members employed in the ICU (n = 54) received this education on the new protocol and practice change. The in-service training sessions were repeated on multiple days and at different times to accommodate nursing staff members across all shifts and work schedules. The PI maintained a roster of participating staff nurses in the ICU. After each in-service training session, the participating nursing staff member was given a written explanation of the project and signed an acknowledgment form, which was retained by the PI.

Although the new protocol was implemented across the unit, only patients who received enteral nutrition support for at least 24 hours in the ICU during the preselected 60-day period were included in the study. The preintervention period was designated as the first 30 days of the previously determined time period (August-September 2019), and the postintervention period was designated as the second 30 days (September-October 2019). Patients in the preintervention group continued to receive routine GRV assessments by the nursing staff via direct aspiration, whereas those in the postintervention group were treated according to the newly implemented guidelines and protocol.

Patient medical records were reviewed to collect the following data: patient age, patient admitting diagnosis, BMI, ethnicity, diabetes status, malnutrition status as measured by the MST, type of enteral nutrition access, total volume of enteral nutrition received daily compared with medical provider–ordered volume as documented by nursing staff members, number of incidents of emesis while receiving enteral nutrition, incidence of aspiration pneumonia as diagnosed by the attending physician, length of stay, discharge status, intravenous narcotic administration, and intravenous vasopressor administration. To determine nursing compliance with the new protocol, the medical records were also evaluated for the number of GRV assessments and whether measurements were aligned with the protocol. Data were collected from the electronic medical record using PowerChart (Cerner).

After data collection, all information was coded and entered into IBM SPSS Statistics, version 25, for analysis of descriptive frequencies. The statistical significance level was set at P < .05. Groups were compared using independent-samples t tests for normally distributed continuous variables and Mann-Whitney tests for nonnormally distributed variables. The primary outcome, total volume of enteral nutrition intake, was assessed in terms of mean percentage of ordered volume.

The institutional review boards at Atrium Health Navicent Medical Center and Georgia College and State University, Milledgeville, completed review and granted exempt approval of this quality improvement project. There were no financial expenditures associated with the design and implementation of this practice change.

Data were collected from 25 patients before and 22 patients after the change in GRV measurement practice. At baseline, the groups did not differ significantly in sex, age, ethnicity, BMI, diabetes status, malnutrition risk status as measured by the MST, primary admitting diagnosis, or enteral nutrition feeding route (Table 2). The most common admitting diagnoses were respiratory dysfunction (acute or chronic respiratory failure, acute respiratory distress syndrome), sepsis, acute renal failure, and multiorgan dysfunction. Orogastric feeding was most common in both groups, followed by feeding via percutaneous endoscopic gastrostomy and nasogastric tube. Of those in the preintervention group, 40% and 64% received vasopressors and narcotics, respectively. Of those in the postintervention group, 68% and 32% used vasopressors and narcotics, respectively, with the difference between groups not reaching statistical significance (P = .06).

The primary outcome was enteral nutrition adequacy, defined as total enteral nutrition received as compared with enteral nutrition prescribed measured as a daily percentage over the ICU stay. Mean (SD) enteral nutrition adequacy in the preintervention group was 60% (18%), with a range of 29% to 94%. In the postintervention group, mean (SD) enteral nutrition adequacy was significantly higher, at 81% (12%; P = .002), with a range of 50% to 95% (Figure 4).

Secondary outcomes explored included the occurrence of emesis or pneumonia, length of stay, discharge setting, and adherence to protocol changes. No significant differences in emesis or pneumonia between groups were found (Figure 5). Before the intervention, emesis occurred in 5 patients (20%). After elimination of routine GRV measurements, 1 patient (5%) had an emesis event. Differences in rates were nonsignificant (P = .12). Aspiration pneumonia events were uncommon, occurring in 1 patient before and no patients after the intervention (P = .35).

Before the intervention, the mean (SD) hospital length of stay was 18.5 (18.0) days. After the intervention, the mean (SD) hospital length of stay was 14.9 (16.3) days (Figure 6). The difference was not statistically significant (P = .34). After hospitalization, 28% of the preintervention patients were discharged home and 48% were discharged to a skilled nursing facility. In the postintervention group, 27% were discharged home and 41% were discharged to a skilled nursing facility. Mortality rates were 24% (n = 6) in the preintervention group and 32% (n = 7) in the postintervention group. The discharge outcomes did not differ significantly between the groups.

Adherence to the protocol change was high. Only 8 GRV assessments took place during the postintervention period. Of these, 5 were performed according to protocol (owing to the patient having ileus) and 3 were performed without cause.

In this single-center pre-post study, eliminating routine GRV monitoring was associated with a 21-percentage-point increase of enteral nutrition delivered to critically ill patients without increased occurrence of aspiration or emesis. The most consistent benefit of not monitoring GRV reported in the available literature is a significant increase in achievement of nutritional goals. In a large, multicenter randomized clinical trial (RCT) by Reignier et al,¹⁵  more patients receiving mechanical ventilation met 100% of their goal caloric intake when GRV was not monitored. In another RCT, Ozen et al¹⁶  found that, without GRV monitoring, ICU patients achieved enteral nutrition targets more quickly, with a lower calorie deficit accumulated over 5 days. After cessation of routine GRV monitoring, Poulard et al¹⁷  reported that median daily volume of enteral feeding delivered increased and Bruen et al¹⁸  reported that ICU patients met higher percentages of caloric (93.4% vs 69.6%, P = .01) and protein goals (71.8% vs 55.9%, P = .01). Similar to the present study, Wiese et al¹⁹  reported a 21.4-percentage-point increase in patients meeting more than 90% of goal enteric nutrition volume over the ICU stay (P = .001). Although these results are consistent with those of previous studies, there are notable differences in study design that demonstrate the unique addition of the present work to the literature.

The first novel aspect of this study was that at baseline of this intervention, the routine practice was to measure GRV every 4 hours and withhold enteral nutrition if GRV exceeded 400 mL. Previous studies either stopped enteral nutrition or reduced the enteral nutrition rate at lower thresholds (ranging from 200 to 250 mL).^15–19  Therefore, this work demonstrates that there may be an advantage to eliminating routine GRV monitoring even in institutions currently using higher thresholds for GRV.

Additional unique aspects of this study include the design of the intervention and the measurement of nursing compliance. This intervention followed a standardized approach (the 7 steps of evidence-based practice of Melnyk et al¹⁴ ) to inform the training and inclusion of nursing staff and to document compliance. Compliance was found to be high. It is unclear whether previous studies in this area documented compliance to ensure that routine GRV monitoring had, in fact, been discontinued.

Beyond the primary outcome of enteral nutrition delivery, secondary outcomes were also measured, including emesis and aspiration pneumonia rates, hospital length of stay, and discharge status. We found that emesis and aspiration pneumonia were not more common when routine GRV monitoring was ceased. Multiple studies have also reported no difference in incidence of vomiting events^16,19 ; however, a large RCT by Reignier and colleagues¹⁵  reported more vomiting events when GRV was not routinely monitored. In fact, a meta-analysis of 5 studies including that by Reignier et al¹⁵  indicated that although the incidence of vomiting was significantly higher when GRV was not monitored, the incidence of feeding intolerance was significantly reduced.²⁰ 

Although this study and several others found no clear advantage to routine GRV monitoring, this practice continues to be a common reason for withholding enteral nutrition. Heyland et al²¹  reported audit data from 785 ICUs and found that 74% used GRV values to inform enteral nutrition feeding practices. The reported GRV threshold used to indicate enteral feeding intolerance increased over time and ranged from 50 to 501 mL, with an average of 259 mL. Interestingly, if elevated GRV was the initial reason for enteral feeding intolerance, patients had a statistically significant hazard ratio of 1.4 for having another day with enteral feeding intolerance. Therefore, Heyland et al²¹  concluded that GRV monitoring may be indicated but primarily at higher thresholds.

Limitations of this work should be considered during interpretation of results and used to inform future trials in this area. Although the baseline demographics indicated no significant differences between groups, this research is limited by the observational rather than randomized design. This intervention was conducted at a single institution in a single ICU, and severity of illness was not accounted for, limiting generalizability of the results. As stated previously, definitions of GI intolerance vary and sometimes include symptoms such as diarrhea and abdominal pain.¹³  However, because of inconsistencies in definitions and documentation within the electronic medical record, these symptoms were not included in our analysis. Additionally, diarrhea was not incorporated into the definition of GI intolerance because it may be related to several other variables (eg, Clostridium difficile infections, sorbitol-containing medications, use of antibiotics).

Although anecdotal observations would suggest that elimination of routine GRV monitoring would be a time-saving intervention, we did not conduct a time study to confirm this presupposition. Calorie and protein deficits were also not assessed in our study.

Another limitation of the study is that Sequential Organ Failure Assessment scores were not part of the original analysis. Inclusion of these scores, although not definitive, could have provided further insight into enteral nutrition tolerance of this particular patient population. Although intravenous vasopressor administration was higher for the postintervention group, the difference between the groups was not statistically significant (P = .055). Decreases in enteral nutrition provision can be related to increased intravenous vasopressor administration for some patient populations, although this study indicated higher enteral nutrition delivery despite this difference.²⁰ 

A final limitation is the relatively small sample size of the cohort and the limitation of follow-up to only 30 days after the intervention. Although this limited period reduced the likelihood of other protocol changes, the long-term implementation and compliance with this protocol could not be reported here. Future research involving larger sample sizes is needed and could benefit other institutions that are creating new or modified protocols related to the delivery of enteral nutrition.

In this study, the elimination of routine GRV monitoring in adult ICU patients was associated with higher delivery of ordered enteral nutrition. This higher rate was achieved without increased rates of emesis or aspiration pneumonia or length of stay. These findings indicate that the use of a protocol for practice change as well as mandatory in-service training can effect changes in nursing practice.

The authors are grateful to Debby MacMillan, Sterling Roberts, and Leslie Culpepper for their assistance in the development of this project.