Diet-related complications according to the timing of enteral nutrition support in patients who recovered from out-of-hospital cardiac arrest: a propensity score matched analysis

Ethics Statement

The study design was approved by the Institutional Review Board of Wonju Severance Christian Hospital, Yonsei University (No. CR320098). The requirement for obtaining informed consent was waived by the Institutional Review Board because of the retrospective nature of the study.

Study Design

This retrospective observational study was conducted in patients with return of spontaneous circulation after non-traumatic OHCA and who were admitted to the ICU of Wonju Severance Christian Hospital between October 2019 and June 2020. Inclusion criteria were age ≥19 years and admission to the ICU for post-cardiac arrest care. Patients who did not respond to verbal instructions after initial resuscitation continued to receive proper care in the ICU. Patients who had intestinal hemorrhage, were presumed to have ileus or intestinal obstruction on plain radiography, or had other reasons for an inability to tolerate enteral feeding were excluded.

Definition of Early Nutrition Support

The early nutrition support (ENS) group was defined as the group of patients who started enteral tube feeding within 48 hours after admission. Other cases of enteral nutrition support after 48 hours were defined as the delayed nutrition support (DNS) group.

Study Setting and Data Collection

When the patients with OHCA recovered spontaneous circulation after cardiopulmonary resuscitation in the emergency room, intensive care including targeted temperature management was continued in the ICU. Clinicians tried to prevent fever by continuous temperature monitoring of the patients who could not apply target temperature management for other reasons. A Levin tube was inserted as far as the stomach to provide early nutritional assistance to patients. For patient safety, enteral nutrition was given with the patient's head up and in a 45° angle. It was a continuous feeding method through the nasogastric pathway, and it was applied slowly at an injection rate of lower than 50 ml per hour using an infusion pump. The initial enteral feeding volume in the ICU was implemented within 100–150 ml per day. In addition, a slow infusion method was maintained for a few hours, rather than the bolus injection method [15,16]. At the time of hospitalization, information about the age, sex, weight, and height of the patients was recorded in the electronic medical record, and body mass index (BMI) was calculated based on these data. Moreover, information such as past medical history, first monitored rhythm, basic life support time, and advanced cardiovascular life support time was collected from relatives or paramedics. The length of hospital stay, type and frequency of complications, and neurologic prognosis were also collected from the medical records.

Study Outcomes

The primary outcomes were diet-related complications during intensive care, including aspiration pneumonia, gastric regurgitation, ileus, fever, hypoglycemia, and hyperglycemia after feeding. Aspiration pneumonia was defined as when gastric or pharyngeal substances entered the lower respiratory tract and adversely affected the lungs after resuscitation. The detection of aspiration pneumonia was based on anteroposterior chest X-ray and clinical signs of infection. Ileus was defined as a temporary pause or decrease of intestinal movement. The detection of ileus was based on anteroposterior abdominal radiography. Gastric regurgitation was defined as the reflux of gastric acid contents into the esophagus. Hypoglycemia was defined as a blood sugar level lower than 70 mg/dl, and hyperglycemia was defined as a blood sugar level higher than 200 mg/dl. Such complications, including gastric regurgitation, fever, and hypo/hyperglycemic events, were recorded by observation or opinions of the medical staff during intensive care. The secondary outcomes were survival rate and favorable neurologic status after post-cardiac arrest care. A favorable neurologic state was defined as a cerebral performance category (CPC) score of one or two points.

Statistical Analysis

Continuous variables are described as mean±standard deviation and were compared using the Mann-Whitney U-test based on the normality assumptions from the Kolmogorov-Smirnov test. Nominal data were calculated as a percentage of the frequency of occurrence and compared using the chi-square or Fisher’s exact test, as appropriate. The confounding variable was corrected by propensity score matching. The propensity score was estimated using a logistic regression model. Confusing variables in propensity score matching included sex, age, advanced cardiovascular life support time, targeted temperature management, and use of inotropics. The ENS (n=23) patients were matched 1:1 with DNS patients using greedy matching in caliper 0.2. After the propensity score matching, there was no significant difference in sex, age, advanced cardiovascular life support time, targeted temperature management, and use of inotropics between the groups. To compare the clinical complications between the ENS and DNS groups, a Poisson regression model was applied, and the relative risk and 95% confidence interval of each complication, survival rate, and neurologic prognosis were quantified. A P-value <0.05 was considered to indicate statistical significance. A certified statistician conducted all statistical analyses using SAS analytics software version 9.4 (SAS Institute, Cary, NC, USA).

Baseline Characteristics

Between October 2019 and June 2020, 110 OHCA patients were treated in the emergency department. Sixteen patients aged <19 years or with non-medical etiologic cardiac arrest were excluded. In addition, seven patients were excluded because of inadequate information about enteral nutrition (Figure 1). Of the 87 patients, 39 were classified into the DNS group and 48 into the ENS group. Propensity score matching was used to select subjects with similar major characteristics between the two groups, and a total of 46 patients were enrolled.

Figure 1.

Study enrollment chart. OHCA: out-of-hospital cardiac arrest; ED: emergency department; CPR: cardiopulmonary resuscitation.

The mean age of the patients was 67.4 years, and 33 (71.7%) were men. The mean BMI was 22.6 kg/m² and the Acute Physiology and Chronic Health Evaluation (APACHE) II score was 19.2. Patients had the following pre-arrest comorbidities: hypertension, 22 patients (47.8%); diabetes, 12 (26.1%); chronic kidney disease, 8 (17.4%); malignancy, 3 (6.5%); acute coronary syndrome, 4 (8.7%); chronic obstructive pulmonary disease, 3 (6.5%); cerebrovascular disease, 3 (6.5%); pulmonary thromboembolism, 1 (2.2%); and hyperlipidemia, 7 (15.2%). There was no difference in prior comorbidities between the two groups. There was no significant difference between initial rhythm (P=0.491), basic life support time (P=0.060), advanced cardiovascular life support time (P=0.888), inotropic use (P=1.000), and other laboratory findings. The mean feeding volume for the initial 5 days in the ENS group was significantly higher than in the DNS group (1,363.0 ml vs. 100.9 ml) (Table 1).

Table 1.

Baseline characteristics of the enrolled patients before and after 1:1 propensity score matching

Primary Outcomes

The DNS and ENS groups had similar ICU hospitalization day stays (5.9 vs. 9.7, P=0.052). The APACHE II score was also statistically similar between the DNS and ENS groups (P=0.209). Regarding complications caused by nutrition support in the ENS group, aspiration pneumonia was present in four cases, gastric regurgitation in two, ileus in none, fever in 10, hypoglycemia in 8, and hyperglycemia in 11 cases. However, there was no significant difference in clinical complications between the ENS and DNS groups (Table 2).

Table 2.

Clinical outcomes according to timing of enteral nutrition before and after 1:1 propensity score matching

Secondary Outcomes

The survival-to-discharge rate was higher in the ENS group than in the DNS group (86.9% vs. 39.1%, P<0.001). Among the CPC scores 1 and 2 between the ENS group and the DNS group, 13 patients (56.5%) were in the ENS group, and four patients (17.4%) were in the DNS group. There were significantly more patients with a good neurologic prognosis in the ENS group (Table 2).

Relative Risk of the Clinical Complications According to Timing of Enteral Nutrition

The researchers calculated relative risk by checking for the presence of complications according to the timing of enteral nutrition. Comparing the ENS group with the DNS group, the relative risk ratios for major complications were similar: aspiration pneumonia 0.95-fold, gastric regurgitation 1.09-fold, ileus 0.96-fold, fever 0.92-fold, hypoglycemia 1.20-fold, and hyperglycemia 1.17-fold. The occurrence of secondary outcome-related variables showed a higher relative ratio of 1.90 for good neurologic recovery and 4.67 for survival in the ENS group (Table 3). In addition, a comparative subgroup analysis was performed for 35 patients without initial shockable rhythms; there was also no significant difference in diet-related complications during post-resuscitation ICU care (Supplementary Tables 1-3).

Table 3.

Relative risk analysis of the patients with early enteral nutrition compared to those of delayed enteral nutrition