In 1996, the mean hospital stay for women who underwent cesarean deliveries was 3.1 days.1 Hospital discharge depends primarily on the absence of febrile morbidity and the return of normal bowel function. Traditionally, a patient who has undergone a laparotomy is prescribed a graduated dietary regimen to decrease the likelihood of developing a postoperative ileus. Initially, the patient is given “nothing by mouth” for 12–24 hours, advanced to a clear liquid diet on postoperative day 1, and then given to a full liquid diet on postoperative day 2. Solid food is prescribed if the patient tolerates the full liquid diet, or passes flatus.2,3 Rarely, when a patient is suspected to have severe ileus or bowel obstruction, dietary modification or abdominal radiographs may be required.
Because most cesarean deliveries are performed under regional anesthesia, require little intestinal manipulation, and involve younger patients, some investigators have suggested that women who have undergone cesarean can safely receive solid food as early as 4–8 hours after surgery.4,5 These investigations have not shown that early feeding has impacted the length of the hospital stay. Our study was designed to evaluate the rate of mild ileus symptoms and hospital course of early-fed patients as compared with those placed on a traditional feeding schedule.
MATERIALS AND METHODS
Women admitted to the labor and delivery area and those scheduled for elective cesarean were recruited for the study. The Institutional Review Board at our hospital approved the study protocol, and written informed consent was obtained from each participant. Patients who underwent cesarean delivery under regional anesthesia between February 1, 1999, and March 31, 2000, were enrolled. Subjects were excluded if they received general anesthesia, magnesium sulfate, underwent intra-operative bowel surgery or had a bowel injury, or had any gastrointestinal and/or medical conditions that precluded the early consumption of solid food. Assignment to one of the two study groups, “early feeding” or “traditional feeding,” was accomplished by a computer-generated random number list before placing the group name inside consecutively numbered, opaque envelopes. The list was then secured until completion of the study. The sealed envelopes were placed in the recovery room so that neither the surgeon nor the other operating room personnel were aware of the study assignment.
Women assigned to the early feeding arm were offered solid food within 8 hours of surgery. Traditionally fed women were given nothing by mouth for 12–24 hours after surgery, and then advanced to a clear liquid diet on postoperative day 1. On postoperative day 2, if the clear liquid diet was tolerated and the woman reported passage of flatus or having a bowel movement, a regular diet was given. If she tolerated a clear liquid diet but did not report passage of flatus, a full liquid diet was given. On postoperative day 3, if she tolerated the full liquid diet, a regular diet was prescribed. Patients were eligible for hospital discharge if they were able to tolerate solid food without emesis, passed flatus or had a bowel movement, and demonstrated no febrile morbidity for at least 24 hours.
The time of onset of surgery was designated as zero hour. The day of surgery was considered to be the first 24 hours; the second postoperative day encompassed hours 24–48, and the third postoperative day consisted of hours 48–72. The operative time was defined as the time from the onset of the surgery to the completion of skin closure. The length of hospital stay was defined as the time interval between the onset of surgery to hospital discharge. Mild ileus symptoms included symptoms of anorexia, abdominal cramping, or nonpersistent nausea and/or vomiting. A severe ileus was defined as abdominal distention, greater than three episodes of emesis in a 24-hour period, and the inability to tolerate oral liquids; patients who required a nasogastric tube or abdominal radiographs were also designated as having a severe ileus. Postoperative febrile morbidity was considered to be oral temperatures equal to or exceeding 100.4F on two or more occasions, at least 6 hours apart, occurring greater than 24 hours after the surgery.
The patients were examined at least twice a day by a third or fourth year obstetrics-gynecology resident, and when possible, they were re-examined by the same physician. Each resident was instructed to elicit and document the outcome information in an unbiased and uniform fashion, using a standard questionnaire.
Outcomes assessed included the symptoms and signs of mild and severe ileus, length of hospital stay, postoperative time interval to bowel movement, febrile morbidity, and hospital readmission. Other statistical investigations should be considered exploratory. Categorical data were analyzed by χ2. Continuous data were analyzed with two-tailed tests. Normally distributed data were reported by mean and standard deviation measurements, and the t test was chosen to compare means. When data were not normally distributed, median and interquartile ranges were reported, and the Mann Whitney U test was used. True Epistat Software (Epistat Services, Richardson, TX) and SPSS 10.0 (SPSS, Inc., Chicago, IL) were used, with a P value ≤ 0.05 chosen as significant.
The primary outcome parameter was the rate of mild ileus symptoms. Experience from a pilot study revealed that 19% of patients delivered by cesarean developed mild ileus symptoms. We calculated that to detect twice the rate of ileus symptoms of 38% with an alpha error of 0.05 and a power of 0.80, a total of 108 patients would be required. A target sample of 124 patients was established to account for attrition or protocol breaches.
A total of 124 women were enrolled into the study, two of whom later refused to participate, and two of whom were excluded because of incomplete documentation of outcome parameters. Sixty women were assigned to each dietary regimen. Women in both groups had similar demographic characteristics, cesarean indications, intra-operative findings, mean intra-operative blood loss estimations, and mean durations of surgery (Tables 1–3). Women in the control arm used more oral analgesic medication postoperatively per 24-hour period, but this was not statistically significant (4.9 ± 4.3 tablets versus 4.1 ± 3.8 tablets).
Compared with those in the traditional feeding group, early-fed women were given solid food sooner after surgery, 5.0 ± 1.2 hours versus 40.0 ± 10.6 hours. Mild ileus symptoms occurred in 19 (31.7%) early-fed patients and 16 (26.7%) traditionally fed women, which was not statistically different (Table 4). One woman, who was on the traditional feeding regimen, developed a severe ileus requiring abdominal radiographs; her symptoms resolved with conservative management. Neither the presence of intra-operative adhesions nor the amount of estimated blood loss correlated with the development of ileus symptoms. Early-fed women had a shorter mean postoperative hospital stay and a shorter mean postoperative time interval to bowel movement.
The majority of postoperative complications were postpartum endometritis. In the early-fed group, one woman was readmitted because of a wound infection, and among those who were traditionally fed, one patient had suspected late-onset preeclampsia, and one patient was diagnosed with a wound seroma.
Within the early-fed group, 37 women had operative times exceeding 40 minutes, and 23 patients had operative times less than or equal to 40 minutes. Significantly more early-fed women had mild ileus symptoms when their surgeries exceeded 40 minutes (40.5% versus 13.4%, P < .01). The surgical time was not found to be associated with estimated blood loss, presence of intra-abdominal adhesive disease, anesthetic medications, prior cesarean delivery, parity, or elective cesarean delivery. Multiple linear regression analysis, while controlling for surgical duration, revealed that the hospital stay for early-fed women was still significantly shorter than that of the control arm (adjusted correlation coefficient 0.655, P < .001). In the traditional feeding women, the duration of surgery had no effect on the incidence of ileus symptoms.
Early feeding with solid food after laparoscopic surgery has been widely accepted as safe.6 Recently, studies have suggested that early feeding seems to be well tolerated and even beneficial to patients who have undergone laparotomy and cesarean delivery.4,5,7 When compared with women who were given nothing by mouth for 12 hours after cesarean delivery, those who were fed early with solid food required less injectable narcotic pain medication.5 Although several investigators have shown that women may tolerate early solid food after cesarean delivery, none have demonstrated an impact on the duration of hospital stay.4,5,7
Our study confirmed that the early consumption of solid food in women after cesarean is well tolerated. There were no significant increases in the incidence of mild ileus symptoms or postoperative complications in the early-fed group. The one patient who developed a severe ileus was on the traditional feeding regimen. The extent of intra-abdominal adhesions or amount of estimated blood loss did not seem to correlate with the development of ileus symptoms. Because the rate of ileus symptoms in the control group was 26.7%, which exceeded our assumed rate of 19%, our sample size only reached a power of 55% to demonstrate a 20% difference. To detect a difference between the rates of 27% and 47% of ileus symptoms, we calculated that a sample size of 202 subjects would be required.
We found that women in the early feeding arm had a more rapid return of bowel function, with a substantially shorter mean postoperative time interval to bowel movement (34.5 hours versus 51.0 hours). Our study demonstrated that women who were given solid food shortly after cesarean had a shorter mean hospital stay, by nearly 24 hours (49.5 hours versus 75.0 hours). The economic impact of early dismissal from the hospital after an uncomplicated cesarean delivery cannot be overlooked. The average charge for hospitalization was $1262 less for early-fed women than those fed traditionally ($6602 versus $7804). Furthermore, in contrast to other interventions aimed at facilitating hospital discharge, we found that women who were offered food early were generally pleased with the prospect of eating earlier.
One unexpected finding was that in early-fed women, the incidence of mild ileus symptoms seemed to be related to the length of surgery. As compared with early-fed women whose cesarean sections lasted less than or equal to 40 minutes, early-fed patients whose surgeries exceeded 40 minutes were more likely to have mild ileus symptoms, 40.5% versus 13.4%. The same association was not found in traditionally fed women. There was no correlation between the length of surgery and the presence of intra-abdominal adhesions, the estimated blood loss, the experience of the surgeon, the reason for cesarean, or intra-operative or postoperative medications. We did not find an explanation for this finding, which may very well be spurious; however, the duration of surgery may possibly in some way predispose early-fed women to mild ileus symptoms. Also, the importance of mild ileus symptoms is unclear. Within the early-fed group, the surgical times did not appear to influence the hospital stay. The question of whether surgical duration is associated with ileus symptoms in women prescribed solid food soon after cesarean delivery needs further investigation.
Our study did not have a sufficient sample size to elucidate the effect of early feeding on mild ileus symptoms, or rare complications such as severe ileus. Also, although dietary prescriptions were offered to patients, the actual of amount of food ingested was not quantified by the use of a food diary or by calorie count. Because of the nature of the intervention, neither the patient nor the practitioner was blinded to the feeding assignment; this fact may affect the assessment of subjective complaints such as abdominal cramping and nausea. However, the return of bowel function documented by a bowel movement or passage of flatus was the limiting factor for discharge in most cases. Because the criteria for hospital discharge were delineated clearly and fairly objectively, it is not likely that the practitioner's being nonblinded played a major role in affecting the hospital course. Finally, because our study took place in a metropolitan private community hospital, the results from this study may not be generalizable to other populations.