You are on page 1of 5

International Journal of Gynecology and Obstetrics 126 (2014) 115119

Contents lists available at ScienceDirect

International Journal of Gynecology and Obstetrics

journal homepage:


A randomized comparative trial of early initiation of oral maternal

feeding versus conventional oral feeding after cesarean delivery
Shabeen N. Masood a,, Yasir Masood b, Uzma Naim c, Muhammad F. Masood d
Sobhraj Maternity Hospital, Karachi Metropolitan Corporation, Karachi, Pakistan
Department of Surgery, Ziauddin Medical University, Karachi, Pakistan
Department of Surgery, Wayne State University, Detroit, USA
Department of Cardiothoracic Surgery, University of Pittsburgh Medical Center, Pittsburgh, USA

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To compare the effects of two maternal feeding policiesearly versus conventional oral feedingafter
Received 2 October 2013 cesarean delivery. Methods: This prospective multicenter randomized comparative trial was conducted at tertiary
Received in revised form 21 February 2014 care hospitals in Sindh, Pakistan, from 2010 to 2012. Women with an uncomplicated cesarean delivery under spi-
Accepted 20 April 2014 nal anesthesia were allocated to an intervention of early (after 2 hours) or conventional (after 18 hours) initiation
of oral feeding. Outcomes included maternal ambulation, maternal satisfaction, gastrointestinal functions, and
length of hospital stay. Results: In total, 1174 women (n = 587 per group) were included in the nal analysis. Gas-
Day-care cesarean delivery
Early ambulation
trointestinal complications were not signicantly different between the two groups. Lower intensities of thirst and
Early oral maternal feeding hunger and a higher rate of maternal satisfaction were observed in the early feeding group (P b 0.05), and 53.8% of
Length of hospital stay women in this group were able to ambulate within 15 hours of surgery, compared with 27.9% of women in the
Maternal satisfaction conventional feeding group. The frequencies of readmission, febrile morbidity, and wound infection were insignif-
Maternal thirst and hunger icant. Conclusion: Early oral dietary initiation after cesarean delivery resulted in early ambulation, greater maternal
satisfaction, and reduced length of hospital stay, with no detrimental outcomes, making this practice cost-
effective. Hence, day-care cesarean delivery might be an option in resource-constrained settings.
Trial Registration number: ChiCTR-TRC-13003651,
2014 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction recovery, reduces the degree of negative protein balance, and makes
women feel energetic.
There are certain trends in postoperative dietary management that The objective of the present study was to determine the effects of
most health providers follow. One common practice is to withhold early initiation of oral feeding versus conventional feeding in women
uids and foods after cesarean delivery until the return of bowel sounds with an uncomplicated cesarean delivery under spinal anesthesia on
or the passage of atus. This practice is based on the belief that all intra- time to ambulation, discontinuation of intravenous uids and catheter-
abdominal surgical procedures are followed by postoperative ileus [1]. ization, maternal satisfaction, thirst and hunger, gastrointestinal func-
Women who are potential candidates for cesarean delivery are usually tions, and length of hospital stay.
kept fasting in the labor ward for a long period of time; however, this
practice is not scientically tested. Evidence supports early oral feeding
after surgery with no signicant increase in gastrointestinal complica- 2. Materials and methods
tions [2]; this is well tolerated with many nutritional advantages over
standard intravenous uids. The present study was a prospective multicenter randomized
Delayed immobilization after surgery is an important reason for trial that compared two active feeding policiesearly and conven-
prolonged hospital stay after a cesarean delivery. By contrast, early dis- tional maternal feedingafter cesarean delivery. The study centers
charge from hospital after an uncomplicated cesarean delivery may re- were chosen based on convenience and number of obstetric patients
duce the nancial burden and could make cesarean delivery a day-care and comprised the departments of Obstetrics and Gynecology at
procedure. Early intake of an oral diet, as opposed to the conventional Countess of Dufferin Fund Hospital in Hyderabad, Liaquat University
practice of withholding an oral diet for 1224 hours, promotes prompt Hospital in Hyderabad Noor Hospital in Karachi, and Fatima Bai
Hospital in Karachi, all in Pakistan. The study was conducted from
Corresponding author at: U-19, Hasan Apartment Extension, Hasan Square, Gulshan-e-
March 1, 2010, to May 31, 2012. Ethics approval was obtained from
Iqbal, Block-13D, Karachi 75300, Pakistan. Tel.: +92 3332233414; fax: +92 2134856118. the Ethics Review Board (ERB) of Liaquat University of Medical and
E-mail address: (S.N. Masood). Health Sciences, Jamshoro, Pakistan.
0020-7292/ 2014 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.
116 S.N. Masood et al. / International Journal of Gynecology and Obstetrics 126 (2014) 115119

The study consisted of three consecutive phases, which were all con- better compliance and motivation. Women in the early feeding (EF)
ducted at the recruiting hospitals. In phase I, research staff conducted group were given 200250 mL of liquid (water, tea, milk, juice in
focus group discussions on feeding practices after cesarean delivery tetra/tinned packs) 2 hours after surgery. If the liquid diet was well tol-
with the stakeholders (all pregnant women who attended the hospitals erated, a solid diet was started 4 hours after surgery. An interval of
prenatal clinics, especially those who had previously had a cesarean de- 2 hours was chosen before initiation of the liquid diet to allow time
livery or who were scheduled for elective cesarean delivery, and staff of for the fourth stage of labor (1 hour) and for logistic adjustments such
the maternity unit) during prenatal clinics. The responses from the par- as transfer of the patient from the operating room to the ward
ticipants indicated a positive perception toward early oral maternal (1 hour). Women in the conventional feeding (CF) group received
feeding. In phase II, the doctors and nurses of the recruiting hospitals clear uids after 12 hours and a solid diet after 18 hours. The decision
were trained before implementation of the research protocol. Phase III to deviate from the allocated feeding regime (for example because of
was the implementation phase of the study. food intolerance, mild/severe ileus, or clinical deterioration) remained
The primary outcome measures were the time to ambulation and at the clinicians discretion. Any inevitable deviation was recorded.
the level of maternal satisfaction. The secondary outcomes included The duration of surgery was dened as the time from skin incision to
gastrointestinal complications and duration of hospital stay. skin closure. All other durations were recorded as hours after the com-
Initially, 4610 women gave their consent to participate in the study. pletion of surgery. Mild ileus was dened as nausea and abdominal pain
The actual study sample was selected based on nonprobability purpo- in the rst 12 hours, and severe ileus was dened as persistent nausea,
sive sampling. The estimated prevalence of cesarean delivery in vomiting (more than four times), abdominal pain, and distention of
Pakistan is 27%45% [36]. Assuming a 45% prevalence of cesarean de- more than 6 cm at the level of the umbilicus within 1224 hours post-
livery with an 80% power and a 4% margin of error, the estimated sample operatively. The intensities of thirst and hunger and the level of mater-
size was approximately 1200. This was calculated with the WHO sample nal satisfaction were measured on a visual analog scale (VAS). In both
size calculator. After matching for age, parity, and number of previous groups, intravenous hydration was discontinued and the Foley catheter
cesarean deliveries, a computer-generated randomization list was was removed after the woman had taken her rst full meal without
used to allocate the women to the early feeding or conventional feeding nausea or vomiting. Approximately 1.01.5 L and 3.03.5 L of uid
group. The research staff at the recruiting hospitals received sealed were given postoperatively in the EF and CF groups, respectively. All
envelopes; color-coded les were used to identify the study groups women received three doses of intravenous antibiotics. Mobilization
(red for early feeding and green for conventional feeding). The data was encouraged in both groups. The study did not alter/interfere with
for 587 women per group were included in the nal analysis (Fig. 1). any treatment/care given routinely to the women except for the inter-
The study included women who gave informed consent and who ventions to be evaluated.
had an uncomplicated cesarean delivery under spinal anesthesia with- Demographic data were collected at recruitment. The research staff
out comorbidity (medical, obstetric, or surgical history of any serious ill- or the duty doctor collected the remaining data during the womens
ness). Informed consent was obtained preoperatively from all eligible stay in the postoperative ward and at the 1-week follow-up at the hos-
women and their spouses. Once a woman had been recruited, she con- pital, using structured questionnaires. The duty doctor was expected to
tinued to be in the trial. report serious adverse events.
After cesarean delivery, the women were transferred to separate The data were analyzed with SPSS version 15.0 (IBM, Amonk, NY,
wards according to their feeding group allocation, in order to obtain USA). Continuous variables were reported as mean standard

Enrollment Assessed for eligibility (n=4610)

Excluded (n=2792)
Did not meet inclusion criteria (n=2770)
Vaginal birth (2728)
Intraoperative complications (n=42)
Refused to participate (n=22)

Randomized (n=1818)


Allocated to early feeding intervention (n=909) Allocated to conventional feeding intervention (n=909)
Received allocated intervention (n=596) Received allocated intervention (n=851)
Did not receive allocated intervention (n=313) Did not receive allocated intervention (n=58)
Noncompliance of patient or attendant (n=56) Noncompliance of patient or attendant (n=9)
Noncompliance of staff (n=15) Noncompliance of staff (n=7)
Allocation not followed (n=242) Allocation not followed (n=42)


Lost to follow-up (n=9) Lost to follow-up (n=20)

Could not be contacted (n=7) Could not be contacted (n=14)
Not satisfied with the intervention (n=2) Not satisfied with the intervention (n=6)


Analyzed (n=587) Analyzed (n=587)

Excluded (n=244)

Fig. 1. Study ow. a No match by age, parity, and number of previous cesarean deliveries.
S.N. Masood et al. / International Journal of Gynecology and Obstetrics 126 (2014) 115119 117

deviation and compared with the t test. Categorical variables were re- [2,9], there were also no signicant differences in the rates of abdominal
ported as number and percentage and analyzed using the 2 test. The distention, ileus, and wound dehiscence and in the times to passage of a-
Fisher exact test was used if there were fewer than ve occurrences in tus and return of bowel sounds, despite an increase in postoperative nau-
any group. P b 0.05 was considered statistically signicant. sea and vomiting after early introduction of oral feeding.
Cesarean delivery is the most commonly performed operation
3. Results worldwide [10]. It generally involves minimal bowel manipulation
and the duration of surgery is relatively short, allowing a rapid postop-
In total, data for 1174 women were available (EF group, n = 587; erative return of bowel functions even before this is clinically detectable
CF group, n = 587). The mean age of the study participants was [11]. Early oral feeding is associated with less depletion of protein
25.90 4.42 years and the parity was 2.22 1.57 (Table 1). Elective stores, a decreased infection rate, and improved wound healing, and if
cesarean delivery was performed in 971 (82.7%) of the women, and adequate calories are administered, it prevents weight loss by maintain-
203 (17.3%) of the women underwent emergency cesarean delivery. ing serum electrolytes and a positive nitrogen balance [12].
A history of previous cesarean delivery was present in 619 (52.9%) of Lower intensities of thirst and hunger with higher rates of maternal
the women. The frequencies of nausea, vomiting, mild or severe satisfaction were observed in the EF group in the present study, without
postoperative ileus, and abdominal distention were not statistically an increase in vomiting. An audit of an early feeding program in France
signicant between the two groups. Nausea was found in 40 (6.8%) [13] showed that postoperative thirst was a bigger problem for patients
versus 27 (4.6%) women in the early and conventional feeding than hunger. After the introduction of early oral feeding, signicantly
groups, respectively (Fig. 2). Vomiting was observed in 14 (2.4%) fewer mothers experienced hunger and thirst on the rst two postoper-
women in each group, mild ileus was noted in 58 (9.9%) versus 55 ative days [13]. Maternal satisfaction was higher in the EF group
(9.3%) women in the early and conventional feeding groups, and (48.2% vs 34.4%), because the women were less hungry and thirsty
severe ileus in 5 (0.9%) versus 7 (1.2%) women. Abdominal disten- given that feeding was introduced earlier. Similar rates of maternal sat-
tion was present in approximately 12 (2%) of the women in each isfaction have been reported in most other studies [14,15].
group (P N 0.05). Women in the EF group were given 1.01.5 L of uid postoperatively
The time to rst auscultation of bowel sounds, time to rst percep- compared with women in the CF group, who were given 3.03.5 L be-
tion of bowel sounds by the woman, time to passing atus, and abdom- fore they were allowed full meals. The early introduction of uids and
inal girth at the level of the umbilicus 1 hour after the cesarean delivery food was associated with an increased rate of early ambulation, earlier
were similar in the two groups (Table 2). However, discontinuation of discontinuation of intravenous uids, and earlier removal of the cathe-
intravenous uid and removal of the catheter occurred signicantly ear- ter. Guidelines from the National Institute for Health and Clinical Excel-
lier in the EF group than in the CF group (P b 0.05). lence [16] recommend that the urinary catheter should be removed
Twice as many women in the EF group were able to ambulate once a woman is mobile after regional anesthesia. The fact that EF
15 hours after surgery compared with the CF group (316 [53.8%] vs kept women in their hospital bed for a shorter period of time may result
164 [27.9%]) (Fig. 3). The intensities of thirst and hunger were signi- in a signicant reduction in both tangible costs (cost of medications and
cantly higher in the CF group (P = b 0.001 and P = b0.001, respectively). hospital utilities) and intangible costs (nursing care and paramedical
One week after discharge, 4 (0.7%) women in the EF group had been services). Similar ndings were reported in a study from Mulago,
readmitted (Fig. 4); 2 because of wound infection and 2 because of high- Kenya [17], where women in the EF group were also able to leave the
grade fever secondary to endometritis. In the CF group, one patient had bed earlier (15.1 hours vs 17.8 hours).
been readmitted because of severe secondary postpartum hemorrhage. Length of hospital stay and cost-effectiveness are important factors
Wound infection was rare (EF group, 2 [0.3%]; CF group, 0 [0.0%]. Febrile to take into account when considering the possibility of making cesare-
morbidity was seen in 18 (3.0%) women in each group. an delivery a day-care procedure. The authors of a Canadian study [18]
recommended early feeding because it was associated with a shorter
4. Discussion hospital stay, compared with delayed feeding; the median length of
hospital stay was 6.0 days for the traditional feeding group and 4.0
In the present study, 53.8% of the women in the early feeding group days for the early feeding group (P b 0.001). Similarly, a study from
were able to ambulate within 15 hours postoperatively compared with the USA [2] found that women who received EF had a shorter hospital
27.9% of the participants in the conventional feeding group. Similar re- stay, and this had a signicant impact on hospital charges. Studies
sults were reported in a study conducted in India [7]. from India [19] and Nigeria [20] also reported shorter hospital stays
Although the frequency of nausea was higher in the EF group, no sig- with early feeding (mean 59.0 7.3 hours versus 88.0 9.5 hours, re-
nicant differences between the groups were observed in terms of ab- spectively; P b 0.001) [19]. In the present study, the lengths of hospital
dominal distention, paralytic ileus, and wound dehiscence. Similarly, a stay in the two groups differed by only 3 hours. The women in the EF
study from Thailand [8] found no signicant difference in gastrointestinal group could have been discharged sooner, but they were kept in hospi-
adverse effects between women with early (after 8 hours) versus delayed tal for the same duration of time as the women in the CF group to ob-
(after 24 hours) rst meal intake after cesarean delivery. In other studies serve the effects of early oral diet introduction.

Table 1
Demographic and other characteristics of women with uncomplicated cesarean delivery.

Characteristic Early feeding Conventional feeding Total P value

(n = 587) (n = 587) (n = 1174)

Age, y 26.07 4.51 25.72 4.33 25.90 4.42 0.18a

Parity 2.26 1.61 2.17 1.52 2.22 1.57 0.36a
Previous cesarean delivery(s)
Yes 312 (53.2) 307 (52.3) 619 (52.9) 0.791b
No 275 (46.8) 280 (47.7) 555 (47.3)
Elective cesarean delivery 486 (82.8) 485 (82.6) 971 (82.7) 0.938b
Emergency cesarean delivery 101 (17.1) 102 (17.4) 203 (17.3)
t test.
2 test.
118 S.N. Masood et al. / International Journal of Gynecology and Obstetrics 126 (2014) 115119

Fig. 2. Gastrointestinal outcomes after early or conventional feeding.

Table 2
Postoperative ndings among women with an uncomplicated cesarean delivery.

Postoperative event Early feeding Conventional feeding Total P valuea

(n = 587) (n = 587) (n = 1174)

Time to rst auscultation of bowel sounds, h 17.09 12.9 16.15 11.15 16.62 12.05 0.19
Time to rst perception of bowel sounds by the woman, h 18.33 11.87 17.29 10.74 17.81 11.32 0.13
Time to passing atus, h 19.78 11.57 19 10.35 19.4 10.99 0.25
Abdominal girth at the level of the umbilicus 1 hour after cesarean delivery, cm 82.65 17.5 83.45 17.12 83.05 17.31 0.475
Time to discontinuation of intravenous uids, h 25.27 8.75 30.24 14.21 27.75 12.06 b0.001
Time to removal of catheter, h 13.21 2.78 33.71 6.35 23.46 11.37 b0.001
Length of hospital stay, h 74.63 13.85 77.4 16.13 75.83 15.08 0.006
t test.

No signicant detrimental effects on postoperative clinical outcomes In the present study, a larger number of women was recruited than
were seen among women who were allowed early oral feeding. Early in previously published trials [21,22] and no signicant maternal ad-
postoperative recovery is an important factor that can enable women verse event was seen as a result of post cesarean early feeding. In a pre-
to ambulate earlier and reduce their hospital stay. Therefore, initiation vious study [23], 142 (38.9%) providers of obstetric care said that EF is
of EF seems to be safe for women with an uncomplicated cesarean deliv- cost-effective, whereas 106 (29%) participants felt that it may increase
ery and is more acceptable to women and their families. costs because of complications or readmissions. In the present study,

Fig. 3. Factors inuencing maternal satisfaction with early or conventional feeding. a P b 0.005.
S.N. Masood et al. / International Journal of Gynecology and Obstetrics 126 (2014) 115119 119

Fig. 4. Outcomes 1 week after discharge from hospital.

postoperative follow-up after 1 week did not show a signicant increase [8] Kovavisarach E, Atthakorn M. Early versus delayed oral feeding after cesarean deliv-
in the rates of maternal readmission and infectious morbidities. ery. Int J Gynecol Obstet 2005;90(1):314.
[9] Teoh WH, Shah MK, Mah CL. A randomized controlled trial on benecial effects of
The present participants came from a wide range of socioeconomic early feeding post-caesarean delivery under regional anesthesia. Singapore Med J
backgrounds; this further strengthens the external validity of the trial. 2007;48(2):1527.
The present study is limited by a high rate of protocol violations by [10] Gibbons L, Belizn J, Lauer J, Betrn A, Merialdi M, Althabe F. The global numbers and
costs of additionally needed and unnecessary caesarean sections performed per year:
patients and their attendants, the use of self-reported outcome mea- overuse as a barrier to universal coverage. Geneva: WHO; 2010.
sures, and compliance difculties among the attendants related to the healthsystems/topics/nancing/healthreport/30C-sectioncosts.pdf.
early initiation of oral feeding. [11] Casto CJ, Krammer J, Drake J. Postoperative feeding: a clinical review. Obstet Gynecol
Surv 2000;55(9):5713.
In conclusion, the early initiation of oral feeding after cesarean [12] Weinstein L, Dyne PL, Deurbeck NB. The PROEF dieta new Postoperative regimen
delivery under spinal anesthesia resulted in early ambulation and for oral early feeding. Am J Obstet Gynecol 1993;168(1 Pt 1):12831.
greater maternal satisfaction. This feeding regime can reduce the [13] Benhamou D, Tcsy M, Parry N, Mercier FJ, Burg C. Audit of an early feeding pro-
gram after cesarean delivery: patient wellbeing is increased. Can J Anesth
length of hospital stay without any detrimental effects on postoper- 2002;49(8):8149.
ative clinical outcomes in uncomplicated cesarean deliveries. Early [14] Chantarasorn V, Tannirandorn Y. A comparative study of early postoperative feeding
oral feeding also reduces the tangible and intangible costs of services versus conventional feeding for patients undergoing cesarean section; a randomized
controlled trial. J Med Assoc Thai 2006;89(Suppl. 4):S116.
provided to the patients. National recommendations regarding early
[15] Izbizky GH, Minig L, Sebastiani MA, Otao L. The effect of early versus delayed
oral feeding after cesarean delivery are required. Day-care cesarean postcaesarean feeding on women's satisfaction: a randomised controlled trial.
delivery could become a future option in resource-constrained set- BJOG 2008;115(3):3328.
tings if mothers are allowed to drink and have oral food sooner [16] National Institute for Health and Clinical Excellence. Caesarean section: NICE
clinical guideline 132.
after surgery. 57163.pdf. Published November 2011. Updated August 2012. Accessed
May 11, 2013.
Conict of interest [17] Adupa D, Wandabwa J, Kiondo P. A randomised controlled trial of early initiation of
oral feeding after caesarean delivery in Mulago Hospital. East Afr Med J
The authors have no conict of interest to declare. [18] Steed HL, Capstick V, Flood C, Schepansky A, Schulz J, Mayes DC. A randomized con-
trolled trial of early versus "traditional" postoperative oral intake after major ab-
dominal gynecologic surgery. Am J Obstet Gynecol 2002;186(5):8615.
References [19] Mehta S, Gupta S, Goel N. Postoperative Oral Feeding After Cesarean Section-
Early Versus Late Initiation: A Prospective Randomized Trial. J Gynecol Surg
[1] Holte K, Kehlet H. Postoperative ileus: a preventable event. Br J Surg
[20] Orji EO, Olabode TO, Kuti O, Ogunniyi SO. A randomized controlled trial of early
[2] Patolia DS, Hilliard RI, Toy EC, Baker B. Early feeding after cesarean: randomized trial.
initiation of oral feeding after cesarean section. J Matern Fetal Neonatal Med
Obstet Gynecol 2001;98(1):1136.
[3] Bilal N, Yasmin F, Akhtar S. Frequency and indication of caesarean section in a tertia-
[21] Adeli M, Razmjo N, Tara F, Ebrahimzadeh S. The effect of early post cesarean feeding
ry care maternity unit. J Postgrad Med Inst 2005;19(4):3925.
on womens satisfaction. J Fam Reprod Health 2010;4(2):7982.
[4] Karim F, Ghazi A, Ali T, Aslam R, Afreen U, Farhat R. Trends and Determinants of Cae-
[22] Bar G, Sheiner E, Lezerovizt A, Lazer T, Hallak M. Early maternal feeding following
sarean Section. J Surg Pak (International) 2011;16(1):227.
caesarean delivery: a prospective randomized study. Acta Obstet Gynecol Scand
[5] Yusuf F, Haider G, Shafqat G, Haider A, Nasiruddin M. An audit of Cesarean Sections
in a teaching hospital. PAFMJ 2009(5).
[23] Masood SN, Siddiqui IA, Masood MF, Masood Y, Mumtaz SN. To assess the practices
[6] Shamshad. Factors leading to increased caesarean section rate. Gomal J Med Sci
about initiation of Oral Maternal Feeding after Cesarean Section under Regional An-
esthesia. Pak J Med Sci 2011;27(5):9961000.
[7] Malhotra N, Khanna S, Pasrija S, Jain M, Agarwala RB. Early oral hydration and its im-
pact on bowel activity after elective caesarean sectionour experience. Eur J Obstet
Gynecol Reprod Biol 2005;120(1):536.