|Year : 2019 | Volume
| Issue : 1 | Page : 31-37
Post-operative management in uncomplicated caesarean delivery: A randomised trial of short-stay versus traditional protocol at the Lagos University Teaching Hospital, Nigeria
Nuvie Oyeyemi1, Lawal Oyeneyin2, Ayodeji Oluwole1, Abisoye Oyeyemi3, Bosede Afolabi1
1 Department of Obstetrics and Gynaecology, Lagos University Teaching Hospital, Lagos, Lagos State, Nigeria
2 Department of Obstetrics and Gynaecology, Mother and Child Hospital, Ondo, Ondo State, Nigeria
3 Department of Community Medicine, Faculty of Clinical Sciences, Niger Delta University, Wilberforce Island and Niger Delta University Teaching Hospital, Okolobiri, Bayelsa State, Nigeria
|Date of Web Publication||12-Mar-2019|
Department of Obstetrics and Gynaecology, Mother and Child Hospital, Ondo, Ondo State
Source of Support: None, Conflict of Interest: None
Context: Caesarean section (CS) is the most common major obstetric operation. There has, therefore, been an increasing interest in issues pertaining to the management and length of hospital stay following the procedure. Aim: This study aimed to evaluate morbidity outcomes as well as incurred costs between traditional and short-stay protocols, following uncomplicated CS deliveries. Settings and Design: This was a randomised controlled trial conducted among booked antenatal patients who had elective CS at the Lagos University Teaching Hospital. Materials and Methods: Using a parallel study design, patients were randomised into short-stay and traditional protocols. Patients in the short-stay group were ambulated and graded oral intake initiated from 6 h post-operation. Their urethral catheters were discontinued at 12 h, and subsequent discharge was at 3rd day post-op. Those in the traditional group were ambulated from 12 h, graded oral intake initiated and urethral catheters removed at 24 h, then the patients were discharged on the 5th day post-operation. Pain scores of all the patients at 72 h, fever in the first 10 days (excluding the first 24 h), clinical signs of wound sepsis, urinary tract infection and puerperal sepsis in the first 14 days post-op were recorded. Statistical Analysis: Descriptive statistics were used to summarise the quantitative variables. The association between categorical variables was tested using Chi-square test, and differences in group means were assessed using t-test. The confidence level was 95%, and the level of significance was set at P < 0.05. Results: There were no significant differences in febrile and infective morbidities between the two groups. However, women in the short-stay group had significantly lower pain scores (t = 4.75, P < 0.001) and hospital expenses (t = 5.53, P < 0.0001) than women in the traditional group. Conclusions: The short-stay protocol following uncomplicated CS delivery was safe and more cost-effective than the traditional protocol.
Keywords: Caesarean section, post-op complications, post-op management
|How to cite this article:|
Oyeyemi N, Oyeneyin L, Oluwole A, Oyeyemi A, Afolabi B. Post-operative management in uncomplicated caesarean delivery: A randomised trial of short-stay versus traditional protocol at the Lagos University Teaching Hospital, Nigeria. Niger Postgrad Med J 2019;26:31-7
|How to cite this URL:|
Oyeyemi N, Oyeneyin L, Oluwole A, Oyeyemi A, Afolabi B. Post-operative management in uncomplicated caesarean delivery: A randomised trial of short-stay versus traditional protocol at the Lagos University Teaching Hospital, Nigeria. Niger Postgrad Med J [serial online] 2019 [cited 2019 Mar 24];26:31-7. Available from: http://www.npmj.org/text.asp?2019/26/1/31/253975
| Introduction|| |
There is an increasing interest in patients' management following caesarean section (CS), with tendency towards early discharge, quick recovery and return to function.,,, However, the duration of hospital stay following CS still varies from centre to centre., Professional recommendations imply that, in uncomplicated deliveries, the length of hospital stay post-partum should range from about 48 h for vaginal delivery, to about 96 h for caesarean deliveries. Previous studies have not detected an increased rate of morbidity associated with early discharge following CS.,, Furthermore, it is anticipated that early post-partum discharge would save cost, promote family bonding as well as reduce maternal and neonatal exposure to nosocomial pathogens.,,,
The present trend is towards reducing the length of hospital stay post-CS., This is particularly important in parts of the world where misgivings about this mode of delivery still exist.,, However, it is important to ensure and document that such a practice is safe and does not compromise the patient in any way.
The traditional 'protocol' for post-op care at the Lagos University Teaching Hospital (LUTH) has been for at least 5 days after CS, during which the stitches are removed or the wound is inspected, prior to discharge. As short-stay protocols of 48–72 h have been found to be safe and cost-effective in other countries, we decided to examine our practice at LUTH. The aim of this study was, therefore, to evaluate morbidity outcomes as well as incurred costs between short-stay and traditional protocols, following uncomplicated caesarean deliveries.
| Materials and Methods|| |
Ethical clearance for the study was obtained from the Health Research and Ethics Committee of LUTH (Private Mail Bag 12003, Lagos) on 17 April 2012. The protocol number is ADM/DCST/HREC/VOL.XVI/APP. This was a randomised clinical trial carried out among booked antenatal patients, women of reproductive age (15–49 years) scheduled for elective CS for various indications at LUTH between 1 June and 31 December 2012 and who have consented to participate in the study. Exclusion criteria were women who had medical conditions that required post-op management by other specialities, those with pre-planned prolonged hospital stay and immunosuppressed patients. Using the elective CS list, a total of 200 eligible consenting patients were randomised into equal numbered short-stay (A) and traditional (B) protocol groups, of 100 each. Consent was taken by resident doctors when the patient was being wheeled into the labour ward theatre.
The null hypothesis for this study states that there is no difference in the incidence of wound infection between short-stay and the traditional protocols after uncomplicated caesarean delivery. This study takes into account the incidence rate for wound infection (which is the primary outcome variable) in women in a previous study comparing short-stay and prolonged hospital stay following uncomplicated caesarean delivery.
The minimum sample size was determined using the following formula:
- n = minimum sample size of each group
- u = one-sided percentage point of the normal distribution corresponding to 100% the power, for example, if power = 90%, u = 1.28
- v = percentage point of the normal distribution corresponding to the (two sided) significance level, for example, if significance level = 5%, v = 1.96
- μ1= rate of wound infection in the longer hospital stay group = 10%
- μ0= rate of wound infection in the short hospital stay group = 6%
- μ1− μ0= difference between two rates – in these cases, the sample size refers to the same units as used for the denominator of the rates. For example, the rate here is expressed as the number of women who had wound infection (3 for short and 5 for longer stay, hence difference is 5–3 = 2). The formula gives the number of women required for observation.
Thus, the minimum sample size per group
To allow for attrition and uncompleted follow-ups, the sample size was increased by about 20%. Therefore, a sample size of 50 should be used in each group. This was doubled in order to increase the power of the study.
The randomisation process applied computer-generated random numbers (unstratified balanced allocation from WinPepi), with random allocation concealment using sealed opaque envelopes numbered from 1 to 200. An independent observer (resident doctor) picked the envelopes consecutively at the time the patient was being wheeled into the operating suite. The selected sealed envelope was placed within the patients' case notes, together with the partially filled questionnaire (containing patient data) and a pain score sheet. Both the surgeon and the patient were blinded to this allocation at the time of surgery. The envelopes were opened by the nursing staff as soon as the patient got to the lying-in-ward from the recovery room. In line with standard protocol, all the surgeries were performed by consultants or 2nd-year senior registrars in the Department of Obstetrics and Gynaecology of LUTH.
Post-op management was according to the protocol for each group, with all records and findings entered into the data collection forms. The patients were then followed up, evaluated at 2 as well as 6 weeks following delivery and all the information was also recorded in the data collection forms. Post-op notes were written and then placed in the case notes of all randomised patients, irrespective of the mode of anaesthesia. The predetermined intervention checklist for Group A was placed on the noticeboard at the nursing station on the ward for attention of the staff. Those in Group B were managed according to the existing protocol. The sealed opaque envelopes containing allocations were only opened in the lying-in ward, and all the post-op interventions were made by the nursing staff according to the respective groups. Observations were made and documented in the questionnaires by trained observers (resident doctors). The outcomes were marked as 'yes' if any of the features were present, and 'no' if absent. Pain scores were assessed using Visual Analogue Charts and recorded by the principal investigator.
All patients in Group B were sat out of bed from 12 h post-op. Twenty-four hours post-op, they were commenced on graded oral sips as tolerated – either with passage of flatus or presence of normoactive bowel sounds, ambulated around the ward and had urethral catheters discontinued. Intravenous fluids such as dextrose/saline alternating with normal saline or dextrose water were given at the dosage of 1 L 8 hourly for at least the first 24 h after surgery until oral intake was well established. Intramuscular pentazocine 30 mg combined with promethazine 25 mg 4–6 hourly, diclofenac 75 mg 12 hourly and paracetamol 600 mg 8 hourly were all administered for the first 24 h after the surgery. Intravenous co-amoxiclav 1.2 g (or cephalosporin) 12 hourly (after peri-operative dose) and metronidazole 500 mg 8 hourly were also given for the first 24 h. Antibiotics were converted to the oral equivalent when intake was well established, whereas analgesics were converted to oral paracetamol 1 g and diclofenac 50 mg 8 hourly. The wound was inspected on the 5th day, and if healing was well and there were neither complaints nor complications, the patient was subsequently discharged home on the same day.
All patients in Group A were commenced on graded oral sips as tolerated 6 h post-op, and then intravenous fluids were discontinued within 12 h once intake was established. They sat out of bed from 6 h and ambulated from 12 h post-op. Their urinary catheters were also discontinued 12 h post-op. Analgesics and antibiotics were administered as in Group B. The wound was inspected on the 3rd day after the operation day, and the patient was subsequently discharged home on the same day. The wounds were cleaned using methylated spirit before discharge to home in both groups, with instructions to continue the care twice daily at home until the 2-week postnatal visits. Records of any temperature spikes on admission were made from in-patient charts. The participants were given and taught how to use digital thermometers to check as well as record their temperatures daily or whenever they perceived a fever, from the time of discharge till 10 days after delivery. They were also informed to report to the accident and emergency unit of the hospital for any complaint; otherwise, they were seen finally at the 2-week postnatal visit when all recordings were done on the pro forma.
Relevant details about the operation notes and post-op management were obtained from the caesarean delivery records, while pain was assessed using the scoring sheet 72 h post-op (in which the patients were asked to score their pain on a scale of 0–10, where 0 was no pain, 1 was mildest and 10 was the most severe pain). The total hospital expense for each patient was calculated and recorded at discharge by the principal investigator. Other independent observers (house officers) blinded to the group each participant belonged to were also trained to administer the remaining portion of the questionnaires at the postnatal clinics. Patients that did not attend the clinic for evaluation were contacted through phone calls.
Data analysis was done using the EPI-INFO version 3.5.3 statistical software (Centre for Disease Control and Prevention, Atlanta, Georgia, USA). Both intention-to-treat analysis and per-protocol analysis were done. The intention-to-treat analysis disregarded the patients that dropped out and analysed as if they remained in their respective groups as initially randomised. The per-protocol analysis was done for only patients that remained after dropout. The analysis included use of descriptive statistics such as mean, proportion and standard deviation to summarise the quantitative variables. Association between categorical variables was tested using Chi-square test, and differences in group means were assessed using t-test. A confidence interval of 95% was used with the level of significance set at P < 0.05. The primary outcome variable was the rate of wound infection/sepsis (painful redness and swelling or induration of wound edges, sero-purulent/purulent discharge or dehiscence). The secondary outcome variables were pain scores, hospital bills at discharge and febrile and other infectious morbidities (clinical and/or laboratory features of urinary tract infection as well as puerperal sepsis).
| Results|| |
In this study, the sociodemographic, physical and obstetrics characteristics of the patients are depicted in [Table 1]. The important pre-, intra- and post-operative variables are shown in [Table 2] and [Table 3]. Three patients developed complications (1.5% complication rate) intraoperatively and were excluded from the study. Eight women (six in Group B and two in Group A) were lost to postnatal follow-up and could not be reachable by telephone. The results in the intention-to-treat analysis [Table 4] were comparable to that in the per-protocol analysis [Table 5].
|Table 1: Sociodemographic, physical and obstetric characteristics of the patients|
Click here to view
|Table 4: Post-operative management outcome variables for safety (intention-to-treat analysis)|
Click here to view
|Table 5: Post-operative management outcome variables for safety, per protocol|
Click here to view
Eleven of the patients initially randomised into the short-stay protocol group dropped out (they did not follow through with the earlier documented protocol). All members of the traditional group remained and followed the allocated protocol till discharge. One of the patients lost to postnatal follow-up in the short-stay group had a wound infection. The observation was initially made in the ward before she was eventually lost to follow-up (this patient also dropped out of short-stay group). This made the total number of patients observed for wound sepsis (98) in the short-stay group (instead of 97 after removing the complicated case and the 2 lost to follow-up). There was no case of complete wound dehiscence. [Figure 1] illustrates the patients' flow from randomisation to analysis.
The mean hospital bill for the short-stay group was = N = 80,930 ± 10.979 ($224.80 ± 0.03) and median = N = 83,300 ($231.39). The mean hospital bill for the traditional group was = N = 92,636 ± 17.956 ($257.32 ± 0.05) and median = N = 90,300 ($250.83). In the intention-to-treat analysis, this difference was statistically significant (t = 5.53, P < 0.0001). In the per-protocol analysis, the mean hospital bill for the short-stay group was = N = 80,536 ± 10.998 ($223.71 ± 0.03) and median = N = 83,200 ($231.11). The mean and median for the traditional group remained the same as with previous analysis. This difference was also statistically significant (t = 5.47, P < 0.0001).
| Discussion|| |
The study showed that the short-stay protocol was preferable to the traditional protocol in terms of post-operative pain and hospital costs. Both protocols were comparable in terms of other safety parameters such as wound and other infection rates. These findings are important in our environment where there is so much aversion for CS with the belief that it is painful, costly, unsafe and leads to prolonged hospital stay. Although some studies have clearly shown hospital cost savings in short-stay and early postnatal discharge, they were, however, uncertain about maternal and neonatal outcomes and express concern about the possibility of compromise to quality of care.,, This study, however, showed that, though the short-stay protocol provided a significant reduction in the cost of healthcare, it in fact also showed a reduction in post-operative pain and no compromise to the quality of care in terms of safety parameters.
There were no significant differences in physical and sociodemographic characteristics as well as parity distributions between the two groups. This lack of differences could be explained by the randomisation process which would have eliminated selection bias. The difference in mean gestational age at delivery between the two groups was found to be statistically significant, but values were at term for both groups. This minimised the significant occurrence or difference in the rate of preterm birth, which is a factor associated with an increased risk of post-op infection.
The various indications for the caesarean deliveries were comparable in both groups. Although more consultants performed the surgeries and general anaesthesia was more frequently used in the traditional group than the short-stay group, the differences were not statistically significant. It has been stated that delivery by an inexperienced surgeon was associated with an increased risk of post-op infection.,, In this study, however, all the other surgeons were experienced 2nd-year senior registrars at the very least. The anaesthetic technique used for most of the participants was the regional subarachnoid block which is the safest as it encourages early post-op ambulation so as to minimise the risk of venous thrombosis., Studies have also shown that, when compared to general anaesthesia, the regional method for caesarean delivery seemed to be associated with shorter duration of hospital stay. In this study, general anaesthesia featured in only five participants; three who had failed subarachnoid block; another with major placenta praevia and questionable haemodynamic stability as well as a case with a co-existing huge right ovarian mass in whom additional procedures (right ovariectomy and peritoneal washings) were anticipated.
Most of the participants had modified Pfannenstiel or Joel–Cohen incisions, while six had subumbilical midline ones. All those who had the latter incision were done along a previous midline scar. The Joel–Cohen-based techniques for caesarean delivery, which include blunt separation of tissues along natural tissue planes with minimum sharp dissection and non-closure of both layers of the peritoneum, have advantages over subumbilical midline in terms of short-term outcomes such as operation time, fever and estimated blood loss., Some studies have examined and suggested this technique as a component of packages for enhanced recovery after elective caesarean delivery, which may reduce the length of hospital stay.,, From the post-op wound inspection in this study, it is likely that surgeons used the Joel–Cohen incisions but recorded them as Pfannenstiel. Nevertheless, the difference in the types of skin incision between the two groups was not statistically significant. The difference in the extent of adhesions (anterior abdominal wall and intraperitoneal), surgery duration and estimated blood losses was also not statistically significant. Prolonged duration of surgery has been associated with post-operative morbidity and blood loss at surgery with an increased risk of post-operative infection.,, Most of the women had closure of the subcutaneous layer, although routine closure of the subcutaneous tissue space was found not to reduce the incidence of wound infection., The difference between the two groups was not statistically significant.
The multiple-drug and multiple-dose regimen for prophylactic antibiotics (including metronidazole) administered intravenously was used for all the patients because it is the routine practice in our environment. The study was not controlled for the other types of antibiotics used. Antibiotics were often commenced in the immediate post-op period with only a few pre- or intra-operatively. Previous studies have included post-operative antibiotic prophylaxis in the package for enhanced recovery following elective CS to reduce the length of hospital stay and stressed that the absence of peri-operative complications will enhance early discharge, as well as reduce the risk of prolonged length of stay.,, There is the possibility that some participants might have subclinical medical conditions which may have manifested post-operatively and resulted in the dropouts. To minimise this, a brief medical history was taken at the point of recruitment for the study.
A limitation was in the ability to verify phone numbers of the participants at the point of recruitment for the study because most of the participants were not with their phones. This made follow-up with phone calls difficult in those who did not report for postnatal follow-up and had wrong phone numbers. In addition, there was an inability to predict incidental or accidental complications which occurred post-operatively that might prolong hospital stay (such as post-operative adhesive intestinal obstruction from one or more previous abdominal surgeries). Febrile morbidity occurring at home after discharge might be subjective because the participant might not have checked her temperature with a thermometer. For this, digital thermometers were provided for all the participants, and they were told to record the temperatures obtained and report abnormal measurements.
| Conclusions|| |
In this study, following uncomplicated CS, the short-stay protocol had lower post-op pain scores and more cost-effectiveness when compared to the traditional protocol. Safety outcomes (febrile and infectious morbidity) were, however, the same for both protocols. Bearing in mind that CS remains the most common major obstetric procedure, it is recommended that hospitals adopt the short-stay post-op management protocol. This may result in increased acceptability in the general population and reduction in workload for caregivers as well as intrinsic costs for patients.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
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 2003;80:345-50.
Buchmann EJ, Pillay N, Katusiime C. Early discharge from hospital after caesarean section at Chris Hani Baragwanath Hospital. S Afr J Obstet Gynaecol 2011;17:17-8.
Corso E, Hind D, Beever D, Fuller G, Wilson MJ, Wrench IJ, et al.
Enhanced recovery after elective caesarean: A rapid review of clinical protocols, and an umbrella review of systematic reviews. BMC Pregnancy Childbirth 2017;17:91.
CAESAR Study Collaborative Group. Caesarean section surgical techniques: A randomised factorial trial (CAESAR). BJOG 2010;117:1366-76.
Campbell OM, Cegolon L, Macleod D, Benova L. Length of stay after childbirth in 92 countries and associated factors in 30 low-and middle-income countries: Compilation of reported data and a cross-sectional analysis from nationally representative surveys. PLoS Med 2016;13:e1001972.
Bravo P, Uribe C, Contreras A. Early postnatal hospital discharge: The consequences of reducing length of stay for women and newborns. Rev Esc Enferm USP 2011;45:758-63.
Bowers J, Cheyne H. Reducing the length of postnatal hospital stay: Implications for cost and quality of care. BMC Health Serv Res 2016;16:16.
Coates E, Fuller G, Hind D, Wrench IJ, Wilson MJ, Stephens T, et al.
Enhanced recovery pathway for elective caesarean section. Int J Obstet Anesth 2016;27:94-5.
Blumenfeld YJ, El-Sayed YY, Lyell DJ, Nelson LM, Butwick AJ. Risk factors for prolonged postpartum length of stay following cesarean delivery. Am J Perinatol 2015;32:825-32.
Kim SJ, Kim SJ, Han KT, Park EC. Medical costs, cesarean delivery rates, and length of stay in specialty hospitals vs. non-specialty hospitals in South Korea. PLoS One 2017;12:e0188612.
Fasubaa OB, Ogunniyi SO, Dare FO, Isawumi AI, Ezechi OC, Orji EO, et al.
Uncomplicated caesarean section: Is prolonged hospital stay necessary? East Afr Med J 2000;77:448-51.
Kirkwood BR, Sterne JA, editors. Calculations of required sample size. In: Essential Medical Statistics. 2nd
ed. Oxford: Blackwell Science Ltd.; 2003. p. 413-28.
Abramson JH. WINPEPI (PEPI-for-windows): Computer programs for epidemiologists. Epidemiol Perspect Innov 2004;1:6.
Otubu JA. Caesarean section. In: Agboola A, editor. Textbook of Obstetrics and Gynaecology for Medical Student. 2nd
ed. Ibadan: Heinemann Educational Books (Nigeria) Plc.; 2006. p. 216.
Hema KR, Johanson R. Techniques for performing caesarean section. Best Pract Res Clin Obstet Gynaecol 2001;15:17-47.
Gibb D. Operative Obstetrics. In: Maxwell DJ, editor, 1st
ed. Surgical Techniques in Obstetrics and Gynaecology. London: Churchill Livingstone; 2004. p. 57-80.
National Collaborating Centre for Women's and Children's Health. Caesarean section. London: RCOG Press; 2004. p.35-90.
Abalos E. Alternative techniques and materials for caesarean section: The WHO Reproductive Health Library. Geneva: World Health Organisation. 2011.
Villar J, Valladares E, Wojdyla D, Zavaleta N, Carroli G, Velazco A, et al.
Caesarean delivery rates and pregnancy outcomes: The 2005 WHO global survey on maternal and perinatal health in Latin America. Lancet 2006;367:1819-29.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]