Fixed time interval compared with on-demand oral analgesia protocols for post-caesarean pain: a randomised controlled trial

DOI: 10.1111/1471-0528.14546 www.bjog.org General obstetrics Fixed time interval compared with on-demand oral analgesia protocols for post-caesarean pain: a randomised controlled trial E Yefet, a H Taha, a R Salim, a,b J Hasanein, c Y Carmeli, d N Schwartz, e Z Nachum a,b a Department of Obstetrics and Gynecology, Emek Medical Center, Afula, Israel b Rappaport Faculty of Medicine, Technion, Haifa, Israel c Neonatology Department, Emek Medical Center, Afula, Israel d Obstetrical Anesthesiology Unit, Department of Anesthesiology, Emek Medical Center, Afula, Israel e Research Authority, Emek Medical Center, Afula, Israel Correspondence: E Yefet, Department of Obstetrics & Gynecology, Emek Medical Center, Yitzhak Rabin Boulevard 21, Afula 18101, Israel. Email enavy1@gmail.com Accepted 25 December 2016. Published Online 25 February 2017. Objectives To compare the efficacy, safety and satisfaction from two modes of oral analgesia administration for the treatment of post-caesarean pain in the first 48 h following surgery: ondemand versus fixed time interval administration. Design Open label parallel-group, randomised-controlled trial from February to December 2013. Setting University-affiliated hospital in Israel. Population Two-hundred women who underwent caesarean delivery with regional anaesthesia. Methods Patients were randomly assigned to receive predetermined combinations of tramadol, paracetamol and diclofenac either following patient demand or at predetermined 6-h intervals for the first 48 h. If the patient requested additional analgesia, Percocet (oxycodone and paracetamol) was given as a rescue treatment. Main outcome measures Pain intensity and satisfaction were selfevaluated with visual analogue scale of 0 (no pain/least satisfaction) to 10 (worst pain/highest satisfaction). Breastfeeding, need for supplemental formula, and maternal and neonatal adverse effects were also evaluated. Results The fixed time interval group, compared with the ondemand group, had lower mean pain score (2.8 0.84 versus 4.1 0.48, respectively; P < 0.0001), higher satisfaction rate (9.1 1.2 versus 8.3 1.5, respectively; P < 0.0001), more breastfeeds (23.7 6.5 versus 19.2 6.2, respectively; P < 0.0001) and less use of supplemental formulas (8.2 5.2 versus 11.9 6.5, respectively; P < 0.0001). The number of times that drugs were given was slightly higher in the fixed time interval group without an increase in maternal adverse effects, which were mild. No adverse effects were reported for the neonates. Conclusion Administration of oral analgesia in fixed time intervals is superior to drug administration following patient demand without increasing maternal or neonatal adverse outcomes. Keywords Breastfeeding, caesarean sections, oral analgesia, postcaesarean pain. Tweetable abstract Oral analgesia in fixed time intervals is superior to analgesia following demand. Linked article: This article is commented on by A Butwick, p. 1071 in this issue. To view this mini commentary visit https:// doi.org/10.1111/1471-0528.14597. Please cite this paper as: Yefet E, Taha H, Salim R, Hasanein J, Carmeli Y, Schwartz N, Nachum Z. Fixed time interval compared with on-demand oral analgesia protocols for post-caesarean pain: a randomised controlled trial. BJOG 2017;124:1063 1070. Introduction Caesarean sections (CSs) are one of the most prevalent surgeries worldwide. 1 3 Inadequate maternal pain treatment was suggested to impair maternal bonding, and decrease breastfeeding and infant care. 4 Also, it impairs maternal This study was presented as an oral presentation at the Society for Maternal- Fetal Medicine 35th Annual Meeting, San Diego, California, February 2015. Trial registration ClinicalTrials.gov Identifier: NCT01764048 mobilisation leading to increased risk for thromboembolism and other postoperative complications. 5,6 Several national agencies have laid out goals for postoperative pain scores stressing the importance of postoperative analgesia to be optimised. Those recommendations include pain score <3 at visual analogue scale (VAS) by the Joint Commission 7 and 90% patients having a pain score <3 and being satisfied with their pain management by the Royal College of Anaesthetists. 8 Several methods for post-cs pain management are available, including intravenous opioids ª 2017 Royal College of Obstetricians and Gynaecologists 1063

Yefet et al. and continuous epidural analgesia. These methods increase patient immobility due to cumbersome equipment and may impair maternal function as a result of adverse affects of leg numbness, nausea, vomiting, constipation, pruritus, urinary retention, respiratory depression and sedation. 9,10 Administration of spinal morphine during CS is effective, but it is associated with adverse effects such as pruritus, nausea or vomiting, and it is not applicable in cases of general anaesthesia. Oral analgesia may have the advantages of ease of administration, need for less cumbersome equipment, low cost and ability for use after general anaesthesia. Although a few studies demonstrated oral analgesia to be effective for post-cs pain, 11,12 several issues still exist. First, it has become more acceptable to use a strategy for pain prevention rather than pain treatment in protocols for pain management. This can be achieved by giving pain medications in predetermined fixed time intervals rather than only following patient demand. The data regarding the preferred method for the treatment of post-cs pain either in fixed time intervals or following patient demand are inconclusive and are based on studies with small sample size and contradictory results. 13,14 Second, adequate pain relief, defined as VAS for pain below 3, was not achieved in previous studies. 13,14 This may be due to the use of a single analgesic rather than a combination of agents that tend to have an advantage of better efficacy and fewer adverse effects via synergism. Finally, the compatibility of pain medications with breastfeeding was not sufficiently addressed. We hypothesised that treatment of post-cs pain with oral analgesia comprised of a combination of agents given at fixed time intervals will be superior to administration only following patient demand. In the present study, we aimed to compare the efficacy, safety and satisfaction from a combination of oral pain relievers given in two modes for the treatment of post-cs pain in the first 48 h following surgery: on patient demand versus fixed time interval. Methods Design An open-label parallel-group, randomised-controlled trial was conducted at Emek Medical Center, a universityaffiliated hospital in Israel. This study was authorised by the local review board at Emek Medical center (ID of the approval EMC-120-12) and was registered on www.clinical trials.gov (NCT01764048, principal investigator EY, registration: 1 February 2013). Participants provided written informed consent. We recruited women before CS delivery with regional anaesthesia from February to December 2013. Eligibility and all baseline measures were taken before randomisation. We excluded women who suffered from chronic pain, women using chronic pain medications, known allergy to any drug used in the study, women who were scheduled or eventually underwent general anaesthesia during the surgery, who delivered vaginally, or women with abnormal liver functions. Randomisation of the groups was performed in a 1:1 ratio in blocks of 10 using a computer randomisation sequence generation program, and the randomisation results were kept in the labour and delivery ward in a closed study box in sealed opaque envelopes. The randomisation sequence was concealed until intervention was assigned. Patients were enrolled and interventions were assigned by physicians that were listed as investigators in this study. Following randomisation, the patients were allocated into two groups receiving pain relievers either in predetermined fixed time intervals or only following patient demand. Those protocols were followed from the time the patient arrived at the maternity ward, which was 2 h after surgery, for 48 h. Interventions All the study participants received spinal anaesthesia with fentanyl 25 lg and Bupivacaine 10 mg (isobaric) for the surgery. In the recovery ward, the patients received one tablet of Percocet (oxycodone 5 mg and paracetamol 325 mg). No cases of patients inability to receive enteral analgesia were reported. Treatment protocols were as follows. 1 Fixed time interval group once the patient arrived at the maternity ward she received intravenous tramadol hydrochloride 100 mg (the only time an intravenous medication was used), a tablet of paracetamol 500 mg and a tablet of diclofenac 100 mg. Six hours after patient arrival and every 6 h the patient received two tablets of Zaldiar (each tablet contained paracetamol 325 mg and tramadol 35.5 mg). The patient also received a tablet of diclofenac 100 mg 12, 24 and 48 h from arrival. 2 On-demand group patients allocated to this group received the same medications in the same combinations and order as described in the fixed time interval group protocol, only patients in this group received pain treatment only following demand, and the time intervals described above were considered as the minimal time for giving the next combination of drugs. In both groups, if the patients required additional pain relievers, they were given a tablet of Percocet (oxycodone 5 mg and paracetamol 325 mg) as necessary up to four times per day. In the on-demand group, this treatment was given if the patient requested additional pain relievers prior to 6 h past the last treatment. 1064 ª 2017 Royal College of Obstetricians and Gynaecologists

Oral analgesia for post-caesarean pain Study outcomes The primary outcome was the mean pain intensity during 48 h following surgery. Pain intensity (taken at rest) was measured using the self-reporting VAS, in which 0 represents no pain and 10 represents the worst pain. Pain scores were measured by nurses that prior to the study had a comprehensive guidance regarding the study protocol, including times and method for pain score measurements. Pain scores were taken at rest for standardisation. Pain measurements were taken at least every 4 h, before each time pain reliever was administered and an hour afterwards. We also calculated the sum of time-weighted VAS pain scores as an area under the curve (AUC) during the 48 h post-cs for each participant. Those curves were used to calculate the mean AUC for each group. In addition to the mean pain score during the 48 h post-cs, we also calculated the mean pain scores in blocks of 6 h by addressing the available measurements in each period. [i.e. from patient arrival (time 0) to the first 6 h following patient arrival (0 6), 6 12, 12 18, 18 24 h, etc.). Patients without VAS measurements were excluded from the analysis. At the end of the study period (i.e. 48 h), each participant was asked to rank her degree of satisfaction from the pain management protocol using a scale of 0 (least satisfied) to 10 (most satisfied). Additional outcomes were the number of times that pain medications were given, the number of times that rescue doses were given, the rate and type of maternal adverse effects such as gastrointestinal discomfort, nausea and vomiting, and altered consciousness. During this study, neonatal outcomes were also assessed. Those included adverse effects with possible relation to the pain medications or dehydration due to inadequate breastfeeding. The neonatal outcomes were documented by a physician from the department of neonatology who was an investigator in this study, without knowledge regarding which group the mother was allocated to. Neonatal outcomes included altered consciousness, irritability, drowsiness, maximal neonatal bilirubin (mg/dl) and need for phototherapy. Finally, the number of breastfeeds and supplemental formula used during the 5 days post-surgery were evaluated. This was done by a nursing staff of the neonatology department that was not aware of which group the mother was allocated to. Statistical analysis Sample size Internal data suggested that the mean VAS pain score of patients receiving analgesia following demand was 4 [standard deviation (SD) 2]. A sample size of 85 + 10% dropout for each group is required, assuming a mean VAS pain of 4 (SD 2) versus 3 (SD 2) for the on-demand group and the fixed time interval group, respectively, with twosided alpha of 0.05 and a power of 90%. This difference was shown to be clinically significant. 15 This sample size is also appropriate for mean satisfaction score of 7 2 versus 8 2 for the on-demand group and the fixed time interval group, respectively (two-sided alpha of 0.05 and a power of 90%). Study outcomes were compared using the Student s t-test (or the Wilcoxon two-sample test) for continuous variables and v 2 [or Fisher s exact test (two-tailed)] for categorical variables. For each comparison of VAS for pain intensity, the 95% confidence interval is also presented. Mean AUCs of each group were compared using the Student s t-test. A mean pain score was calculated for each participant at the following time intervals: 0 6, 6 12, 12 18, 18 24, 24 30, 30 36 and 36 42 h post-cs. In order to demonstrate and compare the mean VAS for pain along the hours post-cs between the study protocols, the locally scatterplot smoothing (LOESS) non-parametric regression model was utilised; 95% confidence intervals of the LOESS curves were also presented. 16 The statistical analyses were performed using SAS 9.2 software (SAS Institute, Cary, NC, USA). Statistical significance was determined if P < 0.05. Results Figure 1 shows the patient flow chart. Of the 254 patients who were screened, 252 were eligible to be included in the study. Of them, 214 underwent randomisation: 108 to the fixed time interval group and 106 to the ondemand group. Only 38 patients refused to participate in this study, and thus the recruitment rate reached 85%. After randomisation, seven patients in the fixed time interval group and five patients in the on-demand group were excluded from the study as they met the exclusion criteria (Figure 1). The rest of the patients: 100 in the fixed time interval group and 100 in the on-demand group received the treatment protocol and were included in the analysis. There were 107 and 106 newborns in the fixed time interval group and on-demand group, respectively (including seven and six pairs of twins in the fixed time interval group and on-demand group, respectively). One patient in each group was not interested in breastfeeding and therefore their babies were excluded from the analysis. The baseline characteristics of the mothers as well as the newborns are presented in Table 1. The baseline VAS score for pain intensity at patient arrival at the maternity ward in the on-demand and fixed time interval groups was comparable [4.3 (SD 2.7) versus 3.8 (SD 2.4), respectively; P = 0.3]. ª 2017 Royal College of Obstetricians and Gynaecologists 1065

Yefet et al. 254 Individuals assessed for eligibility 2 Spinal anesthesia was contraindicated 38 refused to participate 214 were randomized No. Assigned to Receive fixed time interval protocol 108 No. Assigned to Receive On demand protocol 106 101 Received fixed time interval protocol 101 Received On demand protocol 2 eventually delivered vaginally 2 eventually delivered vaginally 5 converted to general anesthesia 2 converted to general anesthesia 1 Patient reported about allergy to diclofenac only after randomization 100 Included in the analysis 1 Excluded from analysis due to the absence of VAS measurements 100 Included in the analysis 1 Excluded from analysis due to the absence of VAS measurements Figure 1. Patients flow chart. The mean VAS pain intensity during the whole study period was lower in the fixed time interval group compared with the on-demand group [Figure 2; 2.8 (SD 0.84) versus 4.1 (SD 0.48), respectively; P < 0.0001]. The mean AUC for pain scores across the 48 h post-cs was significantly lower for the fixed time interval group compared with the on-demand group [141 (SD 52) versus 206 (SD 53), respectively; P < 0.0001]. The mean VAS for pain intensity during the whole study period in blocks of 6 h of the two study groups are presented in Figure 2 and Table S1. In all of the blocks, the mean VAS for pain intensity was lower in the fixed time interval group compared with the ondemand group (P < 0.0001). Mean VAS 1 h after medication administration was also lower in the fixed time interval group compared with the on-demand group [1.5 (SD 0.5) versus 2.2 (SD 0.6), respectively; P < 0.0001]. The satisfaction score was higher in the fixed time interval group (Table 2). The number of times that drugs were given was slightly higher in the fixed time interval group. The rate of adverse effects was low and comparable between the groups (Table 2). One woman in each group developed a skin reaction suspected to be allergic following administration of either paracetamol (in the fixed time interval group) or Zaldiar (in the on-demand group). One patient in the on-demand group with known cardiac disease had bradycardia. Because the differential diagnosis included tramadol administration, it was discontinued. However, because the heart rate did not change following discontinuation, tramadol administration was ruled out as the cause of this condition. The other reported adverse effects in this study were nausea and vomiting [7 (7%) and 10 (10%) in the on-demand group versus the fixed time interval group, respectively; P = 0.45]. In the fixed time interval group, the women breastfed more and the babies needed less supplemental formulas (Table 2). No neonatal adverse effects were reported (Table 2). 1066 ª 2017 Royal College of Obstetricians and Gynaecologists

Oral analgesia for post-caesarean pain Table 1. Demographic and obstetric characteristics of mothers and neonates Characteristics On-demand Fixed time intervals Maternal N = 100 N = 100 Age (years) 31.5 (5.3) 31.5 (5.1) Age 35 years 29 (29%) 33 (33%) No. of previous CS 2.2 (1.1) 2.2 (1.0) First CS 26 (26%) 30 (30%) Emergent CS 19 (19%) 16 (16%) Gestational week 38.5 (1.0) 38.4 (1.3) Gestational week <37 3 (3%) 4 (4%) No low-transverse CS scar 0 2 (2%) Body mass index 25.1 (4.9) 25.9 (5.6) (pre-pregnancy) Diabetes mellitus 11 (11%) 13 (13%) Gestational diabetes mellitus 10 (10%) 12 (12%) Pre-gestational diabetes mellitus 1 (1%) 1 (1%) Pre-eclampsia 3 (3%) 3 (3%) Chronic hypertension 0 1 (1%) Neonatal N = 106 N = 105 Birth-weight (g) 3181 (567) 3232 (570) Apgar score 1 min from delivery 9.5 (0.7) 9.4 (0.6) Apgar score 5 min from delivery 9.9 (0.3) 10 (0.2) Umbilical artery ph 7.30 (0.04) 7.30 (0.06) CS, caesarean section. Values are given as mean (SD) or number (percent). Discussion Main findings In the present study, we compared the efficacy, safety and satisfaction of two strategies of oral analgesic administration for the treatment of post-cs pain in the first 48 h following surgery: on patient demand versus fixed time interval. Our results demonstrate that oral analgesia administered at a fixed time interval protocol is superior to administration only after patient demand according to pain intensity, patients satisfaction scores and the quality of breastfeeding. The maternal adverse effects profile was low and comparable in both methods of administration. Finally, no adverse effects related to the study medications were reported for the neonates. Strengths and limitations The strengths of this study are the high recruitment rate, which decreased the selection bias; the use of multiple aspects for evaluating the preferred protocol, including pain relief efficacy, patient satisfaction, safety issues and breastfeeding; and the prospective randomised controlled design. The limitations of this study are that it was a single site study, the focus on short-term outcomes and indirect assessment of patient mobility. Because mobility is an important factor in preventing post-cs complications, this objective should be the focus of future studies. In addition, patients and providers were not blinded, which might subject the results to bias. An additional limitation is the Table 2. Maternal and neonatal outcomes On demand Fixed time interval P value N = 100 N = 100 # of times the medications were given 6.1 (1.6) [6, 5 7] 7.2 (1.6) [8, 7 8] 0.038 # of rescue treatments 0.90 (1.1) [1, 0 1] 0.94 (1.0) [1, 0 2] 0.6 Patients satisfaction score from the treatment protocol (scale 0-10)* 8.3 (1.5) [8, 7 10] 9.1 (1.2) [10, 9 10] <.0001 # of patients who reported adverse effects during hospitalization, 9 (9%) 11 (11%) 0.6 # of breastfeeds during hospitalization 19.2 (6.2) [19, 14 23] 23.8 (6.5) [24, 20 27] <.0001 # of supplemental formula given during hospitalization 11.9 (6.5) [11.5, 8 15] 8.2 (5.2) [7, 4 11] <.0001 Maximal neonatal bilirubin (mg/dl) 8.4 (4) [9, 5.7 10.9] 9.3 (3.1) [10, 7.6 11.5] 0.06 Neonates needing phototherapy, 12 (11%) 11 (10%) 0.9 Neonatal irritability or altered consciousness 0 0 Change in neonatal weight at the time of discharge (grams) 179.2 (204) [185.5, 101 250] 185 (131) [205, 125 250] 0.2 VAS, visual analog scale. *Ninety nine and ninety one women gave satisfaction scores in the on demand and fixed time interval groups, respectively. Details regarding the adverse effects are elaborated in the result section. Relative risk (RR) = 1.22 [95%CI: 0.53 2.82]. Neonatal outcomes N = 105 and N = 106 newborns in the on demand group and fixed time interval group, respectively (one patient in each group was not interested in breastfeeding and therefore their babies were excluded from the analysis). RR = 0.95 [95%CI: 0.44 2.06]. Continuous variables are presented with mean (SD) [median, IQR]. ª 2017 Royal College of Obstetricians and Gynaecologists 1067

Yefet et al. 7 LOESS/SMOOTH=0.3: Time (hours) since surgery Vs. VAS 6 5 4 VAS 3 2 1 0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 Time (hours) since surgery Protocol On-demand Fixed time intervals Figure 2. The reported post-cs pain in the on-demand group compared with the fixed time interval group. LOESS curve representing the mean pain intensity using the VAS. The mean VAS was calculated for every 6-h interval since patient admission until the end of the study period. Baseline VAS at patient arrival to the maternity ward was comparable between the groups (P = 0.3). After CS, mean VAS scores were significantly lower in the fixed time interval group compared with the on-demand group (P < 0.0001; table S1). anaesthetic and post-analgesic regimens used in this study that may not be in line with those used in other institutions. For example, in this study patients did not receive intrathecal long-acting narcotics (e.g. morphine, diamorphine), which were shown to substantially improve pain scores and reduce the need for additional analgesia post-cs compared with no long-acting narcotic. 17 19 Moreover, some centres may advocate protocols that are not based on opiates, as data suggest that a combination of tramadol and non-steroidal anti inflammatory drugs (NSAIDs) has a similar analgesic profile to acetaminophen/nsaid combination post- CS, but those receiving tramadol have a sevenfold higher incidence of nausea. 20 Nevertheless, the principle of using analgesia in fixed time intervals rather than only following patient demand is probably relevant to institutions that use different medications to alleviate post- CS pain, as in this way pain sensation can be better prevented. Interpretation Oral analgesia to treat post-cs pain was suggested previously. In a non-randomised trial that examined post-cs pain, treatment with oral analgesia comprised of immediate-release morphine sulphate with dipyrone as a rescue analgesia and vice versa, both protocols received high satisfaction scores. 21 VAS scores of patients receiving oral oxycodone were comparable to intravenous Piritramide, with a trend toward additional rescue medications in the intravenous group. 12 In the study by Davis et al., oral oxycodone acetaminophen administration was superior to a morphine patient-controlled analgesia device for pain relief. The oral analgesia group also had less nausea and drowsiness at 6 h and slightly more nausea at 24 h, suggesting a tolerable adverse effect profile. 22 A combination of oral oxycodone, paracetamol and tramadol was superior to epidural analgesia (0.1% ropivacaine + 0.1 lg/ml sufentanil) with less pruritus and better satisfaction scores. 11 In our study, the fixed time interval protocol was superior to the on-demand protocol for both pain and satisfaction scores. An important aspect is the result that pain intensity was also lower 1 h following administration of analgesia medications, suggesting that medications were more effective when given in a scheduled protocol. Additional advantages of a scheduled protocol are that the nursing staff can be better organised to execute predetermined orders and that the patient is less preoccupied with the need to set the time for the next medication administration. In this context, on-demand protocols may be less effective due to a lack of knowledge of the patient regarding the medications in use and the proper way to use them in combination to manage post-cs pain efficiently. Our results are supported by the non-randomised prospective study by Jakobi et al., in which fixed time interval analgesia administration yielded better pain relief and satisfaction scores than when drugs were administered only after patient demand. 21 An advantage for a scheduled 1068 ª 2017 Royal College of Obstetricians and Gynaecologists

Oral analgesia for post-caesarean pain analgesia protocol was also demonstrated in the study of Valentine et al., in which scheduled acetaminophen resulted in decreased opioid use and more consistent acetaminophen intake compared with acetaminophen administered as needed via combination acetaminophen opioid analgesics, without compromising analgesia. 23 On the contrary, in the study by Sammour et al., fixed time interval administration of oral analgesia (naproxen or tramadol) yielded comparable results to on-demand oral analgesia in term of VAS pain scores. 14 It should be noted, however, that this study comprised four groups receiving either naproxen or tramadol in fixed time intervals or following patient demand. The sample size in each group was small (N = 30) and the treatment was heterogeneous; thus, the interpretation of the results is limited. In our study we used a combination of pain relievers. This strategy enabled us to give more analgesic agents in lower doses and, therefore, possibly increasing efficacy and decreasing adverse effects via synergistic effects. This way the mean VAS pain score in the fixed time interval group was below 3, the recommended VAS intensity, making the difference with the on-demand group clinically significant. Lower VAS scores were also recorded in previous studies that used a combination of agents, 11,22 compared with other studies that used primarily single medication protocols. 13,14 We also evaluated breastfeeding by counting the number of breastfeeding events and the use of supplemental formulas. In our area it is customary for mothers to breastfeed at least in the first weeks following delivery. Except for one patient in each group, all the other mothers chose to use breastfeeding as the primary feeding route. In the fixed time interval group, the mothers breastfed more compared with the on-demand group. Such a result may suggest that the mothers suffered less pain, making breastfeeding more comfortable. On the other hand, there may have been more breastfeeds because each individual feed was less effective. In order to distinguish between the two possibilities, we also evaluated how much supplemental formula the babies consumed in each group. The use of a lesser amount of supplemental formula together with the comparable infant weight difference of the two groups suggests that the breastfeeding quality was better in the fixed time interval group. Information regarding the breastfeeding quality in different pain reliever protocols is scarce, as most of the studies were satisfied with a statement that the research medications are approved in breastfeeding. In the study by Sammour et al., they used a breastfeeding score of 1, 2 and 3 for no breastfeeding, partial breastfeeding (supplementary formula feeding), and full breastfeeding (no formula feeding), respectively. 14 The breastfeeding scores were comparable between the fixed time interval and ondemand groups. Information regarding women s preference was not provided. Also, using a three-level scale might not have sufficient resolution for identifying differences between the groups, which had small sample size to begin with in this study. Conclusion Our results support the use of analgesia in oral administration to be effective and safe for the treatment for post- CS pain. Administration in fixed time intervals should be preferred over administration only following patient demand. This study emphasises the concept that pain management protocols should focus on pain prevention instead of pain relief. Disclosure of interests None declared. Completed disclosure of interests form available to view online as supporting information. Contribution to authorship EY, HT, ZN, RS, NS, JH and YC all made substantial contributions to the concept and design of the study. EY, NS and ZN analysed the data. EY and ZN drafted the article, and HT, RS, NS, JH and YC provided critical revisions of the article for intellectual content. All authors approved the final manuscript version. Details of ethics approval This study was authorised by the local review board at Emek Medical Center (protocol EMC-120-12; 21 January 2013). Funding None. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Mean VAS scores for pain intensity during the first 48 hours post cesarean surgery.& References 1 MacDorman MF, Menacker F, Declercq E. Cesarean birth in the United States: epidemiology, trends, and outcomes. Clin Perinatol 2008;35:293 307. 2 Sebastiao YV, Womack L, Vamos CA, Louis JM, Olaoye F, Caragan T, et al. Hospital variation in cesarean rates: contribution of individual and hospital factors in Florida. Am J Obstet Gynecol 2016;214:123.e1 18. 3 Boyle A, Reddy UM. Epidemiology of cesarean delivery: the scope of the problem. Semin Perinatol 2012;36:308 14. ª 2017 Royal College of Obstetricians and Gynaecologists 1069

Yefet et al. 4 Karlstrom A, Engstrom-Olofsson R, Norbergh KG, Sjoling M, Hildingsson I. Postoperative pain after cesarean birth affects breastfeeding and infant care. J Obstet Gynecol Neonatal Nurs 2007;36:430 40. 5 Leung AY. Postoperative pain management in obstetric anesthesia new challenges and solutions. J Clin Anesth 2004;16:57 65. 6 Toglia MR, Nolan TE. Venous thromboembolism during pregnancy: a current review of diagnosis and management. Obstet Gynecol Surv 1997;52:60 72. 7 National Pharmaceutical Council. Pain: Current Understanding of Assessment, Management, and Treatments (Monograph). Washington, DC: Joint Commission on Accreditation of Healthcare Organizations, 2001. 8 Pickering EH. A pain relief after Caesarean section. In: Kinsella M, editor. Raising the Standard: A Compendium of Audit Recipes, 2nd edn. London: Royal College of Anaesthetists; 2006. pp. 168 9. 9 Gadsden J, Hart S, Santos AC. Post-cesarean delivery analgesia. Anesth Analg 2005;5(Suppl):S62 9. 10 Lowder JL, Shackelford DP, Holbert D, Beste TM. A randomized, controlled trial to compare ketorolac tromethamine versus placebo after cesarean section to reduce pain and narcotic usage. Am J Obstet Gynecol 2003;189:1559 62. 11 Zhong TD, Liu Q, Zhao JN, Wang HW, Konstantatos A. A randomized trial to compare pain control using oral analgesia with epidural analgesia after cesarean section following combined spinalepidural anesthesia. Genet Mol Res 2014;13:7086 93. 12 Dieterich M, Muller-Jordan K, Stubert J, Kundt G, Wagner K, Gerber B. Pain management after cesarean: a randomized controlled trial of oxycodone versus intravenous piritramide. Arch Gynecol Obstet 2012;286:859 65. 13 Jakobi P, Solt I, Tamir A, Zimmer EZ. Over-the-counter oral analgesia for postcesarean pain. Am J Obstet Gynecol 2002;187:1066 9. 14 Sammour RN, Ohel G, Cohen M, Gonen R. Oral naproxen versus oral tramadol for analgesia after cesarean delivery. Int J Gynaecol Obstet 2011;113:144 7. 15 Gallagher EJ, Liebman M, Bijur PE. Prospective validation of clinically important changes in pain severity measured on a visual analog scale. Ann Emerg Med 2001;38:633 8. 16 Jacoby WG. Loess: a nonparametric, graphical tool for depicting relationships between variables. Elect Stud 2000;19:577 613. 17 Dahl JB, Jeppesen IS, Jorgensen H, Wetterslev J, Moiniche S. Intraoperative and postoperative analgesic efficacy and adverse effects of intrathecal opioids in patients undergoing cesarean section with spinal anesthesia: a qualitative and quantitative systematic review of randomized controlled trials. Anesthesiology 1999;91:1919 27. 18 Lane S, Evans P, Arfeen Z, Misra U. A comparison of intrathecal fentanyl and diamorphine as adjuncts in spinal anaesthesia for Caesarean section. Anaesthesia 2005;60:453 7. 19 Palmer CM, Emerson S, Volgoropolous D, Alves D. Dose-response relationship of intrathecal morphine for postcesarean analgesia. Anesthesiology 1999;90:437 44. 20 Mitra S, Khandelwal P, Sehgal A. Diclofenac-tramadol vs. diclofenacacetaminophen combinations for pain relief after caesarean section. Acta Anaesthesiol Scand 2012;56:706 11. 21 Jakobi P, Weiner Z, Solt I, Alpert I, Itskovitz-Eldor J, Zimmer EZ. Oral analgesia in the treatment of post-cesarean pain. Eur J Obstet Gynecol Reprod Biol 2000;93:61 4. 22 Davis KM, Esposito MA, Meyer BA. Oral analgesia compared with intravenous patient-controlled analgesia for pain after cesarean delivery: a randomized controlled trial. Am J Obstet Gynecol 2006;194:967 71. 23 Valentine AR, Carvalho B, Lazo TA, Riley ET. Scheduled acetaminophen with as-needed opioids compared to as-needed acetaminophen plus opioids for post-cesarean pain management. Int J Obstet Anesth 2015;24:210 16. 1070 ª 2017 Royal College of Obstetricians and Gynaecologists