Kintu et al.

BMC Health Services Research (2019) 19:68

https://doi.org/10.1186/s12913-019-3911-x

RESEARCH ARTICLE Open Access

Postoperative pain after cesarean section:

assessment and management in a tertiary
hospital in a low-income country
Andrew Kintu1* , Sadiq Abdulla2, Aggrey Lubikire1, Mary T. Nabukenya1, Elizabeth Igaga1, Fred Bulamba3,
Daniel Semakula4 and Adeyemi J. Olufolabi5

Abstract
Background: There is little information about the current management of pain after obstetric surgery at Mulago
hospital in Uganda, one of the largest hospitals in Africa with approximately 32,000 deliveries per year. The primary
goal of this study was to assess the severity of post cesarean section pain. Secondary objectives were to identify
analgesic medications used to control post cesarean section pain and resultant patient satisfaction.
Methods: We prospectively followed 333 women who underwent cesarean section under spinal anesthesia. Subjective
assessment of the participants’ pain was done using the Visual Analogue Scale (0 to 100) at 0, 6 and 24 h after surgery.
Satisfaction with pain control was ascertained at 24 h after surgery using a 2-point scale (yes/no). Participants’ charts were
reviewed for records of analgesics administered.
Results: Pain control medications used in the first 24 h following cesarean section at this hospital included diclofenac
only, pethidine only, tramadol only and multiple pain medications. There were mothers who did not receive any
analgesic medication. The highest pain scores were reported at 6 h (median: 37; (IQR:37.5). 68% of participants reported
they were satisfied with their pain control.
Conclusion: Adequate management of post-cesarean section pain remains a challenge at Mulago hospital. Greater
inter-professional collaboration, self-administered analgesia, scheduled prescription orders and increasing availability of
analgesic drugs may contribute to improved treatment of postoperative pain with better pain scores.
Keywords: Cesarean section, VAS, Postoperative pain, Post-caesarean section pain, Mulago hospital

Background In low-income countries, postoperative pain manage-

Inadequately treated postoperative pain can contribute ment can be particularly challenging for several reasons
significantly to morbidity of surgical patients, resulting including the expectation of postoperative pain by
in the delay of patients’ recovery and ability to return to patients (thereby making no effort to request for pain
daily functional activities [1]. Early recovery is especially relief ) and the high patient-to-nurse ratio that limits
important for a patient who is expected to take care of assessment of pain and administration of adequate pain
her newborn shortly after an operative procedure. relief medication [4]. At Mulago National Referral
Evidence from studies done in high income settings has Hospital, a tertiary teaching hospital, it has been shown
demonstrated that inadequately treated pain after that patients have inadequate pain control after ortho-
cesarean section is associated with an increased inci- pedic and general surgery [5], but little is known about
dence of chronic pain [2] and post-traumatic stress syn- the postoperative pain management after obstetric sur-
drome [3]. gery. We hypothesized that women undergoing cesarean
section do not have satisfactory postoperative pain man-
* Correspondence: kintuug2012@gmail.com
agement in the first 24 h after surgery.
1
Department of Anesthesia, Makerere University College of Health Sciences, The primary goal of the study was to determine the
Mulago, Uganda current pain status and management following cesarean
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Kintu et al. BMC Health Services Research (2019) 19:68 Page 2 of 6

section at Mulago National Referral Hospital, Kampala, using hyperbaric bupivacaine 10 to 12 mg. 124 patients
Uganda. The secondary objectives were to determine pa- did not meet inclusion criteria due to refusal to consent,
tient satisfaction with their pain control and document failed regional technique or the occurrence of CS late in
the methods and regimes of analgesia used. the day. Patient data were obtained from interviewing
patients, reviewing medical charts, and nurses’ patient
Methods records. The information was recorded in a pre-tested
We conducted a prospective descriptive hospital based questionnaire and transferred into an electronic data-
survey. The study was conducted in the obstetric ward base. Personal identification information was replaced
at Mulago National Referral Hospital in Kampala, with anonymous participant number to mark the ques-
Uganda between November 2014 and January 2015. tionnaires. Participants who met study criteria had their
Mulago National Referral hospital was built with a 1500 pain assessed using the visual analogue scale (VAS)
bed capacity but currently functions as 2500 bed cap- (using a 0 to 100 scale) with 0 having no pain and 100
acity with approximately 32,000 deliveries a year, which having experienced the worst pain. The pain VAS was
includes a 20% cesarean delivery rate. self-completed by the patient who was asked to place a
line perpendicular to the VAS line at the point that rep-
Study population resented their pain intensity. Using a foot ruler, the pain
Eligible participants included all parturients undergoing score was determined by measuring the distance on the
cesarean section between 0800 and 1800 h, under spinal 100 line between the no pain anchor and the patient’s
anesthesia with an American Society of Anesthesiolo- mark, providing a range of scores from 0 to 100. Pain
gists physical status classification (ASA) of I to III and category was determined as 0–4 no pain, 5–44 mild
able to communicate freely with a non-family member pain, 45–74 moderate pain, and 75–100 severe pain.
interpreter to obtain informed consent. All parturients Pain was assessed as soon as the patient arrived at the
who experienced failed spinal anesthesia and required recovery area (0 h) and at 6 h and at 24 h in the postop-
general anesthesia were excluded from the study. erative obstetric ward.
In addition, at 24 h, all participants were asked to pro-
Sampling vide a yes or no response regarding their satisfaction
For purposes of continued and easy patient follow up, with pain control. A record of all analgesics adminis-
we only screened parturients whose cesarean section tered in the first 24 h after surgery were documented in-
were performed between 0800 and 1800 h due to avail- cluding the time of administration, name of analgesic
ability of staff responsible for consenting patients. Con- and prescriber. During the entire study period, the study
secutive sampling was used to recruit the parturients team did not interfere with pain management of the
into the study. participants.
We calculated an initial sample size of 245 participants
using the Kish and Leslie formula (1965) for a single Variables and measurements
proportion based on a previous study by Ismail et al. [6] Independent variable was the analgesic treatment group
for a power of 80% at 95% confidence. We adjusted this while the dependent variable was VAS scores at the dif-
sample size by an additional 35% to a total of 333 to ferent times of assessment.
compensate for any potential loss to follow-up and
non-response, which we assumed would be high because Statistical analysis
of the level of activity around postoperative patients with The data were exported to Stata statistical analysis soft-
newborns. ware version 12.1 for analysis. Our analysis is based on
data from all the 333 participants enrolled in the study.
Ethical approval Although we noted missing data on different variables,
Approval for this study was granted by the Mulago we determined it would be best not t to exclude any
Hospital Research and Ethics Committee. Written in- participant with missing data but include all in the
formed consent was obtained from parturients with or final analysis. Because of this, some variables have dif-
without the help of an interpreter and was certified by ferent totals.
either the patient’s signature or thumbprint. For patient characteristics, we present frequencies and
proportions. For the primary outcome, we summarized
Study procedure findings at the various time points using medians and
We screened 457 patients in the anesthesia waiting area interquartile range (IQR).
in the obstetric unit at Mulago Hospital and recruited For secondary objectives, we calculated the proportion
333 patients who were scheduled for both elective and of participants in the different treatment groups. In
emergency cesarean section under neuraxial anesthesia addition, we assessed differences in pain scores among
Kintu et al. BMC Health Services Research (2019) 19:68 Page 3 of 6

the different treatment groups and conducted a post-hoc Table 1 Baseline characteristic, indications for cesarean section,
evaluation for pairs of treatment groups found to have anesthesia providers, prescribing staff and patients whose drugs
statistically significant differences in pain scores. Assess- were administered as prescribed
ment of normality was done using graphical plots and Characteristics (N) Frequency Percentage
(n)
the Shapiro-Wilk test. Given that the data were not nor-
mally distributed we used the Kruskal-Wallis test to de- Age (N = 326) ≤20 65 20.0
termine differences in the pain scores among the 21–30 198 61
treatment groups. The data from this analysis is pre- ≥31 63 19
sented as median pain scores, and their IQR. For patient Parity* (N = 325) First time 96 30
satisfaction, a proportion of participants satisfied with pregnancy
their analgesia was estimated. 2 102 31
≥3 127 39
Results
Common indications for cesarean section**
We recruited 333 parturients in the anesthesia waiting
area in the obstetric unit. 43 participants had at least Fetal distress# No 301 90
(N = 333)
one missing variable data. All participants were included Yes 32 10
in the analysis. For clarity, we state the number of par- Obstructed labor*** No 244 23
ticipants with complete data for each outcome. (N = 333)
Yes 89 27
The baseline characteristics of the study participants Previous C-section No 210 63
and care providers are summarized in Table 1. Majority of (N = 333)
Yes 123 37
participants were 30 years or younger with at least two
previous pregnancies carried beyond 28 weeks of gesta- Big baby**** No 309 93
(N = 333)
tion. The commonest indications for cesarean delivery Yes 24 7
were previous cesarean section scar and obstructed labor. Prolonged labor## No 311 93
In this cohort, 72% (237/328) of anesthesia was provided (N = 333)
Yes 22 7
by the anesthetic officer (non-physician provider). For an- Care practices
algesic prescription, complete data was available for 290 @
Anesthesia provider Senior house 57 18
participants. Analgesia was prescribed by surgeons for (N = 328) officer
95% of the parturients (275/290). No analgesic medication @@
Anesthetic 237o 72
was prescribed by a nurse. Only 42% (123/290) of partici- officer
pants received their analgesics as prescribed (Table 1). In Anesthesiologist 34 10
all participants, the median time to first analgesic adminis-
Prescribing staff Surgeon 310 95
tration after leaving the operating room was 241 min (N = 328)
(IQR: 350). At 24 h assessment, 68% (197/290) of partici- Anesthetic 18 5
provider
pants determined pain control was satisfactory.
Patients whose Drugs No 173 58
Figure 1 shows the comparison of the distribution of were administered as
pain scores at the different time points. For this out- prescribed (N = 297) Yes 124 42
come, data was complete for all participants (333) at T0 Key: *Number of pregnancies carried above 28 weeks, **Most of the patients
hours and T6 hours but for only 301 participants at T24 had more than one indication for cesarean section, ***Obstructed labor referred
to failure of labor to progress (no change in cervical dilation and descent of fetus)
h. The median pain scores increased from 8 (IQR: 43) at despite good uterine contractions for more than 4 h, ****Big baby was any fetus
T0 hours to 37 (IQR: 37.5) at T6 hours and reduced to ≥4kgs diagnosed by ultra sound or clinical examination, #Fetal distress referred to
30 (IQR: 35) at T24 h. any fetal heart rate below 120 or above 160, ##Prolonged labor referred to any
patient in active labor (cervical dilation greater than 4) for ≥14 h, @Senior House
At all the three time points, a higher proportion of officer (SHO) is a Masters of Medicine in anesthesia trainee (also referred to as a
participants experienced no pain or mild pain. However, resident), @@Anesthetic Officer a non-physician anesthesia provider with a
diploma training in anesthesia, @@@Anesthesiologist is a physician who has
this proportion reduced markedly between T0 hours completed Masters of Medicine in anesthesia training
77% (257/333) and T6 hours 61% (203/333) and increased
at T24 h 71% (214/301). (47/327) received multiple drugs while 42% (136/327)
On the other hand, the proportion of participants ex- received none. Intramuscular diclofenac was the most
periencing severe pain increased from 11% (37/333) at T0 prescribed analgesic followed by tramadol and pethid-
to 14% (48/333) at T6 and then reduced to 6% (18/301) at ine respectively. No patient received intravenous or
T24 hours. The trend was similar for participants who intramuscular morphine.
experienced moderate pain. 4% (13/328) participants who had complete data on
In the first 24 hours following surgery, 44% (144/327) anesthesia technique received intrathecal opioids (mor-
participants received only one type of analgesic drug, 14% phine) as part of the spinal anesthetic for C- section.
Kintu et al. BMC Health Services Research (2019) 19:68 Page 4 of 6

Fig. 1 A comparative distribution of pain scores in the first 24 h after C-section

Participants were classified into six treatment groups The distribution of the pain scores in the different treat-
according to the analgesic medication they received ment groups at all the three time points did not follow a
during the study period. These groups included: diclofe- Gaussian distribution. The Kruskal-Wallis equality-of-
nac only, pethidine only, intrathecal opioid only, trama- populations rank test for the pain scores at time = T0
dol only, multiple pain medications, and no pain revealed that at least one treatment group had median
medication. The corresponding proportions at T0 h, T6 pain scores that were statistically significantly different
h and T 24 h are shown in Table 2. from those of participants in other groups (X2 (5df)
At T0 hours, the median pain scores were lowest for =16.15; p = 0.006). A further post-hoc analysis using
the group that received pethidine only (1, IQR:14) and Dunn’s pairwise comparison of treatment groups at time
the group that received a combination of analgesics (3, T0, adjusting for multiple comparisons using the Bonfer-
IQR:17) while the group that received tramadol only had roni method revealed that participants on pethidine had
the highest pain scores (31, IQR:60). At T6, the median statistically significantly lower pain scores than those on
pain scores were lowest for the group that received tramadol (Z = − 3.13, p = 0.01) or those who did not
tramadol only (30, IQR:31) while the group that received receive analgesics (Z = − 2.78, p = 0.04), while those who
intrathecal morphine had the highest pain scores (67, received multiple types of analgesics had lower pain scores
IQR:72). At T24, the median pain scores were lowest for than those who received tramadol (Z = 2.9, p = 0.028).
the group that received intrathecal morphine (22, However, there was no statistically significant difference in
IQR:27) while the group that received only tramadol had the median pain scores between the rest of the treatment
the highest pain scores (36, IQR:43) (Fig. 2). group pairs for time T0; and no statistically significant
difference in the median pain scores among the different
treatment groups at time T6 (X2(5df) =1.8, p = 0.87) and
Table 2 Management of post cesarean section pain in the first 24 h T24 (X2(5df) =2.2, p = 0.81).
Treatment group T0 (N = 319) T6 (N = 311) T24 (N = 296)
n % n % n % Discussion
Diclofenac only 80 25 77 25 76 26
This study is an exploratory survey examining the ex-
perience of post C-section pain in patients at a large ter-
Pethidine only 23 7 23 7 21 7
tiary hospital in Uganda. It highlights the gaps in pain
Intrathecal morphine only 3 1 3 1 3 1 management but also the opportunity for improvement.
Tramadol only 35 11 34 11 31 10 Many of the participants in this study did not receive
Multiple analgesicsa 46 15 45 15 43 15 their analgesics as prescribed (42%). A considerable
No analgesics 132 41 129 41 122 41 number did not get any analgesia, similar to the practice
a
Multiple analgesics included anyone who received more than one analgesic
in general surgery patients in the same hospital as well
drugs, of any kind as in some hospitals in Nigeria [5, 7]. In the end, only
Kintu et al. BMC Health Services Research (2019) 19:68 Page 5 of 6

Fig. 2 A box and whisker plot showing the distribution of median VAS scores among groups of treatment at T0, T6, T24

68% reported being satisfied with their pain control. This severe pain [12, 13]. Multimodal analgesia is currently the
discrepancy between prescription and actual administra- recommended practice [14, 15] although median pain
tion of analgesics may be due to irregular and inad- scores for participants who received single analgesic were
equate availability of medications due to hospital stock comparable to those in the multiple analgesic category
out leading to patients having to purchase their own across the three time points, in our study. Our study also
drug, high patient: nurse ratio, poor prescription prac- suggests that different types of analgesic drugs may be
tices and unfamiliarity with some modern analgesic better at different time points. For example, Pethidine
techniques and drugs [4]. The high patient: nurse ratio seem adequate analgesic if administered immediately
makes postoperative pain management challenging in following surgery while tramadol may be better if adminis-
many low-income countries. These findings are consist- tered 4–6 h after surgery. Patients who received intra-
ent with the results of Kiswezi et al, that showed inad- thecal morphine had better pain control towards the end
equate post-laparotomy pain management in the same of 24 hours. This however will need to be explored locally
hospital [5]. In that study, factors that contributed to in future studies.
poor pain management included inconsistent pain Many high-income countries use longer acting intra-
assessment and irregular supplies of prescribed drugs. thecal opioids that reduce postoperative pain in the first
Prescribing orders could also play a role as many doctors 24 hours. Despite the benefit of intrathecal morphine
write pro re nata (prn) orders rather than scheduled towards longer analgesic effects for cesarean sections
orders. Inadequate postoperative pain control seem a [16, 17], its use during the study period was minimal
universal surgical problem demonstrated in other low- possibly due to unfamiliarity of use by anesthesia
income and high-income countries [8–10]. providers. Irregular availability of opioids, persistent
The types of analgesics that were used are similar to concern of the risk of respiratory depression and the in-
those used in other low-income settings for obstetric pa- ability to monitor patients limit their use in low-income
tients [7, 11]. In the first 24 h after C-section, the largest settings. The worst pain experienced in our study popu-
number of participants received only intramuscular lation was 6 hours post- C-section likely due to spinal
diclofenac, followed by multiple drugs and only tram- anesthesia wearing off without further analgesia given as
adol in that order. The most popular route of admin- shown by the time it took for participants to receive
istration in our cohort was the intramuscular route. analgesic drugs after leaving the operating room (241
Studies have shown that use of single analgesics is min (IQR 350, range 0–950). This breakthrough pain
not effective in the management of moderate and experience is not usually experienced with intrathecal
Kintu et al. BMC Health Services Research (2019) 19:68 Page 6 of 6

morphine use. Intrathecal morphine was introduced into Consent for publication
practice at Mulago hospital during the study period and Not applicable.

supplies were irregular. As of 2018, the use of intrathecal Competing interests

morphine at this hospital has markedly increased. The authors declare that they have no competing interests.
Finally, surgeons accounted for 95% of the postopera-
tive analgesia prescriptions. This practice is similar in Publisher’s Note
other low-middle income countries [4, 6, 9, 11]. In Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
high-income countries, immediate postoperative anal-
gesia is managed by anesthesia or acute pain services Author details
1
that are experienced and trained in the use of prescribed Department of Anesthesia, Makerere University College of Health Sciences,
Mulago, Uganda. 2Department of Anesthesia, University of British Columbia
analgesia. We would suggest that this practice needs to and BC Women’s Hospital, Vancouver, Canada. 3Department of Anesthesia,
be reviewed in the light of persistent failure to ad- Busitema University Faculty of Health Sciences, Mbale, Uganda. 4Innovations
equately manage postoperative surgical pain. and Knowledge Translation Office, Makerere University College of Health
Sciences, Mulago, Uganda. 5Department of Anesthesia, Duke University,
A limitation in our study is not exploring other factors Durham, NC, USA.
that influence pain such as cultural and personal percep-
tions of staff and patients. Received: 2 November 2015 Accepted: 17 January 2019

Conclusion References
1. Barash P, Cullen Bruce F, Stoelting Robert K, Cahalan Michael K, Stock Christine
The management of post cesarean section pain in M. Clinical anesthesia. Philadelphia: Lippincott Williams & Wilkins; 2009.
Mulago hospital is inadequate demonstrated by signifi- 2. De Brito Cançado TO, Omais M, Ashmawi HA, Torres MLA. Chronic pain after
cant number of participants who received no analgesic cesarean section. Influence of anesthetic/surgical technique and postoperative
analgesia. Brazilian Journal of Anesthesiology. 2012;62(6):762–74.
medication after leaving the operating room and this 3. Maggioni C, Margola D, Filippi F. PTSD, risk factors, and expectations among
needs addressing. The majority of participants also did women having a baby: a two-wave longitudinal study. J Psychosom Obstet
not receive treatment as prescribed and overall level of Gynecol. 2006;27(2):81–90.
4. Bagi IA, Ahmed ME. Postoperative pain and analgesic prescription in
satisfaction with pain management is relatively low. Khartoum: evaluation of current practice. East Afr Med J. 1993;70(8):502–5.
Awareness and practice change need to occur. Use of 5. Kiswezi AK, Ahmed K. Evaluation of pain control following laparotomy in
multimodal analgesia would be beneficial in our setting, Mulago hospital. [Master’s thesis]. Makerere University School of. Health
Sciences. 2006. http://makir.mak.ac.ug/handle/10570/313.
and as such should be encouraged. Interprofessional col- 6. Ismail S, Shahzad K, Shafiq F. Observational study to assess the effectiveness
laboration, training in pain and pain control, ensuring of postoperative pain management of patients undergoing elective
availability of analgesic drugs, and self-administered cesarean section2012 January. 2012;1:36–40.
7. Apfelbaum JL, Chen C, Mehta SS, Gan TJ. Postoperative pain experience:
analgesia may contribute to improvements in patients’ pain results from a National Survey Suggest Postoperative Pain Continues to be
experience following post C-section at Mulago hospital. undermanaged. Anesth Analg. 2003;97(2):534–40.
8. Soyannwo O. Post-operative pain control--prescription pattern and patients'
Abbreviations experience. West Afr J Med. 1999;18(3):207–10.
C- section: Cesarean section; df: Degrees of freedom; h: Hours; IQR: Interquartile 9. Mwaka G, Thikra S, Mung'ayi V. The prevalence of postoperative pain in the
range; mg: Milligrams; min: Minutes; mm: Millimeters; T: Time; VAS: Visual first 48 hours following day surgery at a tertiary hospital in Nairobi. Afr
analogue scale Health Sci. 2013;13(3):768–76.
10. Woldehaimanot TE, Eshetie TC, Kerie MW. Postoperative pain management
among SurgicallyTreated patients in an Ethiopian hospital. PLoS One.
Acknowledgements 2014;9(7):e102835.
We would like to acknowledge Dr. Naomi Kronitz of the University of British 11. Carpenter RL. Optimizing postoperative pain management. Am Fam
Columbia for her guidance and support. Physician. 1997;56:835–44.
12. Blackburn A, Stevens JD, Wheatley RG, Madej TH, Hunter D. Balanced
Funding analgesia with intravenous ketorolac and patient-controlled morphine
The study had no external funding. following lower abdominal surgery. J Clin Anesth. 1995;7:103–8.
13. McDonnell NJ, Keating ML, Muchatuta NA, Pavy TJ, Paech MJ. Analgesia
Availability of data and materials after caesarean delivery. Anaesth Intensive Care. 2009;37:539–51.
All datasets used and/or analyzed during the current study available from 14. Karaman S, Günüsen I, Uyar M, Biricik VF. The effects of morphine and
the corresponding author on reasonable request. fentanyl alone or in combination added to intrathecal bupivacaine in spinal
anesthesia for cesarean section. Agri. Apr. 2011;23(2):57–63.
15. Weigl W, Bieryło A, Krzemień-Wiczyńska S, Mayzner-Zawadzka E.
Authors’ contributions Comparative study of postoperative analgesia after intrathecal
AK, SA contributed to study design, conducted research, acquisition of administration of bupivacaine with fentanyl or morphine for elective
data, and manuscript preparation. AL, EI, MTN and FB contributed to data caesarean section. Anestezjol Intens Ter. 2009 Jan-Mar;41(1):28–32.
acquisition and manuscript preparation. DS contributed to data analysis 16. Ogboli-Nwasor E, Sule ST, Yusufu LMD. Pattern of postoperative pain
and manuscript preparation. AO contributed to manuscript preparation. management among adult surgical patients in a low-resource setting.
All authors approved the final manuscript. J Pain Res. 2012;5:117–20. https://doi.org/10.2147/JPR.S28198.
17. Chung F. Recovery pattern and home-readiness after ambulatory surgery.
Ethics approval and consent to participate Anesth Analg. 1995;80:896–902.
Study was approved by the Research & ethics Committee of Mulago
National Referral Hospital (MHREC); reference number is MREC 638.