Peripartum traditional medicine use and surgical site infections: A prospective cohort of women delivering via cesarean section in rural Rwanda


 Background Traditional medicine is commonly used in low- and middle-income countries (LMICs). Little is known about the use of traditional medicine among women undergoing cesarean section (c-section) and the association to surgical site infections (SSIs) in LMICs. In this study, we describe peripartum use of traditional medicines and the risk of SSIs among women delivering via c-section in rural Rwanda.Methods This prospective cohort study enrolled women who underwent c-section at Kirehe District Hospital in rural Rwanda between September 2019 and February 2020. We collected self-reported data regarding traditional medicine use before and during pregnancy and after discharge up to postoperative day (POD) 11. On POD 11 (+/- 3 days), the women returned to the hospital for a study follow-up visit. We used Fisher’s exact test to assess the relationship between sociodemographic characteristics and traditional medicine use, and logistic regression to determine the association between traditional medicine use and SSI development while controlling for confounders. Results Of the 841 women enrolled in this study, 45 (5.4%) reported using traditional medicine to get pregnant. Nearly 39% used traditional medicine during pregnancy; the majority (96.9%) for a pregnancy-related reason. Only four women (0.5%) reported traditional medicine use between c-section and the POD 11 study visit. Of the 775 women who responded at all time-points, 341 (44.0%) reported using traditional medicine at some point during pregnancy or c-section recovery. No demographic characteristics were significantly associated with traditional medicine use (p>0.05), except for smoking (p=0.048) and alcohol consumption (p=0.010). Both traditional medicine use during pregnancy (p=0.04, aOR=2.0, 95% CI: 1.05, 3.85) and at any time point (p=0.04, aOR=2.0, 95% CI: 1.04, 3.83) were associated with development of SSI.Conclusions Traditional medicine use among c-section patients was high in the peripartum period, particularly during pregnancy, and was significantly associated with SSI. Knowing patterns of traditional medicine use during the peripartum period can help providers collaborate with traditional healers and give appropriate, culturally-sensitive pregnancy and postoperative care and counseling to patients.


Abstract Background
Traditional medicine is commonly used in low-and middle-income countries (LMICs). Little is known about the use of traditional medicine among women undergoing cesarean section (csection) and the association to surgical site infections (SSIs) in LMICs. In this study, we describe peripartum use of traditional medicines and the risk of SSIs among women delivering via c-section in rural Rwanda.

Methods
This prospective cohort study enrolled women who underwent c-section at Kirehe District Hospital in rural Rwanda between September 2019 and February 2020. We collected self-reported data regarding traditional medicine use before and during pregnancy and after discharge up to postoperative day (POD) 11. On POD 11 (+/-3 days), the women returned to the hospital for a study follow-up visit. We used Fisher's exact test to assess the relationship between sociodemographic characteristics and traditional medicine use, and logistic regression to determine the association between traditional medicine use and SSI development while controlling for confounders.

Results
Of the 841 women enrolled in this study, 45 (5.4%) reported using traditional medicine to get pregnant. Nearly 39% used traditional medicine during pregnancy; the majority (96.9%) for a pregnancy-related reason. Only four women (0.5%) reported traditional medicine use between csection and the POD 11 study visit. Of the 775 women who responded at all time-points, 341 3 (44.0%) reported using traditional medicine at some point during pregnancy or c-section recovery.

Conclusions
Traditional medicine use among c-section patients was high in the peripartum period, particularly during pregnancy, and was significantly associated with SSI. Knowing patterns of traditional medicine use during the peripartum period can help providers collaborate with traditional healers and give appropriate, culturally-sensitive pregnancy and postoperative care and counseling to patients.

Background
While still high, there have been notable declines in maternal mortality ratios in sub-Saharan Africa over the last 20 years, from 870 deaths per 100,000 live births in 2000 to 533 deaths per 100,000 live births in 2017 [1]. Increased access to safe cesarean section (c-section) delivery has contributed to this decline [1,26]. However, c-sections can result in complications, including, most commonly, surgical site infections (SSIs); in sub-Saharan Africa, the reported rate of SSIs after csection has ranged from 6.8% to 26% [4,8,10,12,13,21,27].
The World Health Organization defines traditional medicine as "the knowledge, skills and practices based on the theories, beliefs and experiences indigenous to different cultures, used in the maintenance of health and in the prevention, diagnosis, improvement or treatment of physical and mental illness" [2]. In Africa, up to 80% of people use traditional medicine as a core part of their health care [2]. Factors associated with traditional medicine use in Africa include lower educational status, knowledge of traditional medicine, low socioeconomic status, older age, large family size, unemployment, and rural residence [6,23,19,35] There is limited information about the use of traditional medicine by women delivering via csection in sub-Saharan Africa. Some patients undergoing other surgeries, including orthopedic, general, and gynecology surgeries, in South Africa reported using traditional medicine during recovery [33]. Additionally, some literature in sub-Saharan Africa reports that women use traditional medicines to get pregnant or during their pregnancies, and there has been an association between traditional medicine use and poor pregnancy outcomes [14,16].
However, we are unaware of studies describing the use of traditional medicine in the context of post-c-section care in sub-Saharan Africa or the relationship to the development of SSIs.
In Rwanda, 12% of women nationally and 9% in rural areas deliver via c-section [37]. A study in rural Rwanda estimated post-c-section SSI prevalence at 10.9% [34]. Between 50-80% of women in rural Rwanda reported using traditional medicine during pregnancy both to treat symptoms as well as to maintain the mothers' health during her pregnancy [9]; however, the use of traditional medicine after c-section is unknown.

Methods
This study was nested within a larger prospective cohort study that included all c-section deliveries performed at Kirehe District Hospital between September 2019 and February 2020 [34].

Study setting
Kirehe District Hospital (KDH), located in the rural Eastern Province of Rwanda, is managed by the Rwandan Ministry of Health in partnership with the international nongovernmental organization Partners In Health. The hospital has a catchment area of 19 health centers, of which two are located in the Mahama Refugee Camp, and serves a population of approximately 360,000 [30]. In November 2019, KDH had a 233-bed capacity and was staffed by 136 employees, including one Obstetrician/Gynecologist, 15 general practitioners, 77 nurses, and 11 midwives [34].
Women with uncomplicated labor usually deliver at their nearest health center, but those with complications are transferred to the district hospital for delivery. The majority of c-sections in this study were performed by general practitioners, with specialist intervention when necessary. Most of the patients in the Kirehe catchment area have community-based health insurance; the contribution of community-based health insurance is based on the household's annual income with people in lowest category fully covered without any medical costs and others paying 10% of health care costs.

Study population
This study included all women who underwent c-section at KDH from September 2019 to February 2020 and were permanent residents of Kirehe District. Patients from outside Kirehe District, patients from the Mahama refugee camp, and those who were not discharged from hospital by postoperative day (POD) 10 were excluded from this study. There were no age restrictions.

Data collection
We used multiple sources of data, including patient self-reported data collected through in-person or phone-based interviews, patients' clinical charts, and physical examinations. On POD 1, data collectors consented and enrolled patients in the study and collected information on clinical and sociodemographic characteristics. Data collectors also asked about traditional medicine use before and during pregnancy. On the day of discharge (typically POD 3), data collectors extracted clinical data from the patient's files. On POD 11 (+/-3 days), the patient returned to the hospital for a study specific follow-up visit. Women received a voucher to cover the costs of transport. At the visit, a general practitioner completed a physical examination to detect the presence or absence of SSI. The data collector asked the mother about traditional medicine use since time of discharge.
Data collectors entered data directly into REDCap, and all data were routinely audited and cleaned by the study managers.

Data analysis
We analyzed patients' sociodemographic and clinical characteristics using frequencies and percentages. We assessed the relationship between sociodemographic characteristics and traditional medicine use as well as the relationship between traditional medicine use and SSI development using Fisher's exact tests. We assessed whether the use of traditional medicine was associated with SSI using logistic regression controlling for possible confounders, evaluated at the α= 0.05 significant level. We used Stata version 16 for data cleaning and analysis.

Ethics
To ensure patients' confidentiality, we used patient identification numbers and did not link any personal identifiers to the data. Data were stored in an encrypted, password protected computer.
Prior to the implementation of the study, we obtained ethical approval from the Rwanda National

Results
Of the 841 women enrolled in this study, 249 (29.6%) were 30 years or older (Table 1)    None of the demographic characteristics were significantly associated with traditional medicine use (p>0.05), except for smoking (p=0.048) and drinking (p=0.010) ( Table 3). Both traditional medicine use during pregnancy (p=0.035, aOR=2.0, 95% CI: 1.05, 3.85) and at any time point (p=0.038, aOR 2.0, 95% CI:1.04, 3.83) were associated with development of SSI (Table 4). Table 3: Sociodemographic associated with traditional medicine use in post-Cesarean section Table 4: The association between traditional medicine use after cesarean section and the development of SSI.

Discussion
In our study of women delivering via c-section in rural Rwanda, we found that over 40% of women used traditional medicine peripartum, the majority during their pregnancy for pregnancy-related reasons. A recent systematic review conducted in sub-Saharan Africa found traditional medicine use during pregnancy ranged from 6.5% to 79.9% [15]. High traditional medicine use during pregnancy in Africa has been attributed to pregnant women perceiving that traditional medicine is more accessible and affordable than orthodox medicine [38,39]. In sub-Saharan African, the ratio of traditional healers who operate within rural communities is 1:500, while the ratio of medical doctors is only 1: 40,000, as they primarily reside in urban areas [5] We found that only 5% of women used traditional medicine to get pregnant. Studies carried out in Uganda and Zimbabwe reported high prevalence of traditional medicine use (76.2% and 36.5%, respectively) in the treatment of infertility conditions [20,22]. The higher prevalence in these other studies could be attributed to patient selection bias, as those studies recruited only non-pregnant women seeking care for infertility, as opposed to our study which recruited women post-delivery.
Only four women (0.5%) reported traditional medicine use in the postpartum period, which was considerably lower compared to a study of women who delivered via vaginal and c-section delivery in Zimbabwe (postpartum traditional medicine use of 17.3%) [24]. The difference in prevalence may be due to the different study settings; however, we hypothesize that our traditional medicine use rates were underreported, possibly due to desirability bias when reporting at the study follow-up visit.
We also found that peripartum traditional medicine use was associated with increased risk of SSI, even when controlling for education, smoking, and alcohol use. These results are consistent with findings in Zimbabwe, where 33% of women experienced harmful adverse events due to traditional medicine use during pregnancy [24]. While some research suggests that traditional medicine can support wound healing [40], the specific herbs and traditional medicines must be studied carefully in this context before encouraging its use. We must emphasize that most of the traditional medicine use reported was in pregnancy, and therefore the full mechanism linking traditional medicine use to SSIs, the possibility of underreporting use in the postpartum period, and SSI risk must be further explored.

Limitations
There are several limitations to consider when interpreting these results. First, women reported traditional medicine use while in a clinical setting, in close proximity to nurses and doctors. Given the negative attitudes towards traditional medicine among healthcare professionals, patients may have been reluctant to disclose this information in this setting. During the course of this study, loss to follow up occurred at different data collection time points leading to exclusion of women with missing data in our analysis. Finally, our study was conducted at one district hospital in Rwanda and may have limited generalizability. However, given that most district hospitals in Rwanda and in the region are located in rural settings with similar population characteristics, these results can be extrapolated to other similar settings.

Conclusions
The peripartum use of traditional medicine is high and significantly correlated with development of SSIs among women delivering via c-section in rural Rwanda. Additional studies regarding traditional medicine use should be carried out to better understand these links, ideally collecting data in non-clinical settings to avoid desirability and reporting biases. Knowing patterns of traditional medicine use can help providers collaborate with traditional healers and give appropriate, culturally-sensitive postoperative care and counseling to patients.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.