Materno-fetal and neonatal outcomes associated with caesarean section delivery in private and public hospitals in low- and middle-income countries: a systematic review CURRENT

Background: Advancement in medicine has improved birth assistance. As a result, caesarean section delivery (CSD) has become the most commonly performed surgical procedure. The exponential growth has now skewed toward low- and middle-income countries (LMICs) despite the well-established morbimortality risk and extra costs associated to this procedure. The expansion of private healthcare sector may be playing a significant role. The objective of this review synthesizes knowledge and investigates the difference in materno-fetal and neonatal outcomes of CSD in the perinatal period, between private and public hospitals. Methods: Medline, Embase, Cinhal, Cochrane Database, LILACS, and HINARI were screened for peer-reviewed published studies in English and French, from 1990 to 2019, in human subjects and supplemented by manual searches. The studies included were prospective and retrospective cohort studies, cross-sectional and Delphi studies comparing perinatal outcomes of women whose neonates were delivered by caesarean section and by vaginal delivery in public and private hospitals. In total, the searches yielded 7,762 studies, assessed independently by two assessors. Of these, 26 quantitative studies were included which risk of bias was considered fairly low. Results: Elective or not, CSD is associated with a variety of outcomes, including death for both the mother and the neonate. Low quality of life, postpartum depression, infections, and scars were burdens attributable to CSD in both sectors. CSD is associated with less urinary incontinence compared to vaginal delivery but no difference exists in early skin-to-skin contact or in early breastfeeding introduction. Finally, across continents, Africa leads in terms of adverse consequences. Discussion/Conclusion: facilities surpassed rate were associated with the least severe materno-neonatal like China are

succeeding through robust policies interventions formulation to contain the CSD epidemic and the health issues associated thereto. Background Caesarean (from caesare in Latin, meaning "to cut") as a delivery method was first documented in 1020 [1]. Since this time, unprecedented improvement has been made, including procedures and incision techniques as well as the discovery of antibiotics, which has definitely increased its safety [2]. This leaves more room for the medicalization of childbirth, putting an end to the then common traditional delivery procedures [3].
Undoubtedly, medically justified CSD (or elective CSD) has been associated with tremendous benefits [4], including the prevention of maternal and perinatal morbidity and mortality [5]. However, as a major surgical procedure, CSD is not risk-free, despite the vigilance of professional and scholarly societies in developing guidelines to maximize its safety [6].
A wealth of studies has reported some morbimortality events associated with CSD. In that line, Gilliam [7] laid out uterine rupture. According to other authors, it may result in major short - or long-term health issues, including greater IgE-mediated sensitization [8], a higher risk of developing asthma [9], neonatal lung disease [10], rhinitis or an atopic allergy [11], an increased risk of childhood-onset type 1 diabetes [12], chronic immune disorders [13], celiac disease [14], and modification of the endocrine-immune system in the newborn [15,16]. The very recent landmark systematic review of Keag et al. [17] portrayed long-term risks for the mother-infant dyad.
Apart from CSD rates being higher in private hospitals, a differential in materno-fetal and materno-infant outcomes was found between private versus public hospitals [18,19]. Some associate this difference to hospital culture [20] or routine [21] even though it does not fit the guideline [6]. Other literature pointed out the supplier-induced demand for profit maximization [22,23], which is likely to be more prevalent in the private health sector.
Additionally, the public perception that a CSD is now a nearly risk-free procedure facilitates the request of elective CSD in the absence of clinical indications [2,22,24,25].
To date, CSD prevalence contrasts with the recommendations of the World Health Organization (WHO) of between 10% and 15% (WHO, 1985) as the "ideal" rate, suggested since 1985. The mounting trend was first noticed in high-income countries (HICs) [26,27].  [30], and 55.6% in Brazil, where second deliveries by CSD verge on 80% when the first was by CSD and over 99% for third births, when the first two were by CSD [28,31].
A wealth of reviews has addressed the subject of CSD from various angles. However, despite the topicality of the challenge, the subsequent materno-fetal and materno-infant outcomes of the procedure from the perspective of hospital ownership (public versus private) still remain unclear. Thus, a systematic review may be necessary to gather and assess current global evidence on materno-fetal and materno-infant outcomes associated with CSD in public and private hospitals in LMICs. This may be particularly imperative giving the expansion of the private healthcare sector in LMICs in the last two decades.
Additionally, such a review could generate information for researchers and policymakers alike allowing for evidence-based decision-making. To do so, this review aims to investigate the difference in materno-fetal and neonatal outcomes of CSD in the perinatal period, between private and public hospitals. The review focuses on LMICs with more or less comparable healthcare systems.

Methods
This review is shaped from the published protocol [32], which was registered within the International Prospective Register of Systematic Reviews (PROSPERO 2016: CRD42016036871).The Cochrane Handbook for Systematic Reviews of Interventions [33] was used as a foundation to carry out this systematic review.

Types of Studies Included and Inclusion Criteria
This review included quantitative studies as well as qualitative and mixed-method studies.
Searches included experimental study design, namely Randomized control trials (RCTs), non-RCTs, before - and-after quasi-experimental and interrupted time series designs. Non-RCT studies included case-control, cohort and cross-sectional studies dealing with CSD.
We selected studies that include medically-prescribed planned and unplanned CSDs performed in accredited public or private healthcare settings, as well as elective CSDs. In addition, primiparous as well as multiparous, twin, and breech births were also included.

Search Strategy
The present systematic review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement [35]. A concept plan was built and adapted to keywords and descriptors of the six targeted databases. Using a three-step strategy, we first conducted an exploratory search in Ovid-Medline, testing the strategy (descriptors and vocabulary). Second, the strategy was adjusted to each database  [36], leaving a 5-year period as a time lag for countries to implement it. The publication language was restricted to English and French, and all the records imported in the Rayyan reference screening system [37]. IB and AD, independently conducted the initial stream of screening of all titles and abstracts captured to determine eligibility. Using an agreed-upon extraction grid, IB, AD and TM extracted data that was populated in a customized spreadsheet in Microsoft Excel 2010.

Quality Assessment
We detected bias and graded the quality of studies by employing the Cochrane Risk-of-Bias (ROB) tool for the assessment of possible methodological bias [33]. In order to ensure a consistent overview, cross-sectional studies were assessed with the same tool. IB and AD independently rated the quality of each study as either "low", "unclear", or "high" risk of bias, throughout the six domains of the ROB tool (Fig. 3)..

Types of Comparisons
In the review, we compared the materno-fetal and materno-infant outcomes of CSD in the perinatal period between private and public facilities. For the purpose of this review, private providers were defined as "all organizations and individuals working outside the direct control of the state" [38]. They consist of for-profit (FP) and not-for-profit (NFP) providers; the former are defined as benefit-focused and the latter include philanthropic medical institutions. Institutional stewardship is further considered to complete the definition of the private healthcare sector, because in LMICs, services delivered in the private sector may be publicly financed - in the case of NFP providers [39]. Further, a comparison between continents was conducted.

Outcome Measurement
Based on the targeted outcomes (materno-fetal and materno-infant), the review looks at the key characteristics that may determine the CSD outcomes: (1) the what (i.e. ownership, financing system, payment scheme, and the management system), the where (2) (country and continent), and (2)

Description and Quality of Evidence Reviewed
Summary descriptions of the articles selected are presented in Table 2 The evidence reviewed showed that CSD in LMICs is associated with a range of maternal health complications. Considering mental health-related outcomes, one prospective cohort study conducted in private institutions found that 11.2% of post elective-CSD postpartum depressive symptoms, 5.6% before and after childbirth and 5.6% after childbirth [63]. Two studies discussed postpartum-related infection. Rwabizi et al. outlined a prevalence of 83% of maternal morbidity associated with postpartum infection in a sample of public and private facilities [43]. More specifically, Sharma and Dhakal found an incidence of surgical site infections (SSI) of 7.6% [54].

CSD Outcomes in Public vs. Private Facilities
Studying the LBW-related outcomes and clustering the analysis into private and public university hospitals, Murta et al [60] distinguished, through a multiple logistic regression, that there is a higher risk of LBW among born to women undergoing CSD in private hospitals (AOR = 2.33, 95% CI = 1.19-4.55) than in public ones (OR = 1.4, 95%CI = 0.82 to 2.4). In one study that targeted maternal mortality associated with facility ownership, undergoing CSD in the public sector was 3.3 (95% CI = 2.6-4.3) and was associated with a risk of maternal mortality, although the rate was 32.9% compared to 80.4% in the private sector [59].

Discussion
This review lays out four crucial results. Firstly, there is a huge discrepancy between public versus private hospitals in terms of maternal and perinatal morbimortality.
Secondly, CSD, including elective ones, can result in maternal and perinatal morbimortality. Thirdly, we found a difference between continents in CSD performancerelated outcomes, with African outcomes lagging behind. Finally, the results of vigorous policies revealed ways to critically reduce the CSD upward trend and the worst outcomes associated.
This review laid out the relevance of the facilities' ownership in materno-fetal and neonate CSD-related outcomes. Of the 14 studies targeting both public and private facilities [4,45,48,[52][53][54][55]57,[59][60][61][62]64,65], only Murta et al. [60] and Kilsztajn et al. [59] have frontally compared the CSD outcomes between participants in public and private institutions. In the case of the former, private institutions led statistically to fewer LBW babies (OR = 1.4 vs. 2.33), while in Kilsztajn et al.'s study, the likelihood of dying from CSD was 3.3 times higher [59]. This, as evidenced in Latin America, confirms the wellknown volume-outcome relationship from a great deal of empirical published works [67].
Nevertheless, cross-comparative findings reveal that private facilities offer significantly better outcomes in terms of CSD-related maternal mortality [59] and public health institutions show worse outcomes of severe morbimortality versus VD [43,58]. The comparison features show a similar trend while using the Robinson approach for perinatal mortality, neonatal distress, and maternal mortality ratio [44, 46,50], or maternal CSD complications [49]. This is also true for the incidence of SSI [56], emergency caesarean (versus VD), as well as the hospital level [58]. However, VD stands to be protective for urinary tract-related outcomes (compared to CSD) and for postpartum QOL (using SF12) [51].

The Within-Sector Comparison of CSD Outcomes
There was very little comparative data of relevance to CSD-related outcomes in the private versus public sectors at the within-country level. However, in Brazil, despite the much higher prevalence of CSD in the private sector - up to 90% [28,59,62], the highest in the world [68,69] - the risk of LBW was almost twofold that of the public sector [60]. In regard to SSI, findings in the public sector are very high (23.5%) compared to 0.7%-5.1% reported in Brazil [70,71], but far less compared to subSahara Africa (SSA) situations (15.6%) [72].

Even Elective CSD Leads to Maternal and Perinatal
Morbi-mortality This review underlined that, in all the LMIC settings, emergency CSD protects mothers from death [71]. As for high elective CSD rates, it is protective for fetuses and neonates [48]. Compared to VD, elective CSD resulted in significant odds of mortality in Peru [58], maternal mortality in Africa [48], and depressive symptoms in Brazil [63].
In addition to maternal mortality, elective CSD results infetal death and neonatal death [66]. In all settings, in private and public health institutions, elective CSD leads to a delay in breastfeeding (OR = 10.115) in Africa [48] and in the Middle East [55]. In addition to preterm birth and its correlates, LBW is among the salient findings in Brazil, the epidemic setting [61], in both public and private facilities in Brazil - more in private facilities [60].

Intercontinental Comparison of CSD Outcomes
When comparing continents, irrespective of ownership, the findings of [57,66] showed that Latin America performed better with better maternal outcomes comparing VD to CSD, followed by Asian settings. Africa is lagging behind for both antepartum CSD and intrapartum CSD outcomes [65]. This ranking was however partly demonstrated by Bauserman et al. [64]. Indeed, maternal mortality in Africa - in particular in SSA, which has remained which is indicative of the poor quality of maternal services - progressively declined over the last three decades [73]. We argue that the comparatively high adverse outcome associated with CSD in SSA is associated with vigorous policies, including, for instance, general practitioners, non-MD training [74] or free or the levying of a nominal fee by governments for CSDs [75,76]. Unfortunately, some countries like Morocco, rapidly lead to an alarming peak rate [77] necessitating a retaliate plan [78].
Hopeful Results in the Reduction of the CSD Trend:

Lessons from China
China, in particular, is paving the way to address the CSD epidemic. Very recent initiatives have shown CSD rates decline, particularly in supercities, by as much as 30% between 2008 and 2014 [79]. The literature also indicated that other strategies are feasible. Li et al.'s retrospective cohort study demonstrated that trial of labour after caesarean delivery may be a potential strategy for decreasing the CSD rate [80]. This steady decline of CSD, particularly significant among nulliparous and multiparous births without a uterine scar, which is concomitant with a decline in perinatal mortality from 10.1 per 1,000 births to 7.2 per 1,000 births, is the result of institutional interventions and policy change [81]. While China [83]. The authors suggest the inclusion of CDMR rates in hospitals' performance assessment matrix [82]. Finally, Torloni et al. the systematic review [84] and others literature [85] contended that, the use of Robinson classification as well as clinical and non-clinical classification in intervention policies [83,86], societal approaches [87], and management interventions (health promotion, practitioner training and tightening of hospital regulations) [83] are possible alternatives.

Conclusion
Interestingly, this review revealed that high emergency CSD rates, unlike high elective CSD rates, were associated with poor perinatal outcomes, while the risk of neonatal death was lower in facilities with high elective CSD rates and, fortunately, high elective CSD rates reduced the risk of fresh stillbirths. In contrast, neonatal deaths increased with high emergency CSD rates. Increased emergency CSD rates were associated with fresh stillbirths, neonatal deaths, and severe neonatal morbidity. In contrast, increased elective CSD rates were associated with fewer fresh stillbirths and neonatal deaths. Despite the quality of the studies selected, the majority of these targeted both the public and private sectors of health, but did not distinguish the part played by each sector in terms of outcome, leaving room for extrapolation, based on the magnitude of the CSD rate in each sector. Finally, all of the Latin American studies in this review were conducted in Brazil.
This systematic review portraying the evidence on materno-fetal and neonatal outcomes between private and public hospitals outlines the critical question on how best to shape policies to optimize the quality and safety of CSD. Figure 1 Prisma flow diagram illustrating the search strategy   Bias assessment This is a list of supplementary files associated with the primary manuscript. Click to download.

Supp. File.pdf
Tables.pdf