Caesarean section among referred and self-referred birthing women: A cohort study from a tertiary hospital, northeastern Tanzania

Background: The inequity in emergency obstetric care access in Tanzania is unsatisfactory. Despite an existing national obstetric referral system, many birthing women bypass referring facilities and go directly to higher-level care centres. We wanted to compare Caesarean section (CS) rates among women formally referred to a tertiary care centre versus self-referred women, and to assess the effect of referral status on adverse outcomes after CS.Methods: We used data from 21,011 deliveries, drawn from the birth registry of a tertiary hospital in northeastern Tanzania, during 2000-07. Referral status was categorized as self-referred if the woman had bypassed or not accessed referral, or formally-referred if referred by a health worker. Because CS indications were insufficiently registered, we applied the Ten-Group Classification System to determine the CS rate by obstetric group and referral status. Associations between referral status and adverse outcomes after CS delivery were analysed using multiple regression models. Outcome measures were CS, maternal death, obstetric haemorrhage ≥ 750 mL, postpartum stay > 9 days, neonatal death, Apgar score < 7 at 5 min and neonatal ward transfer.Results: Referral status contributed substantially to the CS rate, which was 55.0% in formally-referred and 26.9% in self-referred birthing women. In both groups, term nulliparous singleton cephalic pregnancies and women with previous scar(s) constituted two thirds of CS deliveries. Low Apgar score (adjusted OR 1.42, 95% CI 1.09-1.86) and neonatal ward transfer (adjusted OR 1.18, 95% CI 1.04-1.35) were significantly associated with formal referral. Early neonatal death rates after CS were 1.6% in babies of formally-referred versus 1.2% in babies of self-referred birthing women, a non-significant difference after adjusting for confounding factors (adjusted OR 1.37, 95% CI 0.87-2.16). Absolute neonatal death rates were > 2% after CS in breech, multiple gestation and preterm deliveries in both referral groups.Conclusions: Women referred for delivery had higher CS rates and poorer neonatal outcomes, suggesting that the formal referral system successfully identifies high-risk birth, although low volume suggests underutilization. High absolute rates of post-CS adverse outcomes among breech, multiple gestation and preterm deliveries suggest the need to target self-referred birthing women for earlier professional intrapartum care. © 2011 Sørbye et al; licensee BioMed Central Ltd. We used data from the medical birth registry at the zonal referral hospital KCMC in northeastern Tanzania to perform a cohort study of 21,011 births and 21,614 newborns from the period January 1st 2000 to August 31st 2007. Births with birth weight ≥ 500 g or gestational age ≥ 22 weeks were included. The birth registry, which has been described in detail elsewhere [15], systematically and prospectively collects information on sociodemographic and basic obstetric indicators, as well as information on delivery modes and pregnancy outcomes. Trained midwives conduct interviews and collect case record information in the days after birth, with a response rate of > 98%. The facility runs as a private/public partnership. The obstetric department receives patients from the local uptake area (Moshi town) in addition to referrals from a larger geographical area. CS is almost exclusively performed at hospitals in Tanzania [10], and most CS deliveries for women living in urban Moshi (Moshi District Council) are carried out at the facility. In Kilimanjaro Region, 70% of births take place at a health facility [6], and in Moshi, 92% deliver at a facility [16]. The site of the present study (a tertiary birth centre) is thus not a population-representative sample, as many women deliver at lower level facilities in the area or at home. There is potential selection towards financially better off women due to the cost-sharing policy gradually introduced for maternity services at KCMC from 2005 onwards. For a normal delivery, out-of-pocket costs are in the range of 5,000-15,000 TZS (5- 15 USD), while a CS has added minimum costs of 25,000-30,000 TZS (25-30 USD) [17]. In comparison, 88.5% of the population in Tanzania lived on less than USD 1.25 a day in 2000 [18]. The national health policy provides exemptions for the poor, but these are incompletely implemented. The Ten-Group Classification System for CS deliveries provides a standardised framework for monitoring of obstetric practice for individual institutions. The classification is meant both for application to existing birth data, and for use as a prospective tool to identify at-risk groups. Contrary to previous classification systems for CS, the Ten-Group Classification System is independent of the medical indication(s) for a CS. Using this standardised classification it is easy to identify which groups are the primary contributors to the overall CS rate, as well as determine CS rates and pregnancy outcomes within the different obstetric groups. CS rates in each group and contributions to overall rate can be compared across different facilities and between different levels of facilities. It has been applied internationally in high-resource settings among equivalent sub-populations [14,19,20]. We applied the Ten-Group Classification System to existing birth data drawn from the medical birth registry at KCMC. We classified women into ten mutually exclusive groups based on four obstetric characteristics: previous obstetric history, gestational age, category of pregnancy and course of pregnancy [14]. The essential information needed to apply the Ten-Group Classification System was available in the registry. We defined the following variables: parity coded as 0 or ≥ 1; multiple gestation coded as yes or no; presentation (at delivery) coded as cephalic, breech or abnormal; previous CS coded as yes or no; induction of labour coded as yes or no; CS coded as elective or non-elective; and gestational age coded as < 37 completed weeks or ≥ 37 completed weeks. We considered elective CS proxy for CS before labour, reflecting the practice at the facility. Gestational age was calculated according to the last menstrual period (LMP) registered on the antenatal card. For the 10% with missing LMP, birth weight ≥ 2,500 g was used as proxy for gestational age ≥ 37 weeks [21]. Information on the other variables necessary to complete the Ten-Group Classification System was missing in less than 1% of the sample. Additional variables used to characterize the sample were: maternal age in years coded as < 20, 20-24, 25-29, 30-35 or > 35; parity coded as 0, 1-4 or ≥ 5; maternal education coded as none, primary (1-7 years), secondary (8-11 years) or higher (≥ 12 years); and current residence coded as rural, urban or semi-urban. Missing data were less than 1%. We selected medical characteristics known to be associated with adverse pregnancy outcome: female genital mutilation (FMG) coded as any type or none; HIV testing coded as recorded or not recorded, HIV status of those recorded coded as positive or negative; antenatal visits coded as 1-3 or ≥ 4; serious maternal morbidity (preeclampsia, eclampsia, abruptio placentae and placenta praevia) coded as yes or no; and low birth weight of < 2,500 g coded as yes or no. We categorized the main admission diagnosis recorded among women delivered by CS. For formally-referred birthing women, this was the referral diagnosis, whilst for self-referred women the receiving midwife noted the main reason for arrival. Selected maternal outcomes were maternal death, prolonged maternal hospital stay as proxy for maternal complications (> 9 days after day of delivery = 97.5 percentile) and major obstetric haemorrhage at delivery (≥ 750 mL). Data on haemorrhage by clinical estimation were available from 2005 onwards. Due to the high prevalence of anaemia among pregnant women in the area, we chose a cut-off of 750 mL as a clinically relevant level of obstetric haemorrhage [22]. Neonatal outcomes were neonatal death (excluding intrauterine death diagnosed before labour), low Apgar score (< 7 at 5 minutes) and postnatal transfer to the neonatal ward. We excluded cases with missing variables such as delivery mode or presentation (2%). The final sample included 20,662 births and 21,255 infants with complete information to enable classification using the Ten-Group Classification System. Women were categorized as formally-referred when they were referred by qualified health personnel from other hospitals or health facilities such as health centres or dispensaries. The criteria for referral of women for hospital delivery from other health facilities in Tanzania can be found in Table ​Table1.1. Women who came directly to KCMC, bypassing referring facilities, were categorized as self-referred birthing women. The hospital charges these women an extra registration fee. Women delivered by CS in a previous pregnancy are routinely asked to register at KCMC for the next birth. These women were categorized as self-referred if not referred for other (medical or obstetric) reasons. Self-referred birthing women thus constituted a case mix of women with a wish to deliver in the facility (and able to pay), women directly seeking emergency assistance for obstetric complications bypassing referral facilities for whatever reason and women recommended for delivery at KCMC due to uterine scar(s) but without other obstetric complications. The hospital provides emergency transport for referrals between the regional birth centre (Mawenzi) and KCMC. From other facilities, transport was not regularly available. There were no community-based referral systems in place during the period. Missing referral status applied to 9.4% of the women. Demographic and obstetric characteristics and pregnancy outcomes for the missing cases were near identical to the total sample average (data not shown). These cases (n = 1950) were excluded from the outcome analysis. Criteria for referral from health facility to hospital-level delivery, Tz† † According to the Reproductive and Child Health Card (RCHC-4), Ministry of Health, Tanzania. *As defined in the RCHC-4 Permission to conduct the study was granted by the National Institute for Medical Research of the Ministry of Health in Tanzania, and the ethics committee at KCMC Hospital. Approval date 2003, reg. NIMR/HQ/R:Sa/Vol. IX/126. We extracted and analyzed data with Statistical Package for the Social Sciences/Predictive Analytics Software (SPSS/PASW) version 16.0. We used the χ2 test to determine trends in the proportion of CS and formally-referred birthing women during the period, and also to determine crude associations between referral status in CS deliveries and maternal/neonatal outcomes such as maternal/neonatal death, prolonged hospitalisation, obstetric haemorrhage, low Apgar score and transfer to the neonatal ward. Crude odds ratios (cOR) with corresponding 95% confidence intervals were estimated. We used a one-step multiple binary logistic regression framework to adjust the odds ratios and corresponding 95% confidence intervals for significant potential confounders such as type of CS, urban/rural residence, parity and low birth weight as proxy for preterm delivery. We considered significance level (p-value) below 0.05 statistically significant.