Low birth weight in a sub-urban area of Cameroon: An analysis of the clinical cut-off, incidence, predictors and complications

Background: The World Health Organisation recommends that each country adopts its own cut-off value of low birth weight (LBW) for clinical use. The aims of this study were to establish a clinical cut-off point for LBW and to determine its incidence, predictors and complications in a sub-urban area’s hospital of Cameroon. Methods: We conducted a study in two phases: a 6-year retrospective phase during which we collected demographic and clinical information from the records of the maternity of the Buea Regional Hospital (BRH) and a 3-month prospective phase during which data were collected from consenting pregnant women using a structured questionnaire, and newborns were examined and followed after birth. Results: A total of 4941 records were reviewed during the retrospective phase and the 10th centile of birth weights was 2600 g. In the 200 pregnant women enrolled during the prospective phase, using this cut-off yielded an incidence of LBW of 19.0 %. Independent predictors of LBW were preterm delivery, hypertensive disorders in pregnancy, HIV infection, maternal age >36 years, maternal height <150 cm and pre-delivery BMI < 25 kg/m2. Neonates with LBW were more likely to have neonatal asphyxia, foetal distress, respiratory distress and neonatal death. Conclusions: Our results suggest that newborns under 2600 g have LBW in sub-urban Cameroon. They represent one out of every five babies, and they deserve close care. Preventive measures targeting the predictors described here are warranted to reduce the incidence and complications. Similar studies in urban areas are required in order to generalize the results. We carried out a descriptive and analytic study with two phases at the Buea Regional Hospital (BRH): a retrospective register analysis used to determine the cut-off of LBW in our population, and a prospective phase used to assess the incidence, predictors and complications of LBW. The BRH is situated in the Buea Health District (BHD) which is part of the Fako Division, South-West Region, Cameroon. The BRH acts as the reference hospital in the health district and also performs more deliveries than all the other health facilities in the district. The hospital has all standard units of general medicine. One gynaecologist-obstetrician and one paediatrician run the maternity and the paediatric units respectively. The institution lacks a neonatal intensive care unit; also, the maternity is situated 114 m apart from the operating rooms (where oxygen is kept). This distance represents a pitfall whenever a newborn requires oxygen. For the retrospective phase, all the records of pregnant women who gave birth during a six year period from the 1st January 2007 to the 31st December 2012 were reviewed. Only the records that clearly stated a gestational age less than 28 completed weeks, reported multiple gestation or had incomplete information were excluded (Fig. 1). Information was said to be complete if it contained: maternal age, marital status, gestational age, type of delivery, birth weight, sex of the newborn and APGAR score at the first minute. These records were collected and carefully stored in a handwritten register by the obstetrician and midwives in the maternity. Reasons for exclusion of records in the retrospective phase. From the 1st of January, 2007 to the 31st of December, 2012, 6001 deliveries occurred in the BRH. From this amount of deliveries, 4941 records were included in the retrospective phase of the study yielding a response rate of 82.3 %. The various reasons for the exclusion of some records included: incomplete records (629), destroyed records (17) and babies born before arrival at the hospital, abortions and multiple gestations (414) For the prospective phase, the target study population included all pregnant women-newborn pairs who attended the BRH during the period of the study; we excluded women who delivered at a gestational age below 28 weeks, those who had multiple gestations and those who did not provide consent to take part in the study (Fig. 2). Reasons for the exclusion of some pregnant women in the prospective phase. From the 2nd of January, 2013 to the 23rd of March, 2013, the period during which the prospective phase of the study took place, a total of 245 deliveries took place in the BRH. The study included 200 of these women and their babies yielding a response rate of 88.9 % from the prospective phase of the study. The various reasons for the exclusion of some pregnant women-neonate pairs were: multiple gestations (15), abortions (5) and refusal to provide consent (25) In the retrospective phase, after obtaining ethical clearance from the Institutional Review Board of the University of Buea, we obtained permission to access the birth records in the BRH from the Regional Delegate of Public Health for the South west region and the Director of the BRH. Socio-demographic and clinical data were collected from the birth records of the hospital. The birth weights of infants born in the hospital during the defined period were recorded. These data had been collected by midwives, trained nurses and doctors who worked in the maternity of the hospital. These data were handwritten into a register stored in the maternity. A total of 4941 records was included in this phase. The prospective phase was carried out over a period of 3 months, from January 2 to March 23, 2013. Pregnant women who were about to put to birth had their labour monitored making sure that they were provided the standard care that is expected. Participants’ hospital records were used to extract information on the following Participants’ weight -to the nearest kg using a Camry bathroom scale-, height -to the nearest cm using a locally made stadiometer calibrated to the Butterfly brand measuring tape- were measured, and a standard physical examination was performed. Throughout the labour, the foetus was monitored for the presence of acute foetal distress (foetal heart rate < 120 beats per minute and/or green meconium-stained amniotic fluid at delivery). After delivery, the neonates were assessed for viability (breathing, beating of the heart, pulsation of the umbilical cord or definite movements of voluntary muscles), and their APGAR scores at the first, fifth and tenth minutes were determined. The baby’s weight was measured to the nearest gram using a Holtex + digital baby scale. The newborns were monitored for the presence of respiratory distress which was identified by the presence of any one of the following: respiratory grunting, nasal flaring, intercostal recessions, sub-xyphoid recession, thoraco-abdominal asynchrony, and respiratory rate < 40 or > 60 breaths per minute. All stillbirths and any in-hospital neonatal death was noted. The mother was then approached and informed of the entire study and her consent requested. Only those who agreed to participate in the study were included after they signed the inform consent form. Minors and their guardians signed assent and guardian consent forms respectively. The newborns and their mothers were observed until the day of discharge. This was usually a period of 3 days for normal deliveries and could extend to 10 days for complicated deliveries including caesarean sections and instrumental deliveries. Two hundred participants were enrolled during this prospective phase. Data were cross-checked for errors before entry into a password-protected personal computer. The data were stored securely in a private location and kept confidential. Participants were referred to only by identification numbers. Identifiable information (consent forms) was kept separate from the data collection forms and it was only possible to link both through a coding sheet which was available solely to the principal investigator. Data were analysed using Epi Info version 3.5.4 and Microsoft Excel 2007. Means (standard deviations) were used to summarise continuous variables, and proportions and frequencies for categorical variables. Frequencies were compared using Fisher’s exact tests. The McNemar test was used to compare the incidence of LBW using the traditional cut-off (2500 g) to that yielded by our newly identified cut-off, and the agreement between the two was also computed. Predictors for LBW were determined using bivariate and multiple logistic regression models. The multivariate model included all variables which p-values in the bivariate analyses were less than or equal to 0.25 [5]. The statistical significance was set at p < 0.05. Ethical approval was obtained from the Institutional Review Board of the Faculty of Health Sciences, University of Buea. The study also received administrative clearance from the South-West Regional Delegation of Public Health and from the Director of the BRH.