Menu
Menu
Menu
Menu
Background: Neonatal sepsis is one of the most important causes of mortality in developing countries and yet the most preventable. In developing countries clinical algorithms are used to diagnose clinical neonatal sepsis because of inadequate microbiological services. Most information on incidence and risk factors of neonatal sepsis are from hospital studies which may not be generalized to communities where a significant proportion of mothers do not deliver from health facilities. This study, conducted in urban Uganda, sought to determine the community based incidence of clinical neonatal sepsis and the factors associated. Methods: This was a cohort of mother-neonate pairs in Kampala, Uganda from March to May 2012. The enrolled neonates were assessed for clinical sepsis and factors associated, and followed up till the end of the neonatal period. STATA version 10 was used to analyse the data. Results: The community based incidence of neonatal sepsis was 11% (95% CI: 7.6-14.4). On bivariate analysis, lack of financial support from the father (OR 4.09, 95% CI 1.60-10.39) and prolonged rupture of membranes more than 18 h prior to delivery (OR 11.7, 95% CI 4.0-31.83) were significantly associated with neonatal sepsis. Maternal hand washing prior to handling the baby was found to be protective of neonatal sepsis (OR 0.41, 95% CI 0.18-0.94). Of the 317 infants who completed the follow up period, one died within the neonatal period giving a neonatal mortality of 0.003%. Conclusion: The high incidence of clinical neonatal sepsis in this urban community with high rates of antenatal care attendance and health facility delivery places a demand on the need to improve the quality of antenatal, perinatal and postnatal care in health facilities with regards to infection prevention including promoting simple practices like hand washing. The astoundingly low mortality rate is most likely because this was a low risk cohort. However it may also suggest that the neonatal mortality in developing countries may be reduced with promotion of simple low cost interventions like community follow up of neonates using village health teams or domiciliary care.
The study was conducted in Kawempe division, an urban community in the northwestern part of Kampala, the capital city of Uganda. Kawempe division has an area of 32.45 km2 and population estimates of 268,659 of which 52% are female. It is densely populated and has areas characterised by uncontrolled developments and slum conditions [13]. It is served by 3 government health facilities, one private-not-for-profit hospital and several privately owned clinics which provide curative services. The sample size was calculated using the modified Kish Leslie formula for sample size estimation; Where: Deff: Design effect taken to be 1.5 Hence, N = 325 neonates The number of clusters, C, that was studied was calculated from the formula by Bennett S et al. 1991 [15] Where: The estimated number of clusters was 33. To allow for non-response a total of 34 clusters was studied. Thus the sample size calculated was 335 households with neonates. This was a population based cohort study with both retrospective and prospective components. The retrospective component consisted of the history of the condition of the neonate from birth to the point of contact with the research team, while the prospective component included the follow up period till the end of the neonatal period. The study participants included mother-neonate pairs living within Kawempe division during the study period who consented to participating in the study. The study enrolled neonates from birth to 28 days of age. Neonates with gross congenital malformation and extremely low birth weight were excluded from the study because their presentation may simulate symptoms of clinical neonatal sepsis. Thirty four out of 119 zones within Kawempe division in Kampala district were sampled using probability proportional to size. The principal investigator and two research assistants (study team) contacted the Local Council 1 chairpersons and village health teams (VHTs) of the zones and held meetings to explain the research. The study team moved with the VHTs in the zones, and in each zone a total of 10 households with neonates were consecutively enrolled in the study. Informed consent was obtained from eligible mothers. A neonate aged 0 to 28 days of age who met the selection criteria was enrolled in the study. A pretested questionnaire was used to obtain history, physical examination and evaluate factors associated with neonatal sepsis. These included maternal factors, delivery and newborn care practices, and household factors. The newborn care practices assessed included cord care, skin care, washing of the hands prior to handling the baby, early initiation and exclusive breastfeeding and thermal protection. The WHO IMNCI criteria were applied to assess babies for clinical sepsis [5]. The IMNCI criteria uses the following clinical features to make a diagnosis of clinical neonatal sepsis: if the neonate had temperature more than 37.5’C or felt hot to touch, convulsions (by history), fast breathing (> 60 breaths/minute), severe chest in drawing, nasal flaring, grunting, bulging fontanelle, pus draining from ear, umbilical redness extending to the skin, feels cold (by history), many or severe skin pustules, difficult to wake up, cannot be calmed within 1 h, less than normal movement, not able to feed and not able to attach to breast or suck. A retrospective review of the history was taken to find out if the neonate had the symptoms suggestive of neonatal sepsis since birth. A conclusion of clinical neonatal sepsis was ascertained if the baby had two or more symptoms of sepsis listed in the IMNCI criteria and had been reviewed or admitted in a health unit. Medical documents from the health units attended were also used to get information on presentation of the patient to the health units and the treatment received. Neonates diagnosed with clinical neonatal sepsis were referred to the emergency unit of the national referral hospital (Mulago hospital). All the mothers enrolled were availed the telephone contacts of the principal investigator and research assistants and informed to call the research team in case of symptoms of neonatal illness. Most of the mothers whose neonates had symptoms suggestive of sepsis took their babies to the national referral hospital. However a few opted for care in private clinics. The study outcome was ascertained after 28 days of life. The study team made another visit to the homes of enrolled infants and inquired if the infants had developed symptoms suggestive of sepsis which were not reported to the study team since the last contact with the research team. Mothers who did not contact the study team when their babies were ill were asked about the symptoms the baby had. The medical records of the babies, where available, were also reviewed. The study team made telephone calls to mothers who had changed location or those not found at home at the end of the follow up period. Questionnaires were checked daily for completeness and correctness. All data was double entered, cleaned, edited, coded and double entered into ACCESS data base 2007 and exported to STATA version 10 for analysis. Univariate analysis was used to get the general description of the data. Categorical variables were summarised into percentages and proportions. The continuous variables were summarised into means, medians, standard deviation and ranges for description. The incidence of clinical neonatal sepsis was obtained by calculating the proportion of neonates with symptoms and signs of clinical neonatal sepsis out of the total number of neonates who completed the study. Bivariate analysis was used to determine association between neonatal sepsis and various independent variables including maternal factors, perinatal factors and the newborn care practices. Continuous independent variables were categorised and associations established using Chi-squared tests. This was similarly done for categorical variables. Odds ratio was used as a measure of strength of association for categorical variables. P-value of less than 0.05 and 95% confidence limit not including one were used as tests for statistical significance. Multivariate analysis was done to assess for interaction and confounding of the independent variables with respect to the main predictor. Factors with P-value of 0.2 or less at bivariate analysis were selected for further multivariate analysis. During the study period a total of 353 neonates were screened and of these 15 were excluded from the study (8 did not consent to participate in the study and 7 planned to move out of the study area before the end of the neonatal period). Of the 338 subjects enrolled, 317 completed the follow up period. Twenty one (6%) of the neonates enrolled were lost to follow up. The demographic characteristics of the neonates who were lost to follow up were not significantly different from those who completed the study. The main reason for the loss to follow up was change in residential location and the absence of a functioning telephone contact.
N/A
