DSpace Community:

The effects of weighting in the regression analysis of survey data collected using non-probabilistic sampling methods: A secondary data analysis

Thu, 01 Jun 2017 00:00:00 GMT

Title: The effects of weighting in the regression analysis of survey data collected using non-probabilistic sampling methods: A secondary data analysis Authors: Mudadi, Leon-Say Abstract: Introduction When surveys are conducted especially for hidden populations, data is rarely collected using random sampling which is the ideal way to collect representative data. However, it is common practice to analyse this data as if it was collected through random sampling ignoring the sampling design. We sought to determine the effects of including weights in the analysis of survey data collected through non-probabilistic sampling methods. Broad objective To assess the effects of weighting on risk taking behaviours associated with STIs among female sex workers (FSW) and long distance truck drivers (LDTD) in Beitbridge using weighted and unweighted logistic regression models. Methods Both inverse probability weighted and unweighted forward selection multivariate logistic modelling techniques were used to determine significant risk taking behaviours associated with STIs in FSW and LDTD. Final models compared magnitude of the difference between odds ratios, the selection of final variables, standard errors,statistical significance of selected variables and the overall fit of the models to determine whether or not we believed weighted models were more appropriate for the analysis of the survey data for FSW and LDTD. Results For risk taking behaviours associated with STIs, inclusion of weights resulted in an increase in the odds ratios, a decrease in the standard errors and narrowing of theconfidence intervals for the parameters in the weighted model for FSW. In the weighted model for LDTD, the odds ratios were higher than in the unweighted model and the confidence intervals were slightly narrow. However, the standard errors were higher in the weighted models. Conclusion Based on the results, we concluded that weighting in the regression analysis of survey data collected using non probabilistic sampling methods helps to improve the precision of the regression estimates; hence weighted models should be used.

Delays in performing emergency caesarean sections at Harare maternity hospital and Mbuya Nehanda hospital: Causes and outcomes

Sat, 01 Jul 2017 00:00:00 GMT

Title: Delays in performing emergency caesarean sections at Harare maternity hospital and Mbuya Nehanda hospital: Causes and outcomes Authors: Muyotcha, Annie Fungai Abstract: Introduction The ideal decision-to-delivery interval (DDI) for emergency Caesarean sections (ECS) quoted in international guidelines is 30minutes. Achieving this DDI is anticipated to improve perinatal outcomes. It has however been found in several institutions that it is not achievable in routine practice. There had not been a study in Zimbabwean institutions to determine our achieved DDI for ECS. In the event that we fail to achieve a 30minute DDI, there was no data on what were the influences or causes of delay in performance of ECS, and the maternal and perinatal outcomes thereof. Objective What is the achievable DDI for ECS performed at Harare Maternity Hospital (HMH) and Mbuya Nehanda Maternity Hospital (MNMH). What are the causes of delay in performing ECS and what are the perinatal and maternal outcomes. Design Hospital based prospective descriptive study. Setting Harare Maternity Hospital and Mbuya Nehanda Maternity Hospital Study population Consenting women that had undergone emergency Caesarean section. Methods Convenience sampling of women who had had ECS and were able to give consent to participate in the study was done. Data was collected by the researcher using a questionnaire on the day after they had ECS. On day 7 after the operation, a follow-up interview was conducted to check on the condition of both mother and baby. Data analysis was done using EPI INFO version 3.22 statistical software. Ethical approval was obtained from the ethics boards of each institution. Main outcomes of measure The indication of the ECS and the achieved DDI were explored. We also looked at the demographics and obstetric history of the participants, the stated causes of delayed DDI, the maternal morbidity and perinatal morbidity and mortality associated with delayed DDI. Results The total number of deliveries performed at both hospitals during the study period was 3 724 of which 1 050 (28.2%) were performed as Caesarean sections. Of all Caesarean sections, 866 were ECS (82.5%). The calculated sample size was 183. The study included 200 participants. The median age of participants was 25.5years. The majority of participants were married (94.5%), educated to secondary level or better (74%), Christian (68.5%), housewives (67.5%). 81 participants were primiparous. 13 participants delivered twins, therefore the total number of delivered infants was 213. Of these, 38 (17.8%) were delivered prematurely and 130 (16.9%) were term. 177 participants (88.5%) had booked their pregnancies. On admission, 147 (73.5%) were referred from within the Greater Harare Maternity Unit (GHMU), 26 (13%) were self-referrals and 27 (13.5%) were from outside the GHMU. The majority of participants had not had previous uterine surgery (75.5%). Most had successful regional anaesthesia (68%). In the study group, 92 participants (46%) had category 1 ECS and 108 (54%) had category 2 ECS. The median DDI for the whole group was 201.5minnutes (3hours 21minutes). Notably MNMH achieved a median DDI which was 1hour less than that at HMH. The top five causes of delay were delays in pre-operative preparation of the patient, theater being otherwise occupied, laboratory delays, delays in accessing blood products and delays in obtaining consent for theater. Of the 200 participants, only 14 (7%) had postpartum haemorrhage, and only 3 (1.5%) of these were still admitted 7days after the ECS while awaiting blood transfusion, the rest were home and in satisfactory condition. Of the 213 infants born, 186 (87.4%) had a 5minute Apgar of >7, 89 (41.8%) were admitted to neonatal unit and 27 (12.7%) suffered perinatal death. Conclusion A 30 minute DDI is not achievable in our institutions but with minimal adverse effects on the parturient or her infant.

Microbiology of puerperal sepsis and its clinical implications among HIV-infected and HIV-uninfected women in a hospital sample in Zimbabwe

Sat, 01 Jul 2017 00:00:00 GMT

Title: Microbiology of puerperal sepsis and its clinical implications among HIV-infected and HIV-uninfected women in a hospital sample in Zimbabwe Authors: Majangara, Rumbidzai Abstract: Title: Microbiology of puerperal sepsis and its clinical implications in a Hospital sample in Zimbabwe Introduction: Puerperal sepsis is infection of the female genital tract occurring at any time between the rupture of membranes or labour, and the 42nd day postpartum. Puerperal sepsis has become the leading cause of maternal death in Harare public health institutions accounting for 19% and 30% of maternal deaths for the years 2010 and 2014 respectively, from being the fourth nationwide cause at 12.3% in the year 2007. The objectives of this study were to determine the identity and antibacterial susceptibility profiles of bacteria colonizing the genital tract and blood stream, and to assess clinical outcomes and association with HIV infection in women with puerperal sepsis. Methodology: A prospective cohort study was conducted at Parirenyatwa and Harare Hospitals between 02 September 2014 and 01 July 2015. Endocervical swabs and blood were collected for culture and susceptibility testing from 151 consecutive women who met the World Health Organisation criteria for puerperal sepsis. HIV sero-status, immunological status and antiretroviral therapy (ART) use were determined. Medical records were reviewed for assessment of clinical outcomes. Proportions, categorical values and means were compared using Z-test, .² test and t- test along with 95% confidence interval (CI) and p-value of <0.05. Results: The mean age was 25.1 ±5.8years and most women were multiparous(53.6%). The majority of women had delivered at a hospital (78.1%) and by caesarean section (57.6%). The commonest bacterial isolates were Escherichia coli (30.6%) and Klebsiella pneumoniae (15.3%). Multidrug resistant organisms (MDRO) accounted for 10.9% of the isolates. MDRO were associated with prolonged mean hospital stay 23.0days (d) compared to 10.5d in women without MDRO (p=0.009). The frequency of genital colonization with Enterobacter species was significantly higher in HIV infected (9.1%) than uninfected women (1.7%) (p=0.04). Among HIV infected women (21.9%), severe immunosuppression (CD4 <200/mm³) was associated with a greater need for laparotomy 42.9% vs 4.5% (p=0.01) and prolonged mean hospital stay 19.0d vs 10.2d (p=0.03) compared to mild-advanced (CD4 count 200-500/ mm³) and insignificant immunosuppression (CD4 >500/mm³). There was a non-significant trend towards,earlier onset of sepsis; and higher rates of pelvic abscess, septic shock, wound dehiscence, peritonitis, death and need for admission into the intensive care unit (ICU) in women with severe immunosuppression. Antiretroviral therapy use did not independently influence outcomes. Puerperal sepsis case fatality rate was 7.3%. Conclusion: Gram negative bacilli, particularly E. coli, are the commonest bacterial isolates in puerperal sepsis. There is emergence of MDRO gram negative bacilli resistant to carbapenems, especially K. pneumoniae. MDRO and HIV associated severe immunosuppression are risk factors for prolonged hospital stay and need for surgery. Robust infection control strategies, emphasis on rational drug use and clinical culture surveillance to identify MDRO and monitor epidemiologic trends is recommended.

Socio-cultural realities of following through with prevention of mother- to- child transmission of HIV programme in Chiota district Zimbabwe: Implications for elimination of paediatric infection

Thu, 01 Jun 2017 00:00:00 GMT

Title: Socio-cultural realities of following through with prevention of mother- to- child transmission of HIV programme in Chiota district Zimbabwe: Implications for elimination of paediatric infection Authors: Nyati-Jokomo, Zibusiso Abstract: The study assessed the social and cultural realities of following through with prevention of mother-to-child transmission of HIV during the postnatal and breastfeeding period in a rural community in Zimbabwe and its implications on elimination of paediatric infection. The assumption was that paediatric HIV infection was not only through mother to child transmission but other social and cultural practices. Following through with PMTCT was conceptualised as the mother’s ability to adhere to ART, exclusive breastfeeding for six months and protecting the baby from getting infected through having protected sex among other factors. The study was conducted in Chiota District, one of the districts with a pronounced HIV burden. A sequential model combining both qualitative and quantitative methods was used for the study. Qualitative data were obtained through in-depth interviews/narratives with mothers on the PMTCT programme from two rural health facilities, (n=15). Focus group discussions were conducted with community members (n=231), and key informant interviews with the health staff (n=8). Quantitative data were collected through a cross sectional survey of breastfeeding women (n=103) accessing PMTCT interventions. Qualitative data were analysed thematically whilst STATA version 11 was used for quantitative data analysis where descriptive statistics, bivariate and multivariate regression analyses were done. Cultural practices, community, self and institutional stigma affected the effectiveness of the PMTCT programme. The prevalence of adherence to ART among the mothers was 82.5%. Only 6.8% of the mothers exclusively breastfed for the first six months. The major reasons for non-exclusive breastfeeding were the mother’s belief that the milk was unsafe (66%), inadequate (55%) and breastfeeding was not practical (67%). Risky traditional practices during PMTCT included ‘treatment’ of fontanelle by inserting the father’s male organ in the mouth of the child, toning of the girl child’s sexual libido through rubbing the father’s penis on the child’s vagina, improvement of eyesight and sense of hearing through use of mother’s milk among other practices. These practices exposed babies to bodily fluids like semen, precum, breastmilk and vaginal fluids, which are known to contain HIV. Culturally embedded practices, self, community and institutional stigma compromised the ability of mothers to adhere to PMTCT. Evidence from this study suggests that culture plays a major role in following through with PMTCT. This calls for taking cognisance of culture in designing HIV programmes. There is a need for further research on PMTCT during the postnatal period. Programmes should be cognisant that a ‘onesize fits all’ approach does not work as women are different.