Volume 123, Issue S2 p. 121-129
Abstracts
Free Access

Global Health (EP6)

First published: 16 June 2016
Citations: 1

EP6.006

Skilled birth utilisation in a rural Nigerian community

Adaji, S1,2; Bawa, U1,3; Jimoh, A1,3; Olorukooba, A1,3; Ismail, H1,3; Adelaiye, H1,3; Garba, C1,3; Lukong, A1,3; Suleiman, I1,3; Shittu, O1,3

1The Population & Reproductive Health Initiative, Zaria, Nigeria; 2York Teaching Hospital, York, UK; 3Ahmadu Bello University Teaching Hospital, Zaria, Nigeria

Introduction Skilled birth attendant (SBA) utilisation is an important factor in maternal health. In many rural communities in low and middle income countries, the availability and utilisation of SBA remains low. A clearer understanding of the levels of and factors influencing SBA utilisation will contribute to designing appropriate interventions to promote maternal health especially in rural communities.

Objective To assess SBA utilisation and demand side influences in a rural community in northern Nigeria.

Methods This was a cross-sectional, community survey in which hand held devices (mobile phones and tablets) running on Android Operating System pre-installed with Open Data Kit (ODK) Collect were used for data collection from 309 women. An 80-item questionnaire, adapted from the Nigerian 2008 Demographic and Health Survey was utilised. Univariate, bivariate and multivariate analysis was applied to data.

Results SBA utilisation rate for antenatal care was 51.8% and 12.5% for delivery. Only a small minority utilised SBAs for postnatal care. Women aged <20 years were more likely to utilise an SBA for prenatal care (Adjusted OR 1.326, 95% CI 0.759–2.318) and intrapartum care (Adjusted OR = 2.104, 95% CI 0.961–4.609). There was a positive relationship between formal schooling of women and SBA utilisation for prenatal care (Adjusted OR = 2.104, 95% CI 0.961–4.609) and intrapartum care (Adjusted OR = 2.104, 95% CI 0.961–4.609), as well as between husband's formal schooling and SBA utilisation for prenatal care (Adjusted OR = 0.879, 95% CI 0.531–1.456) and intrapartum care (Adjusted OR = 0.900, 95% CI 0.412–1.963). While women who engaged in income generating trade were more likely to utilise an SBA for prenatal care, they were less likely to do the same for intrapartum care. The relationship between household mobile phone ownership and SBA utilisation followed a similar pattern. Small sample size limited out ability to further analyse patterns of SBA utilisation for postnatal care.

Conclusion Low levels of SBA utilisation exists in the study community. Girl child education and educating the male child for the purpose of improving SBA utilisation could be rewarding interventions to the study community. The pathways between income earning and SBA utilisation would need further exploration. The links between mobile phone ownership and use and SBA utilisation also require more in-depth understanding.

EP6.011

Can a single visit approach to cervical screening using visual inspection with acetic acid and thermocoagulation be carried out during gynaecological healthcare camps in rural Nepal?

Nunns, D; Hollingworth, A; Devonald, B; Keeling, C; Ferrer, C

Nottingham University Hospitals, Nottingham, UK

Introduction The International Nepalese Fellowship have been running gynaecological camps in the far west of Nepal for over 15 years and have an established model of care where western staff provide a clinical service working with the local Nepali team. Camps provide an opportunity for cervical screening with VIA (Visual inspection with acetic acid) and treatment using a ‘single visit approach’ (SVA) as endorsed by the government. In this setting, a study was carried out to determine (i) the acceptability of the VIA technique, (ii) whether a SVA approach using VIA screening followed by thermocoagulation be implemented, and (iii) determine the point prevalence of colposcopic high grade CIN in women attending the camp.

Methods This study was carried out during a gynaecology camp in the Dialekh region in the far west of the country. VIA was carried as described by the WHO by one UK colposcopist and two UK GPs competent to carry out cervical cytology on all women. Screen positive women were then triaged for an ‘assisted’ VIA screen/‘colposcopy’ by an accredited UK colposcopist (DN) with a local Nepali gynaecologist (assisted VIA screening). The MobileOCT and the Cerviscope were used as aids to assess the cervix. Women with ‘colposcopic’ evidence of high grade CIN underwent treatment with the thermocoagulator.

Results In a 7 day camp, 880 women were seen and all but two consented for screening. VIA screening was carried out on 598 women. Screening was deemed not appropriate in the remaining 280 women for a variety of reasons eg procidentia, menstruation, ‘unmarried’. The VIA positive rate was 9.9% (59/598). All but one of the VIA positive patients had ‘aided’ VIA/colposcopy and 2 patients were treated for colposcopically apparent high grade CIN. There were no complications and clinically HPV was present in 8 patients.

Conclusions The VIA technique is acceptable and deliverable within a population of women attending a rural gynaecological surgical camp. The SVA and thermocoagulator treatment can be implemented safely as long as electricity is available. The point prevalence of colposcopic evidence of CIN (all grades) and HPV was low. Futures studies should focus on the current reality of oncogenic HPV disease prevalence and develop refined optimal cost-effective screening and treatment strategies.

EP6.013

Criteria-based audit of caesarean section in a referral hospital in rural Tanzania

Heemelaar, S1; Nelissen, E2,3; Mdoe, P2; Kidanto, H4; van Roosmalen, J5,6; Stekelenburg, J7,8

1Department of Obstetrics & Gynaecology, Katutura State Hospital, Windhoek, Namibia; 2Department of Obstetrics & Gynaecology, Haydom Lutheran Hospital, Tanzania; 3Department of Obstetrics & Gynaecology, Southmead Hospital, Bristol, UK; 4Department of Obstetrics, Muhimbili National Hospital, Dar es Salaam, Tanzania; 5Department of Obstetrics, Leiden University Medical Centre, Leiden, The Netherlands; 6Athena Institute, VU University Amsterdam, Amsterdam, The Netherlands; 7Department of Obstetrics & Gynaecology, Medisch Centrum Leeuwarden, Leeuwarden, The Netherlands; 8Department of Health Sciences, Global Health Unit, University Medical Centre Groningen/University of Groningen, Groningen, The Netherlands

Background Maternal mortality is still unacceptably high in most low-income countries. The increase in caesarean section (CS) rate in sub Saharan Africa has not improved maternal nor fetal outcome, suggesting that CS are not performed in time or not correctly indicated. However, WHO advises policy makers to use CS rates to monitor progress in maternal health and access to comprehensive emergency obstetric care.

Objective The first aim of this study was to audit indications for CS among women with severe maternal morbidity and mortality. The second aim was to determine the prevalence of substandard obstetric care.

Methods A cross-sectional study was performed from November 2009 to November 2011. Women with severe maternal morbidity and mortality were identified and those who had a CS were included in this audit. Audit criteria were developed based on literature review and (inter)national guidelines. Subsequently, hospital notes were reviewed by Tanzanian and Dutch doctors.

Outcome Primary outcome was the number of CS with an indication that was not in accordance with the audit criteria. Secondary outcome was the prevalence of substandard obstetric care, defined as either (i) delay in performing a CS, or (ii) delay or absence of an intervention to prevent CS, or (iii) delay or absence of an intervention to improve maternal and/or fetal outcome.

Results 216 women with severe maternal morbidity and 32 maternal deaths were identified, of which 82 (33.1%) had a CS. The indication for CS was in accordance with audit criteria for 36/82 (44.0%) cases without delay. In 20/82 (24.4%) cases the indication was correct, however there was significant delay in providing standard obstetric care. In 16/82 (19.5%) cases the indication for CS was not in accordance with audit criteria. During office hours CS were more often correctly indicated compared to unsociable hours, 60.0% versus 36.0% (P < 0.05).

Discussion Preventing unnecessary CS should have the highest priority, especially in low resource settings, considering the short and long-term complications of CS. Maternal and neonatal mortality can be reduced with simple low cost interventions. Despite the availability of these interventions in most low resource settings, our audit showed that they are not used, used too late or used for the wrong indication. This likely contributes to a high number of unnecessary and preventable caesarean sections in our study population. We would therefore advise not to use CS to monitor progress in maternal health and access to comprehensive emergency obstetric care.

EP6.021

IndOSS-Assam: Investigating the feasibility of introducing a simple maternal morbidity surveillance and research system in Assam, India

Nair, M1; Choudhury, M2; Choudhury, S3; Kakoty, S3; Sarma, U2; Webster, P4; Knight, M1

1University of Oxford, Oxford, UK; 2Srimanta Sankaradeva University of Health Sciences, Guwahati, Assam, India; 3F.A.A. Medical College, Barpeta, India; 4Oxford School of Public Health, Oxford Deanery, Oxford, UK

Background India has the highest number of maternal deaths, and within India, Assam (one of eight north-eastern states) has the highest maternal mortality ratio (328/100 000 live births). There are no up-to-date epidemiological data on pregnancy complications to refine existing policies, or develop new ones. Establishing a state-wide obstetric surveillance and research system is a first step towards generating evidence to prevent severe maternal morbidity and to guide management to reduce mortality.

Objective To assess the feasibility of establishing a simple maternal morbidity surveillance and research system in Assam [Indian Obstetric Surveillance System-Assam (IndOSS-Assam)] to investigate the incidence, risk factors, management and outcomes of severe maternal complications.

Methods Five tasks were undertaken from September 2014 to October 2015: (i) setting up a steering committee; (ii) establishing priorities for the region; (iii) mapping of surveillance sites; (iv) piloting the surveillance system in selected centres; and (v) conducting a cohort study in five government medical colleges using anonymous data from hospital records of 1007 pregnant women to examine the association between maternal anaemia and pregnancy outcomes (postpartum haemorrhage (PPH), low birthweight, small-for-gestational-age, and perinatal death).

Results Local stakeholder ownership and a simple uncomplicated anonymous system for case notification were the key strengths of this project. Cases and deaths were reported for six conditions: eclampsia, PPH, puerperal sepsis, septic abortion, uterine rupture and anaemic heart failure. Among 10 475 women delivering over 6 months, 402 had one of these conditions and 66 died (case fatality 16%). 35% (n = 351) of the 1007 pregnant women had moderate-severe anaemia. After adjusting for socio-demographic and pregnancy related characteristics, women with severe anaemia had a higher odds of PPH (adjusted odds ratio (aOR) = 9.45; 95% CI 2.62–34.05), giving birth to low birthweight (aOR = 6.19; 95% CI 1.44–26.71) and small-for-gestational age babies (aOR = 8.72; 95% CI 1.66–45.67), and perinatal death (aOR = 16.42; 95% CI 4.38–61.55).

Conclusion IndOSS-Assam, a collaborative initiative between clinicians in Assam (India) and academics in the UK, was shown to be a feasible and simple hospital-based system for ongoing surveillance of maternal morbidity. Through this platform detailed epidemiological studies can be conducted to identify specific preventive and management factors. IndOSS-Assam is already established in the two medical colleges that participated in this pilot study and there are plans to scale up the project to 14 government tertiary hospitals in the state. In the longer term, the IndOSS-Assam model could be used in other states in India to establish a national system.

EP6.022

Systematic review of the methodological quality of studies aimed at creating gestational weight gain charts

Ohadike, C1,2; Cheikh-Ismail, L1,2; Ohuma E1,3; Giuliani, F4; Bishop, D1,2; Kac, G5; Puglia, F1,2; Maia-Schlüssel, M6; Kennedy, SH1,2; Villar, J1,2; Hirst, JE1,2

1Nuffield Department of Obstetrics & Gynaecology, University of Oxford, Oxford, UK; 2Oxford Maternal & Perinatal Health Institute, Green Templeton College, Oxford, UK; 3Centre for Statistics in Medicine, University of Oxford, Oxford, UK; 4Department of Paediatrics, Turin University, Turin, Italy; 5Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; 6Oxford Clinical Trials Research Unit, Oxford, UK

Introduction A range of adverse outcomes is associated with insufficient and excessive maternal weight gain in pregnancy but there is no consensus regarding what constitutes optimal gestational weight gain (GWG). Differences in the methodological quality of GWG studies may explain the varying chart recommendations.

Methods The goal of this systematic review was to evaluate the methodological quality of studies that aimed to create GWG charts by scoring them against a set of predefined, independently agreed criteria. These were divided into three domains: study design (12 criteria), statistical methods (7 criteria) and reporting methods (4 criteria). The criteria were broken down further into items and studies were assigned a quality score (QS) based on these. For each item, studies were scored as either high (score = 0) or low (score = 1) risk of bias; a high QS correlated with a low risk of bias. The maximum possible QS was 34.

Results The systematic search identified 12 eligible studies involving 2 268 556 women from nine countries; their QS ranged from 9 (26%) to 29 (85%) [median 18 (53%)]. The most common sources for bias were found in study design (i.e. not prospective); objective assessment of pre-pregnancy weight and gestational age; description of weighing protocols; sample size calculation; and taking multiple measurements at each visit. There is wide variation in the methodological quality of GWG studies constructing charts.

Conclusion High quality studies are needed to guide future clinical recommendations. We recommend the following main requirements for future studies: prospective design, reliable evaluation of pre-pregnancy weight and gestational age, detailed description of measurement procedures and protocols, description of sample size calculation and creation of smooth centile charts or z scores.

EP6.031

Are Australian mothers gaining gestational weight in line with IOM recommendations?

Berkemeier, S1,3; Cheney, K2,3; Black, K2,3

1Department of Obstetrics and Gynaecology, Liverpool Hospital, Liverpool, Australia; 2RPA Women and Babies, Royal Prince Alfred Hospital, Sydney, Australia; 3Discipline of Obstetrics, Gynaecology and Neonatology, The University of Sydney, Sydney, Australia

Objective This study sought to examine baseline BMI and gestational weight gain up to the time of the booking in visit in cohort of women attending antenatal clinics. We used the 2009 Institute of Medicine guidelines to assess whether pregnant individuals increase in weight fell above or below recommended levels according to their BMI.

Method This was a secondary analysis of questionnaire based study examining a relationship between BMI and pregnancy intention. The study was conducted across four antenatal clinics in Sydney, Australia. The cross-sectional analysis of 2067 women was conducted using an anonymous, self-completed brief questionnaire including health and socio-demographic data and anthropometry, including self-reported pre-pregnancy weight, and weight and height at the booking in or second antenatal visit.

The primary outcome was the total weekly weight gain (grams) in the second trimester assuming women gained 2 kg or less within the first trimester. This was compared with the IOM recommended rate of weight gain in the second trimester, for each BMI category.

Gestational weight gain and BMI were also examined with respect to other sociodemographic and health determinants.

Results Of the 2067 women who were included in the analysis, according to the IOM BMI classification, 6.4% were underweight, 21.1% were overweight and 11.8% were obese. The mean rate of weight gain for overweight and obese women were 480 and 460 g per week, well above the recommended 280 and 220 g weekly weight gain recommended. Conversely, underweight women on average gained less weight (410 g per week) than the recommended rate of 510 g per week. Of the women who were in the normal weight range, 30% had an excessive rate of weight gain, compared with 45% of overweight and 48% of obese women.

Conclusion A significant proportion of Australian mothers are gaining weight outside the IOM recommendations within the first half of pregnancy. This is particularly prevalent in overweight and obese women. This highlights that more needs to be done pre-conception and in early pregnancy to educate women about overall gestational weight gain in order to reduce complications such as hypertensive diseases in pregnancy and caesarean section rates.

EP6.032

Quality of intrapartum care at an East African referral hospital: A case–control study of stillbirths

Maaløe, N1,2; Housseine, N2; Meguid, T2,3; van Roosmalen, J4; Khamis, RS2; Nielsen, BB5; Bygbjerg, IC1 1Global Health Section, University of Copenhagen, Copenhagen, Denmark; 2Mnazi Mmoja Hospital, Zanzibar City, Tanzania; 3School of Health & Medical Sciences, State University of Zanzibar, Zanzibar City, Tanzania; 4VU University of Amsterdam, Amsterdam, The Netherlands; 5Department of Obstetrics, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark

Objective To analyse contributing factors to stillbirths and related maternal risks.

Design A criterion-based unpaired case-control study.

Setting The East-African low resourced referral hospital of Zanzibar with 13 000 deliveries yearly.

Population From October 1st 2014 till January 31st 2015, all cases of stillbirths with birth weight ≥ 2000 g were included (n = 144), and every tenth delivery if birth weight ≥ 2000 g and Apgar score ≥ 7 as controls (n = 249).

Methods Criteria of realistic best practice were established, and case files were reviewed in accordance with these. Furthermore, maternal characteristics, admission information, and outcome parameters were extracted.

Main outcome measures Percentage of stillbirths with positive foetal heart rate on admission, and prevalence of suboptimal intrapartum management in relation to maternal and foetal condition as well as labour progress.

Results The overall facility-based stillbirth rate was 59 per 1000 total births. The majority had birth weight of ≥2000 g (n = 144, 78%), and in half of these was the foetal heart rate present on admission (n = 74, 51%). Among both cases and controls, major challenges were identified in intrapartum surveillance and decision-making, resulting in delays and inadequate use of non-invasive interventions. This appeared to be the main cause in 99% of the intra-hospital stillbirths, and it furthermore resulted in high maternal risks. Among intra-hospital stillbirths, the mean time from last foetal heart rate assessment till diagnosis of foetal death or delivery was 244 min (total range: 20–960 min). Of intra-hospital stillbirths, 37% received oxytocin augmentation. For the majority, oxytocin was not medically indicated, and the women were sparsely monitored when receiving the infusion. Similarly, management of prolonged labour was insufficient and sometimes led to unnecessary caesarean sections on stillborn babies. Of all included stillbirths, 26% were delivered by caesarean section, and vacuum extraction was only used in a single case. Among women experiencing stillbirth, 20% had severe hypertensive disorders, and an additional 30% did not have their blood pressure recorded during labour and delivery. Compared to the cases, the controls were characterised by later admission and shorter uncomplicated labour durations, while no major differences in management were found.

Conclusion For the women studied, substandard care led to unnecessary maternal and perinatal risks, which furthermore were related to resource draining interventions. Improvement of intra-hospital obstetric knowledge and care, not least at referral hospital level, is warranted to end preventable stillbirths as well as maternal and neonatal deaths and disabilities.

EP6.033

Outcome and maternal perception of vacuum extraction compared to 2nd stage caesarean section in Mulago Hospital, Uganda

Nolens, B1,2; Lule, J1,3; Namiiro, F1; van Roosmalen, J4; Byamugisha, J1,3

1Mulago National Referral Hospital, Kampala, Uganda; 2Canisius-Wilhelmina Hospital, Nijmegen, The Netherlands; 3Makerere University, Kampala, Uganda; 4Athena Institute, VU University, Amsterdam, The Netherlands

Introduction Prolonged second stage of labour is a major cause of perinatal and maternal morbidity and mortality in low income countries. Vacuum extraction is a proven effective intervention, hardly used in Africa. Caesarean section in the second stage of labour (2nd stage CS) is difficult to perform and complications are frequently seen, also in subsequent pregnancies. In 2012, a program to increase the use of vacuum extraction was implemented in Mulago Hospital, Uganda. Since then, over 2000 vacuum extractions have been performed and perinatal mortality decreased significantly. However, a survey amongst health workers showed that they are concerned about neonatal outcome and might therefore choose 2nd stage CS. The objective of this study was to compare outcome of vacuum extraction with 2nd stage CS.

Methods This prospective observational study was performed in Mulago Hospital, the national referral hospital in Uganda with over 32 000 deliveries/year. Mothers were included after delivery with the following inclusion criteria: vacuum extraction or 2nd stage CS and term, singleton, cephalic pregnancy, normal fetal condition at time of decision for intervention. Exclusion criteria were: eclampsia, more than one previous caesarean section. Failed vacuum extraction was analysed in the vacuum extraction group. Data were collected from medical files, interviews and examination on day 1 and 6 weeks after delivery. Outcome measures were perinatal death, maternal death, uterine rupture, decision to delivery time (DDI) and mother's perception.

Results We included 538 mothers: 251 had vacuum extraction and 287 had 2nd stage CS. Groups were comparable for duration of second stage and indication for intervention. Mean DDI was 294 min in the 2nd stage CS group and 72 min in the vacuum extraction group. Perinatal mortality was 12.5% in the 2nd stage CS group and 6.8% in the vacuum extraction group (P = 0.03). There were three maternal deaths and five mothers with uterine rupture in the 2nd stage CS group and nil maternal deaths and one uterine rupture in the vacuum extraction group. These complications combined were 2.8% in the CS group and 0.4% in the vacuum extraction group (P = 0.04). In the vacuum extraction group, 88% of mothers would advice vacuum extraction to a friend.

Conclusion Vacuum extraction has better maternal and perinatal outcome than 2nd stage CS and is well accepted by mothers in this low-income setting. Better accessibility of vacuum extraction might explain the findings. We advice wider implementation of vacuum extraction as a priority to policymakers.

EP6.037

Ten years in Ethiopia: What differences did the Millennium Development Goals really make?

Lattey, K1; Russell, M1; Brown, A1,2

1Maternity Worldwide, Brighton, UK; 2Imperial College, London, UK

Introduction Maternal mortality in Ethiopia is currently estimated to be 353 per 100 000 live births. It is thought that the mortality rate in rural areas far exceeds that of urban areas. Maternity Worldwide, an international non-governmental organisation focusing on maternal health using an integrated approach, started work in Ethiopia in 2002. A detailed evaluative needs assessment was initially performed in 2004 and repeated in 2014 to determine what impact a range of initiatives by Maternity Worldwide and the government have had to address the Millennium Development Goals (MDGs) in the area around Gimbie.

Methods Initial desktop research was performed before each visit to identify Ethiopian national and regional published data and review previous evidence-based interventions. Qualitative and quantitative data were collected in country, from villages and governmental departments. Semi-structured interviews were performed with service providers and users to obtain views regarding maternity services provision.

Results Many changes have occurred in Gimbie over the previous decade. Regarding MDG3 the government has introduced a ‘Women's Developmental Army’ initiative with health extension workers working with families in rural settings delivering health promotion activities. MDG5a has been affected by the building of a governmental hospital alongside Maternity Worldwide's work in the pre-existing Adventist hospital. Four-by-four and motorbike ambulances have been introduced improving access to more rural areas. MDG5b has lead to increased access to long-term contraception. However, determining if this has lead to a decreased maternal mortality rate in the zone of West Wollega is difficult. In 2004, 7.8% of births occurred in all health facilities in West Wollega, this increased to 30.9% in 2013, with the majority occurring in health centres and health posts. In 2013 1.8% of women gave birth by caesarean section, compared to only 0.4% in 2004. The case fatality rate in 2013 was calculated to be 0.08% (2.2% in 2004) based on the figures from the zonal health office; however, qualitative data from service providers and users challenges this figure's reliability.

Conclusion Our data indicate that progress has been made in Gimbie following the effort surrounding the MDGs and Maternity Worldwide's work. However, there are still multiple challenges in the services provided leading to the development of targeted recommendations. These may be of relevance to other charities and global health workers. This work also highlights the difficulty in calculating maternal mortality in rural Africa within the context of increasing political pressure to have met the MDGs.

EP6.040

Maternal mortality in India-A FOGSI survey

Sekharan, PK

PVS Hospital, Calicut, Kerala, India

Maternal mortality continues to be a major challenge to the health system in India. The target for Millennium Development Goal -5, which is a 75% reduction in the maternal mortality ratio (MMR) of 437/100 000 in 1990 to 109/100 000 by 2015. India failed to achieve this target and by reliable estimate the MMR in India is 140/100 000 in 2015.

Though there is a civil registration system for registration of births and deaths, including maternal deaths, we do not have accurate data on maternal mortality ratio of our country. Government of India depends on indirect data on maternal deaths and uses the Sample Registration System (SRS) for estimates of MMR. As per the SRS data available for years 2007–2009, the MMR is 212/10 000 and for 2010–12, 178/100 000.

To have an accurate data on maternal deaths with their cause, India should have the Confidential Enquiry of Maternal Death as it is being carried out in England and Wales for over 50 years.

The Federation of Obstetric and Gynecological Society of India (FOGSI) has conducted a survey of maternal deaths in India through its members who are practicing obstetricians all over India during 2005–2006 and 2010–11. In 2006, 4420 maternal deaths were reported amongst 1 253 696 deliveries, which shows a maternal mortality ratio of 352/100 000. During 2010 all the members of FOGSI were requested again to participate in the survey and to send the maternal death report from the respective hospitals they are working. Details of 5084 maternal deaths were made available for 2010–11 among 1 973 068 deliveries which shows a MMR of 257/100 000.

The FOGSI survey shows that 40% of maternal deaths in India are due to two important causes- PPH and Eclampsia. If we can prevent and treat PPH and eclampsia, 40–50% of maternal deaths could be prevented in India. Anemia is another important cause of maternal death in our country, another avoidable cause of death.

The State of Kerala is having the lowest maternal mortality ratio in India, 61/100 000. This because of the high female literacy rate, economic independence and empowerment of women and nearly 99% institutional deliveries. The Kerala federation of Obstetrics and Gynecology with Government of Kerala and NICE International have developed the Quality standards for the management of postpartum haemorrhage and for pre-eclampsia and eclampsia which are the major cause of maternal death and implementing it in the state maternity services.

EP6.042

Attitudes of healthcare providers towards the provision of Comprehensive Abortion Care (CAC) services in Western Cape, South Africa: The RCOG Leading Safe Choices (LSC) programme

Sheppard, N1; Fiander, A1; Glasier, A2; Simpson, E1; Regan, L1

1RCOG, London, UK; 2University of Edinburgh, Edinburgh, UK

Introduction Health care provider (HCP) bias, as well as a lack of respect for those providing termination of pregnancy (TOP) services and a general shortage of providers, add to the many barriers that women face when accessing TOP services. Despite TOP laws in South Africa being liberal, unsafe and illegal TOPs still remain an important public health problem.

This qualitative research aims to assess the opinion of TOP service providers towards their own role; towards the women they treat and to how they feel perceived within society and health services generally. The objective of this preliminary study was to look at the attitudes of providers before and after short theoretical and skills training courses in Comprehensive Abortion Care (CAC). The course contained training which aimed to upskill health care workers in the provision of TOP services (including counselling for TOP); provide space for reflection on attitudes towards women accessing TOP services and elements that aimed to increase the self-esteem of the HCP.

Methods A qualitative attitudinal questionnaire was constructed using a Likert scale. The questions were informed by focus groups held with CAC providers within the Western Cape.

The questionnaire was piloted in the UK amongst health care professionals.

The questionnaires were administered to participants before and after short 2-day theoretical and skills based training in CAC and analysed using a thematic analysis approach.

Results Opinions differed on the need for CAC services given the availability of effective contraception, as did how comfortable HCP felt about providing more than one TOP for a woman. Opinion was divided about whether a woman pregnant as a result of rape was more deserving of TOP than woman who failed to use contraception. The majority, although not all, HCP were comfortable talking to their friends about their work, however almost all felt there was stigma attached to accessing TOP care. Some of the HCPs were also unwilling to provide MVA TOPs.

Conclusion The findings suggest that even where TOP laws are liberal, HCPs care providers themselves still have negative value judgements towards women accessing HCPs services. However, some attitudes changed as a result of a short training intervention. If HCPs care provision is to be improved, the status of providers needs to be raised in order to attract and retain additional service providers into this vital area of reproductive sexual health.

EP6.045

Symphysiotomy for obstructed labour: a systematic review and meta-analysis

Wilson, A1; Truchanowiczb, E1; Elmoghazy, D2; MacArthur, C1; Coomarasamy, A1

1University of Birmingham, Birmingham, UK; 2Minia University, Minia, Egypt

Introduction Obstructed labour is a major cause of maternal mortality. Caesarean section can be associated with risks, particularly in low- and middle- income countries, where it is not always readily available. Symphysiotomy can be an alternative treatment for obstructed labour and requires fewer resources. However, there is uncertainty about the safety and effectiveness of this procedure.

Objective To compare symphysiotomy and caesarean section for obstructed labour.

Search strategy MEDLINE, EMBASE, Cochrane library, CINAHL, African Index Medicus, Reproductive Health Library, and Science Citation Index (inception- November 2015) without language restriction.

Selection criteria Studies comparing symphysiotomy and caesarean section in all settings, with maternal and perinatal mortality as key outcomes.

Data collection and analysis Quality of the included studies was assessed using the STROBE checklist and the Newcastle Ottawa scale. Relative risks (RR) were pooled using the random effects model. Heterogeneity was assessed using I2 tests.

Main Results Seven studies (n = 1266 women), all of which were set in low- and middle- income countries (as per the World Bank definition) and compared symphysiotomy and caesarean section were identified. Meta-analyses showed no significant difference in maternal (RR 0.48, 95% CI 0.13, 1.76: P = 0.27) or perinatal (RR 1.12, 95% CI 0.64, 1.96: P = 0.69) mortality with symphysiotomy when compared to caesarean section. There was a reduction in infection (RR 0.30, 95% CI 0.14, 0.62) but an increase in fistulae (RR 4.19, 95% CI 1.07, 16.39) and stress incontinence with symphysiotomy (RR 10.04, 95% CI 3.23, 31.21).

Conclusion There was no difference in key outcomes of maternal and perinatal mortality with symphysiotomy when compared to caesarean section.

EP6.046

Optimising informed consent for participants in randomised controlled trials: A comparison of three different methods

Ditai, J1,2; Abeso, J2,3; Mudoola, M1; Faragher, B4; Adengo, M1; Richards, JD4; Carrol, E5; Olupot-Olupot, P6; Storr, J7; Gladstone, M2; Medina-Lara, A8; Weeks, A2

1Sanyu Africa Research Insitute, Mbale, Uganda; 2Sanyu Research Unit, University of Liverpool, Liverpool, UK3Mbale Regional Referral Hospital, Mbale, Uganda; 4Liverpool School of Tropical Medicine, Liverpool, UK; 5Institute of Infection and Global Health, University of Liverpool, Liverpool, UK; 6Mbale Regional Referral Hospital Clinical Research Unit, Mbale, Uganda; 7WHO, Geneva, Switzerland; 8University of Exeter, Exeter, UK

Introduction Descriptions on effectiveness of informed consent remain incomplete. Previous studies have demonstrated poor participant understanding of the research information citing language barrier, lengths of patient information sheet, privacy at the time of consent and study participant condition at the time of consent. This can be particularly a problem in community interventional studies in rural African women where levels of illiteracy, dependency and compliance are high. As a result there have been calls for research into innovations to improve the informed consent process.

Methods This was a study comparing 3 methods of informed consent, nested within the pilot of a cluster-randomised trial. The study was conducted over 3 weeks in the homes of pregnant women from 13 villages within Mbale district, Uganda. Each method of giving study information was used for 1 week for recruitment into the study in addition to the standard patient information sheet. In the first week, all potential recruits received study information using the standard, researcher-read model. During the second week, a researcher gave a verbal presentation to all potential recruits, illustrated by a slide show. The 21 slides were developed by the research team using Microsoft PowerPoint® and shown to women on a Samsung smart phone. In the third week, potential participants were shown a pre-recorded video on the smart phone of one of the research team reading out the study information to the camera.

Two days after giving informed consent, each woman was visited again and her recall of the information was assessed by 5 knowledge-based MCQs. She was also shown the other two methods of information-giving and asked to rate them on a Likert scale and state her preference.

Results A total of 31 pregnant women participated. The majority of women preferred the slide show 61% (19/31), compared with 16% (5/31) and 23% (7/31) for researcher-read and video message respectively. Reasons given included the benefits of having pictures to aid understanding, and the logical progression of the information.

Conclusion The slide show information-giving procedure was the preferred model for recruitment of pregnant women into a randomised controlled trial in rural Ugandan settings.

EP6.048

A pilot cluster randomised trial of alcohol-based hand rub to prevent community neonatal sepsis in rural Uganda

Ditai, J1,2; Abeso, J3; Mudoola, M1; Faragher, B4; Adengo, M1; Richards, JD4; Carrol, E5; Olupot, PO3; Storr, J6; Gladstone, M2; Medina-Lara, A7

1Sanyu Africa Research Insitute, Mbale, Uganda; 2Sanyu Research Unit, University of Liverpool, Liverpool, UK; 3Mbale Regional Referral Hospital, Mbale, Uganda; 4Liverpool School of Tropical Medicine, Liverpool, UK; 5Institute of Infection and Global Health, University of Liverpool, Liverpool, UK; 6WHO, Geneva, Switzerland; 7University of Exeter, Exeter, UK

Introduction Neonatal sepsis is a global problem causing 0.5 million deaths annually, most of which are in low-resource settings. Babies born in African rural homes without running water or toilet facilities are especially vulnerable. Alcohol-based hand rub (ABHR) is widely used to prevent infection. However, it is not known whether supplying it for regular perinatal use can prevent newborn deaths in settings with limited access to running water.

Methods A cluster trial was piloted in 10 villages in Mbale district, Eastern Uganda. Five villages were selected as intervention clusters and Five as controls. Pregnant women of over 34 weeks of gestation were recruited over a 3-month period and followed-up for 3 months postnatally. Data were collected electronically on mobile phones using Open Data Kit (ODK). Women in intervention villages received a clean birthing kit (Maama Kit) along with a supply of ABHR whilst those in the control villages received the Maama kit only. Those with the ABHR were asked to use it at birth, as a single whole-body neonatal wash, on the cord stump and for 3 months postnatal hand hygiene.

Women with concerns about their babies took them to either local health centres or Mbale Regional Referral Hospital. Upon arrival, the infants were screened for sepsis using the WHO/Young Infant Clinical Signs Study Group criteria; a positive response was the study's primary outcome. Secondary outcomes were quantitative and qualitative, and included infant sepsis (physician-defined and microbiological), infant mortality, quality of life (WHOQOL-BREF), and acceptability of study and intervention. Recruitment rates and inter-cluster coefficient were also assessed.

Results Of 175 pregnant women screened in the 10 villages from October to December 2015, 103 women were over 34 weeks and all agreed to participate. One woman in the intervention arm later withdrew her consent. 93 of these have already given birth and of these, 37 infants have completed follow-up. Five infants died (14%) and 15 others self-referred with suspected infection (41%). Follow up of all participants will be finished by April 2016 and full results will be presented at the conference.

Conclusion The trial processes for the cluster randomised trial of community distribution of ABHR to prevent infant sepsis worked efficiently and recruitment rates were excellent. The ODK online data collection functioned well and there was no loss to follow-up. Final analysis of follow-up data is awaited to fully assess feasibility.

EP6.054

Anaesthesia-related maternal mortality in low- and middle-income countries (LMIC): A systematic review and meta-analysis

Sobhy, S1; Zamora, J1,2; Dharmarajah, K3; Arroyo-Manzano, D2; Wilson, M4; Navaratnarajah, R5; Coomarasamy, A6; Khan, K1,7; Thangaratinam, S1,7

1Women's Health Research Unit, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, London, UK; 2Clinical Biostatistics Unit, Hospital Ramon y Cajal (IRYCIS, CIBERESP), Madrid, Spain; 3Northwick Park Hospital, Middlesex, Ukraine; 4Anaesthesia, School of Health & Related Research, University of Sheffield, Sheffield, UK; 5Barts Health NHS Trust, Whitechapel, London, UK; 6School of Clinical and Experimental Medicine, College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK; 7Multidisciplinary Evidence Synthesis Hub (mEsh), Barts and The London School of Medicine and Dentistry, Queen Mary University of London, London, UK

Introduction Anaesthesia is an essential component of emergency obstetric care, however the risk factors contributing to maternal mortality from anaesthesia in low and middle-income countries (LMIC), and the burden of the problem is not known

Methods Major electronic databases until 1st October 2015, for studies reporting risks of maternal death from anaesthesia in LMIC. Two independent reviewers undertook quality assessment and data extraction. We computed odds ratios for risk factors and anaesthesia-related complications, and pooled using a random effects model.

Results Twenty-five studies (414 069 pregnancies) identified anaesthesia related risk factors for complications. Exposure to general anaesthesia increased the odds of maternal and perinatal deaths compared with regional. The rate of any maternal death was also increased when anaesthesia was administered by non-physician, compared to by physician anaesthetists.

Anaesthesia was the primary cause of maternal death in 1.2 per 1000 surgeries (44 studies; 632 556 pregnancies) with the highest rates in sub-Saharan Africa. Anaesthesia accounted for 2.8% of all maternal deaths (95 studies; 32 149 636 pregnancies, 36 144 maternal deaths), 3.5% of direct maternal deaths, and 13.8% of deaths following caesarean section.

The underlying causes were reported for 124 maternal deaths (24 studies). Half of all deaths resulted from airway complications such as difficult or failed intubation, oesophageal intubation, bronchospasm, ventilation difficulties and hypoxia; a third from pulmonary aspiration; a quarter from issues related to staff competency, poor pre-assessment, intraoperative monitoring and equipment failure. Other causes included cardiac arrest at induction or during the procedure (6%), ‘high’ spinal anaesthesia (6%), and drug overdose or adverse reactions (6%).

Conclusion Anaesthesia contributes to a disproportionately large number of maternal deaths in women undergoing surgery in low- and middle-income countries. The current international priority on strengthening health systems should address the factors identified in our review for improving anaesthetic care.

EP6.061

The rate of uptake of the influenza vaccination in pregnancy: A cohort study

Crosby, D1,2; Deleau, D1; Brophy, C1; McAuliffe, F1,2; Mahony, R1,2

1National Maternity Hospital, Dublin, Ireland; 2University College Dublin, Dublin, Ireland

Introduction Influenza is a caused by a highly infectious RNA virus, predominantly influenza A and influenza B, which usually occurs in a seasonal pattern with epidemics in the winter months. Influenza vaccination use in pregnancy is associated with a reduction in maternal and perinatal mortality and morbidity.

Objective The objective of this study was to determine the uptake of the annual influenza vaccine in a pregnant population and ascertain the reasons why some pregnant women did not receive it.

Methods We conducted a prospective cohort study over a 2-week period in January 2016 in the antenatal clinic in the National Maternity Hospital Dublin, a tertiary referral maternity hospital delivering over 9000 infants per year. Patients were recruited from ‘normal risk’ clinics and from the maternal medicine and diabetic service. Patient demographics, medical co-morbidities and whether or not the woman received or did not receive the influenza vaccine were recorded. In those who received the vaccination, the gestational age of receipt was documented. In those who did not receive the vaccination, the reason why not was documented. Data were analysed using SPSS version 23.

Results There were n = 504 women studied over the 2-week period. The mean age was 32.2 years (SD 5.9). There were 239 (47.4%) primiparous women and 265 (52.6%) multiparous women. The mean body mass index (BMI) at the booking visit was 25.4 kg/m2 (SD 4.7) and the mean gestational age when studied was 31.3 weeks (SD 7.5).

Overall, 197 (39.1%) women received the vaccine at a mean gestational age 20.9 weeks (SD 7.0, range 4–36 weeks). Table 1 illustrates the reasons why 307 (60.9%) women did not receive the vaccination. 41.8% of primiparous women and 36.6% of multiparous women received the influenza vaccine. Overall, 48.1% of those classified with a medical co morbidity received the influenza vaccine. When analysed by maternal age less than 30 years (n = 145), there was a 27.6% (n = 40) uptake compared to a 43.7% (n = 157) in those 30 years or greater (n = 359), (P = 0.0008). Of those <30 years of age who did not receive it, 38.1% did not know it was recommended in pregnancy.

Conclusion Given the increased rates of influenza in the community and the associated implications for both mother and infant, it is important that we educate pregnant women regarding the risks of influenza in pregnancy and encourage this cohort in particular to be vaccinated.

EP6.062

The World Health Organisation (WHO) Safe Childbirth Checklist as a tool to understand the ability of health facilities to deliver high quality care: An ethnography from three Malawian health facilities

Merriel, A1; Nambiar, B2; Larkin, M1; Hussein, J3; Makwenda, C4; Malata, A5; Coomarasamy, A1

1University of Birmingham, Birmingham, UK; 2University College London, London, UK; 3University of Aberdeen, Aberdeen, UK; 4Parent and Child Health Initiative, Lilongwe, Malawi; 5Kamuzu College of Nursing, Lilongwe, Malawi

Introduction The WHO launched the safe childbirth checklist to encourage improved quality of care for institutional deliveries. It provides a useful tool with which to consider the performance of a health system throughout a woman's journey of the delivery of her baby.

Objective To ascertain the ability of the health system to deliver high quality maternity care.

Methods Ethnographic observations were undertaken in three settings; high-risk postnatal ward of a large district referral centre; district hospital maternity and postnatal ward; and a community hospital over a period of 7 months. The ethnographic observations were then analysed through the lens of the WHO Safe Childbirth Checklist.

Results There are some areas of the checklist in which all facilities perform well, for example the starting of partograms on admission to the ward, ensuring that a delivery pack is prepared for the mother, and having a hospital attendant on standby to assist if necessary. There is also often quick identification of bleeding after birth and excellent initiation of neonates on antibiotics. All mothers remain in the hospital for at least 24 h and health talks are given daily explaining danger signs and follow-up.

However, there are areas where these three facilities would struggle to implement the Safe Childbirth Checklist. Referral is often driven by resource constraints and safe transfer hindered by financial constraints, such as lack of fuel. The vital signs of patients are seldom taken more than once per day and therefore it becomes challenging to identify when to administer antibiotics or anti-hypertensives. Whilst in some facilities companionship throughout the admission is encouraged, in other facilities it is actively discouraged by the staff due to space constraints and personal preference to ‘not work within guardians (relatives)’.

Conclusion Whilst the facilities already implement areas of the checklist well, the consistent delivery of all aspects of the checklist would require some system level changes. Firstly, there needs to be appropriate resources available to adequately assess and monitor patients. Secondly, and possibly more importantly, there needs to be simple systems for undertaking basic tasks and the subsequent decision making that is required. Finally, in some cases there needs to be a culture shift in the way that staff work with these systems and in terms of companionship at birth.

EP6.065

Diagnosing pelvic tuberculosis – a case series

Bommakanti, L;1 Tailor, V1,2; Robbins T1,2; Surampudi, K1

1Fernandez Hospital, Hyderabad, India; 2NW London O&G Trainee, London, UK

Background Globally the incidence of TB is falling, however female pelvic tuberculosis (PTB), which accounts for approximately 15–20% of extra-pulmonary TB, remains a common cause of infertility in up to 19% of Indian women and <1% in the developed countries. PTB can be asymptomatic, present acutely with PID, menstrual irregularities, or pelvic pain. Diagnosis of PTB can be by histopathology, microscopy for alcohol and acid fast bacilli (AFB), or using molecular methods with PCR assays.

Cases We present a case series of five patients diagnosed with PTB by histopathology at Fernandez hospital, a private women's health hospital. The women, aged 25–28 years, presented to the gynaecology clinic anxious to conceive. Two women were asymptomatic, two had secondary amenorrhoea with preceding reduced menstrual flow, and one experienced lower abdominal pain. Investigations included transvaginal pelvic ultrasonography (TVUS), hysterosalpingogram (HSG) for two women, followed by diagnostic hysteroscopy, laparoscopy and targeted biopsy collection.

On TVUS two women had features of pelvic inflammation with hydrosalpinx and tubo-ovarian masses. For the remaining three polycystic ovaries, thickened endometrium with fluid filled cavity and normal study was reported. One woman with tubo-ovarian mass on TVUS had a HSG demonstrating bilateral tubal blockage and uterine cavity adhesions. The second HSG demonstrated right tubal blockage, however this woman had had a previous right salpingostomy. Only one woman was suspected of having PTB pre-operatively, she had a strongly positive mantoux test, performed due to a family history of TB.

At surgery all cases had findings associated with PTB. Three women had hysteroscopic findings of caseating tissue +/− necrosis or avascular hypoplastic endometrium with ostia stenosis. Each laparoscopy showed features of PTB including caseating granulomas, pelvic and perihepatic adhesions, pyosalpinx, loculated fluid collections and omental thickening. Investigations for all patients included samples for TB PCR, AFB and targeted omental, endometrial and falopian tube biopsies.

In all cases only the histology consistently demonstrated features of TB. The women were given 6 months of TB treatment. To date only two women have conceived following IVF 1 year after diagnosis.

Conclusion PTB has varied and sometimes misleading presentations and investigation findings. A high clinical suspicion for its diagnosis, in both developing and developed nations, especially for women from high risk populations undergoing infertility investigation, is required. Histopathology remains an important diagnostic modality and therefore tissue biopsy should always be sent in suspected cases along with TB PCR and AFB culture to ensure correct diagnosis.