Onsite training of doctors, midwives and nurses in obstetric emergencies, Zimbabwe

Abstract Problem In Zimbabwe, many health facilities are not able to manage serious obstetric complications. Staff most commonly identified inadequate training as the greatest barrier to preventing avoidable maternal deaths. Approach We established an onsite obstetric emergencies training programme for maternity staff in the Mpilo Central Hospital. We trained 12 local staff to become trainers and provided them with the equipment and resources needed for the course. The trainers held one-day courses for 299 staff at the hospital. Local setting Maternal mortality in Zimbabwe has increased from 555 to 960 per 100 000 pregnant women from 2006 to 2011 and 47% of the deaths are believed to be avoidable. Most obstetric emergencies trainings are held off-site, away from the clinical area, for a limited number of staff. Relevant changes Following an in-hospital train-the-trainers course, 90% (138/153) of maternity staff were trained locally within the first year, with 299 hospital staff trained to date. Local system changes included: the introduction of a labour ward board, emergency boxes, colour-coded early warning observation charts and a maternity dashboard. In this hospital, these changes have been associated with a 34% reduction in hospital maternal mortality from 67 maternal deaths per 9078 births (0.74%) in 2011 compared with 48 maternal deaths per 9884 births (0.49%) in 2014. Lessons learnt Introducing obstetric emergencies training and tools was feasible onsite, improved clinical practice, was sustained by local staff and associated with improved clinical outcomes. Further work to study the implementation and effect of this intervention at scale is required.


Introduction
Improving maternity care is a global priority, yet many health facilities in low-income countries are not able to manage obstetric complications adequately. 1 Staff most commonly identified inadequate training as the greatest barrier to preventing avoidable maternal deaths. Training for obstetric emergencies may be part of the solution, but is not always effective: some training programmes either did not have a clinical effect or were associated with increases in perinatal morbidity. In lowincome countries, some studies have reported an increase in knowledge or skills after training, but failed to demonstrate improved clinical outcomes, 2,3 while other interventions failed to demonstrate improved skills. 4 In high-income countries, effective obstetric emergency training is conducted within the clinical setting, involves a high proportion of relevant staff and implements practice-based tools such as emergency equipment boxes. 5 In Zimbabwe, a programme of three-day off-site obstetric emergencies training was introduced in 2006. 6 These courses were mainly attended by senior staff, with junior midwives and doctors, who provide most clinical care, unable to attend due to lack of alternative clinical cover. Staff considered that this training had not significantly improved clinical care. In 2011 a local maternal mortality review meeting at Mpilo Central Hospital recommended that all maternity staff should receive obstetric emergencies training.

Local setting
Maternal mortality in Zimbabwe has increased from 555 to 960 per 100 000 live births from 2006 to 2011 and a national review of maternal deaths deemed that 47% of maternal deaths were avoidable. 7 Mpilo Central Hospital is a public, tertiary referral hospital in Zimbabwe's second largest city, Bulawayo. Mpilo Central Hospital manages approximately 10 000 births per year and provides all of the services defined by the World Health Organization's (WHO's) comprehensive emergency obstetric care services. These include administering antibiotics, uterotonic drugs (oxytocin) and anticonvulsants (magnesium sulfate); manual removal of the placenta, removal of retained products following miscarriage or abortion, assisted vaginal delivery, caesarean sections, safe blood transfusion, basic neonatal resuscitation and provision of care to sick and lowbirth-weight newborns. 8 Problem In Zimbabwe, many health facilities are not able to manage serious obstetric complications. Staff most commonly identified inadequate training as the greatest barrier to preventing avoidable maternal deaths. Approach We established an onsite obstetric emergencies training programme for maternity staff in the Mpilo Central Hospital. We trained 12 local staff to become trainers and provided them with the equipment and resources needed for the course. The trainers held one-day courses for 299 staff at the hospital. Local setting Maternal mortality in Zimbabwe has increased from 555 to 960 per 100 000 pregnant women from 2006 to 2011 and 47% of the deaths are believed to be avoidable. Most obstetric emergencies trainings are held off-site, away from the clinical area, for a limited number of staff. Relevant changes Following an in-hospital train-the-trainers course, 90% (138/153) of maternity staff were trained locally within the first year, with 299 hospital staff trained to date. Local system changes included: the introduction of a labour ward board, emergency boxes, colour-coded early warning observation charts and a maternity dashboard. In this hospital, these changes have been associated with a 34% reduction in hospital maternal mortality from 67 maternal deaths per 9078 births (0.74%) in 2011 compared with 48 maternal deaths per 9884 births (0.49%) in 2014. Lessons learnt Introducing obstetric emergencies training and tools was feasible onsite, improved clinical practice, was sustained by local staff and associated with improved clinical outcomes. Further work to study the implementation and effect of this intervention at scale is required.
Lessons from the field Training for obstetric emergencies in Zimbabwe Joanna F Crofts et al.

Multiprofessional training
Practical Obstetric Multi-Professional Training (PROMPT; http://www. promptmaternity.org) is a programme developed by midwives, obstetricians and anaesthetists that comprises: (i) an evidence-based skills development and training course; (ii) teamwork training incorporated in the clinical training; and (iii) a collection of tested local tools, checklists and local standardization techniques.
PROMPT emphasizes clinical practice over theory, with simulation in local clinical settings using practice based tools to help with correct decisionmaking during emergencies.
The implementation of PROMPT in the United Kingdom of Great Britain and Northern Ireland was reported to have improved knowledge, 9 teamwork 10 and clinical management in both simulation 10 and clinical practice. 11,12 Following the introduction of training, outcomes improved. Neonatal shoulder dystocia was reduced from 30/324 deliveries (9.3%) to 6/262 (2.3%) and hypoxic ischaemic encephalopathy decreased from 27.3 to 13.6 per 10 000 births. 11 Improved outcomes after training were plausible, consistent and sustained. 13 PROMPT has also been associated with improved perinatal outcomes in pilot sites in Australia 14 and the United States of America. 15 In November 2011, PROMPT training was introduced to Mpilo Central Hospital: First, a three-day train-thetrainers programme was held. The training was given by a team of seven staff from the United Kingdom (two midwives, three obstetricians, a paediatrician and an anaesthetist), three of whom were Zimbabwean. Twelve local staff (two midwifery matrons, four labour-ward midwives, two consultant obstetricians, two middle-grade obstetricians, a consultant paediatrician and a consultant anaesthetist) attended the programme to become trainers.
On day one, the team ran a demonstration of the PROMPT course. On day two, the 12 local staff were trained to deliver a local PROMPT course; working through lectures, running simulations on the labour ward and discussing the practicalities of local administration and implementation. On day three the local trained staff delivered an inhospital PROMPT course attended by 15 other local staff. Local staff trainers and participants did not receive per diem payments.
The trainers were provided with the equipment and resources required to run training within their institution including trainer's manuals, course manuals, and a digital versatile disc containing annotated presentations, videos, course timetables, evaluation sheets and certificates, together with two mannequins. One mannequin was used to teach the management of shoulder dystocia and vaginal breech birth (PROMPT birth trainer, Limbs and Things Ltd, Bristol, United Kingdom), and another to teach neonatal resuscitation -NeoNatalie (Laerdal Ltd, Stavanger, Norway).

Outcome of PROMPT
Before the PROMPT course, only 31 (20%) staff of the 153 maternity staff at Mpilo Central Hospital had been to any relevant training; 26 of them had received their training off-site and just 11 staff had received any obstetric emergency training in the preceding 12 months.
PROMPT trained between 18 and 28 staff per one-day training course, with a total of 299 hospital staff trained. Within the first 12 months of training, 138/153 (90%) of staff who worked within the maternity unit had attended a PROMPT course. Midwives, midwifery tutors, nurses, laboratory technicians, pharmacists, junior doctors, obstetricians, paediatricians, anaesthetists and family doctors attended the training.
PROMPT appears to have empowered the staff to make local changes; improving self-reported teamwork and local interprofessional culture. Staff reported more confidence in their management of emergencies and a perception that emergency management has improved overall.
Following the train-the-trainers course, staff developed their own emergency boxes for the management of eclampsia and postpartum haemorrhage, based on those they used during simulated emergencies. Each box contains the equipment required to manage the first 10 minutes of the emergency with evidence-based management protocols. Boxes are kept at the nurse's station and have been integrated into practice; their use is discussed at staff handovers and a member of staff is assigned to ensure the box is restocked after use. Maternity staff designed and implemented their own labour board containing informa-tion on labour progress, risk factors, actions required, and the staff member responsible for care, enabling staff to have an overview of all the patients on the labour ward.
To identify women at risk of complications, colour-coded early warning score charts were introduced. After the score charts were introduced, appropriate action in response to abnormal observations (e.g. starting antihypertensives or antibiotics) had increased from 1/24 (4%) to 11/15 (73%). 16

Clinical outcomes
Local monitoring of clinical outcomes is a key part of improving care quality and assessing the effect of training initiatives. In parallel with the implementation of PROMPT training, a monthly maternity dashboard of key clinical indicators has been introduced. Details of the maternity dashboard have been published. 17 PROMPT training has been associated with a 34% reduction in maternal mortality at Mpilo Central Hospital from 67 maternal deaths per 9078 live births (0.74%) in 2011 compared with 48 maternal deaths per 9884 live births (0.49%) in 2014.

Dissemination of training
Staff developed significant expertise in the adaptation of PROMPT training and tools to the Zimbabwean setting. Presentations of their experiences led to a demand for PROMPT training from other health facilities in Zimbabwe. To meet this demand, PROMPT training is being done in other facilities by volunteer staff from Mpilo Central Hospital with the support of the PROMPT Maternity Foundation. This work will help to define whether similar improvements can be implemented in smaller, rural health centres.

Challenges and solutions
Implementing onsite obstetric emergencies training is challenging, especially in low-resource settings. Releasing staff from clinical work to attend or facilitate training can be difficult. Some staff expected additional payment to attend training, as most off-site training courses provide this. 18 The involvement of Zimbabwean expatriates from the United Kingdom and New Zealand, who had a detailed understanding of Lessons from the field Training for obstetric emergencies in Zimbabwe Joanna F Crofts et al.
health care training in both Zimbabwe and high-resource settings, was crucial in mitigating many issues encountered during the implementation of this training programme. Executive level support was also essential: the hospital executive mandated the participation of all maternity staff and this ensured that staff were released to attend training.
To maintain a sufficient faculty of trainers, six more midwives who demonstrated skill and enthusiasm during their attendance at a PROMPT course have been recruited as trainers by the local team.

Costs
The total cost of providing training materials was 6000 United States dollars. The three most expensive items were a PROMPT birth trainer, a NeoNatalie and a laptop computer. This equipment has formed the basis of the hospital's maternity training department and is also used for other training initiatives.
Once training has been established, the on-going costs of training are low. Training staff on-site eliminates travel, accommodation and hotel venue hire costs. A policy of no per diem payments has also reduced the cost of training. In addition, it may help to reverse a culture where training has become an opportunity to supplement income, rather than an opportunity to develop and disseminate professional skills. 18 Conclusion PROMPT training has acted as a driver for quality improvement; improved the knowledge and skills of staff; selfreported team working and interprofessional culture and provided a place for staff to improve local systems (Box 1). The continuous monitoring and local reporting of clinical outcomes has also reinforced positive changes. Including local outcome data in the training helped to reframe the deaths of mothers and babies as everyone's problem.
It is unlikely that the training alone was responsible for the observed improvements in clinical outcomes. The success of the intervention is more likely to be a combination of factors, with the training being used as vehicle to introduce an improved professional culture and to trial and revise quality improvement tools. The use of these tools in clinical practice makes it easier for practitioners to do the right thing and this is a key factor in improving outcomes.
Training for obstetric emergencies is feasible in a low-resource setting. It empowers staff, improves local culture and can be sustained by local staff. Further work to study the implementation and effect of such interventions at scale and in different contexts is required. ■ Lessons from the field Training for obstetric emergencies in Zimbabwe Joanna F Crofts et al.