Mentoring and measurement for better maternal and newborn survival: developing an intervention to put accountability into practice in Tanzania

Illness and death in pregnancy and childbirth are a major health problems in many low-income countries. In sub-Saharan Africa every year 179 000 women and over one million babies die around the time of childbirth. Maternal mortality ratios are high at 500 per 100 000 livebirths and newborn mortality at 30 per 1000: in Tanzania for every 200 babies born, one woman dies in childbirth and six babies do not survive the first four weeks of life. The most common cause of maternal death is severe bleeding, which can happen quickly just after the baby is born. Other common causes of death are related to hypertension and infection, including sepsis, malaria, HIV and anaemia. Many babies die because childbirth care is poor or because the mother has an infection in pregnancy which means the baby is premature, which means born too soon.

Simple, low-cost ways to tackle these issues are available, and most are affordable even in low-resource settings. However, many women and children simply do not get the interventions they need, particularly during labour, childbirth, and the first few hours of the baby's life. For example, a woman with a life-threatening complication might not find a skilled health care worker to deliver the baby because trained health staff are either absent or busy with other work. Vital equipment might have broken down and be unrepaired, and drugs and supplies may be out of stock. There may be no mechanism to ensure that equipment, drugs and supplies are always available and weak leadership might lead to a culture of frustration or resignation among health staff.

Maternal and perinatal deaths reviews have helped to reduce maternal deaths in the United Kingdom. Death reviews and other quality improvement approaches can stimulate changes in management and support structures so that equipment, supplies and drugs together with skilled professionals are always available for women in need. For more than a decade, the World Health Organization has promoted the use of death reviews to reduce maternal mortality. In Tanzania, however, death reviews are not done systematically or at scale. They are often based on deaths in hospital, while most maternal deaths occur in the community. And reviews often conclude that the woman herself was to blame rather than identifying areas for improvement under the control of the health staff which can be followed up to see whether the change results in improvement. In order to improve quality of care, a review system needs skills including mentoring, communication, analysis and intersectoral collaboration. The new WHO approach of "maternal death surveillance and response" aims to address these weaknesses and emphasizes 1) death surveillance at the community level, 2) analysis of trends, causes, risk factors and underlying causes of deaths and 3) the use of data to adapt local and national strategies. Careful adaptation and a link to mentoring is needed to develop a sustainable surveillance and response approach in Tanzania that is both embedded in the health system and designed for subsequent implementation on a national scale.

We will support the Tanzanian government in preparing, supporting and piloting an adapted and scalable maternal and perinatal death surveillance and response approach based on mentoring and measurement. Our work will include synthesising evidence on direct causes and underlying factors for maternal and newborn deaths, and developing a way to find out about, and act on, all maternal and newborn deaths.

We aim to develop, pilot and assess options for maternal and newborn death surveillance and response in a consultative manner, together with local stakeholders. We will use qualitative research methods including stakeholder interviews to maximise feasibility and acceptability.

In a joint process with the regional health secretariat we will review ongoing surveillance data for maternal and newborn deaths and audits in the facilities and develop a mentoring-based approach to improve quality of care. Previous learning from the EQUIP study and from the "collaborative approach" to quality improvement will guide this work. Alignment to local structures, costs and feasibility will be central to the development work, including national training materials for maternal and perinatal audits prepared by the Ministry of Health. We will also review abstraction tools, and review of matrices for analysis.

1) Ultimately we aim to improve maternal and newborn health and survival. Efforts to reduce deaths in mothers and their newborns will need continuous improvements in the health system and health care delivery. Maternal and newborn audits might have the potential to guide and support this process. However, to have such an effect, the audits will need to include deaths at home as well as deaths happening in facilities. The audits also need to include deaths in pregnancy or some time after birth. Developed strategies need to target the most important underlying factors, thus be well focused, but also implementable within the existing resources. Finally, the strategies need to be implemented, with review to check whether the change resulted in improvement. Our proposed support and mentoring package, Boresha, aims to support the audit process so that all these steps can be achieved.

2) The direct beneficiary of this developmental research project will be the Tanzanian Government and in particular the maternal and child health section of the national Ministry of Health and Social Welfare. Ultimately we aim to provide this beneficiary with a package which is implementable within existing structures.

3) Other beneficiaries will be international and national agencies, other private or public health providers and international non-government organizations interested in quality improvement using audits for maternal and newborn health.

4) Scientists and advocates involved in maternal and newborn health will in particular profit from this study. This study will help to show why many mothers and newborns die even though four of every five mothers deliver in a health facility. Intrapartum-related deaths are a main cause for the high newborn mortality together with premature birth.

We hope to understand why so many intrapartum deaths still occur, despite women delivering in facilities. This might have important implication for international strategies and the best balance between quantity and quality.

5) The development phase should lead to an implementation phase where the effects of the package on utilisation and quality of care will be assessed. This can add to the evidence base on audits.

6) Another impact expected will be further strengthened capacity within Ifakara Health Institute to conduct health systems research in an area relevant for national priorities. The research team has a strong track record in supporting research degree students to link their studies to research studies such as this.