Systematic review on human resources for health interventions to improve maternal health outcomes: Evidence from developing countries

MAIN FINDINGS:

The review included 83 studies (two quasi-experimental, 16 before–after, five cross-sectional and 60 descriptive studies) mostly from sub-Saharan Africa and South Asia. All included effectiveness studies focused on either training, policies, or multiple combined interventions. Overall, the authors found that HRH intervention can improve both health-worker performance and maternal health outcomes. The authors found two quasi-experimental, two cross-sectional and eight before–after studies, mostly from Africa and South East Asia, assessing the effects of HR training interventions. They found that training skilled birth attendants and other health-care workers improved the basic knowledge and skills (such as abdominal examinations and safety measures when taking blood samples) of more than 70 per cent of staff, and reduced maternal mortality in most locations. However, in other performance areas such as obtaining clients’ medical and behavioural histories, and client education, the training interventions improved skills of only around 40 per cent of trained staff. Facilities where training had been provided also observed an increase in the utilisation of services and in the number of deliveries at emergency obstetric facilities. Only one cross-sectional study assessed the effect of task-shifting on maternal outcomes. The study, conducted in Mozambique, assessed the impact of a three-year surgical training of Assistant Medical Officers (AMOs) and found that trained AMOs were generally as effective at performing caesarean-section operations as obstetricians. The study also noted that the training has increased the availability of skilled birth attendants at the referral level in district hospitals. Four before–after studies explored the effects of policy implementation. The authors found that all four policies were associated with positive maternal health outcomes. Two policies (Bangladesh and Nepal) focused on improving emergency obstetric services and facilities. They were associated with increases in professional care seeking, facility deliveries and caesarean sections and reduced maternal mortality in Bangladesh, and increases in met need and births attended by skilled birth attendants in Nepal. In Cambodia, introduction of user fees was associated with increased revenues, bed occupancy, number of outpatients and number of deliveries, and improved services and patient satisfaction. In Ghana, a fee-exemption policy was associated with improved early reporting and handling of complications. Six before–after studies assessed the effects of combined interventions of HRH management, composed of a variety of components including service-provider training, policy and advocacy, partnerships and supervision. The authors found that all studied interventions led to significant reductions in maternal mortality. The authors analysed 60 descriptive studies exploring health-care delivery and the maternal health context in developing countries to assess factors influencing maternal-mortality outcomes. They identified several key factors associated with poor maternal-health outcomes, including lack of skilled birth attendants, failures in referral, absence of transportation systems, and inadequate infrastructure, especially in rural areas. Other factors mentioned in the studies included lack of integration of emergency obstetric care services with other parts of the health-care system; insufficient collaboration between health-care providers; brain drain and other barriers to the recruitment, deployment and retention of skilled health-care professionals; lack of ongoing training; absence of a continuum of care; and limited management capacity. On the demand side, the studies highlighted gender and cultural barriers preventing access to and utilisation of health services. The authors note that future research should focus on evaluating HRH interventions promoting recruitment, deployment and retention of health workers, especially in rural areas, as well as interventions improving the health-care work environment and conditions and HRH information systems.

BACKGROUND:

Reducing maternal mortality has been a key item on the global health agenda. The fifth Millennium Development Goal has set a target of a 75 per cent reduction in maternal mortality by 2015. Nevertheless, progress has been slow, and child birth remains the leading cause of mortality and disability for women of reproductive age in low- and middle-income countries. Human-resource shortages, inequitable access, poor planning, and inefficient management and distribution are some of the most important factors impeding efficient use of human resources for better maternal and child health outcomes. While there is a large body of literature and experience sharing on this topic, there are no studies systematically reviewing this evidence to highlight the important lessons, gaps and recommendations in the literature.

RESEARCH OBJECTIVES:

To review and synthesise existing evidence on the implementation and effects of human resources for health (HRH) interventions on maternal health outcomes, focusing on skilled birth attendants (defined as health personnel with midwifery skills). The review aims to draw out lessons learned, identify gaps and offer recommendations for the improvement of HRH interventions for better maternal outcomes.

METHODOLOGY:

The authors included randomised, quasi-randomised and before–after studies which evaluated interventions in system management of human resources for health (for example, policy, finance, education, partnership and leadership) related to skilled birth attendants in home, community and referral facility settings in low- and middle-income countries. They also included qualitative and observational studies to gather information about the implementation context and the interventions. The authors excluded interventions provided by Traditional Birth Attendants and Community Health Workers. The authors searched peer-reviewed English-language literature published between 2000 and 2010. They searched the electronic database PubMed and the HRH Global Resource Centre, examined cross-references and searched reference lists of identified studies. The authors analysed the results using narrative synthesis and thematic analysis.

QUALITY ASSESSMENT:

The review uses appropriate methods to reduce risk of bias in terms of adopting explicit criteria for study inclusion, screening and data-extraction processes and information about the included studies. However, the review has some major limitations. The search was not sufficiently comprehensive within the grey literature and does not cover literature in languages other than English, so it is not clear that publication bias was avoided. Finally, the review does not assess the quality of the included studies. As a result, the review does not make clear which evidence is subject to low risk of bias in assessing causality and it does not report and analyse findings separately by risk-of-bias status. This is a particular concern, because the review includes a number of study designs with potentially high risk of bias in attributing effects to the intervention. The review does not acknowledge these limitations, but nevertheless draws relatively strong policy conclusions.