A multi-disciplinary team from the Kumuzu University of Health Sciences and LSHTM will address the following interlinked objectives (click on each one for more information):

1) To adapt and validate measures of hospital and district management practices

Formative qualitative research and psychometric performance assessment will generate validated tools to measure management practices.

This objective falls under the formative phase of research. We first identify and describe the types of management-related practices in hospitals and district health offices that may affect the quality of clinical care. Through in-depth interviews with hospital managers, clinicians and district health officials, we are exploring their understanding of what good management looks like in their setting, what practices are expected and implemented, perceptions of what management practices are important for supporting the provision of clinical care, and the attributes of effective leadership in a hospital setting.

An important part of the formative phase is to theorize how management practices may affect quality of care. Here we draw on causal loop diagrams, which are ideal for providing a blueprint for complex systems by representing important, non-linear feedback and relationships that further our understanding of how these systems operate. We undertake “group model building” sessions involving doctors, nurses and managers to map out – with the help of a causal loop diagram – the management-related causes of poor quality of care in the hospital.

Informed by these data and a review of existing instruments, we develop and implement a new survey tool to quantitatively measure the adoption of management practices in 36 hospitals. The survey includes a combination of interview with hospital managers, direct observation, and review of facility records. Using the cross-sectional survey data on management practices collected in the hospitals, we use psychometric methods to assess the acceptability, reliability and validity of the tool.

2) To examine the association between management practices and quality of clinical care

Data on management practices linked to patient level measures of clinical quality will provide observational evidence on which practices matter for quality.

This objective examines the association between hospital management practices and clinical quality, and its variation by facility characteristics. Study outcomes include neonatal outcomes, such as all cause in-hospital mortality, and clinical care processes for newborns, such as adherence to essential interventions for sick babies. The observational study makes use of the cross-sectional data on management practices collected in the first objective and links to these to the NEST360 patient level data. The latter is a de-identified prospective dataset on every newborn admitted to a newborn care unit in study hospitals containing information on the newborn’s admission status, maternal, birth and infant details, observations and nursing care, microbiology and labs, interventions and discharge outcome.

Analyses examine associations of hospital and district management scores with clinical quality. The primary outcome is all cause in-hospital mortality. Secondary outcomes are: birth weight standardized mortality, adherence to essential interventions for sick newborns (e.g. neonates <2500g who get kangaroo mother care); and neonates with hospital-acquired infections. further analyses explore the relative importance of individual management practices and sub-group differences with respect to hospital sector and level.

3) To co-design a problem-focused hospital management intervention and scalable delivery model

Stakeholder workshops that draw in national and international expertise will result in a theory-driven intervention to enhance hospital management practices.

Intervention development follows MRC guidance on complex interventions, including the need to identify relevant literature, develop the intervention around a credible theory of change, and consider implementation at an early stage. The intervention development draws on practical guidance on the importance of middle-level principles (e.g. support factors, derailers, safeguards) and more specific theoretical insights from the quality improvement literature (e.g. lean, six sigma, plan-do-study-act-cycles). It is also informed by early findings from objectives one and two – including the causal loop diagram derived from the model building workshops. We facilitate an inclusive process which brings together key stakeholders to co-design the intervention, including the Ministry of Health’s Quality Management Directorate, hospital managers and neonatal unit clinicians, and parent and community representatives. The output from this process is a co-designed hospital management intervention that is based on a well-articulated theory of change, thickened with information about the local setting in Malawi, and developed with scalability in mind.

4) To evaluate effectiveness of hospital management intervention on neonatal mortality and secondary outcomes (including hospital acquired infections)

A cluster RCT will generate evidence on the effects of enhancing hospital management practices on patient outcomes and clinical quality.

We evaluate the effectiveness of the hospital management intervention on neonatal mortality and secondary outcomes through a cluster randomised controlled trial. Since NEST360 is being implemented in all the study hospitals, we are testing the hypothesis that resources (medical devices, clinical training, quality improvement visits) in combination with supportive management practices can improve quality. The study outcomes include in-hospital mortality of newborn admissions (primary) and secondary outcome measures of clinical quality and uptake of management practices.

5) To assess acceptability, fidelity, and mechanisms of effect and estimate cost-effectiveness

A process evaluation and cost-effectiveness analysis will inform scale-up in Malawi and generalisability of the trial findings to other countries.

We conduct a process evaluation based on MRC guidance on complex interventions and realist evaluation principles, to address the following questions. Is the intervention implemented as intended? Is it acceptable to hospital staff? Do the mechanisms that underpin the intervention work, and if not, why not? How does context affect the success or failure of the intervention? Is it value for money? The process evaluation – particularly the focus on mechanisms – will be guided by the theory of change developed in the third objective. We draw on in-depth interviews, routine programme data and economic cost data.