Pathways to equitable stroke care in Sierra Leone and South Africa

AIM: To examine population level inequalities in stroke occurrence, access to and outcomes of acute care in Sierra Leone and South Africa and generate evidence to increase access to equitable optimised stroke care pathways. RATIONALE: Stroke is estimated to be the second leading cause of adult death in SSA. However, the majority of stroke data in SSA comes from hospital-based registers, prone to selection bias, which may mask inequalities in who suffers stroke and the care they receive. Organized stroke care (OSC) has been shown to effective to improve outcomes after stroke, however OSC can come in different forms and levels of care. This research will estimate the effectiveness of different levels of OSC in Sierra Leone and South Africa to reduce poor outcome after stroke. OBJECTIVES: Create population-based stroke registers within two existing and funded surveillance systems in South Africa and Sierra Leone to investigate inequalities in stroke occurrence, care and outcomes (WP1). Use causal inference methods to understand the effects of different levels of OSC in this setting (WP2). Use these findings to identify priority areas for quality improvement, and work with local policymakers to drive improvements and equity (WP3). METHODS WP1: Stroke notification, ascertainment and standardized data collection will be incorporated into pre-existing and funded surveillance sites, in South Africa the Systematic Healthcare Action Research Project (C-SHARP) with a population of 130,000 and in Sierra Leone using the Integrated Disease Surveillance and Response (IDSR) system with a population of 87,000. Multiple overlapping sources of notification will be used with hot and cold pursuit for case ascertainment. Deaths will be captured by verbal autopsy using the WHO tool, adjusted to ascertain first-in-a-lifetime-stroke. All patients with stroke will be followed up at 90 days to collect survival, function, and health-related quality of life data. Sociodemographic, risk factor, care and outcome data from this population based sample will be compared to tertiary hospital-based data in WP2 to assess inequalities in stroke occurrence, access to care and outcomes. WP2: A study nurse will improve data completeness of the existing stroke routine data at Connaught Hospital and Groote Schuur Hospital. This data will be used to emulate a target trial to assess the effectiveness of different levels of OSC categorized using the OSC Index(16), to reduce poor outcome modified Rankin Scale 3-6 at discharge, for adults with acute stroke. An appropriate causal inference method, such as Inverse Probability Treatment Weights (IPTW) or Propensity Scoring methods will be used to reduce confounding bias. WP3: Policy labs will bring patients, health workers and decision makers together, using the Ganz model of public narrative. Data and inequalities on stroke occurrence, access and quality of care and effects of interventions will be presented to guide quality improvement and service redesign. Outputs Population-based estimates of (inequalities in) stroke risk factors, outcomes and care in two SSA countries. Evaluate relative effectiveness of different components and levels of OSC using hospital data. Policy labs will create and quality improvement will drive optimised equitable stroke pathways in both countries.