Approaching quality improvement at scale: a learning health system approach in Kenya

2018 
In 2002, we identified major shortcomings in the management of sick newborns and children at the first referral or district hospital level in Kenya.1 Failure in the dissemination of knowledge and skills (and thus of translation of evidence informed policy) was a fundamental problem. To address this challenge between 2005 and 2012 we developed, implemented and studied: 1. the national evidence-based clinical practice guidelines in the form of protocol booklets that can be disseminated at scale (and have recently described how this process matured over more than a decade)2 3; 2. the Emergency Triage Assessment and Treatment plus Admission Care course4 (that has been updated over time); 3. the standardised medical record forms including checklists of key symptoms and signs that are key elements of the protocols and help define the nature and severity of common illnesses5 (also updated over time). The effect of implementing these tools as part of a multifaceted strategy including outreach, audit and feedback to improve guideline adherence was tested between 2006 and 2009 and proven effective in a cluster randomised trial.6 In recent years, we have been able to document wider adoption of the protocols, training and record forms (including uptake outside Kenya) with some evidence of improvements in the quality of district hospital care, measured as adherence to guidelines, beyond centres directly engaged in research.7–10 In the last 4 years (2013–2017) we have adopted a new strategy, building on these earlier experiences, to continue efforts to improve hospital care for children in Kenya with a focus on adoption of agreed practice guidelines and uptake of basic technologies. At the heart of this new strategy is a Clinical Information Network (CIN). Here we outline the rationale for and philosophy of the CIN and how we suggest it helps Kenya as a low-income country (LIC) meet …
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    34
    References
    25
    Citations
    NaN
    KQI
    []