Developing Clinical Decision Support Instrument for the Point-of-Care in Low Resources Settings 

In a constrained environment: (I). Following the paper-based clinical guideline is a traditional practice and the only choice utmost, (II). The service is challenged to deliver accurate and inadequate evidence for decision making, and (III). The service is suboptimal. Therefore; better and more robust point of care (POC) instrument can bring huge benefits to strengthen primary and first level hospital in low resource setting (LRS).The emergence of clinical decision support system (CDSS)/clinical pathway (CP) at the POC instrument specifically designed for LRS has the potential of paving the way for health care to be delivered more beneficially and cost effectively in developing countries. It comes in support of primary and first level hospital care to provide rapid and more appropriate integrated treatments. Low clinical competence, limited diagnostic capabilities, high turnover and low health professional motivation are some of the challenges that the vast majority of public health facilities in developing countries are facing on a day to day basis. Cost effective clinical pathway at POC can empower local and (unexperienced) care providers and enable them to make better clinical referral decisions towards primary and first level hospital care or not.

"The CDSS point-of-care instrument is designed for low-resource settings by combining knowledge-based and data-driven techniques."

The goal of this Ph.D. research is to provide a clinical decision support instrument to assist the primary care services in a limited resource setting and improve the healthcare quality with affordable cost and accessibility. In all; try to: (I). Introduce an automated, adaptive, interactive and applicable clinical pathway, (II). Assist the organization of the care processes (by generating the pathway or plan of care) for simple treatment or referral service, and (II). Deliver optimal care by: (I). Reducing delay, (II). Minimizing cost, and (III). Improving patient outcomes. Overall, this Ph.D. proposed a novel strategy for building an automated and dynamic generation of clinical pathways (CPs) in LRS using a hybrid (knowledge- and data-driven) algorithm that works with limited clinical input and can be updated whenever new information becomes available. Our proposed approach: (I). tried to reduce arbitrariness in entry point selection through a range of choices such as using evidence from historical records, dominant factors or dynamically initiate based on the signs and symptoms extracted from the clinical guidelines, and (II). dynamically maps and validates the knowledge-based clinical pathways with the local context and historical evidence to deliver a multi-criteria decision analysis (concordance table) for adjusting or re-adjusting the order of knowledge-based CPs decision priority.

Achievements (Honors & Awards) 
  • 2020 – VUB doctrol school – Datacamp training scholarship: Machine learning scientist in python
  • 2019 – MD4SG workshop Full Travel Grant at ACM-EC Economic and Computation, Phoenix, Arizona. USA
  • 2019 – Machine Learning Summer School – MLSS’19, London – Full Scholarship
  • 2018 – NASCERE Full Scholarship
  • 2018 – Deep Learning INDABA Full Scholarship Deep Learning Indaba 2018: 9-14 September, Stellenbosch University, South Africa. Awarded: AlphaGO POSTER PRIZE at DL INDABA2018.
  • INDABA, EPHA, MLDS-Africa, MD4SG, AI in Ethiopia, BLACKinAI