Validation of InstaDx, A clinical decision support tool information technology based application for ischemic stroke

Date of Award


Document Type


Degree Name

Master of Science in Epidemiology & Biostatistics (MSc Epidemiology & Biostats)


Community Health Sciences


Globally ischemic stroke is second leading cause of mortality and morbidity among adult population. Prompt diagnosis of sub-type of ischemic stroke can lead to better management and clinical outcomes. This study aims to adapt, develop and validate the InstaDx; mHealth based android application to assist ischemic stroke sub-type diagnosis for use among neurology residents against Stroke Expert (gold standard). InstaDx is an evidence- based classification algorithm for ischemic stroke, to diagnose the sub-type in the presence of multiple competing mechanisms to minimize misdiagnosis, improve prognosis and preventive management decisions. Methods: The study was conducted in three phases: adaptation and development of InstaDx, validation of InstaDx and, user uptake and feasibility of InstaDx, qualitative feedback. In the first phase, algorithms of InstaDx were created through standard guidelines and these algorithms were transferred into mHealth based android application. In the second phase a validation study was conducted at Aga Khan University Hospital, Karachi from March to August 2017. Total of 228 consecutive patients of age 2 218 years, presenting in emergency department with neurological deficits consistent with stroke were recruited and InstaDx was used for diagnosing sub-type of ischemic stroke (Large or small artery atherosclerosis or cardio-aortic embolism) by the residents. The sensitivity and specificity of InstaDx for sub-type of ischemic stroke was validated against Stroke Expert diagnosis (gold standard). In the third phase, user feedback was assessed through Focus group discussion on neurology residents to evaluate their perspectives and experiences regarding the usage of InstaDx and what further modalities can be added in the application. Results: In adaptation and development phase of InstaDx, few sequential changes were incorporated in the algorithm of ischemic stroke mechanism. According to standard guidelines, contraindications of rtPA were rectified and integrated in InstaDx. As a part of validation phase, 228 Patients with mean age 62.59 14.60 years with ischemic stroke were studied. 11.84% patients receive revascularization with tissue plasminogen activator (rtPA). The sensitivity and specificity of InstaDx for large artery arthrosclerosis was 65.91 % and 73.57 % respectively, Small artery atherosclerosis was 56.25 % and 84.18 % respectively and Cardio aortic embolism was 58.33% and 99.17 % respectively. Focus group discussion feedback identified that InstaDx was well received as an educational tool by the residents and they suggested that further advanced diagnostic capacities be built-in. Conclusion: InstaDx is a valuable sensitive and specific mHealth based application for diagnosing subtype of ischemic stroke. It provides standard algorithm to confidently assess ischemic stroke patient and focuses on important aspects of stroke care.

This document is available in the relevant AKU library