Feasibility and preliminary validity evidence for remote video-based assessment of clinicians in a global health setting.

• Main Authors: Katherine A Smith, Segolame Setlhare, Allan DeCaen, Aaron Donoghue, Janell L Mensinger, Bingqing Zhang, Brennan Snow, Dikai Zambo, Kagiso Ndlovu, Ryan Littman-Quinn, Farhan Bhanji, Peter A Meaney
• Format: Article
• Language: English
• Published: Public Library of Science (PLoS) 2019-01-01
• Series: PLoS ONE
• Online Access: https://doi.org/10.1371/journal.pone.0220565

<h4>Background</h4>Serious childhood illnesses (SCI), defined as severe pneumonia, severe dehydration, sepsis, and severe malaria, remain major contributors to amenable child mortality worldwide. Inadequate recognition and treatment of SCI are factors that impact child mortality in Botswana. Skills assessments of providers caring for SCI have not been validated in low and middle-income countries.<h4>Objective</h4>To establish preliminary inter-rater reliability, validity evidence, and feasibility for an assessment of providers who care for SCI using simulated patients and remote video capture in community clinic settings in Botswana.<h4>Methods</h4>This was a pilot study. Four scenarios were developed via a modified Delphi technique and implemented at primary care clinics in Kweneng, Botswana. Sessions were video captured and independently reviewed. Response process and internal structure analysis utilized intra-class correlation (ICC) and Fleiss' Kappa. A structured log was utilized for feasibility of remote video capture.<h4>Results</h4>Eleven subjects participated. Scenarios of Lower Airway Obstruction (ICC = 0.925, 95%CI 0.695-0.998) and Hypovolemic Shock from Severe Dehydration (ICC = 0.892, 95%CI 0.596-0.997) produced excellent ICC among raters while Lower Respiratory Tract Infection (LRTI, ICC = 0, 95%CI -0.034-0.97) and LRTI + Distributive Shock from Sepsis (0.365, 95%CI -0.025-0.967) were poor. Oxygen therapy (0.707), arranging transport (0.706), and fluid administration (0.701) demonstrated substantial task reliability.<h4>Conclusions</h4>Initial development of an assessment tool demonstrates many, but not all, criteria for validity evidence. Some scenarios and tasks demonstrate excellent reliability among raters, but others may be limited by manikin design and study implementation. Remote simulation assessment of some skills by clinic-based providers in global health settings is reliable and feasible.