Data-driven analysis of operational innovations in healthcare delivery

Sonmez, Nazli (2021) Data-driven analysis of operational innovations in healthcare delivery. Doctoral thesis, University of London: London Business School.

Abstract

Operational innovations significantly impact the effectiveness of healthcare delivery. In this thesis, I take a data-driven approach to advance rigorous evidence on shared medical appointments (SMAs), which are an innovative and interactive alternative to traditional one-on-one appointments for routine care of chronic diseases. Under this innovative approach, a group of five to fifteen patients with similar chronic conditions (such as asthma, diabetes, glaucoma, etc.) meet with a doctor simultaneously. Each patient, in turn, receives full one-on-one attention including diagnosis and prescription, while the others listen. In contrast to one-on-one appointments, SMAs compromise privacy, while enabling peer interactions. The first essay focuses on empirical analysis of the effectiveness of SMAs as an alternative to traditional one-on-one appointments for glaucoma patients, through a randomized controlled trial. The trial was conducted with 1,000 patients over four consecutive medical appointments, at the Aravind Eye Hospital, in Pondicherry, India. Using causal inference methods, it is found that SMAs can significantly improve patient satisfaction, knowledge, and medication compliance rates, with no significant differences in follow-up rates or disease progression. The second essay focuses on whether the diminished privacy in SMAs might hinder patient engagement, undermining the quality of care. Despite accumulating evidence of their effectiveness, many patients and doctors remain wary of SMAs. There is concern that diminished privacy in SMAs may prevent information sharing related to sensitive medical issues, thus reducing patient engagement, and compromising long-run outcomes. Therefore, it is an empirical question whether the diminished privacy in SMAs might hinder patient engagement, undermining the quality of care. This question is addressed in the second essay, through analysing engagement data collected during the same randomized controlled trial. The empirical approach developed to analyse shared appointment interactions and one-on-one appointment interactions was based on videotaping and coding patient behaviours for every appointment in the trial (20,000 minutes of video). The results suggest that despite the relative lack of confidentiality, patients in SMAs exhibit higher levels of both verbal engagement (asked more questions, participated more in conversation) and non-verbal engagement (attentiveness, positivity, head wobbling, eye contact and end of appointment happiness) during their appointments. The thesis is concluded with a discussion of some learnings from conducting field research and some future research directions related to the two essays.