Data-driven models of healthcare delivery operations

Jonasson, Jonas Oddur (2016) Data-driven models of healthcare delivery operations. Doctoral thesis, University of London: London Business School.


Operational implementation significantly impacts the effectiveness of healthcare delivery. In this thesis a data-driven approach is taken to analyze the impact of operational decisions on performance in two large-scale healthcare delivery systems, with an aim to generate managerial insights and test policy interventions. Each study is practice-based in the sense that it relies on detailed understanding of the problem context through proprietary data-sets and close collaboration with field managers. In the first study we model a supply chain for early infant diagnosis (EID) of HIV in sub-Saharan Africa. Since roughly 50% of untreated, infected infants die before they reach the age of two years, swift diagnosis is imperative. The diagnostic system consists of a network of clinics and laboratories, where blood samples are taken at clinics and diagnosed at specially equipped labs. We develop a rigorous two-part, data-driven modeling framework for designing EID networks, with the aim of maximizing the number of surviving infants starting treatment. On the one hand we develop a detailed discrete event simulation model for performance evaluation and on the other a optimization model for network design. The key decisions we consider include the allocation of diagnostic capacity to labs and the assignment of clinics to labs. The results suggest that for the case of Mozambique better operational use of existing resources would shorten delays by 22% and increase treatment initiation by up to 7%. The second study presents empirical analysis of the performance of London Ambulance Service crews. Using exogenous variation in partner assignments we demonstrate how high prior partner diversity among new recruits improves operational performance. Specifically, we find that prior partner diversity has an immediate positive effect on performance in a lessstandardized process (at the scene) whereas the impact is delayed for a more standardized process (at the hospital). The thesis is concluded by a discussion of some lessons learned from my doing practice-based research and some future research relevant to the two main studies.

More Details

Item Type: Thesis (Doctoral)
Subject Areas: Management Science and Operations
Date Deposited: 10 Feb 2022 16:09
Date of first compliant deposit: 10 Feb 2022
Subjects: Operations management
Supply chain management
Last Modified: 16 Feb 2022 02:41

Export and Share


Published Version - Text
  • Restricted to Repository staff only
  • Request a copy


View details on Dimensions' website

Downloads from LBS Research Online

View details

Actions (login required)

Edit Item Edit Item