Healthcare Management Initiative

Harnessing data intelligence for medical diagnostics: The case of Unilabs Portugal 
June 2021

Together with Luis Menezes (CEO of Unilabs Portugal) and Jose Pedro Almeida (Data Intelligence Director at Unilabs Portugal), we will deep dive into Unilabs Portugal transformational journey and explore how data driven medical diagnostics enabled the organization to respond to the COVID-19 pandemic in a timely manner, to generate insights on individual patient responses to medication, and to promote evidence-based treatment to achieve better clinical outcomes. The session is hosted by Stephen E. Chick (INSEAD Professor of Technology Operations and Management ) and Ridhima Aggarwal (Salmon and Rameau Associate Director of the INSEAD Healthcare Management Initiative).

Why attend?
• Explore the critical role of diagnostics in health systems.
• Learn how organizational structure can facilitate digital transformation.
• Understand how data analytics can optimize and improve internal operations such as call centers and purchasing, enhance existing services – such as the prioritization of lab results for critical subpopulations – and develop new services such as AI-driven analytics for imaging and population health.

Click here for more.

Access to Medicines in Emerging Markets: How Executives and Investors Can Help
Hosted by INSEAD Hoffmann Global Institute for Business and Society at the SDG Tent 
April 2021

Abstract: Improving access to medicines and health products in low- and middle-income countries allows pharmaceutical and life science companies to unlock growth by serving underserved populations and contributes to the creation of healthier and more productive societies. It also contributes to preventing disease outbreaks that can have significant health and economic consequences globally. The Access to Medicine Index ranks 20 of the world’s largest pharmaceutical companies on how they contribute to access to medicine in key areas such as R&D, pricing, manufacturing & distribution, and compliance. The ATM Index shows that in the last 10 years large pharma and life sciences companies have made significant progress on many aspects of improving access. The continuation and scale up of such initiatives requires strong commitment from executive management and large investors. This panel discussion will bring together senior executives to discuss the progress made on this front so far, and ways to further bolster efforts to improve access to medicines in low- and middle-income countries.

Managing Two-Dose COVID-19 Vaccine Rollouts with Limited Supply 
Ho-Yin Mak, Associate Professor in Management Science, University of Oxford, Saïd Business School
March 2021

Abstract: We consider the problem of rolling out an approved COVID-19 vaccine (e.g., the one developed by Pfizer) that requires two doses to be spaced apart by a fixed time interval. Facing severely limited supply and mounting pressure to rapidly achieve herd immunity through vaccination, the U.K. and U.S. governments, along with the E.U., face the decision of whether to reserve stocks for the required second doses for returning recipients. We model the vaccine rollout process and complement it with an SIR (susceptible-infected-recovered) model that captures the disease transmission process. For a two-dose vaccine, we model the inventory dynamics under three stocking policies: (1) holding back second doses, (2) releasing second doses, and (3) stretching the lead time between doses. Counterintuitively, we show that even if one intends to release all second doses, only less than half of the available doses can be allocated to first-dose appointments in order to avoid delays in administering the required second doses for returning recipients, no matter how quickly supply grows over time. We show analytically that under a (weakly) increasing vaccine production rate, releasing second doses reduces the number of infections but creates uneven vaccination patterns. Contrary to popular belief that releasing the required second doses will reduce infections significantly, our numerical results reveal the reduction in infections is quite modest. Compared with the policy of releasing second doses, stretching the between-dose lead time flattens the infection curve but leads to a higher total case count. Finally, we consider an alternative single-dose vaccine with a lower overall efficacy (e.g., the Johnson & Johnson vaccine) and show it can be more effective than its two-dose counterparts in slowing down infections and ending the pandemic. As more than one hundred countries have yet to start their vaccination efforts as of February 2021, our research has important policy implications for various national and local governments to develop their vaccine rollout strategies especially when their supply of vaccine will be likely to be limited initially.

Retail Store Customer Flow and COVID-19 Ttransmission
Robert Shumsky, Professor of Operations Management, Faculty Co-Director, Master of Health Care Delivery Science Program, Tuck School of Business at Dartmouth
February 2021

Abstract: We examine how operational changes in customer flows in retail stores affect the rate of COVID-19 transmission. We combine a model of customer movement with two models of disease transmission: direct exposure when two customers are in close proximity and wake exposure when one customer is in the airflow behind another customer. We find that the effectiveness of some operational interventions are sensitive to the primary mode of transmission. Restricting customer flow to one-way movement is highly effective if direct exposure is the dominant mode of transmission. In particular, the rate of direct transmission under full compliance with one-way movement is less than one-third the rate under two-way movement. Directing customers to follow one-way flow, however, is not effective if wake exposure dominates. We find that two other interventions - reducing the speed variance of customers and throughput control - can be effective whether direct or wake transmission is dominant. We also examine the trade-off between customer throughput and the risk of infection to customers, and we show how the optimal throughput rate drops rapidly as the population prevalence rises.

Hospital Readmissions Reduction Program does not provide the right incentives: Issues and remedies
Tolga Tezcan, Professor of Management Science and Operations, London Business School
October 2020

Abstract: The Hospital Readmissions Reduction Program (HRRP) reduces Medicare payments to hospitals with higher-than-expected readmission rates where the expected readmission rate for each hospital is determined based on the readmission levels at other hospitals. Although similar relative-performance-based schemes are shown to lead to socially optimal outcomes in other settings (e.g., cost cutting efforts), HRRP differs from these schemes in three respects: (i) deviation from the targets are adjusted using a multiplier; (ii) the total financial penalty for a hospital with higher-than-expected readmission rate is capped; and (iii) hospitals with lower-than-expected readmission rates do not receive bonus payments. We study three regulatory schemes derived from HRRP to determine the impact of each feature, and use a principal-agent model to show that: (i) HRRP over-penalizes hospitals with excess readmissions because of the multiplier and its effect can be substantial; (ii) having a penalty cap can curtail the effect of financial incentives and result in a no-equilibrium outcome when the cap is too low; and (iii) not allowing bonus payments leads to many alternative symmetric equilibria, including one where hospitals exert no effort to reduce readmissions. These results show that HRRP does not provide the right incentives for hospitals to reduce readmissions. Next we show that a bundled payment type reimbursement method, which reimburses hospitals once for each episode of care (including readmissions), leads to socially optimal cost and readmissions reduction efforts. Finally we show that, when delays to accessing care are inevitable, the reimbursement schemes need to provide additional incentives for hospitals to invest sufficiently in capacity.

Interpretable Analytics for Healthcare Operations: New Algorithms and Implementation  
Jean Pauphilet, Assistant Professor, London Business School
September 2020

Abstract: Analytics could improve healthcare delivery, provided that it answers industry concerns, chief among which is the issue of interpretability. In this talk, we address these challenges by designing new interpretable machine learning techniques and implementing them in a hospital environment. In the first part of the talk, we present novel algorithmic methods for feature selection in statistics. We consider the problem of training a linear predictive model under sparsity constraints. We propose a combination of a first-order heuristic and a cutting-plane algorithm to successfully compute high-quality solutions for problems with up to 100,000s covariates. On extensive numerical experiments, we achieve more accurate feature selection than competitors, with comparable predictive power. In the second part of the talk, we present a real-world application of these techniques for hospital patient flow prediction, where we obtain interpretable insights from using decision tree models and advanced feature selection techniques. Our methods are now fully implemented at a major hospital in Boston and provide daily predictions with 80% accuracy, alongside patient-specific explanations. Empirically, we observe a 4% reduction in patient misplacement due to our tool.