An Inverse Optimization Approach to Measuring Clinical Pathway Concordance

Abstract

Clinical pathways outline standardized processes in the delivery of care for a specific disease. Patient journeys through the healthcare system, however, can deviate substantially from these pathways. Given the positive benefits of clinical pathways, it is important to measure the concordance of patient pathways so that variations in health system performance or bottlenecks in the delivery of care can be detected, monitored, and acted upon. This paper proposes the first data-driven inverse optimization approach to measuring pathway concordance in any problem context. Our specific application considers clinical pathway concordance for stage III colon cancer. We develop a novel concordance metric and demonstrate using real patient data from Ontario, Canada that it has a statistically significant association with survival. Our methodological approach considers a patient’s journey as a walk in a directed graph, where the costs on the arcs are derived by solving an inverse shortest path problem. The inverse optimization model uses two sources of information to find the arc costs: reference pathways developed by a provincial cancer agency (primary) and data from real-world patient-related activity from patients with both positive and negative clinical outcomes (secondary). Thus, our inverse optimization framework extends existing models by including data points of both varying “primacy” and “alignment.” Data primacy is addressed through a two-stage approach to imputing the cost vector, whereas data alignment is addressed by a hybrid objective function that aims to minimize and maximize suboptimality error for different subsets of input data. This paper was accepted by Chung Piaw Teo, Special Issue on Data-Driven Prescriptive Analytics.

Development of population‐level colon cancer pathway concordance measures and association with survival

“…Methods of measuring pathway concordance have previously been described in published literature. [7][8][9][10][11][12][13][14][15] However, few of these methods could be used to measure concordance with a pathway for a population. For instance, studies have compared actual care with pathway recommendations for single episodes of care (eg, an inpatient hospital stay) within single institutions 7,9,10,13 using data collected through clinician reporting 5,8 or manual analysis of hospital data, 11 which are unfeasible at the population level.…”

“…16 Chan et al developed a summary measure of colon cancer pathway concordance that "weighted" the importance of concordance with each pathway event based on the event's impact on patient survival; this measure was also found to be associated with survival. 15 The aim of this work was to determine the feasibility of developing and validating summary measures of pathway concordance at a population level using administrative data. We developed two measures for quantifying concordance with a colon cancer pathway: a cumulative count of concordant events (CCCE) and a variant of the Levenshtein algorithm used in information science and previously applied to stroke pathway concordance measurement.…”

“…Methods of measuring pathway concordance have previously been described in published literature 7‐15 . However, few of these methods could be used to measure concordance with a pathway for a population.…”

“…We previously measured concordance with a population‐level colon cancer pathway using a simple count of pathway‐recommended events received and found that concordance was significantly associated with 4‐year overall survival in Stage III but not Stage II colon cancer patients 16 . Chan et al developed a summary measure of colon cancer pathway concordance that “weighted” the importance of concordance with each pathway event based on the event's impact on patient survival; this measure was also found to be associated with survival 15 …”

Data-driven integrated care pathways: Standardization of delivering patient-centered care

“…Important deviations from the current bases will signal personalization or the need to create new bases. Causal inference methods (21-23), optimization models (24)(25)(26), and machine learning techniques (27, 28) will be used to identify new standard and more adaptable pathways through nature experiments, computer simulations, and real-world data analysis (29, 30). Furthermore, DICP allows for close monitoring of personalized delivery processes, by checking the uniformity of average plans and personalization algorithms.…”