Longitudinal bi-criteria framework for assessing national healthcare responses to pandemic outbreaks

Abstract

Pandemics like COVID-19 have illuminated the significant disparities in the performance of national healthcare systems (NHCSs) during rapidly evolving crises. The challenge of comparing NHCS performance has been a difficult topic in the literature. To address this gap, our study introduces a bi-criteria longitudinal algorithm that merges fuzzy clustering with Data Envelopment Analysis (DEA). This new approach provides a comprehensive and dynamic assessment of NHCS performance and efficiency during the early phase of the pandemic. By categorizing each NHCS as an efficient performer, inefficient performer, efficient underperformer, or inefficient underperformer, our analysis vividly represents performance dynamics, clearly identifying the top and bottom performers within each cluster of countries. Our methodology offers valuable insights for performance evaluation and benchmarking, with significant implications for enhancing pandemic response strategies. The study's findings are discussed from theoretical and practical perspectives, offering guidance for future health system assessments and policy-making.

Keywords: Bi-criteria analysis; COVID-19 pandemic; Data envelopment analysis; Fuzzy clustering; Healthcare performance evaluation; Longitudinal analysis.

Figure 1

Recovery and death rates, and healthcare resources from selected OECD countries: Canada, France, Germany, Italy, New Zealand, South Korea, and the USA: (a) daily cumulative recovery rates, (b) daily cumulative death rates, (c) resources of NHCS.

Figure 2

Conceptual model of an NHCS in response to the COVID-19 pandemic.

Algorithm 1

Qualitative bi-criteria approach.

Figure 3

Principle of longitudinal DEA.

Algorithm 2

The fuzzy clustering algorithm FJM.

Figure 4

Qualitative bi-criteria visual matrix.

Figure 5

Mapping of population vulnerability profiles for each cluster.

Figure 6

Clinical performance comparison among cluster 2 countries after 25 and 111 days of the pandemic.

Figure 7

Clinical performance comparison among cluster 3 countries after 25 and 111 days of the pandemic.

Figure 8

Clinical performance comparison among cluster 4 countries after 25 and 111 days of the pandemic.

Figure 9

Clinical performance comparison among cluster 5 countries after 25 and 111 days of the pandemic.

Figure 10

DEA longitudinal results for clusters 2 to 5.

Figure 11

Longitudinal average efficiency index for each cluster.

Figure 12

Malmquist index for each cluster.

Figure 13

Malmquist index decomposition for selected countries.

Figure 14

Qualitative analysis for selected countries: (a) New Zealand’s NHCS maintained good performance over time but Spain’s was an under performer in cluster 2, (b) Germany’s NHCS outperforms USA’s in cluster 3, (c) Italy’s NHCS and France’s seems to follow similar patterns in cluster 4, and (d) South Korea’s NHCS seems to outperform Canada’s in cluster 5.