COVID-19 reflection through Total Quality Management to prepare for future pandemics: lessons from Japan, United States, Taiwan, and India Open Access

Emerging stronger from COVID-19 requires diverse approaches tailored to the unique characteristics of each country or region. These include transforming healthcare systems (National Academy of Medicine, 2023; Madara et al., 2021; Lee et al., 2022; Isasi et al., 2021; Balser et al., 2021; Clancy et al., 2021), enhancing political stability and international relations (Ryu, 2021), raising public awareness through the arts (Ong and Deanna, 2022), and considering cultural factors to support the psychological well-being of young adults (Germani et al., 2020).

This study applies the principles of Total Quality Management (TQM), which is defined by the Japanese Society for Quality Control (JSQC) as an organization-wide approach to achieving long-term success through customer and societal satisfaction. TQM emphasizes customer focus, process orientation, system integration, continuous improvement, fact-based decision making, and inclusive participation across all organizational levels. These principles provide a suitable foundation for analyzing pandemic responses, where the goal is to continuously improve social systems to protect public health. By integrating TQM into a structured problem-solving framework — linking scientific leadership, risk communication, and digital tools — this study contributes a novel perspective on managing complex public health crises.

This study distinguishes itself from existing literature in the following four aspects. Firstly, it is an integration of TQM with public health and digital governance. Unlike previous studies that applied Lean, Six Sigma, or Business Continuity Planning (BCP) to internal operations in hospitals or healthcare systems (Hung et al., 2022; Prastiwi and Ayuningtyas, 2023; Jadhav et al., 2023), this paper develops a TQM-based structured problem-solving process tailored to pandemic management at a policy and systems level. It explicitly incorporates cross-national public health outcomes, particularly infection and mortality data, as a foundation for evaluating the effectiveness of governance structures. Second, the study proposes a problem-solving model with four critical perspectives: (1) science-based leadership, (2) process management of social systems; (3) infection prevention through digital risk communication, and (4) targeted protection of high-risk populations. These perspectives are derived from comparative epidemiological analysis across United States, Japan, Taiwan, and India, and have not been jointly conceptualized in existing quality management literature. Third, the study introduces a regionally adaptive digital system for risk communication, based on Device Tokens (DTs) and QR codes, designed to provide individuals and communities with real-time, context-specific infection risk ratings while preserving privacy. This platform aims to bridge the gap between epidemiological intelligence and both individual behavioral guidance and policy-level decision-making. Fourth, this paper adopts a holistic, cross-sectoral perspective that integrates TQM, public health policy, digital infrastructure, and societal behavior. This integration enables both proactive prevention and real-time feedback which are key features for enhancing societal resilience in future pandemics.

The COVID-19 pandemic has resulted in approximately 6.95 million deaths globally (0.088 percent of the total population) and over 770 million confirmed cases (9.8 percent) as of August 2023. However, these aggregated statistics conceal significant differences in outcomes across the countries/regions studied. For instance, United States reported one death per 300 people, while Japan, Taiwan, and India showed widely varying patterns of infection and mortality over time. Figure 1 illustrates the weekly mortality and infection trends per one million population in these four regions, highlighting diverse trajectories of pandemic control despite common exposure to the virus (Table 1).

Such variation invites fundamental questions: Why did some regions succeed in minimizing fatalities and social disruption, while others faced prolonged waves of infection? How should these differences be interpreted and translated into lessons for future pandemic preparedness?

A closer examination reveals recurring systemic issues that transcend national contexts. In United States, the early phase of the pandemic was marked by political conflicts that undermined science-based leadership, resulting in shortages of protective equipment, inconsistent messaging, and delayed interventions. In contrast, Taiwan demonstrated the effectiveness of pre-established governance structures and centralized decision-making, which enabled rapid containment and built public trust.

India’s severe mortality, though relatively lower in per capita terms, highlights challenges in health infrastructure and public communication. Japan, while initially showing restraint in case numbers, experienced a consistent rise in deaths across successive waves (Figure 2), raising concerns about its ability to protect high-risk populations. According to data compiled by the National Institute of Population and Social Security Research (IPSS), based on municipal government reports, as of September 19, 2022, Japan had recorded 43,840 officially confirmed COVID-19 deaths. Of these, 87.2 percent were individuals aged 60 or older, highlighting the disproportionate impact of the pandemic on older adults and underscoring the critical importance of targeted protection for high-risk populations (IPSS, 2022).

From these observations, four interrelated challenges emerge: (1) the failure of top leadership to act on scientific evidence, (2) insufficient integration of infection control into daily social processes, (3) delays in initial response and digital utilization, and (4) lack of targeted strategies to protect vulnerable populations. These challenges form the foundation for the four analytical perspectives developed in this study, aimed at constructing a structured, TQM-based framework leveraging digitalization for managing future pandemics.

The analysis is built on a TQM framework that conceptualizes pandemic control in three steps.

• Step I: social daily life, comprising routine economic and social activities such as work, education, mobility, and interpersonal interactions.

• Step II: testing system, represented by PCR tests and other surveillance methods.

• Step III: medical system, for treating infected patients and preventing fatal outcomes.

This three-step structure mirrors the basic model of quality assurance, in which quality is built into upstream processes before defects are detected downstream. Applied to COVID-19, TQM prioritizes upstream quality assurance at Step I, aiming to reduce infection risk before problems arise, rather than relying solely on testing and medical treatment as downstream corrective actions.

To operationalize Step I in a way that is analytically tractable, this study introduces a set of five “quality elements” that collectively determine whether a society remains “coronavirus-free” or enters a “coronavirus epidemic” state. These elements, denoted as Z and A-D, were reconstructed from prior models by Kano et al. (2020), Suzuki et al. (2022) and are grounded in TQM’s principle of managing quality at the process level. They translate the rather abstract idea of “quality in social daily life” into concrete, observable drivers of infection risk:

• Z: Virus characteristics and immunity status (e.g. variant type, vaccination status)

• A: Mobility patterns (cross-border, domestic, local)

• B: Mode of communication (online vs. face-to-face)

• C: Presence of vulnerable populations (e.g. elderly, chronically ill, institutional settings)

• D: Behavioral compliance with basic preventive measures (e.g. masking, ventilation, distancing, hand hygiene)

These quality elements fulfill three key conditions to serve as an effective framework for process-level pandemic prevention.

• 1.

  Sufficiency:

The five elements collectively cover all critical drivers of infection risk embedded in daily life. Together, they account for both structural (Z, A, C) and behavioral (B, D) factors, thus providing a holistic framework for understanding how risk emerges.

• 2.

  Criticality:

Each element has been shown, through empirical studies and pandemic policies, to directly influence transmission outcomes. For instance, high-density face-to-face interactions (B) in the absence of mask compliance (D), or insufficient risk mitigation among vulnerable groups (C), have been strongly correlated with infection surges.

• 3.

  Exclusiveness:

The elements are mutually non-overlapping in their operational definitions. For example:

• Element Z (virus and immunity) pertains to biological characteristics.

• Element A focuses on population movement.

• Element B relates to communication style and interaction modality.

• Element C identifies high-risk individuals based on health/vulnerability status.

• Element D captures hygiene and behavioral compliance.

While interrelated in practice, these categories are designed to diagnose and intervene along distinct dimensions of risk within a TQM framework, akin to how product defects are classified by type in quality engineering.

This paper focuses on the following four key perspectives:

This perspective addresses how leadership, grounded in scientific evidence, influences the timing, consistency, and effectiveness of pandemic response. Leadership is one of the most critical drivers in the structured problem-solving framework, especially in the prevention phase. It shapes institutional coordination, risk framing, public trust, and the integration of expertise into decision-making.

In United States, leadership conflict emerged between the Centers for Disease Control and Prevention (CDC) — a science-driven institution — and the White House under President Trump, who prioritized economic and electoral considerations ahead of the November 2020 election. This discord delayed border controls, hampered the procurement of personal protective equipment (PPE), and disrupted public guidance on social distancing and masking. Compounding these challenges, the decentralized federal system led to heterogeneous responses across states. The outcome was marked by high death rates and inconsistent application of protective measures, often split along partisan lines (Figure 5 and Figure 6).

In India, the government initially downplayed the severity of COVID-19, declaring on March 13, 2020, that it was not a health emergency. However, this position was reversed abruptly with a nationwide lockdown imposed from March 25, 2020. The opaque and centralized nature of India’s policy formulation — largely concentrated within the Prime Minister’s office and select bureaucrats —limited scientific input and contributed to erratic responses, such as sudden lockdowns without sufficient preparation for migrant workers and healthcare logistics.

By contrast, Taiwan exemplified a proactive, evidence-based leadership model. Drawing lessons from the 2003 SARS outbreak, the government activated the Central Epidemic Command Center (CECC) on January 20, 2020 — well before widespread global escalation. The CECC, though non-permanent, was endowed with strong cross-ministerial authority, integrated military and private-sector collaboration, and emphasized transparent communication. Daily press briefings by Health Minister Chen Shih-chung from January through June 2020 helped maintain public trust and compliance.

These contrasting cases illustrate that scientific leadership is not merely about technical expertise but also about institutional preparedness, communication consistency, and inclusive governance. The failure to integrate science at the top level can fragment responses and erode public confidence, whereas science-aligned leadership enables timely interventions and whole-of-society mobilization. Table 2 summarizes these country/region-level contrasts.

This perspective addresses the importance of process-oriented thinking in quality assurance, especially in the context of pandemic response. According to TQM, building quality into everyday operations requires a proactive and systemic approach to managing risks.

In Japan, although systems such as the ‘Act on the Prevention of Infectious Diseases and Medical Care for Patients with Infectious Diseases’ had been enacted following the 2009 influenza pandemic, implementation of such principles was limited during COVID-19. This resulted in insufficient allocation of hospital beds and unclear decision criteria for activity restrictions. In contrast, Taiwan leveraged lessons from SARS to institutionalize rapid, coordinated responses, highlighting the importance of process governance in ensuring societal resilience.

The contrast illustrates how embedding preventive mechanisms into everyday governance processes — through legal infrastructure, inter-agency coordination, and operational readiness — can support consistent and effective pandemic management across regions and sectors (Table 2).

Building on TQM’s emphasis on timely feedback and data-driven action, this perspective emphasizes the role of digital tools in enabling real-time communication and autonomous behavioral guidance.

Unlike traditional approaches that rely solely on top-down alerts, digital systems based on device tokens and QR codes allow individuals to access personalized infection risk assessments, informed by crowd density, ventilation, and preventive compliance data. These systems also enhance administrative situational awareness, enabling dynamic adjustments in policy and resource allocation.

In Taiwan, QR-based entry logs facilitated centralized health monitoring and swift regulation updates. In contrast, Japan’s COCOA app struggled due to poor uptake and insufficient integration with public policy. This demonstrates that well-integrated digital risk communication not only guides individual behavior but also serves as a foundational decision-support infrastructure for leadership and health authorities (Table 2).

TQM emphasizes equity and customer orientation, which, in the pandemic context, translates to protecting the most vulnerable populations. This perspective focuses on applying stratified risk models to identify and prioritize support for high-risk individuals, including the elderly, immunocompromised, and frontline workers.

In countries/regions like India and United States, the lack of targeted protections contributed to disproportionate outcomes in marginalized communities. In contrast, Taiwan proactively shielded vulnerable groups by integrating risk data into their response strategy (Table 2). This approach aligns with quality management principles by focusing resources where they can produce the greatest value —saving lives and reducing disparities. Digitally enabled stratification — via health records, exposure history, and demographic data — can further refine this process, ensuring precise, timely, and equitable protection.

To evaluate the applicability of the proposed TQM-based framework across democratic contexts, this study selected Japan, United States, Taiwan, and India as focal cases. These four countries/regions were chosen not only for their contrasting pandemic outcomes but also because they represent diverse institutional pathways for embedding TQM principles — scientific leadership, process discipline, digital integration, and risk-based protection — within mature democratic systems. While all share relatively high degrees of political institutionalization and stability, their modes of implementing quality-oriented governance differ significantly, providing a robust basis for comparative analysis.

Japan exemplifies a process-oriented administrative culture with strong institutional capacity but a cautious, bureaucratic decision-making process that delayed timely action. Taiwan demonstrated agile, science-based leadership supported by cross-sectoral coordination, digital innovation, and public trust, enabling swift and consistent response. United States displayed fragmented leadership between federal and state levels, which hindered coherent policy implementation despite advanced digital capabilities. India, although a democracy with a long participatory tradition, faced socio-economic constraints and limited health infrastructure that restricted consistent process control and feedback-based learning.

This diversity within a broadly comparable democratic framework offers a rich testbed for assessing the transferability of TQM-based problem-solving principles to public health governance. The contrasts are reflected not only in infection and mortality data (Figures 1 and 2, Table 1) but also in how leadership, process alignment, and digital communication interacted to influence behavioral compliance (Table 2). The comparative perspective thus reinforces the generalizability and practical relevance of TQM as a framework for adaptive governance under uncertainty.

Furthermore, it is instructive to consider the case of United Kingdom, which, despite its highly institutionalized democracy, advanced healthcare infrastructure, and substantial public resources, encountered serious challenges in the early stages of its COVID-19 response. Delays in lockdown implementation, hesitancy toward mask mandates, and inconsistent public communication contributed to high mortality and widespread confusion. These challenges reveal that institutional maturity alone does not guarantee effective crisis management. Rather, the key lies in timely coordination, cross-agency integration, and sustained behavioral compliance — core elements of process quality and risk communication emphasized in this study’s TQM-based framework. The UK experience thus reinforces the central argument that even in well-established democracies, proactive leadership and structured, feedback-driven social systems are indispensable for ensuring coherence and effectiveness under conditions of uncertainty.

This study has proposed a structured problem-solving process based on TQM to address key challenges in pandemic preparedness and response. The framework integrates statistical comparisons of public health outcomes across four countries/regions (Japan, United States, Taiwan, and India) with a novel risk communication system grounded in digital infrastructure. This integrated approach is both theoretically informed and practically actionable.

The novelty of this framework lies in its application of TQM principles — customer focus, process orientation, system integration, continuous improvement, and fact-based decision making — to pandemic risk management at the societal level, rather than within individual organizations alone. While many prior studies have applied Lean, Six Sigma, or Business Continuity Planning (BCP) to support operational continuity during COVID-19, these approaches often focused on internal institutional improvements. In contrast, our study uses TQM to link epidemiological data with system-level policy evaluation and citizen behavior.

The four perspectives (1) leadership grounded in science, (2) process assurance in daily life, (3) digital risk communication, and (4) risk-based protection for vulnerable populations — are mapped onto a three-phase TQM-based structure: upstream prevention (I), real-time detection and communication (II), and targeted response (III). These dimensions reflect the critical components necessary for managing pandemics as complex, dynamic quality problems involving diverse stakeholders and feedback loops.

The above three-step TQM-based structure is summarized in an integrated framework in Figure 3. On the left side of Figure 3, risk inputs in social daily life are represented by five quality elements —Z (virus characteristics and immunity), A (mobility), B (mode of communication), C (vulnerable populations), and D (preventive behavior) — which together capture how everyday activities and social structures generate infection risk. In the upper right, a digital risk rating engine aggregates data from device tokens or QR codes, visualizes real-time trends, and computes risk levels for different regions and population groups. In the lower right, these risk ratings are translated into targeted interventions, such as prioritizing protection for high-risk individuals, early treatment, priority admission, and online work assignments. In this way, Figure 3 connects upstream risk in daily life, digital information processing, and downstream policy actions within a single TQM-based problem-solving model. These steps are further operationalized into a structured, iterative problem-solving process shown in Figure 4.

Among the perspectives, Perspective 3 introduces a digital risk communication model, which represents a key contribution to this study. Unlike conventional contact tracing applications such as Japan’s COCOA or India’s Aarogya Setu, the proposed system focuses not on retrospective identification but on proactive, context-aware risk visibility. Using technologies such as Device Tokens (DT) or QR code platforms, the system enables individuals and institutions to assess infection risk in real time based on (1) environmental context (location, ventilation, crowding), (2) behavioral context (masking, distancing), and (3) local epidemiological trends.

This system offers three main innovations:

1. Prevention-oriented design: It emphasizes safety confirmation type logic—action is taken only when safety is assured, in contrast to abnormality detection models which act after danger is visible.

2. Two-way communication: Citizens receive alerts based on data, while authorities gain aggregate-level insights to inform timely interventions.

3. Risk stratification: With user consent, the system could incorporate health data to protect high-risk groups and prioritize resource allocation accordingly.

The integration of these digital tools into public health decision-making would enhance leaders’ situational awareness, enable precise communication strategies, and build public trust. Taiwan’s success in using centralized QR-based logs to localize outbreaks stands in contrast to Japan and India, where the lack of real-time, usable data delayed responses. Thus, the proposed framework not only explains past failures but offers a practical path toward resilient, feedback-driven governance.

In summary, this paper advances the field by combining a comparative epidemiological lens, a TQM-inspired structure for phased prevention, detection, and response, and an innovative digital platform design for risk communication. This synthesis supports a system-level shift from reactive to preventive pandemic management.

As an example of analysis on COVID-19 infection from the four perspectives, this section focuses on Perspective 1: the commanding role of a top leader and Perspective 2: the importance of mask wearing and vaccination as key COVID-19 quality factors. In this analysis, the differences among the policies of the 50 states and the District of Columbia are examined.

The data used here are based on mask-wearing rates and mobility from the Institute for Health Metrics and Evaluation (IHME, 2022) COVID-19 projections, the number of positive cases in each state (including the District of Columbia) from CDC (2023), and the timeline of statewide mask requirements from Ballotpedia (2022).

For the analysis of mask mandates up to 30 September 2020, the 50 states and Washington, D.C. were first divided into two categories: states with a statewide mask mandate and those without. States with statewide mandates were then classified into four groups: Group A (statewide mandate before 5 June 2020), Group B (statewide mandate after 5 June 2020), Group C (≥30 percent of the population covered only by partial mandates), and Group D (<30 percent covered only by partial mandates).

The date 5 June corresponds to the World Health Organization’s recommendation to wear masks. When the total number of positive cases per million population in September 2020 is compared across these groups, states in Group A, which introduced statewide mask mandates earliest, show substantially fewer infections than the other groups. Overall, states that implemented mask mandates earlier recorded fewer positive cases in September.

Furthermore, an analysis of the relationship between the mask-wearing rate on 30 September 2020 and the number of positive cases per million population during September found a clear negative association. The correlation coefficient was r = −0.69 (p = 0.0001) for the test of no correlation; when the state of Wyoming (WY) was excluded — Wyoming has a population of approximately 600,000, the lowest among the 50 states, and the second-lowest population density — the correlation became r = −0.79. In other words, the higher the mask-wearing rate, the fewer the number of positive cases. This quantitative relationship reinforces the importance of timely and science-based policy decisions and behavioral compliance, which are central themes of the TQM-based framework proposed in this study.

The scatterplot of the mask effect shown in Figure 5 is stratified according to the winning party in the 2020 presidential election (close races: states with 5 percent or less difference in the number of votes). The differences in mask-wearing rates by supporting party are clear, indicating the presence of political factors in the behavioral change of wearing masks.

In the 23 states where the effects of mobility restrictions (e.g., stay-at-home orders) were controlled for, the relationship between mask wearing and the number of positive cases in September remained significant (Suzuki et al., 2021).

As discussed in the previous section, focusing on the 50 states and the Washington D.C. in United States, the relationship between the number of deaths at the start of the fourth wave (2021.8.17) and the vaccination that may affect the number of deaths during this period is examined. A lag time of 37 days is considered, consisting of 25 days from vaccination - appearance of vaccination effect - exposure - positive result, and 12 days from positive result to death (Linton et al., 2020; Xin et al., 2021; Nature News, 2021). Figure 6 shows the number of deaths on August 17 per million population on the vertical axis and the vaccination rate on July 11, 2021, on the horizontal axis (Our World in Data, 2024b), stratified by the party, Democratic or Republican.

• A)

  The low vaccination rate group with high Republican support has a strong correlation with a high number of deaths.

• B)

  The effect of vaccination is significant with fewer deaths in the high vaccination rate group with high Democratic support.

The above result indicates that there is a clear difference in vaccination rate and the number of deaths depending on the political party supported as well as the correlation between the start date of mask mandate and the extent of its implementation and the number of positive cases. As can be seen the presence of political factors between behavior change and the number of deaths, the role of the top leader is important.

Drawing on lessons from four countries/regions, including Japan, United States, Taiwan, and India, this study identified four critical perspectives: (1) science-based leadership, (2) process management of daily social systems, (3) infection prevention through digital risk communication, and (4) protection of high-risk populations through risk rating, and proposed a structured problem-solving framework for future pandemic preparedness by integrating TQM principles with comparative epidemiological analysis and digital risk communication. Unlike conventional studies that separately address healthcare capacity, public behavior, or institutional responses, this study provides a cross-cutting framework rooted in TQM. It reconceptualizes pandemic response as a quality assurance process comprising three sequential steps: quality built into daily life (Step I), quality of detection and testing (Step II), and quality of treatment systems (Step III). Through this lens, the proposed digital risk communication system — leveraging device tokens (DT) and QR-based feedback loops — not only supports individual decision-making but also enhances leadership situational awareness and process transparency.

Furthermore, the quality elements (Z, A-D) derived from TQM principles and mapped onto social behaviors (e.g., mobility, face-to-face interactions, preventive compliance) provide a diagnostic framework for risk management that is both comprehensive and operationalizable. These elements explain how differences in public health outcomes across the countries/regions studied can be traced back to structural and behavioral inputs.

By bridging statistical analysis with a system-level quality management approach, the study contributes a new interdisciplinary model that integrates digital infrastructure, behavioral science, and policy design. Future studies may build upon this model to develop simulation tools and implementation strategies tailored to regional contexts.

The author would like to express sincere gratitude to Professor Emeritus Noriaki Kano, Professor Tomonori Hasegawa, and Mr. Yoshihisa Okamoto for their valuable guidance, and to Professor Emeritus Hiroe Tsubaki, Professor Tomoko Matsui, Professor Daisuke Murakami, and all members of the Institute of Statistical Mathematics ‘COVID-19 Response Project’ for their support and insightful suggestions.