Sage Journals: Discover world-class research

Abstract

Objectives

The ongoing COVID-19 pandemic has led to an unprecedented loss of life and a severe economic downturn across the globe. Countries have adopted various social distancing and vaccination policies to reduce the spread of the disease and lessen the impact on healthcare systems. The world should work together to confront the disaster and challenge of COVID-19.

Methods

This study uses stochastic frontier analysis to measure the efficiency and influencing factors of the global response to COVID-19 epidemics and to provide follow-up strategies and reference guidelines.

Results

The results of this study show that (1) the average efficiency of the global response to COVID-19 is not good, with significant space for improvement of up to 60%; (2) adequate medical supplies and equipment can reduce mortality; (3) the initial implementation of social distancing policies and wearing masks can effectively reduce the infection rate; and (4) as infection rates and vaccination rates increase so that most people have basic immunity to COVID-19, the epidemic will gradually be reduced.

Conclusions

As the world becomes more aware of the COVID-19 disease, humans will gradually return to normal social interaction and lifestyles. The results of this study are expected to provide a reference for the future direction of the global fight against epidemics and the improvement of public health policies.

Keywords

COVID-19 pandemic death efficiency stochastic frontier analysis

Introduction

In December 2019, a highly contagious virus was first detected in Wuhan, China. The virus was named “SARS-CoV-2,” and the infectious disease was named “COVID-19.” On 11 January 2020, China announced that the first man to contract the new virus had died on 9 January. Countries worldwide were made aware of the COVID-19 epidemic because Wuhan, China, reported its second death to the World Health Organization (WHO) on 17 January 2020. The United States health service announced that three major domestic airports had started screening passengers from Wuhan. Airports across Asia also enforced mandatory screening of passengers arriving from high-risk areas of China. On 23 January, China declared a lockdown of Wuhan, imposed a nationwide quarantine, and suspended flights and trains. The WHO stated the COVID-19 outbreak was not a global emergency, and there was “no evidence yet” of human-to-human virus transmission outside China. After two days of debate at the meeting of the WHO emergency committee, the committee decided not to declare the rising COVID-19 outbreak a Public Health Emergency of International Concern (PHEIC) for the time being.¹

However, on 30 January 2020, one month after the first case, the WHO did declare COVID-19 a PHEIC. It was confirmed as a pandemic by the WHO on 11 March 2020. Within months of the first COVID-19 outbreak, cases appeared on every continent and in most countries. On 4 April 2020, the WHO reported that more than 1 million COVID-19 cases had been confirmed globally, a more than 10-fold increase in less than a month. The COVID-19 pandemic has led to unprecedented loss of life and severe economic recession worldwide.^2–4 Globally, a cumulative 760,897,555 confirmed cases of COVID-19, including 6,874,585 deaths, have been reported to the WHO as of 15 March 2023. The global distribution of confirmed COVID-19 cases is shown in Figure 1. The United States has the highest number of confirmed deaths at 1,113,229, followed by Brazil with 699,310, and India with 530,789 deaths.⁵

Figure 1.

Global distribution of confirmed COVID-19 cases (through March 2023).⁶

The early days of the COVID-19 pandemic have led to a variety of prevention and mitigation measures around the world, from the closure of international borders to various social distancing and isolation measures at the local level, to reduce the spread of the disease and reduce the burden and impact on healthcare systems. China locked down Wuhan, a metropolis of 12 million people, on 23 January 2020 to contain the COVID-19 outbreak. Similar lockdowns and clearances followed in all 15 other cities in Hubei province, which was home to approximately 57 million people, while other Chinese cities followed with varying degrees of restrictions. Wenzhou in Zhejiang province, for example, limited one person per household to one outing every two days. Restrictions and lockdowns in China appeared to curb the spread of COVID-19 effectively. However, these measures have greatly disrupted people's work and lifestyle and may also have a significant impact on their health and well-being.⁷ New Zealand implemented the strictest lockdown measures in the world, raising COVID-19 prevention measures to the highest level of alert level 4 on 25 March 2020 and initiating a nationwide lockdown, border controls, and a complete halt to international travel. New Zealand's national lockdown and clearance policy was once the poster child for the COVID-19 response, as the government's rapid response to the pandemic and geographic isolation kept the country out of the pandemic until the end of 2021. However, due to this strict blockade, its economy suffered a huge loss of approximately US$10 billion.⁸ Japan issued measures for COVID-19 on 25 February 2020, urging people not to rush into hospitals without consultation to reduce the risk of exposure and cross-infection in healthcare facilities. In order to prevent large gatherings and crowded areas, measures such as telecommuting, staggering commuting times, leaves of absence when someone has cold-like symptoms, and closure of primary, middle, and high schools were advocated.⁹ Some countries declared a public health emergency due to shortages of protective gear, medical supplies, and equipment but did not consider lockdowns or quarantines as options for responding to the spread of COVID-19 in order to avoid severe economic impacts, which led to increasing COVID-19 cases and deaths.¹⁰ The Singapore government was among the first to close borders, quarantine all overseas travelers, and actively trace contacts. Singapore introduced a mandatory mask policy in April 2020, requiring everyone to record their movements by scanning bar codes.¹⁰

The United States began implementing various social distancing policies in mid-April 2020 to reduce the burden of COVID-19 on the healthcare system and reduce its spread. These included shelter-in-place orders (SIPO), remote offices, school closures, a ban on large gatherings, closure of entertainment venues, seating restrictions in bars and restaurants, and mask mandates.^9,11,12 Without vaccines and drug treatments at the onset of the pandemic, the aforementioned social distancing policies became a first-line response.¹² However, the literature on the effectiveness of social distancing and control policies in curbing the spread of COVID-19 was unclear and unconfirmed at this early stage. Some studies have suggested that early social distancing policies, especially the ban on large gatherings, can indeed reduce the transmission and diagnosis of COVID-19, but do not have a significant impact on mortality and cause significant losses to the overall economy.^8,12,13 Other studies, such as by Friedson et al.,¹⁴ who focused on the impact of the 29-day SIPO policy in California, found that the control policy was effective and reduced the number of confirmed COVID-19 cases per 100,000 people by 125.5–219.7 despite the economic cost and reduced the number of COVID-19-related deaths by 1661, with the largest reduction in COVID-19-positive cases in countries that introduced lockdowns earlier.¹⁴ The above studies show that COVID-19 epidemics could be slowed by an aggressive lockdown policy in the early days of the pandemic without a vaccine.¹⁰

The relatively high admission rate of patients infected with COVID-19, especially those requiring treatment in intensive care units (ICU), has resulted in a significant impact on medical services and medical personnel, as well as on medical services for other diseases.^15–18 Countries such as Germany, Austria, Portugal, Spain, Switzerland, and Italy had strong public and universal health services but had suffered health budget cuts in recent years due to the 2008 financial crisis, which reduced gross domestic product. They had a particularly successful response to the COVID-19 pandemic and low fatality rates in early 2020, but health services would become overstretched. By the end of 2020, the situation had rapidly transformed into a shortage of ICU beds, a surge in infection and fatality rates, and one of the most serious COVID-19 hotspots in the world.^19–22

Three years after the start of the COVID-19 pandemic, the world has implemented countermeasures and mitigation policies. The early social distancing and lockdown control policies did work but, in the long run, they caused great losses to countries, societies, and economies, while the reduction of interpersonal contact impacted physical and mental health.²³ Earlier studies by Karako et al.⁹ suggested that, in Japan, reducing the time spent in crowded areas to less than 4 hours could suppress the spread of infection, and so they suggested that the countermeasures to control the spread of COVID-19 should include avoiding crowded places as much as possible. Hâncean et al.²⁴ conducted a study involving 6895 patients in Bucharest, Romania, from August to October 2020, along with 13,272 individuals who interacted with them. They recorded the participants’ gender, age, and occupation and constructed a statistical model on the network effects of human-to-human transmission in order to investigate potential infection routes of COVID-19. Their findings indicated that a medical occupation had no significant effect on the spread of the virus. Instead, a transmission chain was observed among private sector workers, as well as between spouses, siblings, and older relatives. Some studies have shown a positive correlation between population density and mortality because a high population density facilitates the spread of the virus through close person-to-person contact. Transmission of COVID-19 is more likely to occur in areas with higher population density.^25–28

When someone infected with COVID-19 coughs, sneezes, talks, sings, or breathes, the virus may be transmitted from the mouth or nose of the infected person in the form of small liquid particles, which range from larger respiratory droplets to smaller aerosols.²⁹ Guidelines from the Centers for Disease Control and Prevention (CDC) and the WHO recommend wearing masks to prevent the spread of COVID-19.³⁰ According to the epidemiological data, the universal mask measures implemented in Japan, Hong Kong, Singapore, South Korea, and Taiwan were most effective in initially reducing the spread of COVID-19.¹⁶

On 5 January 2021, the WHO's Strategic Advisory Group of Experts on Immunization (SAGE) met to discuss the Pfizer/BioNTech vaccine, and it became the first vaccine to receive WHO validation for emergency use for COVID-19.¹ Several studies and observations have shown that vaccines are the most effective and cost-effective way to prevent and control infection, but herd immunity is still needed to end the COVID-19 pandemic.^31,32 In addition, patients who are fully vaccinated (FV) against COVID-19 reduced length of hospital stay, risk of ICU admission, and risk of death compared to those who are not vaccinate, and these vaccines protect against SARS-CoV-2 infection.^33,34 However, the risk of a breakthrough case of severe COVID-19 following vaccination remains, especially among people at higher risk of severe disease.^35,36

Historically, there have been few studies utilizing techniques like stochastic frontier analysis (SFA) to investigate the association between disease occurrence and mortality. In 2000, Sindo et al.³⁷ used SFA to construct an estimation model for pneumonia and influenza mortality in Japan. They aimed to mitigate the underestimation of actual death counts by utilizing excess deaths as a measure. The numerical findings revealed a substantial increase in influenza-related mortality compared to previous estimates. It is hoped that this mortality rate estimation model has the potential to enable timely prediction of influenza outbreaks in the future. Over the past three years, much literature has explored the impact of the COVID-19 pandemic on various global conditions. However, thus far, most have focused on the effectiveness of medical treatment of the virus, while only a few studies of short-term durations of several months to one year have addressed the topic of global effectiveness in combating the COVID-19 pandemic.²² Gearhart et al.¹² used order-m efficiency estimates to investigate the efficiency of the COVID-19 social distancing policy in the United States during the first 100 days of 2020 and found that a more effective strategy would have been to implement the policy in densely populated areas. While banning large gatherings is the most effective policy, shelter-in-place orders have less impact than nonessential business closures. The preliminary indications are that early withdrawal of social distancing policies can lead to surges in COVID-19 cases. Breitenbach et al.³⁸ applied data envelopment analysis (DEA) to explore the adverse output of mortality and infection rates in 220 countries by November 2020 and took the numbers of doctors and nurses and healthcare spending per 100,000 people as inputs. Their findings found that most countries were inefficient in fighting the COVID-19 pandemic. Pereira et al.²² explored 55 member states, associate members, partners, and other countries of the Organization for Economic Cooperation and Development (OECD) with DEA. The outputs of the study included population, infectious diseases, triage, hospitalizations, and ICU admissions, and the inputs included health expenditure and cost. Estonia, Iceland, Latvia, Luxembourg, the Netherlands, and New Zealand were found to be efficient countries. Lupu and Tiganasu³⁹ used DEA to assess the effectiveness of the European health system in combating COVID-19 during 2020, with inputs including COVID-19 cases, doctors, nurses, hospital beds, health expenditures, and outputs including COVID-19 deaths. They found that Italy, Belgium, Spain, and the United Kingdom were less efficient than other countries in the first phase of the pandemic. Ahmad et al.⁴⁰ used SFA to explore the efficiency of countries in Southeast Asia in response COVID-19, finding them to be Thailand, Malaysia, India, and the Philippines in order of least to most efficient. Maity et al.⁴¹ used SFA to explore the efficiency of combating COVID-19 in the states of India. The outputs included the recovery rate of COVID-19, and the inputs included the numbers of doctors, nurses, and isolation beds and exogenous variables such as the elderly population, gender, literacy rate, and urbanization rate. The results showed that investments in health infrastructure, such as doctors, nurses, police, isolation beds, and ICU beds, contributed to increased rates of COVID-19 recovery. An aging population can lead to inefficiency and population density hurts efficiency. Kınacı et al.⁴² used DEA and SFA to discuss a study period from 3 March 2020 after the first detected case in Turkey by using the numbers of COVID-19 cases, deaths, and beds in ICU and other efficiency analyses. Their analysis showed that efficiency in the early epidemic stage was not good, but there has been a gradual improvement since then. However, due to the study's short period for data collection, the change was very different from the actual epidemic.

In conclusion, the outputs used in studies on COVID-19 efficiency have been primarily the numbers of deaths or confirmed cases, while the inputs have been primarily the numbers of doctors, nurses, hospital beds, and beds in the ICU. Most methods used to study efficiency have been DEA or SFA. Both used the production frontier (efficiency frontier) to measure efficiency. DEA is a nonparametric method that uses linear programming to estimate frontier production functions, measure multiple inputs and outputs, and calculate the relative efficiency of each evaluated unit. SFA is a parametric approach involving a theory of econometric models and microeconomics that requires combining data to estimate production functions regarding hypotheses or statistical tests.⁴³ This study used SFA because the estimation results adopt absolute efficiency, which is more stable than DEA. SFA can also consider random error terms, which are more suitable for situations with large differences in sample data size, and it is not susceptible to the influence of outliers, which may cause calculation results to deviate from the actual situation. The results of SFA will not have the same efficiency value, that is, 1, of multiple decision-making units (DMUs) as the results of DEA.⁴⁴ SFA is more suitable for exploring the changing trend of global efficiency in combating COVID-19 year by year. In addition, this research will take the number of COVID-19 deaths as undesirable output and divides the three years of the COVID-19 pandemic (the time of data collection is from January 2020 to December 2022) into two stages for analysis. The first phase is 2020 when the COVID-19 vaccine was not yet completed. The second phase is 2021–2022, when the vaccine had begun to be administered. This study measures the effectiveness and impact factors of the global response to the COVID-19 pandemic over three years. This study aims to measure which strategies or policies are most effective in combating the COVID-19 pandemic in countries and regions worldwide and which need to be improved in terms of their performance to address and mitigate the ongoing COVID-19 disease and mortality. The study results are also expected to provide a reference for the community to consider how to reduce and propose appropriate measures to improve their public health policies to eliminate the PHEIC of COVID-19. In the future, it can also be used as a reference and effective guide for policymakers and health units in countries worldwide in case of epidemics or health crises.

Methodology

SFA method

Efficiency evaluation measures the operational performance of a DMU against the operational space that the DMU can improve. Efficiency is the problem of measuring the optimal efficiency of input and output under the goal of maximum output or minimum cost. Farrell⁴⁵ argued that nonparametric and parametric methods can be used to measure efficiency and proposed an efficiency measurement method that has become the pioneering technique boundary analysis.⁴⁶ DEA is the most commonly used nonparametric method, while SFA is the most representative parametric method. Both methods are commonly used to measure various performances.

SFA is a method proposed by Aigner et al.⁴⁷ and Meeusen and van den Broeck.⁴⁸ The main concept of SFA is to connect the most efficient input–output combination points to form a production efficiency boundary. Only the enterprises at the leading edge of efficiency have technical efficiency. Battese and Coelli⁴⁹ proposed a Cobb–Douglas stochastic boundary production function model for continuous intertemporal data. Considering that environmental characteristics influence COVID-19, exogenous variables are included to estimate the efficiency of the global response to COVID-19 and analyze the impact of exogenous variables on COVID-19 deaths. The Cobb–Douglas stochastic frontier model takes the form⁴⁶:

l n Y_{i t} = β_{0} + \sum_{j} β_{j} l n X_{i t} + v_{i t} - u_{i t}, i = 1, \dots N, t = 1, \dots T

(1)

u_{i t} = δ_{0} + \sum_{j} δ_{j} l n Z_{i t} + \dots + ε_{i t}

(2)

where Y_it represents the appropriate type of output of the i-th manufacturer at time t; X_it represents the i-th vendor input at time t; β represents the individual input coefficient of the production function; Z_it is the exogenous (environmental) variable; and v_it represents the symmetric interference term of random variation of the production function. Here,

v_{i t} \sim i d d N (0, σ_{v}^{2})

and u_it are independent of each other. And u_it is the inefficiency factor of manufacturer i at phase t. It represents the degree of production technology inefficiency and is the only nonnegative random variable. Finally, δ₀ is the intercept term, and ε_it is a random error and a nonnegative truncated normal distribution.⁴⁶

Samples and data sources

The samples and data used in this study are from the University of Washington's School of Medicine. The Institute for Health Metrics and Evaluation (IHME)⁵⁰ provides the data on COVID-19 modeling estimates. The IHME collected COVID-19 data from various sources provided by more than 200 countries and territories, including local and national governments, hospital networks and associations, the WHO, and the data reports from Our World. Most came from a Johns Hopkins University (JHU) data repository. Data on COVID-19, including cumulative (or daily) cases, deaths, etc., were collected by JHU as a free repository and were verified by the CDC. One reason the number of COVID-19 deaths and confirmed cases observed in the data provided by the IHME may differ from the figures reported by governments is the differences in how data are estimated and identified by various countries and regions. The IHME is designed to address irregularities in daily data on deaths or confirmed cases of COVID-19. IHME uses a hybrid modeling approach to generate our forecasts, which incorporates elements of statistical and disease transmission models. IHME estimate past daily infections in a modeling framework that leverages data from seroprevalence surveys, daily cases, daily deaths, and, where available, daily hospitalizations.⁵⁰

This study collated data on COVID-19 from countries and regions worldwide from 2020 to 2022 in IHME's COVID-19 database. After deleting countries and regions with incomplete data, data from 136 countries and autonomous regions were obtained, as shown in Table 1.

Table 1.

136 countries and autonomous regions.

No.	Location	No.	Location	No.	Location	No.	Location	No.	Location	No.	Location	No.	Location
1	Afghanistan	21	Bulgaria	41	Finland	61	Israel	81	Morocco	101	Republic of Korea	121	Taiwan
2	Albania	22	Canada	42	France	62	Italy	82	Mozambique	102	Republic of Moldova	122	Thailand
3	Algeria	23	Central African Republic	43	Gambia	63	Jamaica	83	Myanmar	103	Romania	123	Togo
4	Andorra	24	Chad	44	Georgia	64	Japan	84	Nepal	104	Russian Federation	124	Trinidad and Tobago
5	Angola	25	Chile	45	Germany	65	Kenya	85	Netherlands	105	Rwanda	125	Tunisia
6	Argentina	26	China	46	Ghana	66	Kuwait	86	New Zealand	106	San Marino	126	Turkey
7	Armenia	27	Colombia	47	Greece	67	Latvia	87	Niger	107	Saudi Arabia	127	Uganda
8	Australia	28	Costa Rica	48	Guam	68	Lebanon	88	Nigeria	108	Senegal	128	Ukraine
9	Austria	29	Côte d'Ivoire	49	Guinea	69	Libya	89	North Macedonia	109	Serbia	129	United Arab Emirates
10	Azerbaijan	30	Croatia	50	Guyana	70	Lithuania	90	Norway	110	Sierra Leone	130	United Kingdom
11	Bahamas	31	Cyprus	51	Haiti	71	Luxembourg	91	Pakistan	111	Singapore	131	United States of America
12	Bahrain	32	Czechia	52	Honduras	72	Madagascar	92	Panama	112	Slovakia	132	Uruguay
13	Bangladesh	33	Democratic Republic of the Congo	53	Hong Kong	73	Malawi	93	Papua New Guinea	113	Slovenia	133	Venezuela
14	Barbados	34	Denmark	54	Hungary	74	Malaysia	94	Paraguay	114	South Africa	134	Vietnam
15	Belgium	35	Ecuador	55	Iceland	75	Maldives	95	Peru	115	Spain	135	Zambia
16	Belize	36	Egypt	56	India	76	Mali	96	Philippines	116	Sri Lanka	136	Zimbabwe
17	Bolivia	37	Equatorial Guinea	57	Indonesia	77	Malta	97	Poland	117	Sudan		-
18	Bosnia and Herzegovina	38	Estonia	58	Iran	78	Mexico	98	Portugal	118	Suriname		-
19	Botswana	39	Ethiopia	59	Iraq	79	Mongolia	99	Puerto Rico	119	Sweden		-
20	Brazil	40	Fiji	60	Ireland	80	Montenegro	100	Qatar	120	Switzerland		-

Variables and empirical model

Based on the severity of the risk of illness and death caused by COVID-19 over the past three years and the issue of how to reduce the characteristics related to diagnosis and death, this study selected data of inputs, outputs, and exogenous variables to explore the effectiveness of the global fight against SARS-CoV-2 infection and treatment. The research results are expected to serve as a reference for all fields of research to study the direction and strategy of reducing or relieving the PHEIC of SARS-CoV-2.

This study collated data from 2020 to 2022 from the COVID-19 database of IHME. Cumulative deaths (CD; raw data with an excess mortality scalar applied) from COVID-19 were used as undesirable output. The WHO⁵¹ proposed that measuring excess mortality is essential for understanding the pandemic's impact. The changes in mortality trends can provide information for policymakers to reduce excess mortality and effectively prevent future crises. Excess mortality can be used as an indicator of disease prevalence to estimate an epidemic's duration, region, and scale. The true extent of excess deaths is often obscured because of the limited investment in data systems in many countries. These new estimates use the best available data and are generated with a sound methodology and in fully transparent manner. Therefore, the cumulative number of COVID-19 deaths due to adverse outcomes adopted in this study is the excess mortality rate measured from the original data, which can better reflect the actual situation of COVID-19 deaths. Inputs in this study include cumulative infections (CI; mean estimate), cumulative COVID-19 hospital beds needed (bed; mean estimate), and cumulative COVID-19 ICU beds needed (ICU; mean estimate). The undesirable outputs and inputs above can be substituted into Equation (1) to obtain Equation (3):

\ln (C D_{i t}) = β_{0} + β_{1} \ln (C I_{i t}) + β_{2} \ln (B e d_{i t}) + β_{3} \ln (I C U_{i t}) + V_{i t} - U_{i t}

(3)

In this study, exogenous variables are divided into the two phases. Exogenous variables for phase I (2020) are population density (PD), the total number of beds that exist at a location (EBed), the total number of ICU beds that exist at a location (EICU), and the percentage of population reporting always wearing a mask when leaving home (Mask). The major difference between the two phases is the addition of cumulative all FV in the second phase (2021–2022). Equations (4) and (5) can be obtained by substituting the exogenous variables of the two stages into Equation (2):

u_{i t} = δ_{0} + δ_{1} l n P D_{j t} + δ_{2} l n E B e d_{j t} + δ_{3} l n E I C U_{j t} + δ_{4} l n M a s k_{j t} + \dots + ε_{i t}

(4)

and

u_{i t} = δ_{0} + δ_{1} l n P D_{j t} + δ_{2} l n E B e d_{j t} + δ_{3} l n E I C U_{j t} + δ_{4} l n M a s k_{j t} + δ_{4} l n F V_{j t} + \dots + ε_{i t}

(5)

The computer software used in this study was Frontier Version 4.1, which was provided free of charge by Coelli⁵² to estimate Equations (3) and (4) and Equations (3) and (5). This study constructed the Cobb–Douglas stochastic boundary model to explore the relationship between inefficiencies and impact factors of health policies in combating COVID-19 in 136 countries and regions worldwide. The estimated inefficiency value will be between 0 and 1. The closer the value is to 1, the higher the number of deaths caused by the COVID-19 epidemic. It indicates that the degree of overall control of the COVID-19 epidemic in the country is poor, that is, it is inefficient. The statistical data of the variables in this study are shown in Table 2.

Table 2.

Descriptive statistics of variables.

Items	Variables	Samples	Mean	Std.	Min.	Max.	Median
Output item	Y = cumulative deaths (CD)	408	41,672	147,578	2	2,365,183	7334
Input item	X₁= cumulative infections (CI)	408	29,051,023	90,503,623	3525	1,275,765,000	5,889,667
	X₂= cumulative COVID-19 hospital beds needed (bed)	408	1,433,883	5,181,169	652	83,221,732	238,856
	X₃= cumulative COVID-19 ICU beds needed (ICU)	408	338,127	1,156,919	99	17,397,229	45,152
Exogenous variable	Z₁= population density (PD)	408	282	925	2	7916	86
	Z₂= total number of beds that exist at a location EBed)	408	120,356	384,878	147	3,775,717	22,346
	Z₃= total number of ICU beds that exist at a location (EICU)	408	3743	11,521	2	92,532	439
	Z₄= percent of population reporting always wearing a mask when leaving home (%) (mask)	408	46.48	21.80	1.38	95.12	46.89
	Z₅= cumulative all fully vaccinated (FV)	272	28,730,402	109,169,526	21,532	1,026,599,143	4,445,417

Results

Table 3 and Figure 2 show that the average inefficiencies of various countries and regions in the global fight against the COVID-19 pandemic in 2020, 2021, 2022, and 2020–2022 are 0.602, 0.666, 0.530, and 0.599, respectively. These values indicate considerable room for improvement by up to approximately 60%. The top 10 countries in the three-year average efficiency ranking (with their inefficiency value) from 2020 to 2022 are Venezuela (0.107), China (0.108), Qatar (0.127), Singapore (0.128), Kuwait (0.223), Mongolia (0.250), Sri Lanka (0.264), Taiwan (0.287), Thailand (0.304), and Saudi Arabia (0.308). The bottom five countries are Finland (0.867), Mozambique (0.867), Denmark (0.845), Bolivia (0.844), and Zambia (0.810) in order. The average inefficiency value was 0.599, which indicated a poor global response to the COVID-19 pandemic, with 59.9% room for improvement. The top 10 in terms of efficiency in 2022 were China (0.006), Myanmar (0.054), Venezuela (0.054), Kuwait (0.093), Qatar (0.101), Saudi Arabia (0.133), Singapore (0.133), Sri Lanka (0.151), United Arab Emirates (0.171), and Chad (0.199). The last five were Malawi (0.893), Madagascar (0.883), Finland (0.860), Bahamas (0.852), and Mozambique (0.846) in order. The average inefficiency value was 0.530, indicating poor global effectiveness in combating the COVID-19 pandemic, with 53% room for improvement. The inefficiency values and efficiency rankings of all the countries and regions in 2020 and 2021 are shown in Table 3.

Figure 2.

Distribution of three-year mean inefficiencies in the global response to the COVID-19 pandemic in 2020–2022.

Table 3.

Estimates of the COVID-19 pandemic inefficiencies of SFA for 136 countries and regions.

No.	Location	Year						2020–2022 Mean of inefficiency value	2020–2022 Rank of average efficiency
No.	Location	2020 Inefficiency value	2020 Rank of efficiency	2021 Inefficiency value	2021 Rank of efficiency	2022 Inefficiency value	2022 Rank of efficiency	2020–2022 Mean of inefficiency value	2020–2022 Rank of average efficiency
1	Afghanistan	0.649	61	0.847	132	0.733	106	0.743	109
2	Albania	0.802	116	0.849	133	0.723	103	0.791	124
3	Algeria	0.563	46	0.393	13	0.376	37	0.444	22
4	Andorra	0.687	74	0.622	41	0.502	60	0.604	63
5	Angola	0.861	127	0.770	95	0.790	118	0.807	131
6	Argentina	0.664	66	0.629	43	0.405	45	0.566	50
7	Armenia	0.547	44	0.754	90	0.393	43	0.565	49
8	Australia	0.787	112	0.602	33	0.488	57	0.626	70
9	Austria	0.778	107	0.744	89	0.310	26	0.611	67
10	Azerbaijan	0.672	72	0.802	106	0.581	73	0.685	87
11	Bahamas	0.790	114	0.695	62	0.852	133	0.779	121
12	Bahrain	0.640	58	0.714	72	0.483	56	0.612	68
13	Bangladesh	0.226	12	0.570	27	0.253	19	0.350	13
14	Barbados	0.651	63	0.770	96	0.841	131	0.754	114
15	Belgium	0.777	105	0.516	18	0.243	17	0.512	36
16	Belize	0.751	92	0.723	75	0.681	94	0.718	98
17	Bolivia	0.900	134	0.821	121	0.811	123	0.844	133
18	Bosnia and Herzegovina	0.677	73	0.726	77	0.586	75	0.663	82
19	Botswana	0.465	34	0.821	120	0.823	125	0.703	92
20	Brazil	0.668	69	0.712	68	0.393	42	0.591	56
21	Bulgaria	0.764	101	0.776	99	0.568	70	0.703	91
22	Canada	0.844	125	0.561	23	0.356	33	0.587	54
23	Central African Republic	0.898	133	0.684	56	0.824	126	0.802	129
24	Chad	0.734	83	0.611	38	0.199	10	0.515	39
25	Chile	0.700	77	0.625	42	0.450	51	0.592	57
26	China	0.229	13	0.090	2	0.006	1	0.108	2
27	Colombia	0.709	78	0.743	86	0.458	53	0.637	72
28	Costa Rica	0.496	35	0.550	21	0.365	35	0.470	28
29	Côte d'Ivoire	0.351	24	0.687	58	0.716	101	0.585	51
30	Croatia	0.746	88	0.738	85	0.606	81	0.697	90
31	Cyprus	0.441	32	0.542	20	0.554	66	0.512	37
32	Czechia	0.628	56	0.714	71	0.292	23	0.544	44
33	Democratic Republic of the Congo	0.700	76	0.514	17	0.725	104	0.646	76
34	Denmark	0.902	135	0.838	129	0.793	119	0.845	134
35	Ecuador	0.779	108	0.711	66	0.340	30	0.610	65
36	Egypt	0.355	25	0.637	45	0.652	89	0.548	45
37	Equatorial Guinea	0.763	100	0.673	52	0.368	36	0.601	62
38	Estonia	0.845	126	0.834	127	0.348	32	0.676	83
39	Ethiopia	0.722	80	0.691	61	0.710	98	0.708	93
40	Fiji	0.048	4	0.814	118	0.664	91	0.509	34
41	Finland	0.902	136	0.839	130	0.860	134	0.867	136
42	France	0.500	36	0.479	15	0.210	11	0.396	16
43	Gambia	0.806	119	0.798	105	0.800	121	0.801	128
44	Georgia	0.342	22	0.711	67	0.491	58	0.515	38
45	Germany	0.733	82	0.700	63	0.360	34	0.598	60
46	Ghana	0.606	52	0.713	70	0.627	86	0.648	78
47	Greece	0.787	111	0.684	57	0.706	97	0.726	102
48	Guam	0.606	51	0.805	107	0.766	114	0.725	101
49	Guinea	0.519	37	0.798	104	0.825	128	0.714	95
50	Guyana	0.784	110	0.805	108	0.807	122	0.799	127
51	Haiti	0.520	39	0.688	59	0.723	102	0.643	74
52	Honduras	0.828	123	0.879	136	0.684	95	0.797	126
53	Hong Kong	0.366	27	0.229	6	0.753	109	0.449	23
54	Hungary	0.563	45	0.734	81	0.330	28	0.542	43
55	Iceland	0.387	29	0.056	1	0.592	76	0.345	12
56	India	0.535	41	0.596	31	0.241	16	0.457	25
57	Indonesia	0.335	21	0.744	87	0.834	130	0.638	73
58	Iran	0.690	75	0.621	39	0.556	68	0.622	69
59	Iraq	0.744	86	0.676	53	0.534	64	0.651	79
60	Ireland	0.751	93	0.658	49	0.378	38	0.596	59
61	Israel	0.409	30	0.584	29	0.326	27	0.440	20
62	Italy	0.776	103	0.689	60	0.611	83	0.692	89
63	Jamaica	0.750	90	0.774	97	0.788	117	0.771	118
64	Japan	0.729	81	0.632	44	0.671	92	0.677	84
65	Kenya	0.753	94	0.796	103	0.711	99	0.754	113
66	Kuwait	0.248	15	0.326	10	0.093	4	0.223	5
67	Latvia	0.783	109	0.832	126	0.419	47	0.678	85
68	Lebanon	0.820	122	0.838	128	0.761	112	0.806	130
69	Libya	0.734	84	0.808	111	0.777	115	0.773	119
70	Lithuania	0.762	99	0.810	112	0.621	84	0.731	103
71	Luxembourg	0.645	59	0.670	51	0.517	62	0.611	66
72	Madagascar	0.667	67	0.807	109	0.883	135	0.786	123
73	Malawi	0.671	71	0.791	101	0.893	136	0.785	122
74	Malaysia	0.118	9	0.811	115	0.463	54	0.464	26
75	Maldives	0.163	10	0.603	34	0.215	13	0.327	11
76	Mali	0.774	102	0.755	91	0.757	110	0.762	116
77	Malta	0.869	129	0.684	55	0.648	88	0.733	106
78	Mexico	0.803	118	0.819	119	0.541	65	0.721	99
79	Mongolia	0.008	1	0.527	19	0.216	14	0.250	6
80	Montenegro	0.792	115	0.781	100	0.638	87	0.737	108
81	Morocco	0.751	91	0.738	84	0.712	100	0.733	105
82	Mozambique	0.878	130	0.810	113	0.846	132	0.845	135
83	Myanmar	0.446	33	0.682	54	0.054	2	0.394	15
84	Nepal	0.589	49	0.713	69	0.286	22	0.529	41
85	Netherlands	0.885	131	0.735	82	0.274	21	0.631	71
86	New Zealand	0.523	40	0.315	9	0.476	55	0.438	19
87	Niger	0.588	47	0.599	32	0.570	71	0.586	53
88	Nigeria	0.610	53	0.568	26	0.235	15	0.471	29
89	North Macedonia	0.757	96	0.828	125	0.738	108	0.774	120
90	Norway	0.623	54	0.561	24	0.596	77	0.593	58
91	Pakistan	0.366	26	0.661	50	0.379	39	0.469	27
92	Panama	0.543	43	0.594	30	0.449	50	0.529	40
93	Papua New Guinea	0.183	11	0.731	78	0.560	69	0.491	32
94	Paraguay	0.638	57	0.758	92	0.573	72	0.656	80
95	Peru	0.864	128	0.812	117	0.496	59	0.724	100
96	Philippines	0.369	28	0.638	46	0.517	61	0.508	33
97	Poland	0.667	68	0.732	80	0.586	74	0.662	81
98	Portugal	0.833	124	0.842	131	0.705	96	0.793	125
99	Puerto Rico	0.741	85	0.744	88	0.660	90	0.715	96
100	Qatar	0.082	6	0.199	5	0.101	5	0.127	3
101	Republic of Korea	0.298	17	0.385	12	0.420	48	0.367	14
102	Republic of Moldova	0.604	50	0.714	73	0.308	24	0.542	42
103	Romania	0.778	106	0.736	83	0.309	25	0.608	64
104	Russian Federation	0.349	23	0.719	74	0.382	40	0.483	30
105	Rwanda	0.747	89	0.701	64	0.761	113	0.736	107
106	San Marino	0.333	20	0.758	93	0.556	67	0.549	46
107	Saudi Arabia	0.670	70	0.122	3	0.133	6	0.308	10
108	Senegal	0.649	62	0.822	122	0.603	78	0.691	88
109	Serbia	0.625	55	0.710	65	0.603	79	0.646	75
110	Sierra Leone	0.789	113	0.368	11	0.214	12	0.457	24
111	Singapore	0.009	2	0.241	8	0.133	7	0.128	4
112	Slovakia	0.519	38	0.765	94	0.390	41	0.558	47
113	Slovenia	0.802	117	0.606	36	0.396	44	0.601	61
114	South Africa	0.815	120	0.811	114	0.671	93	0.766	117
115	Spain	0.756	95	0.551	22	0.454	52	0.587	55
116	Sri Lanka	0.058	5	0.584	28	0.151	8	0.264	7
117	Sudan	0.649	60	0.724	76	0.760	111	0.711	94
118	Suriname	0.715	79	0.812	116	0.730	105	0.752	112
119	Sweden	0.819	121	0.622	40	0.607	82	0.683	86
120	Switzerland	0.898	132	0.808	110	0.527	63	0.744	110
121	Taiwan	0.026	3	0.231	7	0.604	80	0.287	8
122	Thailand	0.089	7	0.479	16	0.343	31	0.304	9
123	Togo	0.758	97	0.607	37	0.781	116	0.715	97
124	Trinidad and Tobago	0.541	42	0.823	123	0.831	129	0.731	104
125	Tunisia	0.658	64	0.823	124	0.798	120	0.760	115
126	Turkey	0.238	14	0.606	35	0.625	85	0.489	31
127	Uganda	0.746	87	0.776	98	0.736	107	0.752	111
128	Ukraine	0.312	18	0.731	79	0.250	18	0.431	18
129	United Arab Emirates	0.589	48	0.568	25	0.171	9	0.442	21
130	United Kingdom	0.777	104	0.649	48	0.268	20	0.565	48
131	United States of America	0.663	65	0.644	47	0.449	49	0.585	52
132	Uruguay	0.422	31	0.459	14	0.333	29	0.405	17
133	Venezuela	0.090	8	0.176	4	0.054	3	0.107	1
134	Vietnam	0.262	16	0.863	135	0.408	46	0.511	35
135	Zambia	0.759	98	0.849	134	0.823	124	0.810	132
136	Zimbabwe	0.327	19	0.793	102	0.824	127	0.648	77
	Mean	0.602		0.666		0.530		0.599

From Table 3, the top countries with a better efficiency ranking over the three years (2020–2022) were mostly able to maintain that ranking. High-ranked countries such as China, Qatar, and Singapore implemented strict and comprehensive quarantine measures in early 2020 and implemented a national COVID-19 vaccination policy in 2021. China implemented a zero-clearance policy from the early stage of the COVID-19 outbreak in 2020 until the public demonstrations in December 2022 due to the people's frustration in the three-year fight against COVID-19. After China ended its policy under pressure from economic recession and the public, it caused a shortfall in medical services, a sharp rise in deaths, and shortages of medicines. The number of confirmed cases surged for a short term as a result of the sudden lifting of lockdown measures before gradually decreasing. Therefore, strict infectious disease prevention and control measures against COVID-19 have been lifted since January 2023.^53,54 The number of reported deaths from COVID-19 in Qatar has been low, likely because of the robust healthcare system and the average age of its young and healthy residents. In the beginning, the Qatari government also implemented lockdown measures; closed mosques, shops, restaurants, and other places; and strictly enforced the mandatory wearing of masks in public places. People who test positive for COVID-19 have immediate, free access to quality health care.⁵⁵

In January 2020, Singapore set up an interministerial anti-epidemic task force in response to the epidemic. In the beginning, strict clearance policies, wearing masks, and entry and exit tracking and control measures were implemented. Singapore understands the importance of early access to internationally certified vaccines and vaccination to reduce deaths and contain the spread of COVID-19. However, after the outbreak of the Omicron variant in September 2021, the number of confirmed cases and deaths increased significantly. Nevertheless, with the high vaccination rates for COVID-19 and estimates that nearly 90% or more of the population has contracted COVID-19, Singapore has achieved a strong “mixed immunity.” In March 2022, Singaporean authorities announced that social control measures would be further relaxed, and Singapore would enter the final phase of the plan to live with COVID-19.⁵⁶ Singapore's policy is to transition from eliminating COVID-19 to living with COVID-19. This policy requires high vaccination rates and well-communicated transition plans, with a gradual reduction of mitigation measures over time.⁵⁷ Slowing the spread of the disease, maintaining functioning healthcare systems, and vaccination to reduce severe illnesses and deaths are the main policies and goals of the COVID-19 response in these countries, which show good results.

Maximum-likelihood estimation (MLE) is used for the SFA analysis for the two phases of this study. The estimated results of the first phase in 2020 are shown in Table 4. The second phase is 2021–2022. The difference between the two phases was addition of the exogenous variable of the cumulative all FV. The estimation results are shown in Table 5. Table 4 shows that in the first year of the COVID-19 pandemic in 2020, cumulative infections (estimate: 0.335155) and the cumulative number of hospital beds required (estimate: 0.712924) are both very significantly positively correlated with the efficiency of cumulative deaths. In other words, the higher the number of cases and hospitalizations, the higher the number of deaths from COVID-19. In terms of exogenous variables in 2020, there is a significant positive correlation between cumulative death inefficiency and population density (estimate: 1.211623), indicating that a higher population density is likely to lead to more deaths from COVID-19. Reducing the risk of close contacts, such as at large gatherings, can reduce the infection rate. Second, cumulative death inefficiencies showed a very significant positive correlation with the number of ICU beds available (estimate: 1.979876), suggesting that sufficient and improved access to care and equipment could reduce the mortality of COVID-19. Third, cumulative death inefficiency showed a significant positive correlation with the percentage of people who wore masks outside the home (estimate: 3.475606), suggesting that wearing masks outside the home in the early stages of COVID-19 could indeed reduce confirmed deaths. Fourth, there was a significant negative correlation between cumulative death inefficiency and the number of hospital beds available (estimate: −2.09578), indicating that countries that are unable to provide a sound public health system and adequate medical services to combat COVID-19 suffered an increased number of deaths.

Table 4.

Estimates of the stochastic frontier function of 2020.

Variable description	Coefficient	Estimate	Standard error	t-ratio
Constant	β₀	−4.129872	0.41020557	−10.067811
lnCI_it	β₁	0.335155***	0.058716	5.708111***
lnBeb_it	β₂	0.712924***	0.095652	7.453288***
lnICU_jt	β₃	−0.01528	0.080094	−0.19083
Constant	δ₀	−18.6907	10.82577	−1.7265
lnPD_jt	δ₁	1.211623**	0.493429	2.455518**
lnEBed_jt	δ₂	−2.09578***	0.741401	−2.82678***
lnEICU_jt	δ₃	1.979876***	0.746813	2.6511***
lnMask_jt	δ₄	3.475606*	1.980559	1.754861*
$σ_{s}^{2} = σ_{u}^{2} + σ_{v}^{2}$	$σ_{s}^{2}$	6.927344	2.31684	2.989996
$r = σ_{u}^{2} / σ_{s}^{2}$	r	0.9852	0.006232	158.0866
Log-likelihood function = −138.0855

Note: 1. ***, **, and * represent significance at the 1%, 5%, and 10% levels, respectively.

2. $σ_{u}^{2}$ and $σ_{v}^{2}$ represent the inefficiency error variance and random error variance, respectively; $σ_{s}^{2}$ is the total variance; r is the proportion of inefficiency error variance ( $σ_{u}^{2}$ ) to total variance ( $σ_{s}^{2}$ ).

Table 5.

Estimates of the stochastic frontier function of 2021–2022.

Variable description	Coefficient	Estimate	Standard error	t-ratio
Constant	β₀	−3.50778	0.466087	−7.52602
lnCI_it	β₁	0.228937***	0.032829	6.97367***
lnBeb_it	β₂	0.596755***	0.089343	6.679375***
lnICU_jt	β₃	0.18662**	0.083828	2.226227**
Constant	δ₀	−24.335	45.50468	−0.53478
lnPD_jt	δ₁	0.632337	1.066054	0.593157
lnEBed_jt	δ₂	−1.76614	3.161276	−0.55868
lnEICU_jt	δ₃	1.079071	1.863247	0.579135
lnMask_jt	δ₄	−1.24631	2.090154	−0.59628
lnFV_jt	δ₅	1.805235	3.221939	0.560295
$σ_{s}^{2} = σ_{u}^{2} + σ_{v}^{2}$	$σ_{s}^{2}$	6.953734	11.67598	0.595559
$r = σ_{u}^{2} / σ_{s}^{2}$	r	0.971656	0.045211	21.49173
Log-likelihood function = −297.19237

Note: 1. *** and ** represent significance at the 1%, and 5% levels, respectively.

Table 5 shows that both cumulative infections (estimate: 0.228937) and the cumulative number of hospital beds required (estimate: 0.596755) during the COVID-19 pandemic in the second phase (2021–2022) are very significantly positively correlated with the efficiency of cumulative deaths. In other words, the higher the number of cases and hospitalizations, the higher the number of deaths from COVID-19. Secondly, the cumulative number of ICU beds required (estimate: 0.18662) showed a significant positive correlation with the efficiency of cumulative deaths. The higher the number of confirmed ICU patients required, the higher the number of COVID-19 deaths. In addition, exogenous variables such as population density, number of inpatient beds available, number of ICU beds available, the wearing of masks outside the home, and full vaccination were not significantly associated with cumulative deaths in the second phase (2021–2022), possibly due to the increased duration of the epidemic resulting in higher rates of confirmed COVID-19 cases and vaccinations. Most people have basic immunity to COVID-19. Because population density promotes the spread of the virus through close contact between people, lockdown policies, school closures, and wearing masks are used to reduce the spread of close contact, so the findings are consistent with related studies. However, relevant studies^9,23 also pointed out that the blockade policy would cause social and economic losses, impact interpersonal relations, and affect physical and mental health. The results of this study are similar to those speculated by Kim et al.⁵⁶ and Owens and Parry⁵⁷ and are similar to the recent situation in Singapore, China, and many other countries, where most people have obtained a good “mixed immunity.” The current stage is important for the world to gradually relax COVID-19 control measures over time and to transition to living with COVID-19.

Discussion

This study measured the effectiveness of the global response to the COVID-19 pandemic in 136 countries and territories from 2020 to 2022 and explored the global COVID-19 outbreak and response strategies. In this study, cumulative deaths due to COVID-19 were adopted as undesirable output and excess mortality was applied to measure the data of original cumulative deaths because this data could be closer to and better reflect the actual number of COVID-19 deaths. This study uses the Cobb–Douglas stochastic frontier model and can be divided into two phases to analyze and discuss the COVID-19 pandemic. The first phase was 2020, when there was no vaccine for COVID-19. The second phase was 2021–2022, from when there was a vaccine up until the immediate future. The results of this study show that the global average inefficiencies in combating the COVID-19 pandemic in countries and regions in 2020, 2021, 2022, and 2020–2022 are 0.602, 0.666, 0.530, and 0.599, respectively, with a significant room for improvement of up to 60%. This result shows that the global dilemma in facing the COVID-19 pandemic is the variety of uncertainties about the occurrence, prevention and treatment, and future trend of this disease. Currently, we can only coexist with COVID-19, and the disease cannot be eradicated until a fully effective vaccine or therapeutic drug can be found. Over time, there has been a continuous rise in the number of COVID-19 infection cases alongside increasing vaccination coverage on a global scale. As a result, countries worldwide are gradually easing social restrictions and control measures. However, as anticipated, SARS-CoV-2 variants are still prevalent, posing a significant risk to older adults and individuals with underlying health conditions who, upon infection, may experience severe symptoms or even fatalities. Achieving global success in overcoming the COVID-19 pandemic necessitates the coordination and cooperation of countries worldwide, while striving for low incidence rates and implementing moderate nonpharmaceutical interventions alongside progressive social and economic policies.^58–60

Limitations

This study has certain limitations. First, the use of SFA requires assumptions about the functional form prior to estimation and restricts analysis to a single output item. Second, the data obtained from the IHME were provided by each country and region, which may introduce variations due to the different estimation methods employed. Although the IHME utilized a mixed modeling approach to generate estimated data, there may still be deviations from the actual data. Third, the exogenous variables influencing the COVID-19 mortality rate extend beyond the five variables considered in this study. This study did not focus on the biology or the spread of different variants and its impact on mortality. This study did not analyze or discuss various variant viruses, so future research could explore the inclusion or exclusion of variables to gain further insights. Lastly, the current research methodology may possess some statistical biases, and future studies can explore more effective research methods to replicate the findings. It is anticipated that the outcomes of this study will not only contribute to addressing and improving management of the COVID-19 pandemic but also have potential applications in future large-scale outbreaks.

Conclusions

The empirical findings of this study are consistent in both phases: the more COVID-19 infections, the more hospitalizations required, and the more ICU beds required, the higher the number of deaths. The initial results of the COVID-19 epidemic in Phase I (2020) are as follows: (1) mortality can be reduced with adequate medical services and intensive care in an ICU and (2) bans on large gatherings, lockdown measures, and wearing masks outside the home effectively reduced the number of infections initially. In the second phase (2021–2022), the results are as follows: (1) exogenous variables such as population density, mask-wearing outside, and complete vaccination had no significant relationship with cumulative deaths, indicating that most people already had preliminary antibodies for COVID-19, and the policy of opening borders, lifting lockdown measures, and mask exemption should be conducted gradually. (2) As infection and vaccination rates increase, mortality rates decrease. By measuring the effectiveness of the global response to the COVID-19 pandemic and the results of its impact factors, this study is expected to provide a reference for the community to explore how to reduce and propose countermeasures in order to improve public health policies to mitigate the COVID-19 pandemic. In the future, when faced with a sudden pandemic or health crisis, it can provide global and national policymakers and health units with reference indicators and effective guidance. As humanity faces many unprecedented epidemics and challenges in the future, it is hoped that the world will learn from its mistakes and innovate strategies to improve the response and public health policies while it works together to combat the COVID-19 pandemic.

Footnotes

Acknowledgements

The free software “Frontier Version 4.1” can be used to estimate the equations, which was kindly provided by Professor Coelli.⁵²

Contributorship

Single author JSW researched and edited the manuscript and approved the final version of the manuscript.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Informed consent

This article does not contain any studies with human participants.

Guarantor

JSW.

ORCID iD

Jih-Shong Wu

References

World Health Organization. Timeline: WHO's COVID-19 response, https://www.who.int/emergencies/diseases/novel-coronavirus-2019/interactive-timeline#! (2023, accessed 16 March 2023).

World Health Organization. Statement on the second meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV), https://www.who.int/news-room/detail/30-01-2020-statement-on-the-second-meeting-of-the-international-health-regulations-(2005)-emergency-committee-regarding-the-outbreak-of-novel-coronavirus-(2019-ncov) (2020, accessed 5 March 2023).

Aydin

Yurdakul

. Assessing countries’ performances against COVID-19 via WSIDEA and machine learning algorithms. Appl Soft Comput 2020; 97: 106792.

Wang

Cauchemez

, et al. Risk for transportation of coronavirus disease from Wuhan to other cities in China. Emerg Infect Dis 2020; 26: 1049–1052.

World Health Organization. WHO Coronavirus (COVID-19) Dashboard, https://covid19.who.int/ (2023, accessed 16 March 2023).

Worldometer. COVID-19 Coronavirus Pandemic, https://www.worldometers.info/coronavirus/ (2023, accessed 16 March 2023).

Zhang

Wang

Rauch

, et al. Unprecedented disruption of lives and work: health, distress and life satisfaction of working adults in China one month into the COVID-19 outbreak. Psychiatry Res 2020; 288: 112958.

Gibson

. Government mandated lockdowns do not reduce COVID-19 deaths: implications for evaluating the stringent New Zealand response. N Z Econ Papers 2022; 56: 17–28.

Karako

Song

Chen

, et al. Analysis of COVID-19 infection spread in Japan based on stochastic transition model. Biosci Trends 2020; 14: 134–138.

10.

Nawi

MAA

Zainon

WMN

Noor

NFM

, et al. Measurement of relative efficiency levels of southeast Asian countries in dealing with the Covid-19 pandemic over one year. J Biostat Epidemiol 2022; 7: 403–412.

11.

Bilinski

Emanuel

. COVID-19 and excess all-cause mortality in the US and 18 comparison countries. JAMA 2020; 324: 2100–2102.

12.

Gearhart

Sonchak-Ardan

Michieka

. The efficiency of COVID cases to COVID policies: a robust conditional approach. Empir Econ 2022; 63: 2903–2948.

13.

Williams

Crookes

Glass

, et al. COVID-19 mortalities in England and Wales and the Peltzman offsetting effect. Appl Econ 2021; 53: 6982–6998.

14.

Friedson

McNichols

Sabia

, et al. Did California’s shelter-in-place order work? Early coronavirus-related public health effects (No. w26992). National Bureau of Economic Research 2020, https://www.nber.org/papers/w26992.

15.

Flynn

Moloney

Bhattarai

, et al. COVID-19 pandemic in the United Kingdom. Health Policy Technol 2020; 9: 673–691.

16.

Prather

Wang

Schooley

. Reducing transmission of SARS-CoV-2. Science 2020; 368: 1422–1424.

17.

Remuzzi

. COVID-19 and Italy: what next? Lancet 2020; 395: 1225–1228.

18.

Sen-Crowe

Sutherland

McKenney

, et al. A closer look into global hospital beds capacity and resource shortages during the COVID-19 pandemic. J Surg Res 2021; 260: 56–63.

19.

Berardi

Antonini

Genie

, et al. The COVID-19 pandemic in Italy: policy and technology impact on health and non-health outcomes. Health Policy Technol 2020; 9: 454–487.

20.

Desson

Lambertz

Peters

, et al. Europe's Covid-19 outliers: German, Austrian and Swiss policy responses during the early stages of the 2020 pandemic. Health Policy Technol 2020; 9: 405–418.

21.

Ferreira

Marques

. Do quality and access to hospital services impact on their technical efficiency? Omega 2019; 86: 218–236.

22.

Pereira

Dinis

Ferreira

, et al. A network data envelopment analysis to estimate nations’ efficiency in the fight against SARS-CoV-2. Expert Syst Appl 2022; 210: 118362.

23.

Caroppo

Mazza

Sannella

, et al. Will nothing be the same again?: Changes in lifestyle during COVID-19 pandemic and consequences on mental health. Int J Environ Res Public Health 2021; 18: 8433.

24.

Hâncean

Lerner

Perc

, et al. Occupations and their impact on the spreading of COVID-19 in urban communities. Sci Rep 2022; 12: 14115.

25.

Meijer

Kejs

Stock

, et al. Population density, socioeconomic environment and all-cause mortality: a multilevel survival analysis of 2.7 million individuals in Denmark. Health Place 2012; 18: 391–399.

26.

Wong

. Spreading of COVID-19: density matters. Plos One 2020; 15: e0242398.

27.

KTL

White

Nichols

. Population density and basic reproductive number of COVID-19 across United States counties. PloS One 2021; 16: e0249271.

28.

Rubin

Huang

Fisher

, et al. Association of social distancing, population density, and temperature with the instantaneous reproduction number of SARS-CoV-2 in counties across the United States. JAMA Network Open 2020; 3: e2016099–e2016099.

29.

World Health Organization. Coronavirus disease (COVID-19): How is it transmitted?, https://www.who.int/news-room/questions-and-answers/item/coronavirus-disease-covid-19-how-is-it-transmitted (2023, accessed 16 March 2023).

30.

Ueki

Furusawa

Iwatsuki-Horimoto

, et al. Effectiveness of face masks in preventing airborne transmission of SARS-CoV-2. MSphere 2020; 5: e00637–20.

31.

Rémy

Largeron

Quilici

, et al. The economic value of vaccination: why prevention is wealth. J Mark Access Health Policy 2015; 3: 29284.

32.

Cihan

. Forecasting fully vaccinated people against COVID-19 and examining future vaccination rate for herd immunity in the US, Asia, Europe, Africa, South America, and the world. Appl Soft Comput 2021; 111: 107708.

33.

Maltezou

Gamaletsou

Giannouchos

, et al. Timing of last COVID-19 vaccine dose and SARS-CoV-2 breakthrough infections in fully (boosted) vaccinated healthcare personnel. J Hosp Infect 2023; 132: 46–51.

34.

Whittaker

Kristofferson

Salamanca

, et al. Length of hospital stay and risk of intensive care admission and in-hospital death among COVID-19 patients in Norway: a register-based cohort study comparing patients fully vaccinated with an mRNA vaccine to unvaccinated patients. Clin Microbiol Infect 2022; 28: 871–878.

35.

Brosh-Nissimov

Orenbuch-Harroch

Chowers

, et al. BNT162b2 Vaccine breakthrough: clinical characteristics of 152 fully vaccinated hospitalized COVID-19 patients in Israel. Clin Microbiol Infect 2021; 27: 1652–1657.

36.

Hippisley-Cox

Coupland

Mehta

, et al. Risk prediction of covid-19 related death and hospital admission in adults after covid-19 vaccination: national prospective cohort study. Br Med J 2021; 374. 10.1136/bmj.n2244

37.

Sindo

Ohkusa

, et al. A newly developed Japanese pneumonia and influenza mortality model and statistical analysis of excess mortality by stochastic frontier estimation. Discuss Paper-Osaka Univ Inst Soc Econ Res 2000; 501, https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=1b1b47ef169f0bb99b3cff5813ce288bb8d24309 .

38.

Breitenbach

Ngobeni

Aye

. Global healthcare resource efficiency in the management of COVID-19 death and infection prevalence rates. Front Public Health 2021; 9: 638481.

39.

Lupu

Tiganasu

. COVID-19 and the efficiency of health systems in Europe. Health Econ Rev 2022; 12: 1–15.

40.

Ahmad

WMAW

Zainon

WMNW

Nawi

MAA

, et al. Malaysia's efficiency in dealing with COVID-19 outbreaks compared to other Asian countries by using stochastic frontier analysis (SFA). Ann King Edward Med Univ Lahore Pak 2020; 26: 324–329, https://pesquisa.bvsalud.org/global-literature-on-novel-coronavirus-2019-ncov/resource/pt/covidwho-923086 .

41.

Maity

Ghosh

Barlaskar

. Interstate disparities in the performances in combatting COVID-19 in India: efficiency estimates across states. BMC Public Health 2020; 20: 1925.

42.

Kınacı

Ünsal

Kasap

. A close look at 2019 novel coronavirus (COVID 19) infections in Turkey using time series analysis & efficiency analysis. Chaos Solit Fract 2021; 143: 110583.

43.

. Applying frontier approach to measure the financial efficiency of hospitals. Digital Health 2023; 9: 1–13.

44.

. Measuring economic development and carbon dioxide emissions inefficiency. SAGE Open 2023; 13.

45.

Farrell

. The measurement of productive efficiency. J R Stat Soc 1957; 120: 253–290. 10.2307/2343100

46.

Coelli

Rao

DSP

Donnell

, et al. An introduction to efficiency and productivity analysis. 2nd ed. New York: Spinger, 2005.

47.

Aigner

Lovell

CAK

Schmidt

. Formulation and estimation of stochastic frontier production function models. J Econom 1977; 6: 21–37.

48.

Meeusen

van den Broeck

. Efficiency estimation from Cobb-Douglas production functions with composed error. Int Econ Rev 1977; 18: 435–444.

49.

Battese

Coelli

. Prediction of firm-level technical efficiencies with a generalized frontier production function and panel data. J Econom 1988; 38: 387–399.

50.

Institute for Health Metrics and Evaluation (IHME). COVID-19 estimate downloads. https://www.healthdata.org/covid (2023, accessed 9 January 2023).

51.

World Health Organization. Home/News/14.9 million excess deaths associated with the COVID-19 pandemic in 2020 and 2021. https://www.who.int/news/item/05-05-2022-14.9-million-excess-deaths-were-associated-with-the-covid-19-pandemic-in-2020-and-2021 (2023, accessed 16 March 2023).

52.

Coelli

. A guide to FRONTIER Version 4.1: A computer program for stochastic frontier production and cost function estimation (CEPA Working Paper 96/7). Department of Econometrics, University of New England, https://iranarze.ir/wp-content/uploads/2017/07/7209-English-IranArze.pdf (1996).

53.

Chen

. China can prepare to end its zero-COVID policy. Nat Med 2022; 28: 1104–1105.

54.

Reuters. China declares ‘decisive victory’ over COVID-19, https://www.reuters.com/world/china/china-declares-decisive-victory-over-covid-19-2023-02-17/ (2023, accessed 27 March 2023).

55.

Varma

Dergaa

Ashkanani

, et al. Analysis of Qatar's successful public health policy in dealing with the COVID-19 pandemic. Int J Med Rev Case Rep 2021; 5: 6–11.

56.

Kim

Goh

Kang

JHB

. Moving toward a common goal via cross-sector collaboration: lessons learned from SARS to COVID-19 in Singapore. Global Health 2022; 18: 1–18.

57.

Owens

Parry

. COVID-19: What can China learn from Hong Kong and Singapore about exiting zero covid? Br Med J 2022; 379. 10.1136/bmj.o3043

58.

Priesemann

Balling

Bauer

, et al. Towards a European strategy to address the COVID-19 pandemic. Lancet 2021; 398: 838–839.

59.

Calero Valdez

Iftekhar

Oliu-Barton

, et al. Europe must come together to confront omicron. Br Med J 2022; 376. doi: 10.1136/bmj.o90

60.

Czypionka

Iftekhar

Prainsack

, et al. The benefits, costs and feasibility of a low incidence COVID-19 strategy. Lancet Reg Health-Eur 2022; 13: 100294.

Measuring efficiency of the global fight against the COVID-19 pandemic

Abstract

Objectives

Methods

Results

Conclusions

Keywords

Introduction

Methodology

SFA method

Samples and data sources

Variables and empirical model

Results

Discussion

Limitations

Conclusions

Footnotes

Acknowledgements

Contributorship

Declaration of conflicting interests

Funding

Ethical approval

Informed consent

Guarantor

ORCID iD

References