key: cord-0312176-p576v9du
authors: Haider, N.; Hasan, M. N.; Khan, R. A.; McCoy, D.; Ntoumi, F.; Dar, O.; Ansumana, R.; Uddin, M. J.; Zumla, A.; Kock, R. A.
title: The Global case-fatality rate of COVID-19 has been declining disproportionately between top vaccinated countries and the rest of the world
date: 2022-01-21
journal: nan
DOI: 10.1101/2022.01.19.22269493
sha: ea247c8f5750fe69abedeb0b1dcda60db914125e
doc_id: 312176
cord_uid: p576v9du

Globally 58.83% human population received at least one dose of the COVID-19 vaccines as of 5 January 2021. COVID-19 vaccination rollout is progressing at varied rates globally and data on the impact of mass vaccination on infection and case-fatality rates require definition. We compared the global reported cumulative case-fatality rate (rCFR) between top-20 countries with COVID-19 vaccination rates (>125 doses/100 people) and the rest of the world, before and after commencement of vaccination programmes. We considered the 28th day of receiving the first vaccine in the world as a cut-off to compare the pre-vaccine period (Jan 1, 2020 - Jan 5, 2021) and the post-vaccine period (Jan 6, 2021 - Jan 5, 2022). We used a Generalized linear mixed model (GLMM) with a beta distribution to investigate the association between the CFR and potential predictors of each country and reported the relative risk (RR) of each variable. The mean rCFR of COVID-19 in the top-20 countries with vaccination rates was 1.83 (95% CI: 1.24-2.43) on 5 Jan 2021 and 1.18 (95% CI: 0.73-1.62) on 5 Jan 2022. The CFR for the rest of the world on 5 Jan 2021 was 2.32 (95% CI: 1.86-2.79) and 2.20 (95% CI: 1.86-2.55) on 5 January 2022. In Sub-Saharan Africa, the CFR remained roughly unchanged at 1.97 (95% CI: 1.59-2.35) on 5 Jan 2021 and 1.98 (95% CI:1.58-2.37) on 5 Jan 2022. The GLMM showed vaccination (/100 population) (RR:0.37) and Stringency Index (RR:0.88) were strong protective factors for the country's COVID-19 CFR indicating that both vaccination and lockdown measures help in the reduction of COVID-19 CFR. The rCFR of COVID-19 continues to decline, although at a disproportionate rate between top vaccinated countries and the rest of the world. Vaccine equity and faster roll-out across the world is critically important in reducing COVID-19 transmission and CFR.

Diseases (COVID-19) 28 February 2020 report indicated a crude fatality ratio of 3.8% among the first 91 55,924 laboratory-confirmed cases 1 . Subsequently, systematic reviews on the case-fatality rate (CFR) 92 of COVID-19 reported an estimated CFR between 2.3-3.6% 2-5 . The global cumulative reported case 93 fatality rate (rCFR) of COVID-19 has increased up until the 17 th epidemiological week (April 22-28, 94 2020) of detection of SARS-CoV-2 in Wuhan China at 7.2, and then started to decline steadily up 95 until 31 December 2021 at 2.2 6 . The decreasing rate of CFR has been explained by an increased rate 96 of infection in the younger population or by the improvement of health care management, shielding 97 from infection, and/or repurposing of several drugs that had shorted both hospitals stays and saved 98 lives 6 . However, during the last quarter of the year, 2020 different variants of concern/Interest 99 (VOC/VOI) of SARS-CoV-2 started to emerge with increased transmissibility 7 . Of them, Alpha (first 100 detected in the UK in September 2020), Beta (first documented in South Africa in May 2020), 101

Gamma (first detected in Brazil in November 2020), and Delta (first detected in India in October 102 2020) all reported having higher infectivity and severity than the virus originally detected in Wuhan, 103

China 7,8 . During November 2021, a new variant of concern, Omicron was reported to WHO (first 104 reported from South Africa on 24 November 2021) 9 . Unlike previous VOC, Omicron shows a 105 relatively lower case-fatality rate although the variant has a high transmissibility 10, 11 . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint

Vaccines are not distributed equitably in the world. Although COVID-19 vaccines were developed at 115 an unprecedented rate through the advancement of science and global cooperation, the distribution of 116 the vaccine across the world is questionable 16 . Current global vaccination rates of roughly 6.7 117 million doses per day translate to achieving herd immunity in approximately 4.6 years 16 . Vaccine 118 distribution is absent or very negligible in many of the low-income countries, and experts anticipated 119 that 80% of the population in low-income countries would not receive a vaccine at the end of 2021 16 120 which has been the case. 

The necessary COVID-19 related data, including daily new cases, daily new deaths, total deaths, and 134 total deaths per million inhabitants, vaccination, were collected from the WHO daily COVID-19 135 situation reports of 210 countries from January 01, 2020, to January 05, 2022 18 . On 8 th December 136 2020, the first human in the world received an approved COVID-19 vaccine 19 . We considered 137 the 28th day of receiving the first vaccine in the world as a cut-off to compare the pre-vaccine period 138 (Jan 1, 2020 -Jan 5, 2021) and the post-vaccine period (Jan 6, 2021-Jan 5, 2022). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Three forecasting models (i.e., auto-regressive integrated moving average (ARIMA), automatic time-150 series forecasting model which is also known as 'Prophet model', and simple exponential smoothing 151 (SES)), to identify the global trend of COVID-19 rCFR. We also used the Mann-Kendall (M-K) trend 152 analysis to identify the presence of any trend and the direction of the trend (increasing or decreasing). 153

We developed a generalized linear mixed model (GLMM) with beta distribution to identify whether 154 the explanatory variables have any relationship between the country's rCFR of COVID-19. These 155 methods supported us to make a credible conclusion on the trend of COVID-19 rCFR over the time. 156

All analyses were carried out using the statistical software R version 3.5.2.2 20 and SAS 21 . 157

We selected the SES, ARIMA, and Prophet Models because the key outcome variable (rCFR) is 159 dependent on previous records (time-series events) and all these three models can take this into 160 account. Using the time series models with the reported COVID-19 data, we forecasted trends for the 161 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint (GDP) 25 , worldwide governance indicators (WGI) 26 , and Global Health Security Index (GHSI) 27 , the 170 prevalence of obesity 28 or from "Our World in Data" 29 in our analyses. We also included the 171 country-specific prevalence of diabetes 29 and cardiovascular disease 29 to explain the variation of 172 COVID-19 rCFR. The GHSI index scored between 0 and 100 to indicate the country's capacity for is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint as the second level which group sequential observations within the same area, and independent 198 variables are treated as repeated observations at the lower level. While the location data are assumed 199 to be time-invariant, the independent data are assumed to be universal over the whole study area at a is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. COVID-19 rCFR as of 5 Jan 2022 are Yemen (19.56%), Peru (8.75%), Mexico (7.44%), Sudan 237 (7.07%), and Ecuador (6.09%). The correlating coefficient between vaccination rate (/100 people) and 238 rCFR in different countries of the world on 5 Jan 2022 is estimated as -0.296 (p<0.001) (Fig 2) . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. the total variance. Thus, the spatial unit effects account for approximately 63.4% of the total variance 257 of weekly rCFR, which suggests moderate reliability on location effects on weekly rCFR. It is also to 258 be noted that with the introduction of a random intercept, "vaccination", "population density", 259 "GDP", "weeks" and "stringency index" had significant negative effects on weekly rCFR and "the 260 percentage of people aged 65 and above" have significant positive effects. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. Prophet and SES model. According to the forecast in both models, the ratio of COVID-19 rCFR is 274 expected to decrease considerably in the coming 10 days (Fig 1) . In M-K trend analysis, we found a 275 negative trend of cumulative rCFR (p <0.001 and tau = -0.79). In Sen's slop test, the slope was -0.02 276 (95% CI: -0.03 to -0.01). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. The global rCFR of COVID-19 has been declining since May 2020 and the rate become plateaued or is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint to that of the top vaccinated countries. However, in many countries especially in SSA, the natural 327 infection had reached in a state to limit the infection and reduce the overall burden of the pandemic 38 . 328

In the Republic of Congo 66% of people in Brazzaville 39 in Malawi, 64.9% of blood donors 40 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted January 21, 2022. ; https://doi.org/10.1101/2022.01.19.22269493 doi: medRxiv preprint

Control and Prevention. The epidemiological characteristics of an outbreak of 2019 novel 412 coronavirus diseases (COVID-19) in China

Clinical characteristics of coronavirus disease 2019 (COVID-19) 415 in China: A systematic review and meta-analysis

The Global Case-Fatality Rate of COVID-19 Has Been 417 Declining Since

Covid-19: How many variants are there, and what do we know about them?: Video 419 1

Covid-19: Death rate is 0.66% and increases with age, study estimates

Emergence of new SARS-CoV-2 Variant of Concern 423

529) -highlights Africa's research capabilities, but exposes major knowledge 424 gaps, inequities of vaccine distribution, inadequacies in global COVID-19 response and 425 control efforts

SARS-CoV-2 variants of concern and variants under 427 investigation in England-Technical briefing: Update on hospitalisation and vaccine 428 effectiveness for Omicron VOC-21NOV-01 (B.1.1.529)

Clinical Severity of COVID-19 Patients Admitted to 430 Hospitals in Gauteng, South Africa During the Omicron-Dominant Fourth Wave

BNT162b2 mRNA Covid-19 Vaccine in a Nationwide 433 Mass Vaccination Setting

Oxford-AstraZeneca vaccines on covid-19 related symptoms, hospital admissions, and 436 mortality in older adults in England: test negative case-control study

Interim findings from first-dose mass COVID-19 438 vaccination roll-out and COVID-19 hospital admissions in Scotland: a national prospective 439 cohort study

BNT162b2 COVID-19 Vaccine in Preventing Severe Symptomatic Infection among 442 Healthcare Workers

From Vaccine Nationalism to Vaccine Equity 444 -Finding a Path Forward

Coronavirus disease (COVID-19) Weekly Epidemiological Update and Weekly 447 Operational Update

No TitleCovid-19 vaccine: First person receives Pfizer jab in UK

R: A Language and Environment for Statistical Computing. 2020. 451 21 SAS Institute Inc. Cary, NC U

A comparison of the power of the t test, Mann-Kendall and bootstrap tests for 453 trend detection / Une comparaison de la puissance des tests t de Student, de Mann-Kendall et 454 du bootstrap pour la détection de tendance

Population density (people per sq. km of land area)

The World Bank. Population ages 65 and above, total. World Bank Data

The World Bank. GDP (current US$) | Data

Johns Hopkins Center for Health Security (JHU). GHS Index. 461 2020. 462 28 WHO. Global health services data, obesity and overweight

Stringency Index

Oxford COVID-19 Government Response Tracker. Calculation and presentation of the 465 Stringency Index 4

Forecasting with multivariate temporal aggregation: The case of 467 promotional modelling

Comparison of predictor approaches for longitudinal 469 binary outcomes: application to anesthesiology data

R: A language and environment for statistical computing. R Foundation 471 for Statistical Computing

Differences in the case fatality risks associated with SARS-CoV-2

Delta and non-Delta variants in relation to vaccine coverage: An early ecological study in the 474 United Kingdom

Protection afforded by prior 476 infection against SARS-CoV-2 reinfection with the Omicron variant

Covid-19: Fourth vaccine doses-who needs them and why?

How COVID vaccines are being divvied up around the world

High SARS-CoV-2 Seroprevalence in 482

Eastern Democratic Republic of Congo

Seroprevalence of SARS-CoV-2 among residents of Brazzaville, the 485 Republic of Congo

SARS-CoV-2 exposure in Malawian blood 487 donors: an analysis of seroprevalence and variant dynamics between

Very high relative seroprevalence of

CoV-2 antibodies among communities in Bangui

The positive impact of lockdown in Wuhan on 493 containing the COVID-19 outbreak in China

Lockdown measures in response to COVID-19 in 496 nine sub-Saharan African countries

The Global Health Security index and Joint 498 External Evaluation score for health preparedness are not correlated with countries' COVID-499 19 detection response time and mortality outcome

Factors associated with the difference between the incidence and 501 case-fatality ratio of coronavirus disease 2019 by country

Vaccine Equity Is Crucial to End the COVID-19 503 Pandemic -But It Won't Be Enough

COVID-19-Zoonosis or Emerging 505 Infectious Disease?

Potential transmission chains of variant B.1.1.7 and co-507 mutations of SARS-CoV-2

Evidence for a mouse origin of the 509 SARS-CoV-2 Omicron variant