Technical ReportPDF Available

Sampling Bias: Explaining Wide Variations in COVID-19 Case Fatality Rates

Authors:
Dan Ward at WardEnvironment
Dan Ward
  • WardEnvironment
Download full-text PDF
Read full-text
Download citation

Abstract and Figures

(This technical report is part of a series of reports on sampling bias in COVID-19 testing. Please also see the subsequent report: "Sampling Bias, Case Fatality Rates and COVID-19 Testing: A Further Analysis", also by Dan Ward). This statistical report seeks to explain the wide variation in currently estimated Case Fatality Rates (CFRs) for COVID-19 (coronavirus) across countries. Based on a statistical analysis of 20 countries, it was found that countries that have tested more, relative to the number of deaths, have significantly lower CFR estimates. This is likely to be because these countries are able to detect more people with only mild or no symptoms. Moreover, as these countries are testing more widely, their CFR estimates are also likely to be more reliable, as they are subject to less sampling bias. It was also found, based on an analysis of 60 countries, that countries with higher GDP per capita also have significantly lower CFR estimates. This may be because richer countries are likely to have better health systems, and are thus better able to detect and treat COVID-19 cases. Finally, median age of populations was found not to be currently having a significant impact on CFR estimates across countries, probably because any effect from varying age is being masked by much larger variations in the amount of testing. Key conclusions are that the underlying Infection Fatality Rate (IFR) for COVID-19 will likely be at the lower end of the analysed range (i.e. 0.25%, not 10.1%), and that far more people are likely to have been infected than officially reported numbers of cases suggest: 26 million globally by 9 April 2020, not 1.4 million. It is also concluded that the basic reproduction number (R0) will likely be higher than initially estimated, as many R0 estimates may have been based on under-estimates of numbers of infections. In addition, all countries need to do more testing, especially of those with only mild or no symptoms, as only then will it be possible to detect and quarantine most infected people, and trace and test contacts. Finally, special help needs to be given, in particular, to poorer countries. Many other factors may also be affecting estimated CFRs than were analysed, and the IFR could thus still turn out to be higher than 0.25%. In particular, some under-reporting of deaths may be occurring, possibly by a factor of two. It is thus likely that the IFR will be in the range 0.25% - 0.50%. There is also no room for complacency, as many people are still at risk of dying, even if the IFR is 0.25%, because COVID-19 does seem to be very infectious. The analysis was limited by the availability, accuracy and consistency of official reports.
Figures - uploaded by Dan Ward
Author content
All figure content in this area was uploaded by Dan Ward
Content may be subject to copyright.
Content uploaded by Dan Ward
Author content
All content in this area was uploaded by Dan Ward on Apr 27, 2020
Content may be subject to copyright.
Sampling Bias: Explaining Wide Variations in COVID-19 Case Fatality Rates
*
 ! "
#"$$!!% &&' ()* $+$
$","!! 
!$$($$!
-+  
  - 
are all justified$!
Reducing Confusion
$ , $ 
!!.! /,!,!
0,$! 0,
.!$!.$!!
!! ,
$!/,.
., !)1 !,!
Case Fatality Rates
!!! 
,!23'23$
$!4!!$!235
$ !,0
,.,,+!(&6 
!23,0.25%10.1%"$.!7 
 $$,18
 ,18$
9, $$ $
 ,23 $.,
, !.9,
23$, 
!.23 !
723 ,what this report sets out to do
Hypothesis 1: Sampling Bias
,!$&
!23:!
%  ,*+.,
,, 
.!23
,,.23 
$ !,
,.$!8 ,$!
23$
2$! %
$/;;,,,,!;;+(8(8 1<6
 #!=()
8+(8(8 :",,,,
Disclaimer: $,Email:!>,
Citation: : (8(8#?/@7:23
Hypotheses 2 and 3: GDP per capita, and Median Age
,!,2 !
423 .$
 $$# !!
!$!!23.$!
!239, !$ !
$ ,
Initial Analysis
!,$"
,$!!23"
,.;$!$
$! "$!
!,$""
!$",+
!,.!!!,,
, $!  (&6+
  <# $(8
Initial Results
!,!!!!$
23$,,-!$;,!
'*)A -!,!8&(A"$ ' 
!$,! ,
;!,23
,,;9, ! ,!
!!!,-23,,.$
<B5(*6+5'%(),
:4/51&& 58'
8+(8(8 :",,,, (
' ( (' ) )' 1
888A
(88A
188A
&88A
*88A
888A
(88A
C83;3$
@23 $
Figure 1: Estimated CFR against Log10(Reported Tests/Reported Deaths), by Country
Further Analysis
()! ,, ,,$,
$,!D$"
$) & ,.!/,,,,!;
,' &8
Further Results
,!<B230
$)!!!!$!23$,
,-!$<B,23!'('A -!
,23!()'A 
23 $$
 $$$ ,
!&8,.$! ,!!!
!$,,-23
!!23 .19, !!
!23
Updated Analysis
+$ +,
(&6!(8!$,
$(%!(823
 ,,236 $
$,23 $
,,23all now have less sampling bias
; ,&!%,23
,$,;!! 
$,23  $
across time.7Italy ,23 $
$;$$
$!$23 $$,
$!;
 $!
$$'2 ,+
,across countries ,!!(&60,(8
,.$; ,!!!!
23,-,!,
,$!! ,$
These analyses provide five key conclusions:
1.+.!!!,23,
$
,  $"
 ,$,23
+,!!$!,$
!:4/(5'8* 588%
:4/(5% 5'8623!-5)&)A623!$-51)A
:4/51&8 58&623!-5))A623!$-5*&(A
8+(8(8 :",,,, )
,$'!2323,.$
lower end of the analysed range8('A 8A!  
$!!23 !&
!23$8('A $$!!,%8'8A3
!!*<23!8)%A
! $$ 23,$8('A"8'8A
2. , no room for complacency@!23!$
8('A"8'8A ! (88#,2
88(A6 .,!!(8 
!$$38($!((
$38()$$!! 
,,$!(1)6
$focusing on the wrong problem/!23.$
,! $!38.$
 $  .!#
- $
 test those with only mild or no symptoms'
= 23!8('A"8'8A$,'A8A ,
# ,!$
 $$ $$$
!!$ $
3.! $ 23!$8)%A 
infected many times more people$+
$!$ ,$+ 
 '$! 18 888<$ (&$
! 1D$!,
 ,=$!EF$
!!,$without any explanation
.$$!
“actual numbers of those infected are likely to be much higher” .
 !! 
.$$! $
4+!"$$"!!23$!
$GDP per capita$$(' 
,!! +! 23
!G$.,
5.!not enough testing$ 
!,+ 23.$ $
38.$6 !!.
,  !  3
$!,.(&$!
,,! 
(%#H(*"!,"
$$$!   $$$
$!$,!,! ,
$!$$!!=$!!$
D$+(8(8 !/;;,,,,!;;
8+(8(8 :",,,, 1
Country Tests/Deaths
1'* 1 ( ')) 8*%A % )*) I() %'
*8( ( ( &' 8('A & )8* I%) ())
) '& 1 %) *( 811A ' (%* I%' 1(*
' &'' & (8' 8% 8*A ) )&( I11 *1
 (' (& &* 81%A ( && I(8 (
'&( & % (1 8%A ( **( I() 1**
 (1 ) )&1 1( 1(A ( %*& I( '*
 )( % % ( 8')A ( '%8 I& %(%
1' ( 1 (8* 818A ( 81 I) (88
 (( & ( '&1 )A  *1* I) *%
' *** )1 )' ' 8'*A  8' I1% (&
( 8) () ( **' )A '( I8 *
 8( % ' * &'A 1 I8 %*
(& 8 * 88' )*)A *8 I1 )&1
 (( 1' (' % )1*A '' I)* (1
 *%% 1 & % (*A )' I'% '1'
 &'* '%* 81 *&& 1&A * I) ')(
(8 8&*  88 8 81& '1*A ( I) *(%
%8% 1' ( 1% &)&A 1* I( *(
*8 ') * &' )& 8&8 81A 11 I)( 8)*
Reported
Cases
Reported
Deaths
Reported
Tests
Estimated
CFR
GDP per
capita
,$identify andquarantine! trace
and testIncreased testing capacity! ,
!! $$clinical priority!
 !epidemiological priority!!,
$-. !
 $6 $economic and a health imperative,
$ -! 
$,!($7)8
Table 1: Estimated CFR for Countries, Ranked by Test/Death Ratios, as of 26 March 2020

?) )(
))
=,)1
)'
J)& )%
#)* )
#H18
1
1(
B1) 11
+1' 1&
61% 1*
31
9'8 '
K'(
.') '1
GH'' '&
2'%
B'* '
&8
Table 2: Countries ranked by % change in Tests/Deaths ratios over the last two weeks
Sources:!!EF $=$!EF,.!
,,,,!;;+(8(8072/!!%+(8(8
;.;
;$;$
8+(8(8 :",,,, '
Sources:!!(&;8);(872L/(1;8);(80/(;8);(8#.:
,$$ $,!GDP/,,,,!;;$;
Country
B &)&A 1*A 1* ( ')A
? 8*%A 8'&A % )*)  8 1A
K )1*A *'A '' &( A
 81A (%)A 11 1% &A
6 )A '*A '( 1& A
 8('A 8)&A & )8* ' 111 1A
#H 1(A &A ( %*& ( )18 &A
. (*A 1(A )' (%8 )(A
9 )*)A &%)A *8 1(( 1%A
+ 8'*A (()A  8' 1(* &8A
2 '1*A 1%A ( )( &'A
B )A )(A  *1* &* &%A
 818A 1(A ( 81 &*1 &%A
81%A (8A ( && '1 &*A
# 8%A )*)A ( **( %18 %1A
3 &'A 1%%A 1 (8 %*A
 8*A (1'A ) )&( %(* %*A
=, 811A %1A ' (%*  ( %A
GH 1&A ((&A * )% %A
 8')A )&*A ( '%8 (8( (A
Original
CFR
Current
CFR
Original
Tests/Deaths
Current
Tests/Deaths
Change in
Tests/Deaths
J
Table 3: Estimated CFR for Countries, ranked by GDP per capita
Sources: ,,,,!;;GDP:/;;,,,,!;;$;
Median Age: /;;,,;;;+(%;8);(8 ()'<6
Acknowledgements:.@2 # KB  =:#6M!
$!!N,..
:.,,!D!!$.O
8+(8(8 :",,,, &
Country
C7$  &8' ' 8)A I8' (*8 ))
#,J ( (* () %A I*8 (& 1(1
=, ) %%  8'8A I%' 1(* )(
 ( ( (( 81A I& %(% )&*
G#+ 81 (&  && &)A I' ) )*
. ( 81& '( ('1A I'% '1' 1((
#, ) 8& 8' )1(A I'1 8%' 1(
+ ) )%* ) 8)*A I') *) )*%
= * &8) '1& &)'A I1* %& 1(&
#6 (() ( 1(A I1* 1' 111
+ % &% '* 8%'A I1% (& 11
2  81 % 8&%A I1' %%* 1('
 1 %'% '' &A I11 *1 1((
< '8 *% )' 8&A I11 &*8 1%
? % (*1 (* )%A I1) )(' 11
 ) 8)' ( 818A I1( *'( (
2 )( &1  ' &8'A I) *(% 11
GH 1 '1) %' '((A I) ')( 18'
K  1 1 )(%A I)* (1 1%)
 *& 1*  )1 8'&A I)( 8)* 1''
#H  ))( ) 1A I( '* 1*
# &' % ' )* %*(A I(* %' 1(%
# &)( 1(A I() 1** 11'
B 1 (&* %& %*A I( )& 1((
( (% 8)A I(8 ( 1(
< && (* (8A I (1 11'
  &( ' )8A I* &' )&*
C )'* ' 18A I& %8 1)%
B %*& 1 %*A I' && ((
  &8 ' 8)A '88 )11
+ '* ) ((A I1 '8* )%
9 )88 8 )))A I1 )&1 1()
B  )* & 'A I) *% 18%
3  (( (& (8A I8 %* 1
. ' &* ( &A I8 1* )8
6 ( & (& (8A I8 * (*'
?J ) 1% ( (&A I ** )(
67 '*' * )%A I ((1 (*)
 * )18 ) (( 18'A I* &( )%1
C$ ) * (8'A I% *'% )8'
'* (8 )11A I% (() (*
B &)'  %)A I& %() (*
  )& ' 811A I& '% )%%
$ ') & A I& 1( )8
@  &(% 1 ('(A I& (1 (%%
 )( ))( ( )%* %)'A I' &(* )8)
- 1'* 18 *%)A I' 1 (8
#$ '(* * '(A I1 &( 1(&
+$ *& * 1)8A I1 '( )(
+ 18 (& &)&A I1 81* (*
  81& *% *)(A I) *)% )8(
 ((% % )8*A I) 11 )&
6 )1' () &&%A I) 8*) ()
B *8) '1 &%(A I( *( ()'
G. )8 ' &A I( '( 18&
@ ')& )8 '&8A I( 11 ()
 **% (8 (('A I *8 (%
? 1* ' 81(A I '&1 (&%
B.  )%)  8*8A I 1&% ()*
?.2 *8 '88A I&1( %)
Reported
Case s
Reported
Death s
Estimated
CFR
GDP per
capita
Median Age
(years)
J
3$
-! 
$<B
=5'6@
235()'A
?-! 
$<B
=5'6@
235'('A
P
References:
QGC/ ')1' ')(!$.:Q
,,,,!3+(8(8
(Q,$.Q=?=,*6(8(8
3+(8(8
) ? 9%2$(8(8Q$46@7<$Q=,N.
##=8)&(1))3(+(8(8
1  BQB!,,$G#,.!
Q)6(8(83(+(8(8
'Q!R! @7 @- L2Q@?3+(8(8
& BJ #!06 0 C#$(88*Q#$Q#3/1('*
/81(1;1('*##=1&8&3+(8(8
%Q?!JQ?<&6(8(8##=8%%1&'&3)6
(8(8
*Q/2!Q=C!#S=!,
()6(8(83)86(8(8
 Q=/81!$.EBFKQ
#..$ HJ #H, 6. H HJ6 9!$ #
6 H 6$H. 6# ##J.  NG,$ H. 
37*6(8(83(+(8(8
8 $ 0B H0@ C04 C(6(8(8Q/8 888?
 B6Q<##=8(&)8%%3+(8(8
 9- 6Q/:!!Q,,,$$(+(8(83(+(8(8
(Q36/D.!Q3C=,('6(8(83)86(8(8
)Q#3/#H$.Q3)6(8(83
)86(8(8
1 , K!?03 0+ C0?J 60?. B0 TD0C N06 
6)6(8(8Q!!Q37/
(8(88)'(88)&()/88;(8(88)'(88)&()3(+(8(8
' Q<$23F !@?6 7!G:$$K. 
93+(8(8
& Q4!!!Q@##=88)8&)3+(8(8
% Q@7$R36Q8+(8(8,,,38+(8(8
* Q@$9$#9!!Q,,,<3+(8(8
 Q:!$7! :9Q36(83+(8(8
(8 #Q/9..,.Q#:,!(&6(8(8
3+(8(8
( E?$F # 3+(8(8
(( 9 Q6?,++:Q2$3+
(8(8
() Q#!93(88'@$.!
(8()K(8(8Q,,,,3+(8(8
(1 C 30B #0 ?0# N0U 0N :0# K!!&6(8(8Q#$
!!!#+3#(Q#
/8(&;$$)((##=88)&*8%'B6)(%%8
(' = CQ4+.$$4/#,$+!Q#R+++#'6(8(8
3+(8(8
(&  90? B0K$ CJK?.(*2$(8(8Q
$!,.$!,Q<##=8(&)8%%3+(8(8
(% Q/2!Q=C!#S=!,
%
Limitations:!!$ $!!!
 !!!23
, /,4$
$0$ 0,$
()6(8(83)86(8(8
(* = Q#H:4VQ#R+++#
%6(8(83+(8(8
( Q,$!Q??=,)6(8(83+(8(8
)8 B. <1+(8(8QGH.., Q23+(8(8
) QWXYZ[\]^_`Yabc def]^_gZ^hYQ,,,$+$3(&6(8(8
)( Q/1'A!3?Q<!(&6(8(83(%6(8(8
)) E#3F=!B(&6(8(83(%6(8(8
)1 Q#.(&(8(8Q2.=,3(%6(8(8
)' E(8@GF <!(6(8(83+
(8(8
)& E/Bi.jkl3F6Jkl.m$.J3(&6(8(8
)% Ql.D('nJj.JjjjikRHJJQ,,,.JJJ
(%6(8(83+(8(8
)* QDD.o$#+3#(Q,,,DJ#3(&6(8(8
) Q#7#!Q,,,(&6(8(83(+(8(8
18 E코로나바이러스감염증   국내 발생 현황 ) 
월 (& 
일 8
시 기준 F질병관리본부H(&6(8(83
+(8(8
1 Q#!=B$9@(16Q!9 3$!
(16(8(83+(8(8
1( QG.1'$Jo.QH+(&6(8(83+(8(8
1) QH,,B.Q.,B3(&6(8(8
11 QH,p$VJqr.D p,$JF$-,$JB
(&6(8(83(+(8(8#/QH,BJ$,sJO=:@+=@Q
,,,B(%6(8(83(+(8(8
1' Q=(8Q,,,J<3(&6(8(8
1& Q+:/@G/,,.Q6?=@:#(%6(8(8**$
3(+(8(8
1% QB3HH6/F,,,63(&6(8(8
1* Q,( 8) .()t6@3uQ@=,(1(&6(8(83(+
(8(8
1 Q?!(&8)(8(8Q6#vvvw 333+(8(8
'8 QHkQ.93(&6(8(8
' Q(&x.k!xJy.Jj6JjQ3jp.Jj9(&6
(8(83(+(8(8
'( Q新型コロナウイルス感染症の現在の状況と厚生労働省の対応について（令和２年３月 
(& 
日版） Q,,,,D
K(&6(8(83+(8(8
') QG$Q,,,.3(&6(8(8
'1 3 H6zQ#{.Q#@9|3(%6(8(83(+
(8(8
'' Q=$!.GHQ<GH3(&6(8(8
'& QGHD$' &'*Q@#(&6(8(83(+(8(8
'% EBmm-$  (16(8(8F #m$- 23(&6(8(8
'* EGB+@#==@C3=+3G##@+#@F!93$!B3
(&6(8(8
' Q==G6?@3#/B$#+$!Q<6+=,(&
6(8(83(+(8(8
&8 E+<<3=+6@=(&;8);(8(83@%88FB6 B!6#3‐
(&6(8(8
*
... Still, it might hit hard on those in urban clusters since cases manifested first in the urban areas. 25,26 It was asserted that urban residents might be affected in particular ways, and in some areas, face more unenthusiastic impacts than rural residents. 27 COVID-19 lockdown restricted the movement of many households, preventing them from having regular access to basic needs of life and leaving them abandoned in the house. ...
... The findings are indeed bizarre as they disagreed sharply with previous reports speculating that cumulative emotional impact due to COVID-19 lockdown shall be low in most rural areas but may probably hit high on those in urban clusters. [25][26][27] In application, the findings of this research might help design educational intervention to mitigate the impact of COVID-19 pandemic lockdown with consideration on both urban and rural populations. ...
Emotional impact of COVID-19 lockdown and mitigation options: A cross-sectional survey of households in Ebonyi State, Nigeria
Article
Full-text available
  • Sep 2021
Background Coronavirus disease 19 (COVID-19) has continued to plague households, leading to lockdown problems. Adopting appropriate mitigation strategies can reduce the impact on family members. Purpose To assess the emotional impact of COVID-19 epidemic lockdown and mitigation measures among households in Ebonyi State. Methods Cross-sectional survey design was used to study 516 participants. Emotional impact of COVID-19 lockdown ( r = 0.73) and mitigation options ( r = 0.92) questionnaire was used for data collection. Of the 516 copies of the questionnaire distributed, 493 copies (95.5% return rate) were used for data analysis. Data were analysed using descriptive statistics, standard deviations, and t-tests. Results The data showed the emotional impact of the COVID-19 epidemic was high (2.97 ± 0.48) on households. They embraced friendly communication and communication with their partners, maintaining regular contact with their loved ones by phone, email, social media, or video conference to alleviate the COVID-19 lockdown. No significant differences were found in the emotional impact for location ( p > 0.05). Significant differences were not observed in many gender-based mitigation options. Conversely, a significant difference existed in the mitigation options based on location ( t = 3.143, p < 0.05). However, there was no significant difference in friendly interactions and communication with partners ( t = 0.354, p > 0.05), finding opportunities to develop excellent and promising news and images ( t = 0.770, p > 0.05) and maintaining regular communication with loved ones via phone, email, social media, and video conference ( t = 0.448, p > 0.05). Conclusion The emotional impact of COVID-19 confinement was significant on family life and was more prevalent among men and urban dwellers. There is need to organise an awareness campaign on fundamental ways to overcome emotional distress using media targeting family members to promote emotional health.
View
... However, considerable uncertainty remains as to the relative importance of these and other factors 6 . This report has thus been written to provide the results of an analysis comparing the extent of sampling bias in testing 8 , with the age distribution of COVID-19 cases, which should help provide some much needed clarity. ...
... 9 This has likely occurred in many countries with respect to COVID-19, and variations in sampling bias have already been shown to account for much of the wide variation in Case Fatality Rates across countries. 8,10 Most countries have only tested a small minority of their populations to date (4% on average, for the countries on figure 1), and many tests that have been done, have been prioritised on the seriously ill 11 , due to the clinical priority of diagnosing such individuals (so they can be treated appropriately). Moreover, some countries have had official policies not to test those with only mild or no symptoms 12,13 , whilst it is already wellestablished that most people infected with COVID-19 do not get seriously ill 14,15 . ...
Sampling Bias: Explaining Variations in Age Distributions of COVID-19 Cases
Technical Report
Full-text available
  • May 2020
Why does the age distribution of COVID-19 cases vary so much, and are the elderly really more likely to become infected? This report shows that countries that test more, relative to the number of deaths, have a much higher proportion of young (20-29 year old) cases, and a much lower proportion of elderly (70 years and over) cases, relative to the wider population. Moreover, these countries will likely have much more accurate statistics, because they have much less sampling bias in their testing. These findings thus imply that many countries, and especially those with low tests/deaths ratios, have substantially underestimated the number of young adults infected, in particular. Conversely, the disproportionately high percentage of elderly cases in countries with low tests/deaths ratios is likely to be mainly just an artefact of too little testing in these countries, and is not indicative of a disproportionately high number of actual infections in this age group: i.e. elderly people are not more likely to be infected, although they are more likely to become seriously ill, if infected. It is thus concluded that all countries, and particularly those with low tests/deaths ratios – such as the UK, Algeria, France, Sweden, the Netherlands, Mexico and Brazil – need to do massively more testing urgently, especially of young adults. Any young adults working with the elderly, in particular, need to be tested at the highest priority, in all countries, as this may substantially reduce mortality. Children also need to be specifically tested in all countries because, as they are more likely to be asymptomatic, they are much less likely to have been tested so far. There is thus insufficient information to date to determine infection rates amongst children.
View
... The variation in CFRs can be driven by several factors, among them "real" differences arising from different age-specific mortality risks among the infected (Dowd et al. 2020;Dudel et al. 2020). However, the variation in CFRs may also reflect other factors such as differences in testing intensity (Ward 2020;Hasell et al. 2020) and test specificity (Wölfel et al. 2020;Hasell et al. 2020); variation in the age structure of the confirmed cases (Dudel et al. 2020); and the stage of progress of the COVID-19 outbreak (Lourenco et al. 2020). In addition, variation in the way deaths are classified to COVID-19 versus non-COVID-19 deaths may also explain some of the cross-country differences in CFR. ...
... It is, however, informative to compare the results to those of other studies. Crude infection fatality rates are estimated, for example, by Ward (2020), 0.25%-0.5%, and Russell et al. (2020), 0.2%-1.3%. ...
A demographic scaling model for estimating the total number of COVID-19 infections
Preprint
Full-text available
  • Apr 2020
Background The total number of COVID-19 infections is critical information for decision makers when assessing the progress of the pandemic, its implications, and policy options. Despite efforts to carefully monitor the COVID-19 pandemic, the reported number of confirmed cases is likely to underestimate the actual number of infections. We aim to estimate the total number of COVID-19 infections in a straightforward manner using a demographic scaling approach based on life tables. Methods We use data on total number of COVID-19 attributable deaths, population counts, and life tables as well as information on infection fatality rates as reported in Verity et al. (2020) for Hubei, China. We develop a scaling approach based on life tables and remaining life expectancy to map infection fatality rates between two countries to account for differences in their age structure, health status, and the health care system. The scaled infection fatality rates can be used in combination with COVID-19 attributable deaths to calculate estimates of the total number of infected. We also introduce easy to apply formulas to quantify the bias that would be required in death counts and infection fatality rates in order to reproduce a certain estimate of infections. Findings Across the 10 countries with most COVID-19 deaths as of April 17, 2020, our estimates suggest that the total number of infected is approximately 4 times the number of confirmed cases. The uncertainty, however, is high, as the lower bound of the 95% prediction interval suggests on average twice as many infections than confirmed cases, and the upper bound 10 times as many. Country-specific variation is high. For Italy, our estimates suggest that the total number of infected is approximately 1 million, or almost 6 times the number of confirmed cases. For the U.S., our estimate of 1.4 million is close to being twice as large as the number of confirmed cases, and the upper bound of 3 million is more than 4 times the number of confirmed cases. For Germany, where testing has been comparatively extensive, we estimate that the total number of infected is only 1.2 times (upper bound: 3 times) than the number of confirmed cases. Comparing our results with findings from local seroprevalence studies and applying our bias formulas shows that some of their infection estimates would only be possible if just a small fraction of COVID-19 related deaths were recorded, indicating that these seroprevalence estimates might not be representative for the total population. Interpretation As many countries lack population based seroprevalence studies, straightforward demographic adjustment can be used to deliver useful estimates of the total number of infected cases. Our results imply that the total number COVID-19 cases may be approximately 4 times (95%: 2 to 10 times) that of the confirmed cases. Although these estimates are uncertain and vary across countries, they indicate that the COVID-19 pandemic is much more broadly spread than what confirmed cases would suggest, and the number of asymptomatic cases or cases with mild symptoms may be high. In cases in which estimates from local seroprevalence studies or from simulation models exist, our approach can provide a simple benchmark to assess the quality of those estimates.
View
... Moreover, this confounding effect is not surprising, given that countries varied a lot in how much they tested, 168 and that an infected person could usually only be reported as a case if they were tested. 169,170 In addition, as shown on table 1, testing correlated strongly with many variables, probably because richer countries, with better health care systems, had much more testing capacity. Therefore, if any of these variables were studied in isolation against cases, a very misleading result could be reported. ...
Systematic Review of Covid-19 Cross-Country Comparisons: Why Most Findings Were False
Preprint
Full-text available
  • Apr 2024
During the initial months of the pandemic, there was a pressing need to understand wide variations across countries in Covid-19 cases, deaths and case-fatality rates. However, although many cross-country comparisons were published, in reputable journals, this systematic review shows that most of these were methodologically flawed. As a result, numerous contradictory (and often implausible) explanations were proposed, supported by spurious results, which likely confused and misled many people. 121 cross-country comparisons were reviewed, published between April and August 2020, and listed on PubMed, across which 65 explanations were supported, ranging from air pollution to antibodies, from cabbage to corruption, and from smoking to sandflies. Moreover, of these articles: (1) 50% studied less countries than is normally statistically recommended. (2) 48% did not show their findings were statistically significant (α = 0.05). (3) 60% did not control for the impact of any potential confounding factors. (4) 64% compared at one single time point, despite wide variations over time. 74% of the articles had at least TWO of these easy-to-avoid flaws, each of which have potential qualitative impacts on findings. Most of the articles were thus fundamentally flawed, and most of the proposed explanations therefore likely arose from false findings. Moreover, many of the articles were also very influential, in terms of both citations and Altmetric, and no correlation was found between these metrics and numbers of flaws; i.e. less robust articles were just as influential. Many of the articles thus likely misled many people, increasing confusion at a time of much needed clarity. In addition, more experienced researchers, and more influential journals, were not more careful to avoid the four flaws, and flaws did not decrease over time. Finally, across the most robust articles reviewed, flights, lockdowns, population age, air temperature, humidity, and precipitation were proposed as explanations, all of which have been supported by substantial other research. This implies that cross-country comparisons can provide useful insights, if they avoid such basic flaws.
View
... Therefore, the CFR will increase significantly, which could have adverse consequences, including imposing more force on the health system. Finally, due to the increasing the trend of epidemics in all countries of the world, it is necessary to conduct experiments beyond the clinical priority on a large number of people in the community so that timely diagnosis and control of the disease can significantly reduce CFR (31). ...
An ecological modeling on the adjusted effects of socioeconomic determinants and HLA-DRB1 alleles in fatality of COVID-19 during the early phase of epidemics in a group of countrie
Article
Full-text available
  • May 2022
  • PRZ ELEKTROTECHNICZN
INTRODUCTION. Socioeconomic determinants along with genetic status may affect fatality rate of COVID-19. We intend to investigate the adjusted effects of the HLA-DRB1 alleles and socioeconomic determinants including gross domestic product per capita (GDP cap) and health expenditure per capita (HE cap) in fatality of COVID-19 during the early phase of epidemic in a group of countries. METHODS. As an ecological study, early exposure to epidemics was defined as having more than 5000 confirmed cases of COVID-19 from 1 March 2020 to 1 April 2020. Poisson regression was used to report adjusted incidence rate ratio (IRR) for case fatality rate in this constant time period. RESULTS. Fourteen countries were eligible. Among the alleles, DR7 showed the strongest risk factor (IRR=112.535, P<0.001). Having GDP cap more than 40000$ or having HE cap more than 3000$ was a protecting factor (IRR=0.899, P<0.001, adjusted with allele DR7). Having GDP cap more than 40000$ along with having HE cap more than 3000$ was a protecting factor (IRR=0.471, P<0.001, adjusted with allele DR7). CONCLUSION. Socioeconomic status of the countries may compensate the probable harmful effect of some HLA-DRB1 alleles. This conclusion was limited to a period that all the selected countries had almost similar governmental intervention.
View
... There is a large difference in GDP per capita and its impact on the current CFR estimation among the different countries; so with increasing prevalence, in those with lower GDP per capita CFR can significantly increase because of the infection fatality rate (IFR) consistence for COVID-19 assuming that the number of deaths reported worldwide is true, the actual number of infected people is likely to be much higher than reported, so the CFR will increase significantly, which could have adverse consequences, including imposing more force on the health system. Therefore, due to the increasing the trend of epidemics in all countries of the world, it is necessary to conduct experiments beyond the clinical priority on a large number of people in the community so that timely diagnosis and control of the disease can significantly reduce CFR [39]. ...
An ecological modeling on the adjusted effects of socioeconomic determinants and HLA- DRB1 alleles in fatality of COVID-19 during the early phase of epidemics in a group of countries
Preprint
  • Oct 2020
Introduction: In early exposure to each epidemics, background conditions of the countries are very important. Socioeconomic determinants along with genetic status may affect fatality rate of COVID-19. The present study was designed to investigate the adjusted effects of the HLA-DRB1 alleles and socioeconomic determinants including gross domestic product per capita (GDP cap) and health expenditure per capita (HE cap) in fatality of COVID-19. Methods: As an ecological study, early exposure to epidemics was defined as having more than 5000 confirmed cases of COVID-19 from 1 was regarded. Poisson regression was used to report adjusted incidence rate ratio (IRR) with 95% confidence interval (CI) for case fatality rate in this constant time period. Result: Fourteen countries were eligible. Among the alleles, DR7 showed the strongest risk factor (IRR=112.535, 95% CI=84.496-149.878, P<0.001). Having GDP cap more than 40000$ or having HE cap more than 3000$ was a protecting factor in comparison to less than the thresholds of the both (IRR=0.899, 95% CI=0.868-0.932, P<0.001, adjusted with allele DR7). Having GDP cap more than 40000$ along with having HE cap more than 3000$ was a protecting factor in comparison to less than the thresholds of the both (IRR=0.471, 95% CI=0.456-0.486, P<0.001, adjusted with allele DR7). The model containing all HLA DRB1 alleles without other covariates showed the best fitting among the models (pseudo R 2 =0.990, along with the lowest residues). Conclusion: Although the ecological fallacy remained because of failure to complete the adjustment due to multicollinearity, the current evidence showed that socioeconomic status of the countries may compensate the probable harmful effect of allele DR7. Using complete profile of HLA DRB1 alleles in regression models showed a better fitting for prediction of fatality in comparison to the integration of selected alleles with socioeconomic determinants. This conclusion was for a time period that all the selected countries had almost similar governmental intervention.
View
... However, as regards the impact of GDP, this does not mean that poorer countries will necessarily be spared high mortality, as poorer countries also likely have weaker health systems and also worse population health, both of which will likely increase COVID-19 deaths. In addition, the response of countries to their own statistics of reported cases is problematic, as reported cases are substantially impacted by the amount of testing undertaken: countries that test more, relative to the number of deaths, report far more cases, probably because they are far better able to detect the majority of infected people who only have mild or no symptoms 26,19,20 . It is therefore likely that countries that tested less, detected less cases and thus responded slower to the pandemic, regardless of how many actual infections those countries in fact had. ...
Actions Speak Louder than Age: Explaining Wide Variations in COVID-19 Deaths
Technical Report
Full-text available
  • May 2020
Why do reported COVID-19 deaths vary so much across countries, and is population age really a plausible explanation? This report shows that variations in the extent to which countries responded to the pandemic can explain a lot of the variation in reported deaths, whereas population age cannot. Countries that enacted more stringent policies, at a relatively early stage in the pandemic (when each had one death per 8 million), went on to have a lot less reported deaths per million over the subsequent 6 weeks. Conversely, there was no significant correlation found between population age and deaths per million, across the 26 high income countries analysed, despite the fact that these countries varied substantially in percentage of population over 65 (i.e. 1% to 28%). This is likely to be because any effect from varying population age is being masked by much larger variations in the speed with which governments responded to the pandemic. It is thus concluded that population age cannot account for why some particular countries (such as Italy and Spain) have been worse hit than others by the pandemic, and that having younger populations will not necessarily protect countries against high numbers of deaths. All countries, regardless of population age, thus need to respond appropriately and rapidly to the pandemic. It is also concluded that some experts and publications seem to have over-stated the role of population age in explaining various aspects of the pandemic. The analysis was limited by the availability, accuracy and consistency of official reports.
View
... A statistical report published on 10 April 2020 used ratios between reported tests and reported deaths (tests/deaths ratios) to estimate the amount of sampling bias in COVID-19 across countries. 16 Countries with a high tests/deaths ratio have less sampling bias in their testing, as they are testing more widely, and are thus better able to detect the majority of infected people, who only have mild or no symptoms. Conversely, countries with low tests/deaths ratios are focusing their testing mostly on the seriously ill and dying, and thus have much more sampling bias. ...
Sampling Bias, Case Fatality Rates and COVID-19 Testing: A Further Analysis
Preprint
Full-text available
  • Apr 2020
Why do estimated Case Fatality Rates (CFRs) for COVID-19 vary so much across countries? Many explanations have so far been proposed including varying: health and/or age of populations; access to and/or quality of health care; methods of recording deaths, and; amounts of testing. This analysis shows that variations in the extent of sampling bias in COVID-19 testing can account for a lot of the variation in estimated CFRs, both across countries and across time. Countries that have tested more, relative to the number of deaths, have much lower CFR estimates, and this is highly statistically significant across 60 countries. Moreover, changes in CFR estimates over time are mainly consistent with changes in the extent of sampling bias. Conversely, neither GDP per capita, nor percentage of population over 65, nor median age of population can account for current variations in CFR estimates, likely because any effects from varying age and GDP are being masked by much larger variations in testing. These findings imply that the Infection Fatality Rate (IFR) will likely be at the lower end of the current CFR range (i.e. much closer to 0.14% than 15%), and it is estimated that the IFR will likely be in the range 0.28% to 0.68%. These findings also imply that many, or even most, countries have substantially underestimated numbers of cases of COVID-19, and that the basic reproduction number (R0) may also have been underestimated during the early stages of the pandemic. COVID-19 likely has a lower fatality rate (IFR) than some fear, but it also likely has a much higher infection rate (R0) than many realise. Many people, especially the elderly and those with pre-existing conditions, are therefore still at risk of dying. All countries thus need to do more testing urgently, especially of those with only mild or no symptoms, and this applies particularly to countries with low tests/deaths ratios. Other factors may also be affecting CFR estimates than were analysed, and so the IFR could still lie outside the estimated range (i.e. 0.28% to 0.68%). The analysis was limited by the availability, accuracy and consistency of official reports.
View
... (2020). Covid-19 seems to spread unusually rapid and may be much less fatal than estimated in (e.g., Silverman ea., 2020Ward, 2020), and has often been compared to flu in terms of mortality, as described in the next pages (also see #14). ...
The Pandemic Timescape 2020-2023: A Synthesis
Book
  • Mar 2020
Novel coronavirus (SARS-CoV-2) disease (COVID-19) spread from Wuhan in China (2019) around the globe and this pandemic of 2020 was a salient public health emergency and psychosocial shock event that influenced the lives of almost all humans on Earth. This study synthesizes the pandemic processes over timescape 2020-24, from the >663 million humans who were infected over 2020-21 when >21.3 million died due to Covid and secondary excess mortality, and >70% of all humans were subjected to various societal restrictions and lockdowns, such as ~4.5 billion humans over may 2020 alone. The societal lockdowns and largest global economic peturbation in a century (-3% world GDP, xx% peak inflation, rent from xx to xx, energy prices rose xx%), which coincided with a collective drop in life expectancy and mental health and well-being, especially more anxiety, depression, suicides, and domestic violence against women and children were recorded. Coronavirus exploited social contact and transport, urbanization, liberalism, anxiety, and societal lockdowns. Humans showed marked emotional, cognitive, physical, social, and societal responses over 2020-23, and the collective focused on virus mutations, vaccines, breakthrough infections, long-Covid, and more stress, loneliness, polarization, demonstrations, riots, violence, and conspiracy theories, with more authoritarianism and government coups around the globe (e.g., US Insurrection, 2021). When Russia invaded Ukraine (≥22w8) war returned to Europe, and globally more refugees were observed than since the second world war (WW2). Over 2020-24 we witnessed major individual, sociocultural, geopolitical, and technological changes, from the rise of TikTok and mainstream Artifical Intelligence (ChatGTPo4) to #BLM and second wave of #MeToo, which affected the collective psyche and discourse, and coronavirus and overt nuclear threat coincided with an unprecedented series of natural disasters and record heatwaves that confirmed humans had created a #ClimateCrisis of planetary proportions. Over 2023-24 the planetary climate ventured into uncharted territory driven by the human industrial revolution (>1830). Most humans described the pandemic years 2020-22 as the hardest and most stressful years they had experienced in their lifetime, which affected a generation of young humans. Over 2023 the collective return to “normalcy” was hampered by Ukranian and Israelian wars, cost-of-living crisis, and climate disasters that made humans realize the world had changed into an era of conflict. The pandemic timescape 2020-23 was an unique period in time to learn about human psychology and collective preparedness (a natural experiment), as engagement with natural hazards is a master task of civilizations. Psychological theories could be examined across contexts and in diverse samples, who were often exposed to comparable collective stressors of personal significance. Such perturbations can help to study human stress dynamics and complex social systems, beyond what can be learned from more stable periods. The collective changes and interconnected consequential events over 2020-23 including the global rise in military spending, government control, radical right politics, and a decline in freedom of speech, made this timescape a mini "Axial-period" that shapes the decades ahead.
View
View
View
Cambodia's Coronavirus Complacency May Exact a Global Toll". The New York Times
  • Feb 2020
  • Hannah Beech
Beech, Hannah (17 February 2020). "Cambodia's Coronavirus Complacency May Exact a Global Toll". The New York Times. ISSN 0362-4331. Retrieved 2 April 2020.
Photos of ravaged grocery stores show how people are panic-buying across the US in the wake of the coronavirus pandemic
  • Mar 2020
  • Paulina Cachero
Cachero, Paulina. "Photos of ravaged grocery stores show how people are panic-buying across the US in the wake of the coronavirus pandemic". Insider. 13 March 2020. Retrieved 2 April 2020.
Non-severe vs severe symptomatic COVID-19: 104 cases from the outbreak on the cruise ship "Diamond Princess" in Japan
  • Mar 2020
  • Sakiko Tabata
  • Kazuo Imai
  • Shuichi Kawano
  • Mayu Ikeda
  • Tatsuya Kodama
  • Kazuyasu Miyoshi
  • Hirofumi Obinata
  • Satoshi Mimura
  • Tsutomu Kodera
  • Manabu Kitagaki
  • Michiya Sato
  • Satoshi Suzuki
  • Toshimitsu Ito
  • Yasuhide Uwabe
  • Kaku Tamura
"Non-severe vs severe symptomatic COVID-19: 104 cases from the outbreak on the cruise ship "Diamond Princess" in Japan". Sakiko Tabata, Kazuo Imai, Shuichi Kawano, Mayu Ikeda, Tatsuya Kodama, Kazuyasu Miyoshi, Hirofumi Obinata, Satoshi Mimura, Tsutomu Kodera, Manabu Kitagaki, Michiya Sato, Satoshi Suzuki, Toshimitsu Ito, Yasuhide Uwabe, Kaku Tamura, medRxiv preprint. 18 March 2020. Retrieved 2 April 2020.
Coronavirus: Why death and mortality rates differ
  • Mar 2020
  • Martha Henriques
Henriques, Martha. "Coronavirus: Why death and mortality rates differ". www.bbc.com. 2 April 2020. Retrieved 2 April 2020. 12 "RIVM: verspreiding coronavirus lijkt af te remmen". RTL Nieuws (in Dutch). 25 March 2020. Retrieved 30 March 2020. 13 "Special Report: Italy and South Korea virus outbreaks reveal disparity in deaths and tactics". Reuters. 13 March 2020. Retrieved 30 March 2020.
Global Covid-19 Case Fatality Rates
  • Apr 2020
"Global Covid-19 Case Fatality Rates", Centre for Evidenced Based Medicine, Oxford University. Web blog. Jason Oke, Carl Heneghan. Retrieved 1 April 2020.
World must prepare for inevitable next flu pandemic, WHO says
  • Mar 2019
"Ergebnisse der Heinsberg-Studie machen Hoffnung". www.t-online.de (in German). Retrieved 9 April 2020. 19 "World must prepare for inevitable next flu pandemic, WHO says". Reuters. 11 March 2019. Retrieved 1 April 2020.
Coronavirus: Head-shaking doctors and non-hand-shaking workers". SWI swissinfo.ch
  • Mar 2020
  • Thomas Stephens
Thomas Stephens. "Coronavirus: Head-shaking doctors and non-hand-shaking workers". SWI swissinfo.ch. 26 March 2020. Retrieved 1 April 2020. 21 " Basic reproduction number", ScienceDirect, Retrieved 9 April 2020.
COVID-19) Daily Epidemilogy Update
  • Mar 2019
"Coronavirus Disease 2019 (COVID-19) Daily Epidemilogy Update", Government of Canada. 29 March 2020. Retrieved 1 April 2020.