Table 1.
Eligible studies for the main analysis (those countries that have age-stratified COVID-19 death data and seroprevalence information).
| Country (first author) | Sampling period | Number tested <60 [<70] | Antibody type(s) | Adjusted seroprevalence <60 [<70] (%) | Adjustments made | COVID-19 deaths <60 [<70] (n) | Population <60 [<70] (n) | IFR in <60 [<70] (%) |
|---|---|---|---|---|---|---|---|---|
| Afghanistan (Saeedzai) | 6/1/20 to 6/30/20 | NA [5168] | IgG/IgM | 35.1 [35.2] | NA | 355 [576] | 37,284,465 [38,341,961] | 0.003 [0.004] |
| Andorra (Royo-Cebrecos) | 5/4/20 to 5/28/20 | 49,355 [55,347] | IgG/IgM | 12.3 [12.46] | NA | 2 [6] | 61,881 [70,384] | 0.022 [0.058] |
| Canada (Tang) | 5/1/20 to 9/30/20 | 5789 [7938] | IgG only | 2.12 [2.08] | NA | 280 [915] | 28,346,618 [33,059,361] | 0.039 [0.113] |
| Czech Republic (Piler) | 12/1/20 to 1/31/21 | 5665 [NA] | IgG only | 42.77 [42.77] | NA | 565 [2111] | 7,908,150 [9,232,767] | 0.011 [0.034] |
| Denmark (Espenhain) | 9/11/20 to 12/11/20 (median date 12/16/20) | NA [NA] | IgG/IgM/IgA | 4.64 [4.64] | Test sensitivity and specificity using the Rogan-Gladen estimator | 36 [129] | 4,278,562 [4,933,092] | 0.012 [0.036] |
| England (Ward) | 6/20/20 to 7/13/20 | 77,955* | IgG only | 6.76 [6.25] | Test performance, and weighted to account for sample design and for variation in response rate (age, sex, ethnicity, region and deprivation) to be representative of the England population over 18 years | 2586 [6425] | 42,889,306 [48,870,419] | 0.076 [0.179] |
| Faroe Islands (Petersen) | 11/21/20 to 11/30/20 | 40,467 [46,152] | IgG/IgM/IgA | 0.54 [0.63] | NA | 0 [0] | 40,467 [46,152] | 0 [0] |
| Finland (Melin) | 4/13/20 to 1/25/21 | NA [4887] | IgG only | 0.61 [0.61] | NA | 17 [43] | 3,934,387 [4,646,712] | 0.056 [0.119] |
| France (Warszawski) | Median date 11/24/20 | 48,993* | IgG only | 7.04 [6.61] | Sample design, non-response, census calibration | 1968 [6024] | 47,753,753 [55,518,538] | 0.039 [0.109] |
| France (Carrat) | 5/4/20 to 9/30/20 | 40,193 [56,843] | IgG only | 6.71 [5.9] | NA | 1306 [3684] | 47,753,753 [55,518,538] | 0.029 [0.079] |
| Germany (Neuhauser) | 10/1/20 to 2/28/21 (median date 11/11/20) | 11,302* | IgG only | 1.96 [1.96] | Non-response, test performance and seroreversion | 903 [2708] | 59,792,644 [70,436,786] | 0.077 [0.196] |
| Hungary (Merkely) | 5/1/20 to 5/16/20 | 8088* | IgG only | 0.64 [0.64] | Design weighted. Response sample calibrated to known population counts by region, sex, and age categories. | 27 [95] | 7,076,206 [8,382,638] | 0.057 [0.17] |
| Iceland (Gudbjartsson) | 4/27/20 to 6/5/20 | NA | IgG/IgM/IgA | 0.8 [0.8] | NA | 1 [3] | 267,524 [305,060] | 0.043 [0.113] |
| Ireland (Heavey) | 6/22/20 to 7/16/20 | NA | IgG only | 1.69 [1.69] | Weighted to adjust for varying response rates in age-sex strata | 64 [190] | 4,217,964 [4,779,564] | 0.073 [0.192] |
| Israel (Reicher) | 6/28/20 to 9/14/20 (median date 7/9/20) | 38,673 [47,423] | IgG only | 5.18 [4.95] | Age, sex, time period, RT-PCR status, municipal strata, sampling | 21 [60] | 7,231,052 [7,934,695] | 0.004 [0.012] |
| Italy (Sabbadini) | 5/25/20 to 7/15/20 | NA | IgG only | 2.44 [2.46] | Non-response, region, age, sex, working status, province | 1600 [5112] | 41,830,101 [49,314,963] | 0.135 [0.361] |
| Japan (Yoshiyama) | 6/1/20 to 6/7/20 | 5156 [6476] | IgG/IgM/IgA | 0.13 [0.12] | NA | 37 [122] | 83,064,470 [98,939,705] | 0.033 [0.098] |
| Jersey (Government of Jersey) | Median date 6/16/20 | 1077* | IgG/IgM | 3.65 [3.65] | NA | 1 [4] | 77,734 [87,432] | 0.032 [0.113] |
| Lao PDR (Virachith) | 8/12/20 to 9/25/20 | 2082 [NA] | IgG/IgM | 4.46 [4.46] | Weighted for complex survey sample design, age and sex | 0 [0] | 6,781,408 [7,099,379] | 0 [0] |
| Lithuania (Smigelskas) | 8/10/20 to 9/10/20 | NA | IgG/IgM | 1.38 [1.38] | NA | 5 [17] | 1,974,425 [2,327,236] | 0.013 [0.038] |
| Mexico (Basto-Abreu) | 8/18/20 to 11/13/20 | NA | IgG/IgM/IgA | 25.77 [25.77] | Test performance, and used sampling weights to adjust for selection probabilities and non-response rates (with post-stratification on region, sex and age group) | 36,779 [62,926] | 114,441,068 [122,706,021] | 0.108 [0.172] |
| Nepal (Government of Nepal) | 10/9/20 to 10/22/20 (median date 10/16/20) | NA | IgG/IgM/IgA | 13.54 [13.64] | Survey design weights, age | 340 [504] | 26,615,582 [28,103,660] | 0.008 [0.011] |
| Netherlands (Vos) | 6/9/20 to 8/24/20 (median date 6/14/20) | 4600 [5817] | IgG only | 4.46 [4.56] | Adjusted for survey design, weighted to match the distribution of the general Dutch population (based on sex, age, ethnic background, and degree of urbanization) and controlled for test characteristics | 196 [694] | 12,576,973 [14,706,474] | 0.03 [0.09] |
| Norway (Eik Anda) | 11/25/20 to 2/15/21 (median date 12/20/20) | 22,264* | IgG only | 0.94 [0.94] | Rake weighting for population estimates of seroprevalence by age, sex, place of birth and county based on individual-level data for the invited sample (participants and non-responders) together with the corresponding distributions from the source population, provided by the Norwegian Population Register. Applied propensity scores for nonresponse adjustment and jackknife replicate weights for the raking procedure. Estimates subsequently corrected for test performance | 23 [64] | 4,159,899 [4,746,055] | 0.037 [0.091] |
| Oman (Al Abri) | 11/8/20 to 11/13/20 | NA | IgG only | 22.32 [22.32] | Age group, sex, nationality | 553 [930] | 4,888,809 [5,031,596] | 0.042 [0.068] |
| Pakistan (Ahmad) | 10/21/20 to 11/8/20 | 4022 [NA] | IgG/IgM | 6.33 [6.33] | NA | 3287 [5448] | 206,007,412 [214,909,826] | 0.02 [0.031] |
| Portugal (Canto e Castro) | 9/8/20 to 10/14/20 (>90% of the test were performed 9/8/20 to 9/20/20) | NA | IgG/IgM/IgA | 2.32 [2.32] | Adjusted for test performance, used sample weights and post-stratified by sex to adjust the seroprevalence extrapolating from the strata to the whole population | 89 [257] | 7,202,167 [8,495,991] | 0.04 [0.098] |
| Slovenia (Poljak) | 10/17/20 to 11/10/20 | NA | IgG/IgM/IgA | 5.32 [5.32] | NA | 18 [55] | 1,502,217 [1,787,158] | 0.015 [0.041] |
| Spain (Government of Spain) | 11/16/20 to 11/29/20 | NA | IgG only | 9.58 [9.71] | Characteristics of the random subsample of the fourth round | 2329 [6593] | 34,605,241 [39,945,895] | 0.046 [0.111] |
| USA (Sullivan) | 8/9/20 to 12/8/20 (median date 10/30/20) | 3481* | IgG/IgM/IgA | 16.48 [16.48] | Test performance, design weights | 32,487 [73,947] | 255,284,698 [293,772,868] | 0.077 [0.153] |
| USA (Kalish) | 4/1/20 to 8/4/20 (>90% of the tests were performed 5/10/20 to 7/31/20) | 6785 [NA] | IgG/IgM/IgA | 4.8 [4.8] | Age, region, sex, urban/rural, race, Hispanic, BRFSS survey response, sensitivity, specificity | 16,411 [37,410] | 255,284,698 [293,772,868] | 0.097 [0.192] |
* Number tested <65 years