Read on Twitter
New serology from Iquitos, Peru (~426k) has now been published. The serology covers two periods - a baseline study in July (2 months after the peak of the 1st wave in Iquitos) and a follow up study in August. Baseline survey found 70% sero+ve 1/n https://www.medrxiv.org/content/10.1101/2021.01.17.21249913v1
Sampling for the survey was random and systematic - serological study was carried out with the Biotech Co LTD IC assay. On the 1 month follow up mark the study found ~65% seropositive adjusted 68% to account for those lost to follow up. 2/n
However, this apparent seroreversion may largely be an artifact of the weakening sensitivity of these rapid IgG/IgM kit used for the study and may also reflect some antibody waning. 3/n
Due to the high incidence of dengue in the Iquitos area, there has been some concern on potential cross-reactivity giving out false positive antibody results. However, studies on this have shown very low risk of x-react with the type of assay used here 4/n https://pubmed.ncbi.nlm.nih.gov/32813814/
Oddly, attack rates in 18-29 and 30-59 were significantly lower than <12, 12-17 and >60. A new paper in the Lancet studying household transmissions in Wuhan did suggest >60 yr old are likely at higher risk of infection than other age groups. 5/n https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30981-6/fulltext
For differing attack rates between 18-59 and <17 yr olds. Maybe some form of clustering effect due to higher contacts between <17 and >60? The authors also propose other possible reasons for the results seen 6/n
Iquitos recorded ~2500 confirmed + suspected COVID deaths in its 1st wave suggesting at least ~0.6% of its population died over this period. Maynas province (Iquitos makes up 79% of the population) has so far recorded 7947 excess DPM (0.79% of the population). 7/n
Overall, this is fairly well done study demonstrating very high attack rates and high excess mortality in a single pandemic wave in a LMIC setting. Data from this study partly reflects emerging scenarios seen from serodata from other LMICs regions and LATAM. 8/n
Scenario 1: Final attack rates for COVID19 in these settings appear to be in the 60-80% range - see Barranquilla, Monteria, Delhi, Mumbai etc 9/n https://twitter.com/CovidSerology/status/1347930428754202625
Scenario 2: 30-40% seropositivity in these areas does not appear to be enough to prevent significant second waves in these settings. More examples of this phenomena is seen in Mumbai, Delhi etc 10/n https://twitter.com/CovidSerology/status/1351104889406025728
Scenario 3: excess deaths in LMI areas may run as high as expected ranges in older HIC. Some of this effect maybe due to lower survival rates during periods of overburdened healthcare capacity and contribution from indirect mortality from HC collapse 11/n https://twitter.com/CovidSerology/status/1350430989684596737
Examples of this third effect is seen in e.g Lima (4k), Iquitos (6K DPM), Mexico city (6K DPM), Guayas (4K DPM). Note: worse case scenario for ED in an unmitigated epidemic from the Imperial March report was ~8200 DPM for the much older US/UK pop 12/n https://twitter.com/CovidSerology/status/1351227178759892993
Source:
Excess deaths - https://www.datosabiertos.gob.pe/dataset/informaci%C3%B3n-de-fallecidos-del-sistema-inform%C3%A1tico-nacional-de-defunciones-sinadef-ministerio
13/n
Excess deaths - https://www.datosabiertos.gob.pe/dataset/informaci%C3%B3n-de-fallecidos-del-sistema-inform%C3%A1tico-nacional-de-defunciones-sinadef-ministerio
13/n
