Thread by @apresj20

🔴More than two weeks after our press release regarding the serious statistical and methodological errors of the @JAMAInternalMed #LongCovid study, the authors have still not answered our questions.
We publish a summary of the flaws of the study: tinyurl.com/apresj20ensynd…
⬇️🧵

1- The study is presented as robust because it is based on a large cohort, but it is very unbalanced with very few individuals concerned by Covid-19!

The study covers 27,000 people, BUT less than 6% of the cohort (~1,500 people) were somehow affected by Covid-19, either by a positive serological test or by a reported contamination, limiting the scope of the statistical results of the study.

2- The study’s main hypothesis is biased: “We hypothesized that the belief in having been infected with SARS-CoV-2 would be associated with persistent symptoms while controlling for actual infection.”

➡️ Our objection: serological tests cannot be used to identify infections,
because they have too much uncertainty.

2.1- Not taken into account: large proportion of false positives in the serological tests. The imbalance of the cohort results in 59% false positives among the 1091 positive serologies, yet considered proven infections.

➡️ By dilution, the statistical results of the positive serologies have little relation to Covid-19!

2.2- Not taken into account: large proportion of negative serologies among Covid19 patients
● serological tests are not totally reliable despite presence of antibodies (87%)
● 25% of patients have no detectable antibodies & this rate increases over time as antibodies disappear

➡️SARS-COV-2 may have actually infected a significant proportion of reported Covid-19 patients who are seronegative, with the rate of negative serology among reported patients (50%) being close to that expected for proven patients (>34%).

3- The proportion of anosmia (loss of smell), which is very specific to Covid-19, is very high in
those who report infection, regardless of serology result.
(25 times higher than the general population)

➡️This confirms that a large proportion of persons "believing" they were infected were indeed infected with SARS-COV-2.
➡️The statistical relationship between "belief" and 16 prolonged symptoms seems related to Covid-19.

In the last few days, some of the study’s authors have appeared in the French media, responding to errors which experts have pointed out. Their response raises questions about the principles of scientific validation of data.

Our association, #ApresJ20 Covid Long France, calls for medical care integrating scientific evidence with the lived experience of patients

🔴Let’s be clear: this flawed study should be retracted🔴

Here is our complete press release in english tinyurl.com/apresj20pressr…
#LongCovid

#FlawedJAMAStudy
@EmmanuelMacron
@HAS_sante
@hcsp_fr
@Sante_Gouv
@olivierveran
@lescure_xavier
@MichelZumkeller
@JulienBorowczyk
@MIRALLESMP
@TrastourIsnart
@VillaniCedric
@Assur_Maladie
@Tuffnell17
@mnlienemann
@agenceANRS
@Inserm
@institutpasteur
@SPILF_
@FRM_officiel

@Dr2NisreenAlwan
@elisaperego78
@MartaEsperti
@LongCovidSOS
@long_covid
@ClaireHastie1
@LongCovidKids
@LongCovidScot
@LongCovidWales
@LongCovidCH
@LongCOVIDPhysio
@ahandvanish
@patientled
@selkieO
@LongCovidOT
@itsbodypolitic
@ShaneyWright
@SessionsTlc
@Survivor_Corps

@14daystochange
@UTc19LongHaul
@Vickyvdtogt
@fi_lowenstein
@ZeroCovAlliance
@LongCOVIDEurope
@Tortillou
@touretteturgis
@ClareRayner6
@GreeceLck
@LongCovidRo
@co19longcours
@postCOVIDbe
@hospiphilo
@longcovidirl
@NoLongCovid
@katemeredithp
@K_Bishof
@LCKScotland @stillillcorona1

@JMarieRenaudin
@lilystbarth
@HirschEmmanuel
@grafffio
@Cote_Science
@LehmannDrC
@Le___Doc
@barriere_dr
@DocStrain
@DocArnica
@Curiolog
@james__glen
@Kelsveil
@covidpersisarg
@PRicoLongCovid
@LongCovidNews
@fitterhappierAJ
@Carrie_McGinn
@LongCovidPaper

@longcovidadvoc1
@LongCovidNews
@AcpeCovid
@WesElyMD
@doctorasadkhan
@calirunnerdoc
@keyeri
@darrenabrown
@PutrinoLab
@Daltmann10
@EricSALATprof
@nathanpsmad
@C_A_Gustave
@Sonic_urticant
@BorrazOlivier
@AlexSamTG
@longcovidbr
@ClareRayner6
@LongCovidOT
@callumjodwyer

@Jerome_Larche
@GillesThoni
@longcovidwork
@MillionsMissgFr
@GilbertDeray
@Ecole_Oubliee
@lesleymacniven
@maxwele2
@DrEricDing
@grafffio
@boylandry
@ShaneyWright
@AVervialle
@Fr_Assos_Sante
@WHO
@diazjv
@mvankerkhove
@DrTedros
@profhelenward
@Savolainen_J
@keyeri

#FlawedJAMAStudy

@JAMAInternalMed
@APHP
@MartinHirsch
@lemondefr
@BFMTV
@LCI
@24hPujadas
@LCP
@pascale_santi
@sfoucart
@leQdM
@LaCroix
@jeanneferney