Thread by @bhavani_md

Racial bias in temporal (forehead) thermometers🌡

Out @JAMA_current @JAMANetwork

ja.ma/3cLF53j

23% relative reduction in fevers detected in Black patients w/ temporal compared to oral thermometers

Another medical device that could be failing our patients

(1)

Cohort: All patients w/suspected infection🦠from 4 hospitals

Paired oral & temporal temperatures within 1 hour of each other🌡️

Eval of association between fever (>38°C) & route of measurement adjusting for age hospital & comorbidities in Black patients & White patients

(2)

Figure shows odds of fever w/ temporal thermometers (oral = reference) in Black & White patients

Horizontal line at 1.0 = no difference in fevers w/ temporal vs oral

@ varying fever cutoffs Black patients had⬇️odds of having fever detected w/ temporal thermometers

(3)

Why does this matter? Hospitals are structured around cutoffs linked to alerts

In the ED, 38°C cutoff➡️SIRS criteria➡️ #sepsis alert

In the wards, 38°C➡️physician notification

When there's significant % undetected fevers, this could lead to missed alerts & delays in care

(4)

Next steps are to find out

1. Is this inaccuracy present in other hospital systems?

2. Does this inaccuracy lead to delays in antibiotics?

If yes - let's switch to a different measurement site until the devices are further evaluated tested and fixed

(5)

Thanks to our team @EmoryDeptofMed @EmoryMedicine
Zanthia Wiley Craig Coopersmith @IghoID @DrPhilipVerhoef

Our work follows #PulseOx studies that uncovered medical device bias thanks to Dr.Tobin @iwashyna @msjoding @tsvalley @v_valbuen @AshFawzyMD @AIanWongMDPhD & others

(6)