916,404 records: heat fills the ER in a dozen places, not just the cooling bay

The multiplier that will travel from this paper is the more than 30-fold rise in accidental gunshot wounds above the threshold ^[6], and it is close to useless for anyone sizing a shift. Large ratios sit on small baselines, and the reporting gives multiples without per-category counts, so the added visits behind each figure cannot be recovered from what has been published [c19b]. What can be worked out is the scale of the denominator: 916,404 visits across 2011 to 2023 ^[1] averages about 70,500 a year, roughly 193 a day ^[19]. A common category that doubles off that kind of base moves more beds than a rare one that multiplies by thirty.

The harder problem is that the rising categories do not share a destination. Kidney and renal complaints, skin problems, varicose veins, cannabinoid-related mental health presentations and accidental gunshot wounds ^[5] land in nephrology, dermatology, vascular care, psychiatric holding and a trauma bay. They do not draw on the same staff, the same rooms or the same consult lines. A hospital that answers a heat advisory with cooling capacity and one extra attending in fast-track has provisioned for the part of the curve that already had a code.

That coding point is the design of the study, not a footnote to it. The Northwestern group scanned all records for what became more frequent above 92.6 degrees Fahrenheit (33.7C) rather than searching for known heat diagnoses ^[8][2], which is how more than a dozen broad categories surfaced ^[3]. Dr. Nicholas Cozzi of the Illinois College of Emergency Physicians, who was not involved, put the corollary plainly: heat illness is undercounted when studies look only for explicit codes such as heat stroke ^[9]. Surveillance built on those codes is also the surveillance that justifies staffing budgets.

Two things should slow anyone reaching for the result as a planning input. Senior author Dr. Abel Kho says the work establishes association, not cause and effect ^[10], and his proposed mechanism, that physiologic stress worsens external injury risk, renal complications and drug effects ^[12], is a hypothesis the data are consistent with rather than a proven pathway. The cohort also over-represents Black patients and women relative to the general population ^[13], which University of Washington researcher Kris Ebi treats as a strength given how often those groups are missing from medical datasets ^[13].

Then there is the absence. Cardiovascular problems, the standard heat signal in other literature, do not appear in these results, according to the coverage ^[11]. Either Chicago's record-level picture differs from the mortality and admissions studies that made cardiac events the canonical marker, or a whole-record scan of ER and urgent care encounters is measuring a different layer of illness than those studies did. Both readings argue against importing another city's heat playbook, and one co-author, climate scientist Daniel Horton, frames the stakes as morbidity rather than deaths: the bad things that happen on the way up to dying ^[14].