During the most intense days of Chicago's July 1995 heat wave, the Cook County Medical Examiner certified 485 deaths as heat-related. A statistical comparison of that same week against normal mortality patterns found 739 excess deaths. Those two numbers measure different things. The ME determines whether heat killed a specific person. The statistician asks whether that person would have died that week without the heat. Roughly 254 people fell into the space between those questions. Their death certificates recorded the final mechanism, a cardiac arrest or respiratory failure, rather than the 106-degree afternoon that accelerated it. The CDC acknowledged the problem plainly: deaths where heat was a contributory cause "may not be recorded as heat-related." A standard definition was promised. Criteria were published in 1997. Adoption remained voluntary. Three decades later, every official heat-death toll still functions as a minimum estimate, and the gap between that floor and the real number determines how seriously any city treats the next July.
How the ME certified a heat death (1978–1995): Body temp ≥105°F at death, or circumstantial evidence (closed windows, no AC, high ambient temp), or decomposed remains with no other cause and last seen alive during the heat wave
What excess deaths capture that certificates don't: The statistical spike in cardiovascular and respiratory deaths during extreme heat, from people whose underlying conditions were accelerated by heat but whose paperwork never says so
485 vs. 465: Both are certified-death counts from the same heat wave; 485 covers the peak week (July 14–20), 465 covers a wider window (July 11–27)
No national standard adopted: Heat-death certification criteria remain voluntary. The official count from any heat event is a minimum estimate by construction