Caught in machinery: the US severe injury and amputation numbers

The two OSHA documents cover different things and are easy to mix up. The seven-year lookback covers 2015 to 2021 and breaks injuries down by cause, which is where the caught-in-machinery figure lives. The annual report covers calendar 2024 by sector and rate, and does not repeat the cause breakdown. Neither is a superset of the other.

Coverage is the bigger caveat. These are reports submitted to FEDERAL OSHA, and the lookback states plainly that it covers approximately half of the US workforce. State-plan states run their own programs and are not fully included, so every count here is an undercount of the national total rather than a census.

Counts and rates move differently, and only one of them is comparable across years. Manufacturing amputations fell from a 1,438 annual average to 1,348 in 2024, and the severe-injury rate fell from 29.6 to 25.8 per 100,000 full-time equivalent workers. The rate is the honest comparison, because the count also moves with how many people are employed.

For the machine-guarding case, cite the cause breakdown: caught in running equipment or machinery was the leading cause of severe injury from 2015 to 2021 at 12,930 reports, 18% of the total, and manufacturing reported 69% of them.

For anything about a recent year, use the 2024 annual report and say so: 9,034 severe injury reports, 7,327 hospitalizations, 2,426 amputations.

Compare years with the rate, not the count. Manufacturing at 25.8 per 100,000 FTE in 2024 against a 29.6 nine-year average is a real decline; comparing raw counts confounds it with employment.

State the coverage limit whenever you quote a total. These are federal OSHA reports covering roughly half the US workforce, so they understate the national figure.

Presenting these as national totals. They are federal OSHA submissions covering approximately half the US workforce, not a full national count.

Comparing a count across years without the rate. Employment changes, so the rate is the comparable measure.

Mixing the two reports. The cause breakdown is 2015 to 2021 only; the 2024 report does not restate it, so there is no 2024 caught-in-machinery figure to quote.

Reading amputations and hospitalizations as separate populations. A single severe injury can be reported as both, so the categories overlap and cannot be summed.

Both figures sets come from OSHA's own published severe-injury reporting, retrieved as PDFs and confirmed to contain each figure quoted. Where a figure comes from a table rather than a sentence, we checked the table's column headings before using it, because a number read out of the wrong column is the easiest error to make in these documents.

We report the seven-year lookback and the 2024 annual report as two distinct sources rather than blending them, since they cover different periods and different breakdowns. Any figure on this page can be traced to exactly one of them.

We carry OSHA's own coverage caveat rather than burying it. The lookback states it covers approximately half of the US workforce, and no total here should be read as national.

We looked for a robot-specific severe-injury breakdown within this data and there is not one. OSHA's cause categories stop at 'caught in running equipment or machinery,' which includes robots without separating them, so this page cannot answer how many of these involved a robot.