Heinrich's Triangle is... well, a triangle among other things, however I admit that answer wouldn't earn you much respect if asked in a safety meeting.
The triangle is designed to show the ratios in which unsafe acts, near misses, accidents, fatalities and catastrophes occur, and was designed to be non-specific to any industry. That being said, one can certainly question how the same ratios could possible apply in both an office and an oil rig.
Catastrophic Event- Negligible
400 Reportable Injuries
20,000 Minor Injuries
240,000 Near Misses
2,000,000 Unsafe Acts
At the top of the triangle sits Catastrophic Event, occurring at an unspecified frequency of less than a single fatality. The rest of the triangle lack a degree of specificity too; what is considered a major injury? An unsafe act? A near miss?
A rise of near miss reporting in an organisation could skew the figures, although one could argue that an organisation with a greater amount of near-miss reporting is as such, a safer organisation, would the ratio remain constant throughout?
A final note of criticism is that data could be plugged into the triangle and reviewed by a manager without the slimmest idea of how probability works, who would then assume the organisation wasn't due a fatality until whole bunch more minor injuries occured. But then again, academic models can't necessarily account for idiocy.