OSHA's Severe Injury Clock Starts When a Site Supervisor Finds Out. Your Escalation Policy Starts When Head Office Does.

Most multi-site employers have a serious incident escalation policy. It usually reads well. A site that has a serious injury notifies its regional manager, the region notifies the corporate safety or risk function, legal is looped in, and someone at head office decides whether and how to report to OSHA. The logic is sound on its face: reporting to a federal agency is a consequential act, and a company wants one consistent voice doing it.

The problem is that OSHA's severe injury reporting rule doesn't measure time from when head office knows. It measures from when the employer knows, and it counts knowledge held by the employer's agents. At a single site, that distinction barely matters. Across a portfolio, where the first person to learn of an amputation on a Saturday night shift is a supervisor three reporting layers away from whoever is authorized to call OSHA, it is the whole question.

What 1904.39 requires, and when the clock starts

29 CFR 1904.39(a) sets two deadlines. A work-related fatality must be reported to OSHA within eight hours. An in-patient hospitalization of one or more employees, an amputation, or the loss of an eye, as a result of a work-related incident, must be reported within twenty-four hours. The report can be made by phone or in person to the OSHA Area Office nearest the site of the incident, by phone to OSHA's central toll-free number, or through the online reporting application on osha.gov. If the Area Office is closed, (b)(1) rules out voicemail, fax, and email to that office; it has to be the toll-free number or the online application.

The timing provisions are in paragraph (b). Under (b)(7), if you don't learn of a reportable event when it happens, the eight- or twenty-four-hour period runs from when the event is reported to you "or to any of your agent(s)." Under (b)(8), if you don't learn right away that the event was work-related, the period runs from when you or any of your agents learn that it was. The regulation does not define agent within this section, but the plain effect of the phrase is that the clock does not wait for the specific person or function the company has chosen to make the call. For practical planning, it is safest to assume that a site manager or supervisor who learns of the event has started it.

That is the gap. An escalation policy that gives the site twenty-four hours to notify corporate, and corporate some further period to evaluate, can consume the entire federal window before the one person authorized to report has even seen the facts.

Why the gap is a portfolio problem

A single-location employer usually has one or two people who would hear about a hospitalization, and they are often the same people who would report it. At scale, three things change.

First, the chain gets longer. Every layer an escalation policy adds is a handoff that can stall overnight, over a weekend, or when the next person in the chain is on leave. The regulation makes no allowance for any of those, and the rule's own Area-Office-closed provision shows it anticipates events occurring outside business hours.

Second, the facts arrive in pieces and at different sites. 1904.39(b)(6) says a hospitalization, amputation, or loss of an eye is only reportable if it occurs within twenty-four hours of the work-related incident, while a fatality is reportable if it occurs within thirty days. A worker taken to the emergency room for evaluation may be admitted hours later. Under (b)(9) and (b)(10), in-patient hospitalization means formal admission for care or treatment, and admission only for observation or diagnostic testing is not reportable. Those are exactly the details a site learns first, by a phone call from a family member or a supervisor at the hospital, and exactly the details that don't travel well through three layers of email.

Third, sites classify differently. The definition of amputation in (b)(11) is broader than many people assume. It includes fingertip amputations with or without bone loss, partial amputations, and body parts that were later reattached. It excludes avulsions, deglovings, and chipped teeth. One site's supervisor may call a partial fingertip injury a laceration and never escalate it; another may escalate everything. If the corporate reporting decision depends on how each site characterizes the injury on the first call, the portfolio's reporting behavior will vary by site, whatever the policy says.

The exceptions that sites apply inconsistently

The rule has a small number of carve-outs, and each is a place where sites reach different answers without anyone noticing.

Under (b)(3), a motor vehicle accident on a public street or highway is not reportable to OSHA, unless it occurred in a construction work zone, in which case it is. Operations with both road fleets and construction or roadside work need their sites to know that distinction, not just the general "vehicle accidents don't count" version. Under (b)(4), events occurring on a commercial or public transportation system, such as an airplane, train, subway, or bus, are not reportable. Under (b)(5), a work-related fatality or in-patient hospitalization caused by a heart attack must be reported; the Area Office decides whether to investigate. In every one of these carve-outs, the rule notes that the event may still need to be recorded on the injury and illness log, if the establishment is required to keep one.

The heart attack provision is worth singling out, because it runs against intuition. A site manager deciding whether an event is "a work injury" may reasonably assume a cardiac event isn't. The rule doesn't let that assumption end the reporting analysis.

What the report contains, and why it is site-specific

1904.39(b)(2) lists what OSHA needs: the establishment name, the location and time of the incident, the type of reportable event, the number and names of affected employees, a contact person and phone number, and a brief description. Almost all of that lives at the site. The contact person OSHA calls back will, in practice, need to be someone who can speak to the conditions at that address.

This is also why the reporting destination is local. The Area Office nearest the site of the incident is the one that receives phone and in-person reports, and depending on the event, OSHA may open an inspection or ask the employer to investigate and respond. For a portfolio in several states, add the state plans: states operating their own OSHA-approved programs can set reporting rules that differ from the federal one, and a single national escalation timeline may not match each state's requirement. The corporate procedure has to be checked against the rule that applies at each address.

Building an escalation policy that fits inside the window

None of this argues against central oversight of OSHA reporting. It argues for designing the escalation so that it runs in parallel with the regulatory clock rather than ahead of it.

That starts with naming, at each site, who is authorized to make the report if head office cannot be reached promptly, with that person's alternates covering every shift the site operates. It continues with a decision aid at the site level that uses the regulation's own definitions of in-patient hospitalization and amputation, so that the facts are characterized the same way at every location before anyone escalates. It helps to have the time the site first learned of the event, and the time it learned the event was work-related, captured as fields in the incident record, because those are the two times (b)(7) and (b)(8) actually turn on.

At the portfolio level, the useful review is not whether each site has the escalation policy. It is, for each reportable event in the past year, how long elapsed between the first time anyone at the site knew and the time OSHA was notified. A site that consistently reports late, or a region where events are consistently characterized in a way that avoids reporting, shows up in that comparison and nowhere else.