One Post-Injury Drug Testing Policy, Many Sites: What 1904.35(b)(1)(iv) Asks of a Rule Written at Headquarters

The corporate policy is one paragraph long and has been in the handbook for years: any employee involved in a workplace injury or incident is tested for drugs and alcohol. It reads as cautious. It treats every site the same. It gives the company a clean answer if an insurer or a plaintiff's attorney ever asks what it did after an accident.

At site level, the same paragraph can behave quite differently. At one plant it is applied after every first-aid visit. At another it is applied only after recordable cases. At a third, a supervisor quietly decides that an injury the worker "didn't really need to report" is not worth the paperwork, because reporting it means a test. Nobody at head office sees any of this, because the policy is uniform on paper. The variation lives in how each location applies it, and in what employees at each location have learned about the cost of speaking up.

This post is about one provision of OSHA's recordkeeping rule and what it means for a policy that is written once and carried out in many places. The question here is the one that only arises at scale.

What the rule says, and what it was written to prevent

Section 1904.35 of 29 CFR requires employers to set up a reasonable procedure for employees to report work-related injuries and illnesses promptly and accurately. Under 1904.35(b)(1)(i), a procedure is not reasonable if it would deter or discourage a reasonable employee from accurately reporting. The companion provision, 1904.35(b)(1)(iv), states that an employer must not discharge or in any manner discriminate against any employee for reporting a work-related injury or illness.

Two other subparts in the same paragraph matter for a multi-site operator. Section 1904.35(b)(1)(ii) requires the employer to inform each employee of the reporting procedure, and 1904.35(b)(1)(iii) requires the employer to tell each employee that they have a right to report and that the employer will not retaliate. Those are communication duties, and in a large organization they are a distribution problem: whether every site's orientation material, posters, and supervisor briefings say the same thing is an ordinary audit question.

The rule does not mention drug testing. It does not prohibit it. The connection came from OSHA's own explanation when it adopted the provision: the agency said that blanket post-injury testing policies could deter reporting, and that testing intended as discipline for an injury, rather than as a way to evaluate what caused it, was the problem. OSHA later issued a clarification that most workplace drug testing is permissible under the rule, including testing under a state workers' compensation law, testing required by another federal law, and testing to evaluate the root cause of an incident, so long as the employer is testing to find out what happened and not to penalize reporting. Read both statements together and the law is neither "never test" nor "test everyone." It asks what the test is for and whether a reasonable employee would be put off by it.

Why a uniform policy is the harder case

For one location, the analysis is manageable. A safety manager can look at the testing practice, ask whether it is tied to the facts of the incident, and adjust. For a company with dozens of sites, there are three compounding problems.

The policy is identical, the application is not. A rule that says "all incidents" will be read by each site's management through its own habits. If one plant tests after everything and another tests after serious events, the first plant has a reporting-deterrence problem that the second does not, even though both are following the same text. The wording carries the exposure, but the pattern at each site decides whether an investigator sees a deterrent.

Reporting data can look better than reality. A site where employees have learned that a reported injury means a test may have an excellent injury rate. Corporate sees a low number and may even reward the site for it, which is the same dynamic that makes incentive programs a concern under the same rule. Our post on recordable classification consistency across sites explains how a single inconsistent judgment skews a rollup; a deterrent to reporting skews it in the same direction and with less visibility.

State and contract requirements differ. Some sites test because a state workers' compensation program rewards it. Some are covered by customer or contract requirements. Some employees are subject to federal drug and alcohol testing rules through their job duties. State law varies on what an employer may require, on notice, and on the use of results. A headquarters policy that ignores this variation is probably over-broad in some places and under-compliant in others. Which states do what is a legal question for counsel, and this article does not try to answer it. The point for the person accountable for safety is that one paragraph is not likely to be correct everywhere.

What the pattern looks like to an investigator

Complaints under the rule are typically raised through section 11(c) of the Act, which OSHA enforces on an individual basis. An employee has a limited window to file, and the investigation turns on facts, not on policy text: what the employee reported, what happened next, and whether the employer's action followed because of the report.

In that setting, the handbook paragraph is only the beginning. An investigator is likely to look at who was tested, when, and why. Was the test tied to a reasonable suspicion or to the nature of the incident, such as a vehicle collision or a malfunctioning machine with no obvious cause? Or was it a reflex applied to a worker with a minor cut? Was the same thing done to every involved employee, including the supervisor who set up the task? And what happened to the employee afterward?

An employer that can show the testing decision came from an assessment of the incident, and that the same assessment was applied consistently, is in a different position from one that can show only a rule that says "everyone." Consistency across sites helps here, but only if it is consistency of judgment, not merely consistency of text. Our look at whistleblower complaints and section 11(c) exposure across sites covers what happens once a complaint exists.

What the person accountable for safety can do

None of the following requires legal training or an EHS department. It requires asking the question once, centrally, and getting an answer from each location.

First, find out what each site actually does. Ask for the last several incidents and, for each, whether a test was done, who decided, and what the stated reason was. If sites cannot answer, that is the finding.

Second, separate the testing triggers. A policy that distinguishes testing required by law or contract, testing prompted by specific indicators, and testing used to investigate root cause is easier to defend than one that tests everybody. Where a site is subject to a specific requirement, the policy should say so rather than extend it everywhere.

Third, put the reason in the investigation record. If a test is part of understanding what happened, it belongs in the same file as the incident investigation, with the facts that prompted it. A test with no recorded reason looks like a punishment.

Fourth, check the reporting messages. Read the orientation material and the supervisor talking points at several sites against 1904.35(b)(1)(ii) and (iii). A policy that is supportive in the handbook but intimidating at the morning huddle is the pattern to look for.

Finally, look at the numbers differently. A site with unusually few reported injuries is not automatically a success. A reporting climate that is working usually produces some near-misses and minor cases. Silence is a signal worth a conversation, not a trophy.

The honest limit of a central policy

A headquarters document can set a floor and a purpose, but it cannot make a supervisor at a remote location decide fairly on a bad Friday. What it can do is make the purpose explicit, narrow the triggers, assign who decides, and give the corporate function a way to see what is happening. That moves the question from "do we have a policy" to "is the policy doing what we say it does," which is where an investigator will begin too.

This article is general information, not legal advice. Testing rules differ by state, contract, and job role, so have counsel review any policy before relying on it.