When the Sick Note Becomes a KPI: Rethinking Lost Time Injury Classification

Joe hurts his back moving heavy equipment in a maintenance workshop. He reports it, sees his doctor, and comes back with medication, a few physiotherapy sessions and three days off. Clinically, it is unremarkable. Statistically, it is a Lost Time Injury, and that is where the trouble starts.

Management sees a damaged safety KPI and, with it, reduced performance bonuses. The Risk Manager is told to "do something about it". If there is an occupational health clinic on site, the pressure lands there next: overrule the sick note, or phone the treating doctor and negotiate an immediate return to work. Swap Joe's back for a cut finger or a fractured wrist and the same story plays out in workplaces across South Africa every week.

The problem with the LTI

Employers are rightly held legally and ethically accountable for a healthy and safe workplace, and that performance has to be measured somehow. The Lost Time Injury (LTI) rate remains the best known of those measures because it looks objective, is easy to count and can be compared across business units.

It rests on a flawed assumption: that the severity of an injury, and therefore of the incident, is reliably proportional to the time off the treating doctor prescribes. It is not. The length of a sick note is shaped by the patient's personality and expectations, the doctor's style of practice, the physical demands of the job, and even whether the employee can get transport to work. A sick note is a clinical judgement about one patient, not a repeatable, objective measure of how serious an incident was.

Yet the LTI endures. It is simple and binary, and it hands the judgement to an outsider who appears impartial. Meanwhile, the figures are reported upwards all the way to the board and shareholders, and the pressure to keep them looking good flows straight back down.

Why occupational health must not overrule the sick note

When that pressure reaches the OH clinic, the instruction is sometimes blunt: overrule the prescribed sick leave. There are four good reasons to refuse.

1. Scope of practice. An occupational health nurse may not overrule a medical practitioner's recommendations about a patient's care.
2. Supersession. Even a doctor may not simply take over a case and overrule the treating practitioner without first consulting them. The HPCSA prohibits this.
3. Legal exposure. The COID medical report is a legal document that states the employee's fitness for work. The employer may disagree with it, but it stands. Bringing an employee back against that certification can breach the employer's duty under section 8 of the OHS Act to provide a working environment that is safe and without risk to health, which is a criminal offence under section 38(1). If the employee is then injured and the decision was negligent, section 38(2) raises the penalty to a fine of up to R100 000 or two years' imprisonment. Under section 37, the manager or practitioner who gave the instruction can be charged alongside the employer.
4. Ethics. Placing a company statistic above an employee's wellbeing is ethically indefensible, and a complaint to the HPCSA could put the practitioner's registration at risk.

The damage goes beyond professional risk. When employees feel that a statistic matters more than their wellbeing, their view of the company and its values suffers. Over the years, many employees have told me they felt the company did not care and saw them as just a number. And the OH team, caught in the crossfire, can find its relationship with management seriously, sometimes permanently, damaged.

A better approach: classify on impairment, not absence

The answer is not to fight the treating doctor. It is to change how injuries are classified.

• The treating doctor follows a clinical treatment plan without interference from the employer.
• An appointed medical practitioner, typically an experienced occupational medicine practitioner (OMP), assembles the medical and non-medical facts and proposes a classification against pre-agreed criteria.
• Severity is based on the degree of impairment and on how far that impairment affects the core tasks of the job: in other words, what task restrictions apply. Ideally the nature of the incident is factored in too, although that is harder to quantify.
• Sick leave and other treatment decisions are taken into account, but they no longer define the classification.

The employer keeps a familiar lagging indicator. The process is recorded, auditable and as objective as reasonably possible. Employees and their doctors are not pushed into compromised care, and the company is not seen as uncaring.

The seven-step classification process

For each case, the appointed OMP works through and records seven steps:

1. Mechanism of injury. What happened, and the forces involved that caused the injury.
2. Diagnosis. The clinical picture, drawn from the COID medical report (diagnosis and treatment), the clinical assessment by the employer's appointed medical examiner, and any pre-existing injury to the same part of the body.
3. Impairment. The loss of function or anatomical damage: what limits the injury and its treatment place on movement, strength, endurance, flexibility or cognition. As objective as possible, with subjective input where appropriate.
4. Incapacity. How that impairment affects the employee's ability to do the job, based on a careful look at the tasks required and the job description.
5. Reasonable accommodation. The task restrictions that must be accommodated for the employee to return to work, and whether they are reasonable. A useful test: what percentage of the core job functions can the employee still perform?
6. Classification rationale. The proposed injury-on-duty (IOD) category, for example "Medical Treatment Case (minor 2)", with the supporting clinical explanation.
7. Final classification. The final IOD category, recorded for reporting.

Applied consistently, this gives management a severity measure it can defend, and takes the treating doctor's sick note out of the firing line.

A note on other jurisdictions

The seven-step process works well in South Africa and in jurisdictions with similar arrangements, where occupational health staff have legal and ethical access to the injured employee's COID medical reports. That access is what allows the OMP to assemble the clinical facts and base the classification on evidence rather than on the length of a sick note.

Where the law prohibits employers or their occupational health providers from seeing these records, the process is significantly constrained. The principles still hold, but the clinical information has to come another way: with the employee's informed consent to share specific details, through a functional assessment by the employer's own medical examiner, or from a fitness report by the treating doctor that states restrictions without diagnostic detail. Before adopting this model elsewhere, check what local law allows occupational health to see, and design the process around it.