MFOM Written common mistakes
MFOM Written · Understand the traps and how to avoid them.
Writing the management plan as if the question were a clinical case for the GP - investigations, drug treatment, referral - with no statement of fitness for work, restrictions, timescales or workplace adjustments.
Medical-school conditioning pushes candidates to diagnose-and-treat; occupational advice feels obvious or optional, and under time pressure they write the clinical paragraph first and run out of space.
What to do instead
Make an occupational paragraph compulsory in every scenario answer: fitness conclusion (fit / fit with restrictions / not fit), the specific restrictions and their duration, adjustments or temporary redeployment, review date, who will be told and with what consent, and any statutory notification required.
Taking a generic medical history and never asking about job title, tasks, shift pattern, substances handled, control measures, PPE, second jobs or previous exposures - then attributing symptoms to a single agent named in the stem.
Occupation is not a standard prompt in clinical history templates, and candidates assume the question stem already supplies all the exposure information they need.
What to do instead
Use a fixed occupational history checklist (job and tasks, duration, agents, route of entry, controls/PPE, hygiene facilities, workmates affected, hobbies and second employment, prior exposures) and then state explicitly which exposures the history does and does not support.
Treating a UK Workplace Exposure Limit as the boundary between safe and unsafe, or mixing the 8-hour TWA WEL with the 15-minute STEL, or quoting US TLV/PEL values in a UK COSHH question.
WELs look like thresholds and are memorised as bare numbers without their averaging time or legal status.
What to do instead
Always quote a limit with its averaging time and status - 'WEL, 8-hour TWA, set for control and not a safe/health-based threshold, applies to inhalation' - and pair every numerical limit with the COSHH hierarchy of control and, where appropriate, biological monitoring.
Assuming inhalation is the only relevant route, or concluding from a normal air sample or normal biological monitoring result that exposure is adequately controlled.
Air monitoring dominates teaching and produces reassuring numbers; dermal absorption, ingestion, sampling timing and non-occupational sources are easily overlooked.
What to do instead
For each agent state the relevant routes (inhalation, dermal, ingestion) and the interpretation limits of biological monitoring - sampling time relative to shift, half-life, non-occupational sources such as diet or smoking - before commenting on adequacy of control.
Writing an occupational health report that names the diagnosis or discloses clinical detail to the employer without consent, or copying the full GP/specialist letter into the management report.
The clinical habit of full disclosure plus a wish to be helpful to managers; the split between the confidential clinical record and the fitness-for-work advice is not intuitive.
What to do instead
Apply the rule that the employer receives fitness for work, functional restrictions/capabilities, timescales and adjustments only, with diagnosis disclosed solely where the worker has specifically consented; record consent status, who received the report and the worker's right to see it first.
Getting statutory reporting wrong - missing a RIDDOR-reportable event, or giving the wrong category or deadline (over-7-day versus specified injury versus dangerous occurrence, occupational disease reporting) - or citing non-UK legislation such as OSHA, ADA or FMLA in a UK law question.
RIDDOR categories and time limits are detail-heavy and rarely used day to day by trainees, and international reading sources leak into answers.
What to do instead
Learn the RIDDOR core set (deaths, specified injuries, over-7-day injuries, reportable occupational diseases, dangerous occurrences, gas incidents) with the responsible person and reporting route, and anchor every law answer to the UK statute named or implied in the stem (HSWA 1974, COSHH, RIDDOR 2013, Equality Act 2010).
Declaring a worker 'fit' or 'unfit' with no analysis of whether the impairment meets the Equality Act 2010 definition and no consideration of reasonable adjustments.
The biomedical fit/unfit binary is simpler and safer-feeling, and candidates forget that the duty to make adjustments is anticipatory and is not triggered by a diagnosis as such.
What to do instead
Structure the answer in three steps: does the impairment meet s.6 (substantial, long-term of at least 12 months, effect on day-to-day activities, including progressive and past disabilities); what reasonable adjustments (physical, organisational, hours, equipment, redeployment) would remove the disadvantage; then the fitness conclusion with restrictions.
Confusing ill-health retirement criteria with sickness absence or capability processes, for example asserting 'permanent incapacity for all work' when the pension scheme test is incapacity for the member's own role, or vice versa.
Several overlapping tests exist (pension scheme regulations, permanent health insurance, employer capability policy, fit note) with different definitions and durations.
What to do instead
Name the specific test and decision-maker: pension scheme regulations and the scheme's medical adviser for ill-health retirement, the employer's capability procedure for capability dismissal, the fit note for short-term absence - and state the duration criterion and that reasonable adjustments have been considered.
Mixing up epidemiological measures and designs - calling a risk ratio an odds ratio, calling prevalence an incidence, or judging a screening test by sensitivity alone.
Terminology is memorised as labels, and short written answers invite the first familiar-sounding statistic rather than the one the design actually yields.
What to do instead
Identify the design first and then the measure it produces (cohort: incidence, risk ratio, attributable risk; case-control: odds ratio; cross-sectional: prevalence), and for any screening or case-finding question build a 2×2 table and calculate PPV from the stated prevalence rather than assuming high sensitivity gives high predictive value.
Reading an occupational epidemiology result at face value - accepting a low SMR as proof an exposure is safe, or blaming recent exposure for a current cancer excess - without discussing healthy worker effect, survivor bias, latency or confounding.
Working populations genuinely show lower mortality than the general population, which looks reassuring, and candidates seldom volunteer bias unprompted.
What to do instead
For each study state the likely direction of bias - healthy worker hire and survivor effects pull SMRs down, latency means recent exposure cannot explain a current cancer excess, and the general population may be the wrong comparator - then distinguish confounding from effect modification before drawing a conclusion.