A hemosiderin rim alone does not meet the definition of hemorrhage used in the individual patient data meta-analysis, which requires acute or subacute symptoms with evidence of recent hemorrhage, but presentation with a new focal neurological deficit from a brainstem lesion placed this patient in its highest-risk category [2]. Risk is also front-loaded. In the whole cohort, the annual risk of hemorrhage fell from 6.2% in the first year after diagnosis to 2.0% in the fifth, a decline confined to patients presenting with hemorrhage or a focal deficit [2], and a population-based cohort showed a similar decline in the risk of recurrence [6]. Restrictions justified by the consequences of recurrence should therefore be time-limited and linked to reassessment. Conservative management was consistent with the current consensus guideline, which considers it reasonable after a first symptomatic hemorrhage from a brainstem lesion and suggests that resection be considered after a second symptomatic hemorrhage or, in selected cases, a single disabling hemorrhage [7].

Assuming a constant hazard, a five-year risk of 30.8% corresponds to an annual risk of approximately 7% and to roughly one symptomatic hemorrhage per 15,000 eight-hour periods. Because risk is front-loaded, the early risk exceeds this average, and not every hemorrhage would cause sudden incapacity during a task. Aviation medicine has traditionally accepted an annual risk of sudden incapacitation of 1% for multicrew airline pilots, and relaxation of this limit to 2% has been proposed [8]. By that standard, the risk in this case is not negligible, but the relevant question is whether the task is performed with immediate backup.

The specialist recommendation to limit shift work, physical exertion, and stress was incorporated into the decision, but following it does not imply accepting a causal mechanism that the literature has not established. The consensus guideline notes that no rigorous studies have examined the risks of physical activity [7], and occupational exposures are not among the established predictors of hemorrhage [2,7]. A restriction presented as hemorrhage prevention makes an empirical claim that is difficult to defend if challenged and difficult to revise rationally.

The restrictions rest instead on three defensible grounds, summarized with their review criteria in Table 1. The first is the risk to others, which is inherent in the definition of fitness for work [3]. A recurrent brainstem deficit, such as dysphagia or upper limb dysfunction, during emergency perioperative care, where substitution and task interruption are limited, could compromise both the worker's safety and patient care. This ground explains why elective perioperative work was not excluded and why the restriction should be relaxed as event-free time accumulates. The second is documented residual functional limitation. Right-hand fatigability is directly relevant to medication preparation and airway assistance, but it was self-reported and should be measured objectively at the next review. The third is an individualized rationale for avoiding night work, which is not hemorrhage prevention. Psychomotor vigilance is worse at the end of night shifts than during day shifts [9], and sleep disturbance is a recognized migraine trigger [10]. In shift-working nurses, migraine was associated with shift work disorder and insomnia but not with the work schedule or the number of night shifts [11]. Night work could therefore aggravate migraine in a worker who tolerates disrupted sleep poorly and obscure a change in headache pattern, which was the sentinel symptom in this case. Portuguese law provides that night workers with health problems related to night work should, whenever possible, be transferred to suitable day work [4]. Although she was not a night worker, this provision supports recording a health-based rationale for daytime work. Because she already worked days under a parental entitlement that ends when the child turns 12, the restriction gave a clinical basis to an existing arrangement rather than creating a new accommodation.

Table 1. Restrictions, the grounds on which they rest, and criteria for review^a.