What We Can Conclude—and What We Still Need to Test

This series began with a familiar suspicion: after enough difficult medical judgments, a professional may start choosing the easier path. It ends with a more disciplined question about healthcare professionals making sequential decisions: when a later-day or later-in-sequence pattern appears, what can we say about cumulative effort, and what must remain uncertain? The strongest answer is neither that decision fatigue is obvious nor that fatigue is irrelevant. Time of day is not proof; the tested domain-general behavioral effect was not found in the Swedish triage setting; and a broader, more conditional research question remains open.

The conclusion is narrower—and stronger—than the slogan

The evidence supports a layered judgment.

At the level of experience, sustained medical work can be demanding. The mechanism proposed in the field study is coherent: medical judgment requires retrieving and interpreting information, weighing trade-offs, and considering possible outcomes; prolonged effort may make further information processing more costly; and a decision-maker might then rely more on prior beliefs, familiar defaults, or simpler options. That is a plausible account of how fatigue could affect sequential judgment.

At the level of observation, healthcare decisions can vary across a workday or a sequence. The large telephone-triage dataset showed that the highest urgency rating appeared in 14.8% of calls around 08:00, 19.3% around 18:00, and 21.1% around midnight. A pattern like that can matter operationally. It may change staffing needs, referral flows, or the experience of patients seeking advice. But it remains a pattern in the distribution of decisions, not a diagnosis of its cause.

At the level of causal evidence, time of day is insufficient. The clock also carries changes in patient mix, external service opening hours, shift composition, staffing, and organizational routine. Call order is entangled with the clock rather than an independent confirmation. A later-day urgency increase may be partly or wholly caused by those conditions. The observation can be genuine while the fatigue explanation remains unproven.

The systematic review expands the picture without removing the uncertainty. It searched eight databases, screened 14,740 records, and included 82 studies: 72 quantitative, one qualitative, one review, and eight expert discussions. Its narrative synthesis reported significant decision-fatigue effects in 45% of the cases that quantitatively assessed the hypothesis, across diagnostic, test-ordering, prescribing, and therapeutic decisions. That is enough evidence to take the question seriously. It is not a pooled effect size, a universal probability, or proof that all of those studies measured one mechanism.

The review’s most consequential finding may be its conceptual warning. Decision fatigue was inconsistently defined and inadequately operationalised. The same label can cover time since a break, position in a sequence, a prescribing outcome, a diagnostic choice, or a professional’s account of experience. A positive result in each setting does not automatically add up to one domain-general effect. A number of positive studies can indicate a recurring phenomenon, a collection of context-specific effects, or a collection of confounded proxies. The evidence does not decide among those possibilities without closer study.

The registered field test supplies that closer challenge for one important version of the claim. It examined 231,076 calls involving 174 specialized nurses in a national Swedish telephone-triage service. It compared morning and afternoon workers in a shared 13:00–17:00 overlap and compared intense call sequences before and after an 8–40 minute break. It specified two behavioral signatures in advance: greater alignment with personal defaults and higher urgency ratings under greater cumulative effort.

Neither signature appeared in the predicted direction.