What the Healthcare Evidence Actually Shows

The question now is larger than whether one later-day graph can be explained by fatigue. For healthcare professionals making sequential medical judgments, the 2025 systematic review asks how often research has reported decision-fatigue effects across diagnosis, testing, prescribing, and treatment—and whether those studies are even measuring the same construct. The useful conclusion is not a vote for or against an attractive idea; it is a scale-aware reading of what the evidence can support.

A broad map of a heterogeneous literature

A systematic review adds breadth that a memorable individual study cannot. One study can show an intriguing association in one service, specialty, or decision. A review can ask whether similar claims appear across settings and outcomes, how researchers have defined the phenomenon, and where the evidence repeatedly breaks down. It is therefore a way to map a field, not a machine that turns unlike studies into one definitive result.

The review was preregistered and searched eight databases. From 14,740 records, it included 82 studies: 72 quantitative studies, one qualitative study, one review, and eight expert discussions. The inclusion scope covered health professionals and trainees making medical decisions over time in healthcare contexts, using any design. That scope is broad enough to include several kinds of decision and several ways of estimating what “fatigue” might mean.

Study quality was assessed with the Mixed Methods Appraisal Tool or a relevant Joanna Briggs Institute checklist. That matters because inclusion is not the same as strong support. A study can be relevant to the question and still have weaknesses in its design, measurement, comparison, or interpretation. The review’s job is to bring the evidence into view while preserving those differences; the available indexed abstract does not give enough study-level detail to reconstruct every quality judgment.

The review’s narrative synthesis reported that 45% of cases that quantitatively assessed the decision-fatigue hypothesis provided evidence of significant effects. The reported effects appeared across diagnostic decisions, test ordering, prescribing, and therapeutic decisions. Expert discussions also confirmed that healthcare professionals recognize decision fatigue as an important phenomenon.

Those findings establish three things. First, decision fatigue is not an idea supported only by a single famous example. It has been investigated across many healthcare decisions. Second, positive-looking effects are common enough to warrant serious investigation rather than casual dismissal. Third, the literature is not uniformly positive: if 45% of the quantitatively assessing cases reported significant effects, the remaining cases did not provide that same positive result under the review’s synthesis.

The number is informative, but its meaning is narrower than it may sound. “Forty-five percent of cases” is not a pooled effect size. It is not the probability that a healthcare professional will make a worse decision after a given number of calls. It is not a measure of how large the average effect is, and it does not say that the same 45% would be positive under a common test. The review abstract does not provide the study-level estimates needed to rank the included studies by quality, magnitude, or clinical importance.

The denominator matters too.