How does the number of decisions made in a row during a work shift affect the speed and consistency of later decisions?

Why does deciding all day make later choices harder?

The leading explanation is that sustained decision-making depletes a limited pool of mental resources needed for careful control. One account describes decision fatigue as a state in which the store of mental resources for decision-making is depleted, reducing the quality of subsequent decisions [2]. A more recent neuro-metabolic account proposes that cognitive control produces potentially toxic byproducts that accumulate in a brain region important for control, the lateral prefrontal cortex, and that the brain then adjusts to make further control more costly [3]. In that study, people who did high-demand cognitive work for most of a workday showed a shift toward short-delay, low-effort options, and at the end of the day they had higher levels of a brain chemical called glutamate in that control region compared with people who did low-demand work [3].

This does not mean every decision is equally draining. The neuro-metabolic account is a proposed mechanism, not a settled cause [3]. The nursing commentary frames decision fatigue as a temporary reduction in decision quality when cognitive resources are scarce, and it lists factors such as high workload, shift work longer than 12 hours, and complex or stressful conditions as contributors [2]. So the number of decisions in a row matters, but it appears to interact with how demanding each decision is and the conditions under which it is made.

What actually changes later in a shift: speed, consistency, or both?

In emergency departments, the clearest changes were in workflow speed and admission patterns rather than in the time to make a single admit-or-not decision. Over 12-hour shifts at two freestanding emergency departments, more patients were seen per hour during the first 8 hours than during the last 4 [5]. At one site, throughput time increased from about 110 minutes to about 114 minutes, and door-to-doctor time rose from about 4.9 minutes to 6.2 minutes at the other site [5]. Those are modest but measurable slowdowns in the surrounding process. Yet decision-to-admit time itself did not change significantly at either site [5].

Consistency, in the sense of sticking to a stable rule, also appears to shift. In the finance study, credit loan approvals declined during midday compared with early or later in the workday, which the authors interpret as a preference for the default option—the choice that requires less cognitive effort [1]. That is a change in the pattern of decisions, not necessarily a change in how fast each decision was made. The emergency department study found that at one site admission rates fell from about 6.2% to about 3.9% from the first 8 hours to the last 4, while the other site did not show that admission change [5]. So the direction of the consistency shift can differ by setting.

When does the effect show up, and when does it not?

The effect is not simply a clock ticking down. In the emergency department study, the authors concluded that the accumulation of decision fatigue and its effect on patient flow and admissions may be more related to patient volume than to shift length [5]. That fits the broader picture: more decisions, especially under high demand, are what seem to matter, rather than hours elapsed by themselves. The finance study looked at approvals across the course of a day and found a midday dip rather than a steady decline from morning to evening [1], which also suggests the pattern is not a simple end-of-shift collapse.

The nursing commentary adds that decision fatigue is not inevitable for everyone. It cites a reported prevalence of 16% to 36% among nurses, and 29% among intensive care nurses, with physical fatigue, shifts longer than 12 hours, and dissatisfaction with clinical decisions as associated factors [2]. Those figures describe how often decision fatigue is reported in those groups, not a fixed rate of error. The commentary also notes that frequent contact with emergency nurses for health care reduced the quality of decision-making by 5.5% in one cited study [2]. The size of the effect therefore depends on the population and the workload, and it should not be treated as a universal constant.

What should a person running a shift take from this?

The practical answer is that later decisions in a long run are more likely to be slower in the surrounding workflow and more likely to drift toward the default or low-effort option, but the specific decision time may hold steady. In the emergency department, the slowdown appeared in door-to-doctor and throughput times rather than in decision-to-admit time [5]. In finance, the drift appeared as a midday decline in approvals, consistent with default-seeking [1]. A workplace that wants to protect decision quality should therefore watch both the choices being made and the process around them, not just the clock.

One limit is that the evidence does not give a single number of decisions at which fatigue begins. The finance study covered 26,501 loan applications and quantified the revenue cost of the approval pattern—about $509,023 in extra revenue for one month if all decisions had been made at early-morning approval levels [1]—but that is a cost estimate for one bank, not a threshold. The emergency department study covered 9,848 patient visits over five months at two sites [5]. These are real-world settings with their own workflows, so the size and even the direction of the effect can differ. The safest reading is that repeated decisions can degrade later performance, but the form of degradation—slower flow, more default choices, or both—depends on the work.

Sources used in this answer

Related research questions