0118 Performance on a Computerized Threat Elimination Task in an Animated Environment during Total Sleep Deprivation

Emily Moslener,Kimberly Honn
DOI: https://doi.org/10.1093/sleep/zsac079.116
IF: 6.313
2022-06-04
SLEEP
Abstract:IntroductionMilitary and law enforcement operators must make split-second decisions on whether to shoot during confrontations. Quick responses are crucial when force is necessary, but accurate decision-making is also imperative. Often, decisions are made while fatigued, which could impair speed and/or accuracy. Furthermore, the reliability of background information may impact performance. We investigated performance on a computerized shooting task during a total sleep deprivation (TSD) study.MethodsN=86 healthy adults (39 males, age 21-38) completed a 4-day/3-night in-laboratory study, randomly assigned to a TSD (n=56) or control (n=28) condition. A custom task was administered after 32h or 8h of wakefulness (TSD and control groups). Participants were to shoot enemy robots (press spacebar) and not shoot friendly robots (no response) within 500ms of each robot being revealed inside shipping crates (1-5s inter-trial-interval). The task introduction described which crates would contain enemies, but the intel's accuracy varied across four phases: 100% (20 trials), 80% (120 trials), and 20% (40 trials), then irrelevant in a new environment (60 trials). Reaction time (RT) and accuracy (hits and false alarms (FAs)) were analyzed using 2x4 mixed-effects ANOVAs to determine the effects of condition, phase, and their interaction.ResultsThere was a significant effect of phase on RT (p<0.001); in both conditions, participants reacted faster in phase 1 than all other phases. However, there was no effect of condition (p=0.20) or phase-condition interaction (p=0.080) on RT. There were significant effects of condition on hits (p<0.001) and FAs (p=0.004); TSD had fewer hits and more FAs than the control group. There was an effect of phase on hits (p=0.045), with fewer hits in phase 1, and a condition-phase interaction (p=0.026) showing that the TSD group experienced less improvement in hits. For FAs, there was no effect of phase (p=0.86) or phase-condition interaction (p=0.86).ConclusionThe results suggest a speed/accuracy tradeoff during TSD, where relative to the control group, RTs remained equivalent, but accuracy was worse. In both groups, the RT slowing from phase 1 to subsequent phases, suggests that participants initially used the intel to facilitate quicker decision-making, but disregarded it once it was not completely reliable.Support (If Any)CDMRP grants W81XWH-16-1-0319 and W81XWH-20-1-0442
neurosciences,clinical neurology
What problem does this paper attempt to address?