Contextual factors and use of heuristics in the care of critically ill patients by intensivists during the COVID-19 pandemic: a quasi-experimental vignette study in Chile

Background: Clinical decision-making in intensive care units during the COVID-19 pandemic has been poorly studied. The aims of this study are to assess the variability of intensivists' decision-making in Chile, analyze the association between clinical decision-making and certain contextual factors (time, pandemic wave, patient age, available clinical information), and estimate the frequency of heuristic use (statu quo, representativeness, and availability) in clinical decision-making. Methods: Quasi-experimental, cross-sectional, qualitative-quantitative study using an online questionnaire of clinical vignettes on the care of critically ill patients during the pandemic. Fifty-one intensivists participated, choosing between two alternatives per case, one of which involved using a heuristic. Frequencies were calculated to estimate variability. Intra-individual changes after manipulating contextual variables were evaluated using the McNemar test (mid-P) or binomial tests. Results: There is wide variability in the decision-making process: only 4 out of 16 cases exceeded 82% agreement. Contextual factors had a heterogeneous influence. No differences were observed regarding the time of day when the decision was made. The pandemic wave influenced 2 of the 4 vignettes. Regarding patient age, it was observed that the decision to withdraw treatment in the absence of a response within 30 days is more likely in older patients, but this is not a determining factor for admission to intensive care units. On the other hand, the clinical decision is modified according to the availability of clinical information. The use of availability heuristic (greater in the first wave) and statu quo heuristic is observed in three of the four vignettes studied, but not representativeness heuristic. Conclusions: The participating intensivists showed high variability and sensitivity to specific contextual factors. The use of statu quo and availability heuristics was observed under certain conditions. This analysis could help improve decision-making during health crises.