Skip to main content
Fig. 1 | BMC Health Services Research

Fig. 1

From: The effect of time-varying capacity utilization on 14-day in-hospital mortality: a retrospective longitudinal study in Swiss general hospitals

Fig. 1

Causal DAG for time-varying exposure, time-varying confounders and outcomes at day one and day two. Time-varying exposures are capacity utilization (CU1, CU2), time-varying confounders are patient turnover (PT1, PT2), patient clinical complexity level (PCCL1, PCCL2) and outcome mortality (M1, M2) at day one and day two with other fixed covariates (C) (e.g., age, sex, comorbidity weights) and unmeasured variables (U) (e.g., staffing). The arrows are limited to prevent overcrowding, main issue remained same if we add more arrows and to avoid complexity in DAG, we consider single unmeasured variable. The green path represents a causal path (open), the black path an adjusted (blocked confounder) path and the red path the biasing path as per DAGitty

Back to article page