To the Editor Bhanji et al1 found that in-hospital mortality of pediatric patients undergoing cardiopulmonary resuscitation for cardiac arrest was lower during the day/evening than at night. They suggest a causal effect may be at play, possibly mediated by lower staffing at night, and make efforts to estimate the day/night risk contrast in exchangeable populations. Unfortunately, important sources of potential bias were left unaddressed. Here, we request extra steps to accommodate likely confounders.