The influence of ambulance offload time on 30‐day risks of death and re‐presentation for patients with chest pain

In reply: We thank Mallows and colleagues1 for their interest in our study.2 We agree that assessing the interaction between access block, emergency department (ED) overcrowding, and ambulance offload times would have contributed to improved interpretation of the data. Unfortunately, measures of ED overcrowding are not available in the Victorian Emergency Minimum Dataset. It is important to note that the studies referenced by Mallows and colleagues have similar limitations in either assessing a combined hospital overcrowding measure, or ED overcrowding and access block without adjustment for hospital overcrowding.3,4 Ambulance offload delays, ED overcrowding and access block could all be considered a symptom for hospital overcrowding, with the latter driving poor outcomes. This uncertainty is clearly presented in our manuscript in Box 1.2

The suggestion that tertile 1 patients might be more stable and therefore placed in the ED waiting room is speculative and not consistent with the balanced clinical observations and the more frequent emergent triage categorisation for tertile 1 patients (Box 3 and Supporting Information, table 3).2 Few clinical registries classify comorbid disease severity, and it seems unlikely that inclusion of severity data would influence the primary endpoint in addition to the 23 variables included. Notably, the results remained consistent in unadjusted analyses and across each sensitivity analysis including variables more likely to have an impact on 30‐day mortality, such as clinical observations and final diagnosis. Moreover, the referenced studies regarding ED access block and overcrowding do not adjust for any patient demographic characteristic, comorbid condition or clinical variables outside of age and sex.3,4 Inclusion of the Charlson index in Box 3 was requested during the review process, and while this was added with the mentioned caveats, it is not included in the multivariable model. No statewide population‐based datasets exist in Victoria detailing presenting symptoms and comorbid conditions for direct presentations, and development of a statewide clinical ED registry should be considered.

We agree that any policy decisions should be systemic and address hospital overcrowding, which causes offload delays, access block and ED overcrowding. However, these data are prior to the coronavirus disease 2019 (COVID‐19) pandemic; the current situation is substantially worse, and existing Australian targets for ambulance offload are not in keeping with patient and community expectations nor comparable to other health services internationally.5 The results of this study emphasise why such targets might need reconsideration in light of patient safety and outcomes.