Ambulance ramping and patients with cardiac‐type symptoms: understanding the unloading queue

Prioritisation for unloading, once obvious cardiac emergencies have been excluded, disadvantages women and older people

About 10–20% of people transported by emergency ambulance to hospital have (presumptively cardiac) chest pain.1 Our management of these patients is inevitably geared to the possibility of evolving myocardial infarction, in which case any delay in initiating definitive treatment to restore coronary perfusion will increase the short and long term risks of death.2 The three major sources of delay after the ambulance collects the patient are the time taken to deliver the patient to a suitable hospital, the waiting period outside the emergency department before unloading the patient, and within‐hospital barriers to treatment initiation, such as delays in definite diagnosis and the availability of suitably trained staff for delivering definitive treatment. For most purposes, it is best to consider barriers to treatment as a “series resistance model”: what ultimately matters is to ensure that treatment is delivered as expeditiously as possible.

However, much attention in many countries (including Australia) has been directed over the past five years to a slightly different aspect, that of “ramping”, or prolonged ambulance waiting outside hospitals before being able to unload patients for emergency department management.3 Two obvious adverse consequences of ramping are the inadequate interim treatment of patients (“death in the back of the ambulance”), and reduced ambulance service functionality because of enforced inactivity. There is considerable public concern about delays in ambulance dispatch and arrival at emergencies, cardiac or otherwise, because of ramping, and only limited understanding of the causes of ramping, which is clearly an increasing problem.

Ramping is caused by a mixture of hospital emergency department overcrowding and understaffing;4 patients are left at the hospital entrance because of blockage at this point. However, in the current issue of MJA5 a detailed analysis of variability in the ambulance unloading process indicates, for the first time, that the unloading process is not simply a matter of “first come, first served”. Dawson and colleagues evaluated variability in the time needed in Victoria for the unloading of 213544 patients with chest pain who arrived by ambulance, but not obviously life‐threatening cardiac emergencies, between January 2015 and June 2019. During this period, the median time between ambulance arrival and unloading increased from 21 minutes (interquartile range [IQR], 15–30min) to 24 minutes (IQR, 17–37min), but the median time for the longest tertile (ie, the one‐third of transfers requiring the most time) was 41 minutes (IQR, 33–56min), a very concerning figure indeed.

The authors used multivariate analysis to evaluate the determinants of variability in ramping duration, and examined the relationship between unloading time and death or re‐transportation to hospital by ambulance within one month. They found that ambulance unloading was NOT a random process: unloading time was significantly longer for older patients (over 65 years of age) and women, and at larger hospitals. Further, patients who experienced longer unloading times were significantly more likely to die or to again need ambulance transport within 30 days. The number of deaths in ramped ambulances was not reported, but one presumes that they were few.

While Dawson and colleagues took a novel approach to the analysis of the pre‐hospital period, their findings are subject to several caveats. First, although most patients in their study experienced chest pain, for the vast majority it was mild; the presumptive diagnoses were of cardiac pain in fewer than one‐quarter of patients, and just over one‐half of patients received no definitive diagnosis. Similar findings have been recorded for other studies,6 raising the question of whether well defined follow‐up of patients with “no diagnosis” should be a priority. Second, the frequency of diagnoses of coronary artery spasm7 or takotsubo syndrome8 were not reported. Both are now recognised as frequent causes of chest pain with serious implications for short term complications and long term recurrence. Further, the study period preceded the COVID‐19 pandemic, which has added an extra level of complexity to the determinants of unloading times. Finally, the study was seriously limited in that the relationship between unloading time and states of emergency department occupancy was not examined; this factor may well have partly explained the longer unloading times at larger hospitals. It is also a pity that data on Indigenous and non‐English‐speaking patients were not available.

Despite these limitations, the study by Dawson and colleagues is a landmark, and could serve as the basis of a re‐evaluation of the emergency triage procedure. The authors suggest that reducing ambulance unloading times might save lives, but the reported association between ambulance unloading time and short term mortality does not provide evidence of causation. Further, their findings clearly show that patient prioritisation for unloading from ambulances, once obvious cardiac emergencies have been excluded, is biased against women and older people. It remains to be determined whether this bias extends to the other side of the emergency department doors. Hippocrates would have been concerned.