Modeling factors influencing the demand for emergency department services in ontario: a comparison of methods

Unacceptable selection bias

Thomas Hughes   (2011-09-22 14:56)  John Radcliffe Hospital, Oxford, UK

While reduction in primary care availability is certainly a factor in increasing ED attendance, the assumption that patients who are triaged 4 and 5 who are not admitted and not injured are all potential primary care patients is unacceptable selection bias.

The very fact that a patient arrives at an ED rather than attending primary care means their pre-test probability of having a significant medical problem is much higher than those seen in primary care. The populations are different.

To conclude from the modelling that adding more primary care would reduce ED attendance assumes that primary care would be sited where less healthy and poorer patients would have good access, and then that these patients would use the service instead of, rather than in addition to the ED.
Evidence from the UK 'Walk-in Centres' which tried to do just this is that these centres increase healthcare demand without reducing ED attendance or improving outcomes.

The BEACH study in Australia and the UK Primary Care Foundation report looked at this problem in some detail, but are not listed in the references.

Competing interests

Co-chair UK College of Emergency Medicine Informatics Group
Partner, Circle Healthcare

top