Editor,—Emergency departments in NSW have been collecting computerised data about the patients they see using the Emergency Department Information System since 1994.¹ It is possible to use these data to investigate the extent to which the availability of general practitioner services varies with the rate of primary care presentations to NSW emergency departments. It might be expected that emergency departments would see more primary care patients on weekdays that were public holidays than weekdays that were not, because general practices are usually closed on public holidays. In Australia primary care patients have been defined by the National Health Strategy as those who could be managed by a well equipped general practitioner in their surgery. The National Health Strategy operationalised this as patients in triage categories 4 and 5, the less acute categories.² This approach has been used in this study.

To test this hypothesis logistic regression analyses were used to compare the number of patients seen on weekdays that were public holidays with those seen on weekdays that were not public holidays, controlling for the triage category of the patient, the day of the week, the month of the year and the year and clustering on the hospital from which the data were obtained. Data from patients in triage categories 4 and 5 were analysed separately from those in triage categories 1 and 2 because it might be argued that any differences detected arose from changes in patient behaviour because of the public holiday rather than differences in the availability of general practice services. It would not be expected that the availability of general practice services make any difference to the number of patients seen in triage categories 1 and 2 because these patients have major illnesses that need urgent treatment to prevent serious sequelae and consequently are unlikely to seek care from a general practitioner.

Data were available from over 1.8 million encounters on ordinary weekdays and 93 397 encounters on public holiday weekdays with patients in triage categories 4 and 5 and 158 335 encounters on ordinary weekdays and 6357 encounters on public holiday weekdays with patients in triage categories 1 and 2. No difference was detected between the number of patients in triage categories 1 and 2 seen on public holidays to those seen on working days (p=0.709, OR=0.998, 95% CI 0.986, 1.010), however there was a small but significant increase in the number of patients seen in triage categories 4 and 5 (p<0.0005, OR=1.0199, 95% CI 1.013, 1.027). The sample size used to look for a difference in the number of patients in triage categories 1 and 2 had 99.99% power to detect a difference of the size detected for patients in triage categories 4 and 5.

Findings in Australia suggest that either patients do not perceive emergency departments and GPs to be close substitutes and/or that general practice patients are able to defer their requirements for primary care services until they can receive these from a GP. Is this a common experience in other countries?