Do GPs who are co-located with pathology services order more tests?

Click to estimate the cost of getting sick (US)

I’ve cautioned before about the dangers of physical determinism i.e. glibly assuming that the physical environment (or geography) is the driving force underlying human behaviour. Sometimes it is, sometimes it exacerbates another problem, but more often it’s the symptom rather than the cause.

So it was with interest that I noted a new study reported in The Age today that investigated if Melbourne and Sydney GPs who are co-located with pathology collection centres tend to order more tests than GPs who aren’t. I’ve had a look at the study, undertaken by the Melbourne University School of Population Health, in the latest issue of the Medical Journal of Australia.

The study defines co-location as where a GP and a pathology centre are located in the same premises. This group was compared with GPs located at least 50 metres from the nearest pathology centre.

The hypothesis that co-location might lead to a higher propensity to order pathology services seems plausible given, as the authors say, that many studies have shown how doctors’ contact with the pharmaceutical industry can influence their clinical decision-making:

“repeated visits by sales representatives, the availability of free drug samples, and relationships forged through conferences, educational events and consultancy arrangements have been linked to prescribing patterns. The hypothesised mechanism of action is, of course, the same as the pharmaceutical industry’s rationale for spending vast sums on marketing campaigns directed at doctors — namely, doctors who are regularly exposed to information, people, money and events associated with a particular product are more likely to incorporate that product into their clinical activities”.

However the researchers found that GP’s predilection for using pathology services is not related to the proximity of a pathology centre. But it is related to factors like the size and accreditation of a practice and to characteristics of GPs themselves, including their sex and age. Moreover:

“the GP and practice characteristics found to be associated with test-ordering rates in this study largely echo those detected in earlier Australian research. A 1994 study of Medicare data found that rates of test ordering were higher among female GPs and varied according to year of graduation from medical school. Previous analyses of BEACH data showed higher rates of test ordering among female GPs and in larger practices, and lower rates among older GPs. To those recognised associations, our study adds practice accreditation and high workload (11 + sessions per week) as positive predictors; we also found that test-ordering rates were about 10% higher in Melbourne than in Sydney”.

The authors don’t venture an explanation for why co-location has no effect but I’d suggest it might be because pathology collection centres tend to be relatively low level establishments performing routine operations. For example, there’s a small, two room pathology collection centre in my local Principal Activity Centre that’s staffed by a nurse (it’s not co-located but it’s within 50 metres of three large general practices in the centre). There seems to be little reason why GPs and local pathology centre staff would have meaningful face-to-face contact. In the same vein, I very much doubt that local pharmacies actually influence script-writing rates – it’s “big pharma” that does the influencing, not your local chemist.

I suppose it’s possible that GPs who are relatively remote from pathology centres might be less inclined to order tests for patients who have ambulatory or transport problems, but I’d be surprised. However the researchers didn’t measure the distance of GPs from pathology collection centres (other than to establish that they were at least 50 metres away) so we can’t see if there’s any decay in the tendency to order tests with distance.

The authors are aware that if there is an issue of over-testing it’s more likely to be related to higher order matters:

“our findings do not close the book on questions about the structure of pathology services in Australia generally, or over-testing in particular. Several recent trends are noteworthy, namely: the spread of GP Super Clinics, portending further integration of primary care and diagnostic services; the discontinued agreement between the federal government, pathologists and pathology companies over caps on the growth of pathology services; the federal government’s loosening of pathology collection centre licensing rules; and the looming prospect of larger out-of-pocket pathology costs for patients. Together, these developments suggest that the need to better understand drivers of variation in test-ordering behaviour, and what constitutes cost-effective use of pathology services, is even more pressing today than it was when Deeble and Lewis-Hughes sounded the call two decades ago.

Sometimes physical proximity doesn’t explain much at all.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s