Skip navigation

Doing the right test right

Bandolier is constantly on watch for examples of benefits from doing the right thing in the right way - of getting evidence into practice. So we were delighted when we found an example which brought this together with another theme of diagnostic tests.

Maastricht experiment

No, not something to do with politics, but an experiment conducted with 85 GPs serving a population of 187,000 around this delightful little town [1]. What they did, starting in 1985, was to arrange with local laboratories to provide written feedback twice a year to individual GPs on test ordering and appropriateness of tests requested. They used guidelines from their regional health authority and Dutch College of General Practitioners. Comparison was made with other practices which did not have the feedback.


These were the numbers of tests done each year from 1983 to 1991, together with their costs. The costs of tests for the tests practices were measured against a 7% annual incremental trend and control laboratories.


Costs in the trial practices for all the laboratory tests, including ECGs and radiology actually started falling after the first year. In control practices and laboratories, the costs rose inexorably. These cost changes derived from really large differences in test requesting. Comparing 1991 with 1985, the change in test requesting by the 85 GPs was:

Haematology -39%
Serology -60%
Clinical chemistry -30%
Urine/faeces -48%
Bacteriology/virology -17%
ECG/X-ray/histology 0%
Endoscopy +260%

Because of the rising rate of requesting generally, but the falling trend by the GPs receiving feedback, the cost differential rose each year. By 1991 the test practices were saving just under $400,000 for their 187,000 population - about $2 a year for each person - taking into account the cost of providing the feed-back service.


We should not be surprised at this. Audit Commission Reports showed large reductions in X-ray referrals following the introduction of guidelines (see Bandolier 12 ), and a randomised controlled trial of introducing radiology guidelines showed significantly more referrals were appropriate, together with some fall in total request [2].

Laboratories trying to cope with ever increasing request load would probably appreciate the breather they would get if schemes like this were introduced. The money saved in the Maastricht experiment was not insignificant - just try multiplying £1.50 by the population of a GP practice, of a health authority, or a nation.

Getting on with it

When we know that doing something can produce major savings, and provide a better and more appropriate service without any restriction on anyone's freedom, what stops us from getting on and doing it? Answers on a postcard. Here is just one bit of information, which should not be squirrelled away. Any Chief Knowledge Officer worth his or her salt would be trumpeting this abroad and making sure something was done.

At the very least it would be worth a "fact-finding visit" to see how the Dutch did it (and this group continues to publish useful and interesting stuff). Maastricht is a delightful city, with great hospitals, and close to the beautiful ancient capital of the Frankish Empire - Charlemagne's tomb in Aachen is awesome.


  1. RAG Winkens et al. Routine individual feedback on requests for diagnostic tests: an economic evaluation. Medical Decision Making 1996 16: 309-14.
  2. P Oakeshott, SM Kerry, JE Williams. Randomized controlled trial of the effect of the Royal College of Radiologists' guidelines on general practitioners' referrals for radiographic examination. British Journal of General Practice 1994 44: 197-200.

previous or next story in this issue