Comparison of the use of four desktop analysers in six urban general practices.

There is little data on the advantages and disadvantages of using desktop analysers in general practice. This prospective trial compared four of the analysers available in the United Kingdom, in six urban general practices, over a six month period. Of the 2619 tests where the time was noted, 55.8% were performed outside the hours when routine transport to a hospital laboratory was possible (after 12.00 hours). Of the 3530 tests performed the commonest were measurements of cholesterol (14.4 tests per 5000 patients per 30 days), glucose (6.0 tests) and haemoglobin (5.6 tests). Less than 5% of the tests were performed as an emergency despite the speed at which results are available. The main reasons for requesting the tests were screening or case finding (56.9%), with the remainder for monitoring chronic disease, especially diabetes and hypercholesterolaemia. There was evidence that the use of the machines in the four practices reduced requests for hospital laboratory blood tests by 24-40% of pre-study levels. However, there was a considerable increase in testing for cholesterol (three fold) and haemoglobin (eight fold) on the desktop analysers, compared with the number of laboratory tests requested before the study. The cost per test of using such machines is closely related to the level of activity and probably does not compete favourably with hospital testing unless several tests are performed each day. Quality control tests were within the specified limits on at least 98% of occasions, however these tests also identified the need for laboratory back up where a problem was found.