OBJECTIVE: To assess the clinical and economic impact of surgery based near patient testing in general practice for six commonly used biochemical and bacteriological tests. DESIGN: After four months' monitoring, equipment for two bacteriological and four biochemical tests was introduced without cost into 12 practices using a crossover design. Structured request forms were used to monitor laboratory investigations. SETTING: 12 general practices in west midlands and south west Thames with list sizes above 9000. MAIN OUTCOME MEASURES: Investigation rates per 1000 consultations. Changes from baseline rates. Reasons for requesting investigations and provisional diagnoses. Cost per test and sensitivity of costs to rate of use. RESULTS: Investigation rates for the six tests rose by 16.5% (from 78.6/1000 consultations to 91.6/1000) when equipment was available in the surgery and reverted to baseline rates when it was withdrawn. The average weekly number of tests when equipment was available ranged from 0.5 to 10.5 (mean 9.0). Cholesterol tests were used as an addition to laboratory testing, usually for screening. Midstream urine analysis was often done in the surgery instead of in the laboratory, although 30% of samples were tested by both methods. Doctors' reasons for investigation and conditions tested were largely unaffected by availability of surgery tests. Costs for surgery tests were higher for all tests except midstream urine. CONCLUSIONS: Availability of surgery based testing increased the number of tests performed. It was cost effective only for midstream urine analysis.
OBJECTIVE: To assess the clinical and economic impact of surgery based near patient testing in general practice for six commonly used biochemical and bacteriological tests. DESIGN: After four months' monitoring, equipment for two bacteriological and four biochemical tests was introduced without cost into 12 practices using a crossover design. Structured request forms were used to monitor laboratory investigations. SETTING: 12 general practices in west midlands and south west Thames with list sizes above 9000. MAIN OUTCOME MEASURES: Investigation rates per 1000 consultations. Changes from baseline rates. Reasons for requesting investigations and provisional diagnoses. Cost per test and sensitivity of costs to rate of use. RESULTS: Investigation rates for the six tests rose by 16.5% (from 78.6/1000 consultations to 91.6/1000) when equipment was available in the surgery and reverted to baseline rates when it was withdrawn. The average weekly number of tests when equipment was available ranged from 0.5 to 10.5 (mean 9.0). Cholesterol tests were used as an addition to laboratory testing, usually for screening. Midstream urine analysis was often done in the surgery instead of in the laboratory, although 30% of samples were tested by both methods. Doctors' reasons for investigation and conditions tested were largely unaffected by availability of surgery tests. Costs for surgery tests were higher for all tests except midstream urine. CONCLUSIONS: Availability of surgery based testing increased the number of tests performed. It was cost effective only for midstream urine analysis.
Authors: Alex Faulkner; Nicola Mills; David Bainton; Kate Baxter; Paul Kinnersley; Tim J Peters; Deborah Sharp Journal: Br J Gen Pract Date: 2003-11 Impact factor: 5.386