Leeds area . general practitioners in the prescribing preferences of antibioticsensitivity reporting on Influence of laboratory

technique. This is almost certainly the case. 98-4%, respectively, for cervical specimens Towards the end of our reported study we and from 91-9% to 92-9% and 91-5% to noticed a decrease in the sensitivity of our 97-4% for urethral specimens, respectively. cell culture technique. The study was thereTaylor-Robinson, Hawkins, and fore extended, and discrepant results beThomas,1 cautioned against misinterpreting tween cell culture and enzyme immunoassay immunofluorescence test results that prowere re-examined using a direct duce apparent high identification rates for immunofluorescence technique (Micro Trak C trachomatis. As enzyme immunoassay Syva). Overall, 277 specimens were exammethods are not subjective this should be ined by cell cultures and enzyme immunoless of a problem with these techniques. Senassay, comprising 158 cervical specimens sitivity and specificity with enzyme immuand 119 male urethral specimens. Sensitivity noassay will, however, reflect the optical and specificity for enzyme immunoassay density cut off setting, and some results will were 81-25% and 95-2% for cervical specifall in the "retest range." mens, respectively, and 95 3% and 97-5% It is well accepted that cell culture, even in for urethral specimens, respectively. the best hands, is not 100% successful; but There were 22 discordant results: 13 were until more extensive experience is obtained culture negative and enzyme immunoassay with the new methods, epidemiological and positive and nine culture positive and entreatment studies based solely on antigen zyme immunoassay negative. Of these, 10 detection techniques must be interpreted specimens (nine enzyme immunoassay posiwith caution. tive and culture negative) were available for immunofluorescence testing. In addition, five specimens positive by enzyme immuReference noassay and culture and 51 randomly selected specimens negative by both methods 'Taylor-Robinson D, Hawkins DA, Thomas BJ. were retested using immunofluorescence. Syva Micro Trak stains: their use in a routine Immunofluorescence testing was carried laboratory. J Clin Pathol 1983;38:237. out by placing an aliquot of the remaining GL RIDGWAY enzyme immunoassay specimen (held for a JD ORIEL variable period at 70°C) on a 10mm well G MUMTAZ of a teflon coated slide. After fixation in cold GMELLARS acetone for 30 minutes the preparation was Department of Clinical Microbiology, stained according to the manufacturer's University College Hospital, instructions (Micro Trak Syva). Grafton Way, The Table shows the results. All nine London WCIE 6A V culture negative and enzyme immunoassay positive specimens and the single enzyme immunoassay negative culture positive specimen were immunofluorescence positive. Immunofluorescence results on the 56 Influence of laboratory sensitivity reporting non-discordant specimens agreed with the on antibiotic prescribing preferences of cell culture and enzyme immunoassay general practitioners in the Leeds area findings. Thus when checked against immunofluorescence only one enzyme imA survey made over two periods of time was munoassay result was truly false (negative). used to determine the effect ofa limited antiIncorporating these results into our calcubiotic reporting policy on the prescribing lations the sensitivity and specificity imhabits of general practitioners in the Leeds prove from 81 25% to 83-3% and 95-2% to area. 233