Speech recognition and direct data entry in clinical microbiology.

The aim of this study was to design a system (Vigilante) for the direct entry of urine microscopy, culture and sensitivity test results at bench level using natural continuous speech, removing the need for time-consuming report coding and keyboard entry. Graphical and audio feedback facilitated error recognition and correction, and removed the need for time-consuming report checking and validation stages. The system was evaluated over a three-month period by direct comparison of 2360 urine specimens processed by multiple operators using both Vigilante and traditional methods. The results showed that the Vigilante system reduced the time taken to enter data, improved the accuracy of reporting and reduced the time taken to issue reports. The results suggest that direct data entry using speech offers a practical, reliable and innovative alternative to more conventional means of data entry. The system is now used routinely in this laboratory to report urine microscopy, culture and sensitivity test results, and may be adapted to individual urine methodologies or other data entry tasks.

Apparent failure of brain heart infusion broth to cultivate meningococci from clinically infected patients.

Brain Heart Infusion broth is a widely used medium for the isolation of pathogenic bacteria from body fluids. Cerebro spinal fluid and blood inoculated into this medium from clinically meningitic infants infected with Neisseria meningitidis appeared not to support the growth of this organism. In laboratory experiments the broth did not support the growth of Neisseria meningitidis unless adequately vented and with addition of blood. Thorough mixing was necessary to disperse microcolonies that formed in the base blood layer of blood cultures. When this was done they were isolated using standard sub-culture techniques. Broths used for specimens other than blood must be used with blood or a blood supplement if this organism is suspected, and should be incubated aerobically or vented to 5% carbon dioxide, and thoroughly mixed. Awareness of these deficiencies may prevent false negative results.

Isolation of Vibrio cholerae non-01 from a Somerset farmworker and his tropical fish tank.

Vibrio cholerae non-01 is occasionally isolated from wounds. To our knowledge, this is the first such case to be reported from the U.K. It was associated with isolation of a similar organism from a tropical fish tank.

Rational decisions in managing sore throat: evaluation of a rapid test.

Sixty nine general practitioners recorded what they had prescribed for a total of 1189 episodes of sore throat. Antibiotics were prescribed in 763 (64%) episodes and broad spectrum antibiotics in 161 (21%) of these. If there was dysphagia, hoarseness, cervical adenopathy, and inflamed or purulent tonsils a prescription was more likely to be written. An enzyme immunoassay rapid test was evaluated as a means of rationalizing prescribing. Among 23 general practitioners and 250 patients the sensitivity of the test was 63% and the specificity 91.7% compared with 74% and 58% for clinical assessment alone. Test results rarely caused previous prescribing decisions (34 [corrected] (13%) episodes) to be altered. We suggest that the time is not ripe for the use of the enzyme immunoassay rapid test on a wide scale in the routine assessment of sore throats.