The quality of a laboratory test in general practice: the throat swab example

To measure the repeatability, the flow of information, and the data entry error of the throat swab test from a suburban general practice.

A prospective study of consecutive patients seen in routine consultation for a sore throat as their primary complaint. Repeated throat swabs were taken for each patient. Carbon copies of routine laboratory forms were collected.

The kappa statistic for reproducibility of the throat swab test was 0.63 (95% CI 0.50-0.77). In 5.8% of patients, one swab was positive and the other negative. Information took one day or less to travel to and from the laboratory for 25.7% of patients with a beta-haemolytic streptococcal growth compared to the 9.9% of patients with no growth (chi 2 = 6.17, p < 0.02). The laboratory received 53% of the swabs on the day they were taken. Delays in getting the swabs to the laboratory occurred when swabs were taken in the afternoon session or on weekends. Monday morning gave the quickest return. Data entry errors were found on 15.9% of patient forms.

The quality of the throat swab test needs improvement in the suburban general practice context.