Introduction SIGN recommends rapid virological testing for RSV. There was a perception that in the absence of laboratory confirmation, infants were being increasingly investigated and treated to cover for bacterial pneumonia.

Method A cost analysis comparing two periods when NPA was routinely done and when it was not done was undertaken. A retrospective audit comparing patients diagnosed with bronchiolitis in December 2001 when NPA testing was routinely done and December 2010 when it could only be requested by senior clinicians. This decision was meant to cut costs as it was felt that a positive NPA did not change management.

Results In 2001, 15 patients (75%) were diagnosed with bronchiolitis using NPA techniques compared to two patients (5.9%) in the 2010 cohort. In 2010, 10 patients (29.4%) received chest radiographs compared to 6 patients (30.0%) in 2001. A total of 3 patients (15.0%) received the 9 blood investigations in the 2001 cohort group compared to 10 patients (29.41%) who received the 43 investigations in the 2010 cohort. The combined cost of all investigations during inpatient stay per patient was £24.95 in the 2001 cohort group compared to £10.10 in the 2010 cohort group. NPA tests are considerably more expensive, specifically 1.5 times the cost of chest radiographs and more than 8 times the cost of any blood investigation. The mean length of inpatient stay was similar in both cohorts: 2.9 days in 2001 and 2.7 days in 2010

Conclusion NPA testing may not be a cost-effective matter-of-course investigation. Notably, bronchiolitis requires a predominately clinical diagnosis. The cost of NPA testing, appreciably higher than for any investigation, accounted for almost 75% of the investigation spend in 2001, yet only 14.5% in 2010. However, as the other investigations are not interchangeable with NPAs, there must be a presumption that other legitimate rationales dictated their requirement.