Near patient testing for respiratory syncytial virus in paediatric accident and emergency: prospective pilot studyBMJ 1999; 319 doi: http://dx.doi.org/10.1136/bmj.319.7205.289 (Published 31 July 1999) Cite this as: BMJ 1999;319:289
- Audrey Mackenzie, clinical nurse specialista,
- Nick Hallam, consultant virologistb,
- Elaine Mitchell, staff nursea,
- Tom Beattie, consultant in paediatric accident and emergency medicinea
- a Royal Hospital for Sick Children, Edinburgh EH9 1LF
- b Regional Clinical Virology Laboratory, City Hospital, Edinburgh EH10 5SB
- Correspondence to: Dr Hallam
- Accepted 5 March 1999
Respiratory syncytial virus is the most important respiratory pathogen in young children, causing bronchiolitis and pneumonia. Infection is especially serious for those who are immunocompromised and those with conditions such as bronchopulmonary dysplasia and congenital heart disease.1 The virus is highly infectious, and annual outbreaks cause many hospital admissions, putting great strain on isolation facilities and infection control measures. Nosocomial transmission is well documented.2 Rapid testing for the virus is well established3 and cost effective.4 Near patient testing has the added potential benefits of even faster diagnosis, further cost saving, and improved patient management and infection control. We describe a prospective pilot study of near patient testing for respiratory syncytial virus which was carried out in the accident and emergency department of the Royal Hospital for Sick Children, Edinburgh, between December 1997 and March 1998.
Patients, methods, and results
One hundred and three pairs of nasopharyngeal secretions were obtained from 98 children under 2 years of age presenting with respiratory symptoms (five children presented twice). The first specimen was sent for direct immunofluorescence testing at the Regional Clinical Virology Laboratory, City Hospital, Edinburgh (accredited by Clinical Pathology Accreditation (UK)) using the Imagen respiratory syncytial virus reagent (Dako, Ely, United Kingdom). The second specimen (taken immediately afterwards but often smaller in volume) was tested by staff of the accident and emergency department using an enzyme immunoassay (Abbott TestPack RSV, Abbott Laboratories, North Chicago, IL).
This protocol was adopted to avoid compromising the results of direct immunofluorescence testing (our routine method) while the pilot study was in progress. Staff training and near patient testing were carried out in accordance with published guidelines.5 Patients with positive results by the near patient test were isolated or put with others with positive results; those with negative results were also isolated if possible while awaiting further results.
The table shows the results for 94 specimen pairs. Results for the other nine specimen pairs are not included (in two cases the results of the near patient test were void, in two cases direct immunofluorescence was unsatisfactory, and in five cases no specimen was received at the laboratory). Compared with direct immunofluorescence, near patient testing showed a sensitivity of 79% (95% confidence interval 67.3% to 88.5%), a specificity of 97% (82.8% to 99.9%), a positive predictive value of 98% (89.6% to 99.9%), and a negative predictive value of 70% (53.9% to 82.8%).
Near patient testing proved acceptable to the staff performing it, and it was well received in the hospital wards in terms of facilitating patient management and infection control.
The high specificity (97%) of the enzyme immunoassay in this pilot confirmed results from a laboratory based study.3 The comparatively low sensitivity (79%), although comparable with the sensitivity in another study,3 is partially explicable by our using a second, often smaller, specimen for the test. Our results suggest that a positive result of a near patient test is trustworthy and does not require laboratory confirmation (allowing considerable savings in both time and cost) but a negative result needs further investigation. Only one specimen per patient should routinely be required, but this should be treated aseptically to allow for possible laboratory referral.
Near patient testing for respiratory syncytial virus is a viable option for paediatric accident and emergency with important potential benefits.
Contributors: AM designed the study and oversaw the near patient testing. NH assisted in the study design, oversaw near patient and laboratory testing, and wrote the manuscript. EM oversaw and performed near patient testing. TB assisted in the study design and in writing the manuscript. Staff of the accident and emergency department, Royal Hospital for Sick Children, Edinburgh, and of the Regional Clinical Virology Laboratory, City Hospital, Edinburgh, provided technical help. Dr Peter Mackie (who has pioneered the concept of near patient testing for respiratory syncytial virus) and Pamela Joannidis (infection control nurse) at Yorkhill Children's Hospital, Glasgow, provided helpful advice. NH is guarantor for the paper.
Funding Pump priming, Royal Hospital for Sick Children, Edinburgh.
Competing interests None declared.