Combined clinical and virological surveillance of influenza in winters of 1992 and 1993-4BMJ 1995; 311 doi: http://dx.doi.org/10.1136/bmj.311.7000.290 (Published 29 July 1995) Cite this as: BMJ 1995;311:290
- D M Fleming, directora,
- P Chakraverty, clinical scientistb,
- P Litton, medical laboratory scientific officerb
- aBirmingham Research Unit, Royal College of General Practitioners, Harborne, Birmingham B17 9DB
- bPublic Health Laboratory Service, Central Public Health Laboratory, London NW9 5EQ
- Correspondence to: Dr Fleming.
- Accepted 24 April 1995
Influenza is a major public health problem. Some cases occur every winter; there have been six substantial epidemics in the United Kingdom during the past 25 years; pandemics following major shifts in the influenza virus occur less frequently but are particularly severe.1
Most European countries support programmes of influenza surveillance.2 In some countries—for example, France,3 the Netherlands, and Portugal—clinical and virological surveillance is undertaken in sentinel practices that monitor the incidence of influenza and obtain specimens for virological examination in the major virological laboratories. In the United Kingdom the largest clinical surveillance network is the Royal College of General Practitioner's weekly returns service.4 Virological surveillance is based on the results of virological tests in patients admitted to hospital, with centralised reporting to the Public Health Laboratory Service.
Early identification of the organism responsible for epidemics of respiratory infection is important to facilitate the distribution of specific drugs when appropriate (for example, amantadine to elderly people in influenza A epidemics); vaccinate workers in essential services; to plan for the allocation of hospital beds; to grow new virus strains for the next manufacture of vaccine; and to provide the public with reliable, comprehensible information. We present the results of a project designed primarily to hasten recognition of an influenza epidemic.
Materials, methods, and results
During the winters of 1992-3 and of 1993-4, six practices in the weekly returns service sent nose and throat swabs from patients with new cases of influenzalike illness for virological examination at the Central Public Health Laboratory. The practices were in Northumberland, Merseyside, west midlands, Bedford, Kent, and Somerset and had a combined population of 51000. Samples were sent by first class post and examined by direct immunofluorescence5 and by tissue culture according to standard methods. Specimens with positive results on immunofluorescence (and a sample of negative specimens) were also examined by the polymerase chain reaction.
In the winter of 1992-3, 182 specimens were submitted and 11 viruses cultured. Three were influenza A H, three influenza A H, and five influenza B. Specimens containing influenza B viruses were also positive on immunofluorescence and were all found during the outbreak occurring late in March 1993. In the winter of 1993-4, 279 specimens were submitted; 66 (24%) were positive for influenza A (immunofluorescence or culture, or both). The proportion of positive results was highest in children (35% occurred in children aged 0-4 years, 33% in those aged 5-14, 17% in those aged 15-64, and 22% in those aged >/=65). The figure shows their distribution compared with the reported incidence of influenza-like illnesses in the six practices and in the weekly returns service network of 93 practices. Clinical and virological reports peaked simultaneously in week 45.
The polymerase chain reaction yielded 18 positive results in 20 specimens with positive results on immunofluorescence. The reaction yielded eight positive results in a sample of 41 specimens with negative results on immunofluorescence.
These results show the value of virological surveillance in primary care. The combined nose and throat swabs were easy to take and their quality remained satisfactory after having been posted. The results of immunofluorescence were reported to practices within 24 hours of samples arriving at the Central Public Health Laboratory, providing early diagnostic confirmation for the general practitioners and alerting them to local influenza. The pattern of virological results was contemporaneous with the reported clinical incidence nationally, so they effectively monitored the onset and progress of the influenza A/Beijing/32/92 (H) epidemic late in 1993. The results also validate the clinical data reported in both epidemic and non-epidemic periods. Virology reports to the Public Health Laboratory Service from other sources peaked in week 50, five weeks later than those reported from the study practices, and did not indicate substantial influenza activity until week 46. Deaths peaked in week 48.
Effective management of influenza epidemics, whether caused by an existing or new virus strain, depends on effective surveillance. This study shows an improvement in influenza monitoring when virological specimens are obtained in general practice.
Funding The Public Health Laboratory Service and the Birmingham Research Unit of the Royal College of General Practitioners are both funded by the Department of Health.
Conflict of interest None.