New Estimates of Incidence of Encephalitis in England

Julia Granerod¹

, Simon Cousens, Nicholas W.S. Davies¹, Natasha S. Crowcroft¹, and Sara L. Thomas

Author affiliations: Public Health England, London, UK (J. Granerod); London School of Hygiene and Tropical Medicine, London (S. Cousens, S.L. Thomas); Chelsea and Westminster Hospital, London (N.W.S. Davies); Public Health Ontario, Toronto, Ontario, Canada (N.S. Crowcroft)

Abstract

Encephalitis causes high rates of illness and death, yet its epidemiology remains poorly understood. To improve incidence estimates in England and inform priority setting and treatment and prevention strategies, we used hospitalization data to estimate incidence of infectious and noninfectious encephalitis during 2005–2009. Hospitalization data were linked to a dataset of extensively investigated cases of encephalitis from a prospective study, and capture–recapture models were applied. Incidence was estimated from unlinked hospitalization data as 4.32 cases/100,000 population/year. Capture–recapture models gave a best estimate of encephalitis incidence of 5.23 cases/100,000/year, although the models’ indicated incidence could be as high as 8.66 cases/100,000/year. This analysis indicates that the incidence of encephalitis in England is considerably higher than previously estimated. Therefore, encephalitis should be a greater priority for clinicians, researchers, and public health officials.

Encephalitis is associated with severe illness, appreciable mortality rates, and high health care costs (1), but its epidemiology remains poorly understood (2). The sole previous incidence estimate for encephalitis in England of 1.5 cases per 100,000 population per year was for viral encephalitis only and was based on hospitalization data from 1989–1998 (3). Incidence should be understood; as an increasing number of viruses have been found to cause encephalitis in humans, more cases might be found among the high proportion of cases of unknown etiology (2,4–6). Climate change and increasing international travel raise the possibility of wider geographic spread of microbes, which may have important public health implications. Clarifying incidence is also important clinically with the increasing recognition of novel immune-mediated forms of encephalitis, especially because treatment is available if started early (7–9).
Although encephalitis is a statutorily notifiable disease in England, it is grossly underreported, making notification data unsuitable for incidence estimation (3). Almost all encephalitis case-patients require hospitalization; thus, routinely collected hospitalization data provide a possible source of data from which to estimate incidence. However, diagnosis of encephalitis is complicated by the lack of a standard case definition or pathognomonic symptoms and signs. Many patients with suspected encephalitis ultimately are found to have conditions with neurologic signs that mimic encephalitis. We recently reported on the Public Health England (PHE) study, the largest population-based prospective cohort of encephalitis patients to date in England (10). Data from this study, which included exhaustive multistage diagnostic investigations of cases, provided a unique opportunity to complement routinely collected hospitalization data to enable detailed analyses of encephalitis incidence in England. We linked the 2 data sources and performed capture–recapture analyses to estimate the number of encephalitis cases in England attributable to infectious and noninfectious causes.