Association between reported aetiology of central nervous system infections and the speciality of study investigators—a bias compartmental syndrome?

Abstract Background Conventional descriptions of central nervous system (CNS) infections are variably categorized into clinical syndromes for patient investigation, management and research. Aetiologies of the most commonly recognized syndromes, encephalitis and meningitis, tend to be attributed predominantly to viruses and bacteria, respectively. Methods A systematic review was performed of aetiological studies of CNS syndromes and data extracted on reported author specialities. Results The analysis identified an association between the author’s speciality and the CNS syndrome studied, with a tendency for virologists to study encephalitis and microbiologists to study meningitis. Conclusions We suggest there is bias in study design. Stronger multidisciplinary collaboration in CNS infection research is needed.


Background
Aetiologies of the common central nervous system (CNS) infection syndromes of encephalitis and meningitis have tended to be attributed predominantly to viruses and bacteria, respectively. [1][2][3] In Laos, we have not found an association between syndromes and aetiologies and no significant difference in the proportions of detectable viral and bacterial aetiologies (Dubot-Pérès et al., submitted).
The significance of viral infections in the presentation of meningitis is becoming better characterized. [4][5][6] Data suggest the detection of viruses in at least 50% of patients with clinical meningitis. 7 Conversely, 20% of the 1570 encephalitis patients enrolled in the California Encephalitis Project were diagnosed with a bacterial infection. 1 If virologists are not involved in diagnostic testing, a virus may not be identified. Indeed, it is notable that guidelines for the management of meningitis do not consistently feature virologists, nor do encephalitis guidelines involve microbiologists. 2,3 We hypothesized that the reported aetiology of encephalitis and meningitis may be associated with the speciality of the study investigators. We performed a systematic literature review of studies examining the aetiology of encephalitis or meningitis and extracted data on the speciality of the investigators to investigate if this impression is correct.

Methods
Searches were performed in PubMed, Embase and Cochrane, accessed using Ovid gateway, and references from relevant articles from January 2000 to April 2017 using the MESH and keyword search terms ('encephalitis', 'meningitis', 'brain infection' or 'Central Nervous System infection') and ('epidemiology', 'cause', 'aetiology' or 'etiology'). Selection criteria included English-language epidemiological studies investigating the aetiology of communityacquired meningitis or encephalitis in human patients. Each paper was described as concerning meningitis or encephalitis depending on the presence of these keywords in the MESH and keyword search terms. If both terms were present the paper was excluded, as were case reports (defined as fewer than four patients), studies of meningoencephalitis, specific infectious agents or groups (e.g., case series of Streptococcus suis, HIV or viral meningitis) and brain abscesses. Two authors independently assessed studies for inclusion and extracted data on the speciality of all authors and a third author resolved disagreement. Searches were reported according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
The speciality of each author was derived from his or her departmental affiliation stated in the publication. Specialities were recorded as microbiology, virology, infectious diseases, neurology, paediatrics, research/public health and other. Bacteriology was not described as an affiliation in any paper and we assumed that microbiology did not include virology as a subspeciality. The data were summarized for each study as one point for a speciality if one or more authors reported an affiliation for that speciality. Results were analysed in Excel (Microsoft, Redmond, WA, USA) and Stata 14 (StataCorp, College Station, TX, USA) and presented as the number and percentage of the total included papers. Papers were further categorized as those whose investigators included 'microbiology without author virology input', 'virology without author microbiology input' and 'other'. The association between these three categories and two clinical syndromes, encephalitis and meningitis, as defined in the paper were analysed using a Fischer's exact test and reported with the corresponding p-value.

Results
Sixty-nine studies were included in the analysis, 32 investigating encephalitis and 37 investigating meningitis ( T. Bharucha et al. in the UK and USA, with the remainder performed in diverse countries (Table 1). Seventeen (24.6%) studies received a contribution from a microbiologist author, 6 (8.7%) from a virologist and 16 (23.2%) from an infectious disease specialist (Table 2). There was also considerable involvement of neurologists, paediatricians and specialities such as public health and emergency medicine. However, 38 (55%) of the studies had no apparent microbiology, virology or infectious diseases author listed. Articles

Discussion
This relatively crude analysis is consistent with an association between the study of meningitis by microbiologists and encephalitis by virologists. This probably reflects the historic prioritization of bacterial meningitis for investigation by microbiologists owing to the high untreated mortality and the need for urgent antibiotic therapy. 8 Bacterial meningitis is regarded as more life-threatening than viral meningitis. 3 The significance of viral meningitis, particularly Transactions of the Royal Society of Tropical Medicine and Hygiene the contribution of enteroviruses, is becoming better understood. 4,9 It is increasingly recognized that vituperative pathogens such as the viral haemorrhagic fever group and arboviruses such as Japanese encephalitis virus may present as meningitis. 10,11 Similarly, syphilis, leptospiral and rickettsial pathogens represent potentially treatable aetiologies of encephalitis. 12,13 Limitations of the study include that only 17 years of publications were included and all of these were in English. The structure of the workforce in these specialities and the predominance of microbiologists over virologists worldwide is likely to have affected the results. [14][15][16] Microbiologists are likely to have had training in virology and virologists training in bacteriology. The definitions of encephalitis and meningitis as given in the individual papers were used. The determination of author speciality by their reported affiliations may have led to bias: author affiliation is a crude proxy measure of speciality and some authors are likely to have had multiple specialities. For example, both microbiologists and virologists may work in infectious disease departments and physicians in both specialities may also be certified in infectious diseases. Thirty-eight (55%) studies were classified as 'no infection input', including those that gave hospital affiliation only rather than departmental specialty. While we focussed our analysis on infection input, this issue extends to the importance of involvement of neurologists and paediatricians.
The data questions established norms on the research evidence base for diagnosis and treatment of two important clinical CNS infection syndromes with implications for both clinical practice and research. Syndromic approaches certainly have roles, particularly in low-resource settings with poor access to laboratory diagnostics, but they have important limitations. [17][18][19] These data suggest that encephalitis-virus and meningitisbacteria associations have tended to remain compartmentalized, when in reality both clinical syndromes and aetiologies are parts of overlapping continuous spectra. Stronger multidisciplinary collaboration in the design and interpretation of CNS infection studies would avoid the pitfalls of compartmentalization. Funding: This study was supported by the Wellcome Trust of Great Britain. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Competing interests: None of the authors have any conflicts of interest to declare.
Ethical approval: Not required. T. Bharucha et al.