A systematic review of the reporting of tinnitus prevalence and severity.

INTRODUCTION

There is no standard diagnostic criterion for tinnitus, although some clinical assessment instruments do exist for identifying patient complaints. Within epidemiological studies the presence of tinnitus is determined primarily by self-report, typically in response to a single question. Using these methods prevalence figures vary widely. Given the variety of published estimates worldwide, we assessed and collated published prevalence estimates of tinnitus and tinnitus severity, creating a narrative synthesis of the data. The variability between prevalence estimates was investigated in order to determine any barriers to data synthesis and to identify reasons for heterogeneity.

METHODS

and analysis: A systematic review included all adult population studies reporting the prevalence of tinnitus from January 1980 to July 2015. We searched five databases (Embase, Medline, PsychInfo, CINAHL and Web Of Science), using a combination of medical subject headings (MeSH) and relevant text words. Observational studies including cross-sectional studies were included, but studies estimating the incidence of tinnitus (e.g. cohort studies) were outside the scope of this systematic review.

RESULTS

The databases identified 875 papers and a further 16 were identified through manual searching. After duplicates were removed, 515 remained. On the basis of the title, abstract and full-text screening, 400, 48 and 27 papers respectively were removed. This left 40 papers, reporting 39 different studies, for data extraction. Sixteen countries were represented, with the majority of the studies from the European region (38.5%). Publications since 2010 represented half of all included studies (48.7%). Overall prevalence figures for each study ranged from 5.1% to 42.7%. For the 12 studies that used the same definition of tinnitus, prevalence ranged from 11.9% to 30.3%. Twenty-six studies (66.7%) reported tinnitus prevalence by different age groups, and generally showed an increase in prevalence as age increases. Half the studies reported tinnitus prevalence by gender. The pattern generally showed higher tinnitus prevalence among males than females. There were 8 different types of definitions of tinnitus, the most common being "tinnitus lasting for more than five minutes at a time" (34.3%). Only seven studies gave any justification for the question that was used, or acknowledged the lack of standard questions for tinnitus. There is widespread inconsistency in defining and reporting tinnitus, leading to variability in prevalence estimates among studies. Nearly half of the included studies had a high risk of bias and this limits the generalisability of prevalence estimates. In addition, the available prevalence data is heterogeneous thereby preventing the ability to pool the data and perform meta-analyses. Sources of heterogeneity include different diagnostic criteria, different age groups, different study focus and differences in reporting and analysis of the results. Heterogeneity thus made comparison across studies impracticable.

CONCLUSION

Deriving global estimates of the prevalence of tinnitus involves combining results from studies which are consistent in their definition and measurement of tinnitus, survey methodology and in the reporting and analysis of the results. Ultimately comparison among studies is unachievable without such consistency. The strength of this systematic review is in providing a record of all the available, recent epidemiological data in each global region and in making recommendations for promoting standardisation.