Internet search results correlate with seasonal variation of sarcoidosis

Amanda Stanton,Steven J Katz

doi:10.1186/s12890-021-01602-7

Abstract

BackgroundThe etiology and pathophysiology of sarcoidosis remains unclear, with epidemiologic studies limited by its relatively low prevalence. The internet has prompted patients to seek information about medical diagnoses online; Google Trends provides access to an anonymized version of this data, which has a new role in epidemiology. We hypothesize that there is seasonal variation in the relative search interest of sarcoidosis, which would suggest seasonal variation in the incidence of sarcoidosis.MethodsGoogle Trends was used to assess the relative search volume from 2010 to 2020 for “sarcoidosis” and “sarcoid” in 7 countries. ANOVA with multiple comparisons was performed to compare the mean relative search volume by month and by season for each country, with a p-value less than 0.05 indicating statistical significance.ResultsOur analysis revealed a significant seasonal variation in search popularity in 4 of the 7 countries and in the Northern Hemispheric countries combined. Direct comparison showed search terms to be more popular in spring, specifically March & April, than in the winter. Southern Hemisphere data was not statistically significant but showed a trend towards a nadir in December and a peak in September and October.ConclusionsOverall, these findings suggest seasonal variation with a possible peak in spring and nadir in winter. This supports the hypothesis that sarcoidosis has seasonal variation and is more commonly diagnosed in spring, but more evidence is needed to support this, as well as investigation into the pathophysiology of sarcoidosis to explain this phenomenon.

Highlights

The etiology and pathophysiology of sarcoidosis remains unclear, with epidemiologic studies limited by its relatively low prevalence
Sarcoidosis is an inflammatory condition of unknown etiology, characterized by non-caseating granulomas
More than 90% of patients will have pulmonary involvement, but sarcoidosis can manifest in almost any body system, including the lymph nodes, joints, skin, heart, and GI tract [1]

Summary

Introduction

The etiology and pathophysiology of sarcoidosis remains unclear, with epidemiologic studies limited by its relatively low prevalence. The internet has prompted patients to seek information about medical diagnoses online; Google Trends provides access to an anonymized version of this data, which has a new role in epidemiology. With trillions of Google searches every year, Trends is “one of the world’s largest real-time data sets” [10]. Scaled on a range of 0 to 100” [9]; as such, a value of 100 indicates the data point with the highest relative search volume, necessitating the remainder of the data points be lower than this. This cannot be used to infer absolute search volumes

Objectives

Methods

Results

Discussion

Conclusion