Winter peaks in web-based public inquiry into epistaxis

Epistaxis, commonly known as nosebleed, is the most frequent ear, nose, throat (ENT)-related emergency symptom [1]. Simple first aid intervention (e.g., compression) can stop acute hemorrhage in the majority of cases, resulting in epistaxis being a self-limiting symptom with only 10% of patients seeking medical help [2]. Indeed, it is a common belief that meteorological factors such as air temperature and humidity affect the incidence of epistaxis, with maximal and minimal incidences occurring during the winter and summer months, respectively [1, 3‐8]. However, the scientific literature on this topic is inconclusive. While some studies showed no association between seasonal variation and incidence [4, 9‐11], others have shown the opposite pattern of results [1, 3, 5‐8]. These open issues relate not only to the seasonality of epistaxis incidence, but importantly also to the validity and timing of online educational measures.

In recent years, for better or worse, the World Wide Web has become one of the most important sources of information available to the general population [12]. A report from 2013 showed that no less than 72% of all internet users have already searched for health-related issues. Particularly worthy of mention is Google web search, which represents the most commonly used search engine worldwide and provides the data foundation for further analysis [13]. A popular tool for big data mining is represented by Google Trends (GT), which allows users to explore oscillations in public interest, by selecting filters on geographical location and time [14]. Owing to these circumstances, it is not surprising that many health-related issues have already been associated with internet-search peaks in recent years to complement traditional data sources [15‐22]. Noteworthy, previous studies on nose bleeding and GT suffered from two major methodological shortcomings: First, no systematical approach was chosen for evaluating epistaxis-related search terms that are representative for the general population. Cervellin [12] and Braun et al. [23] used the medical terms [epistaxis] and [epistassi], rather than the country-specific non-medical term [nasenbluten] or [sangue dal naso] as primary search terms. Secondly, the validity of the conclusions was compromised due to the absence of hypothesis testing. Braun et al. [23] investigated the search inquiry for epistaxis-related search terms in Germany, but provided only graphical evidence. Given these limitations, there is currently no clear indication about seasonal variations of public interest in epistaxis-related search terms on a worldwide basis.

Therefore, the aim of this study was to utilize Google Trends to test our hypothesis that search query frequencies of relevant epistaxis-related terms show peaks during winter months, independent from the hemisphere. Additionally, extracted GT data were assessed for validity and reproducibility, particularly in countries with fewer inhabitants. Lastly, country-specific search query frequency was compared with ambient weather factors. The results are discussed in terms of the validity of internet tools to predict peaks in information demand for nosebleeds, and of the measures needed to guide patients to reliable medical tips for stopping epistaxis.

The vast majority of epistaxis cases can be handled with simple measurements. At the same time, epistaxis represents a very frequent symptom in outpatient emergency departments that might also require acute intervention. Although it is generally assumed (and even taught) at university level that the incidence of epistaxis increases during winter months, the literature itself remains contradictory. In this regard, Google Trends has become a valuable tool to expand existing knowledge and provide further information for hypothesis formulation [21, 22]. Here, we showed for the first-time seasonal differences in Google search query frequency for epistaxis-related search terms in various countries, from both hemispheres, indicating that the search query frequency for epistaxis-related search terms follow a seasonal pattern. In addition, these differences were also evident in non-English-speaking countries, which supported our primary findings.

It should be emphasized that although our data showed seasonal variations in Google search volume for epistaxis related terms, the current results must be interpreted cautiously, as the nature of these data do not allow biological causalities to be determined. A general increase in overall medical inquiries during winter weather conditions (e.g., more time at home) may be a relevant confounding factor and has to be taken into consideration in future studies on seasonal differences in Google search volume for health-related issues [32]. However, the finding that Google search query frequency for epistaxis-related symptoms increased during the winter months was not altogether surprising. It has been reported previously that epistaxis is more frequent in winter weather conditions [1, 3‐8, 33]. One possible cause is related to the dry and heated-, or cold, air during the winter months (indoor or ambient), which leads to nasal mucosa desiccation, making it friable and vulnerable. Interestingly, Cruz et al. [34] studied a subgroup of patients that were sensitive to cold and dry air (e.g., having typical rhinitis-associated symptoms such as a congested nose during winter months). It was found that nasal challenge with cold dry air lead to a statistically significant higher count of epithelial cells being shed away (“epithelial shedding”) compared to warm and moist air, thus providing evidence for the mucosa-desiccation theory in this subgroup of patients with non-allergic rhinitis.

In reference to the negative correlation between temperature and search volume of epistaxis related search terms, this was commensurate with previous results relating to meteorological factors and incidence of epistaxis in adults presenting to an outpatient emergency clinic. Comelli et al. [7] identified a strong, negative correlation for temperature, and a moderate, negative correlation for humidity and epistaxis incidence. Similarly, Sowersby et al. [8] also found a strong, negative correlation for ambient temperature, although no correlation was found between humidity and epistaxis incidence. On the other hand, Reddy et al. [4] reported no relevant correlation between daily incidence of epistaxis and weather variables including ambient temperature, atmospheric pressure, and water vapour pressure. Interestingly, we found a negative correlation between vapor pressure and search volume of epistaxis related search terms. It is not yet possible to explain the reason behind this relationship, however, this finding might serve as a novel reference point from which to consider further studies on epistaxis incidence and vapor pressure. Based on current data, the association between meteorological variables and epistaxis incidence can only be speculated upon.

Considering the direct correlation of air temperature and incidence of respiratory tract infection of the upper airway (e.g. “common cold”) it is not surprising that viral infections have already been linked to epistaxis [35‐38]. Other factors that might lead to the higher incidence of epistaxis during the winter months which may overstrain the already stressed mucous membrane include: (i) changes at the cellular level, such as the disruption of the epithelial cell barrier, and (ii) the mechanical effect of increased sneezing or nose blowing when affected by cold [39].

First aid measures in cases of epistaxis had been described previously and outlined in detail by Tunkel et al. [40]. Another plausible explanation for the higher Google search volume in epistaxis-related search terms during the winter months may be related to the self-limiting nature of this condition, as it is assumed that only 10% of all cases will require medical assistance [2]. Considering that (i) internet penetration in the countries surveyed was among the highest in the world [41], (ii) 72% of the general population already searched for health-related information online [42], and (iii) Google represented the most popular search engine (~ 65% of the total volume [13]), it is tempting to speculate the existence of a linear correlation between search query frequency for epistaxis-related search terms and epistaxis incidence.

Regarding the issue of non-English-speaking countries, these also showed a significant seasonality with peaks during the winter months in Europe, which supported our primary hypothesis. A 2013 GT study using [epistaxis] as search term in Germany concluded that the seasonality of epistaxis was coincidental with the winter months [23]. Our present results supported this hypothesis, albeit we additionally demonstrated seasonality for the German word [nasenbluten], representing the epistaxis search term with the highest relative search volume in Germany, based on statistical methods.

The authors of a previous study on association between GT time series data of epistaxis related search terms [sangue da naso] and [epistassi] covering an Italian province of about ~ 450.000 inhabitants and real population epidemiology of epistaxis patients visiting the local emergency department concluded no overlap [12]. From the present perspective, a major factor for this conclusion may have been missed, namely a systematic approach for representative epistaxis-related search terms. Notably, [sangue dal naso] had a ~ 11-fold higher RSV compared to [sangue da naso] during our search queries. Additionally, we support the conclusion of Tran et al. [25] on the reliability of GT time series data retrieved from countries with smaller population sizes, as individual GT time series data were also less reliable in our study. Therefore, we also recommend assessing GT time series data on several days in a row and to verify the reliability of included data, and in cases of poor reliability, to use averaged time series data for analysis purposes. Interestingly, the country with the smallest population size included (New Zealand) also showed the poorest reliability of individual time series and, more importantly, was also the only country undergoing a change in seasonal peak when using averaged time series data, which supported this hypothesis.

Concerning the clinical relevance of this work, our results might have importance at an epidemiological level. Epistaxis requires immediate action by affected individuals, which raises important challenges, related to distribution of reliable and easy understandable advices. The potential danger of misinformation (i.e., the common misconception “Tilt the head back during epistaxis”) underlines the need to make greater efforts in web-based guiding and monitoring of health-information seeking individuals [43]. A further important point is reflected in timing for providing key facts and education to the general public about first aid measures, based on seasonality and incidence peaks. By efficiently distributing and disseminating first aid information among the population, this might reduce the number of hospital visits, ultimately leading to greater financial economy.

Global public interest in seeking epistaxis-related information through the internet displayed seasonal patterns, with the highest interest peaking during winter months. This novel finding may provide further evidence for the seasonality of epistaxis incidence. However, additional clinical and experimental studies will be warranted to explore any causalities with obvious factors such as meteorological variables.