This study represents the first population-based survey on comprehensive health literacy in relation to coronavirus information and the first to adapt the HLS-EU-Q to assess coronavirus-related health literacy. It is also the first study analyzing health literacy in the context of the COVID-19 infodemic in order to gain insights in relation to feeling informed about the coronavirus or confused by the sheer amount of coronavirus information.
4.1. Coronavirus-Related Health Literacy
We found that 50.1% of our sample had “problematic” (15.2%) or “inadequate” (34.9%) coronavirus-related health literacy, whereas 49.9% had “sufficient” health literacy. Mean scores show it was, on average, “easy” for participants to deal with coronavirus information. However, coronavirus-related health literacy should still be increased, as a significant share of participants score below the average health literacy level. Although an adapted questionnaire, the HLS-COVID-Q22, was used, these findings are in line with previous studies on adults’ general health literacy in Germany: The Health Literacy Survey Germany (HLS-GER) [37
] found that 54.3% of the population (n
= 2000) had “inadequate” and “problematic” health literacy, while the German Health Update (GEDA) study [45
] resulted in a slightly lower estimate (44.2% of n
= 4854). Further, our findings are comparable to what has been found in previous studies on adults’ general health literacy in Europe (47%, “inadequate” or “problematic” health literacy [28
]). However, while a social gradient of health literacy has been observed in multiple studies [28
], this was not the case in this study. The missing social gradient could be based on the sample characteristics of the original panel that might not be representative of those with lower social status. However, a missing social gradient could also indicate that ample information is available, and that the given information environment makes it easy to access, understand, appraise, and apply health information in everyday life. During the first wave of the COVID-19 outbreak, when this survey was conducted, adherence to public health interventions and government policies was very high in Germany [55
], i.e., diverse information on coronavirus and recommendations were followed appropriately. This might also explain the missing differences in health literacy regarding sex, age, and state of residence. In addition, information and recommendations were simple and therefore easy-to-understand and easy-to-apply (e.g., hand washing, physical distance, wearing masks, avoid public gatherings, staying home). However, adherence is already declining significantly due to the various adverse effects on the economy, occupation, health, and social life [57
]. This makes lower health literacy a threat to the effectiveness of public health measures to contain the virus the longer the pandemic lasts, with people less likely to follow official public health recommendations.
4.2. Information Tasks and Challenges
Most participants report little difficulty with accessing information about coronavirus and protective behaviors on the internet, or understanding information provided by health professionals, authorities, and family members (items # 1–3, 7–11, 15, 19–20, 22). The most significant association was found for education in relation to tasks associated with all four action areas (# 1, 2, 3, 7, 9, 14, 15, 19).
However, there are also a number of items that highlight critical aspects of dealing with coronavirus-related health information. Between ~20% and 52% of participants report that it is difficult or very difficult to access, understand, appraise, and apply coronavirus information (# 4–6, 12–14, 16–18, 21). Those “difficult” action areas are concerned primarily with medical information regarding infection (# 4, 17), help-seeking (item 5), evaluation of personal risk (# 6, 12, 14, 16), and media information (# 4, 5, 17).
For 20.9%, it is difficult to decide how to protect themselves from infection with the coronavirus based on media information. Even more citizens (32.1%) report that it is difficult to use media information to decide how to act in case of coronavirus infection. Most strikingly, 47.8% of participants report that it is difficult or very difficult to judge whether they can trust media information on the coronavirus, which is similar to findings of the HLS-GER in relation to disease-related media information [37
]. Trustworthy information about the virus is vital for the population’s ability to take part in informed, healthy, and responsible behavior as well as for an effective system-wide response in line with government policies on coronavirus containment and burden reduction for the healthcare system and professionals [19
]. Low trust in health information can cause serious problems during the pandemic, especially regarding the ongoing infodemic and the associated spread of dis- and misinformation regarding COVID-19, potentially leading to irrational behavior endangering the success of prevention measures [19
]. Further, if the information on the pandemic and related preventive measures is not perceived as reliable, communication between national and local governments, including public health and research institutions on the one hand and citizens on the other hand, is impaired, casting doubt on the broad acceptance of future national and evidence-based strategies for protecting public health. Thus, there is a need for measures increasing the reliability of available and accessible coronavirus information and related actions, as well as increasing the trust of the public in such information. Another approach to enabling the general population to judge whether or not they can trust media information on the coronavirus is to promote critical health literacy [22
], which can help people identify reliable sources of information and investigate which information might be misleading.
Accessing, understanding, and appraising information about risks associated with coronavirus, coronavirus infection, and adverse health behaviors pose further problems. More than 20% of participants find it difficult to understand media-based risk information about coronavirus. Deciding which behaviors can help most to lower infection risk is difficult for 22% of the population. More than 30% of adults perceive difficulties judging if they belong to a high-risk group for infection, and 33% even have difficulties finding information on risks. These results call for increased and improved risk communication, adapted to the needs of the population.
Further, it is difficult for almost 40% of the participants to find information on how to recognize a likely infection with coronavirus, and almost 52% say it is difficult to judge whether they may have been infected. These findings are worrying, as knowledge of one’s own infection status may be an important determinant of whether people adhere to distancing rules and other preventive measures [16
]. Although evidence on immunity after being infected with COVID-19 is inconclusive [58
], people that assume to have already been infected, e.g., based on symptoms similar to those of COVID-19, might believe to be immune and put themselves and others at risk by not adhering to preventive measures based on a false sense of safety [19
]. It has been shown that even health professionals and researchers face difficulties providing accurate information on symptoms, which is why controversial information is available on the matter [8
]. In addition, information on COVID-19 changes rapidly, which may contribute to people having difficulties judging what information is reliable [18
]. Going forward, this could potentially jeopardize public adherence to restrictions and preventive measures, especially in the case of a second wave of infections.
Finally, participants also report having difficulties accessing information on how to find professional help in case of coronavirus infection. This can pose a serious threat in emergency situations. Therefore, efforts are needed to improve the transparency and accessibility of the healthcare system regarding diagnosis and treatment options regarding COVID-19.
4.3. Infodemic and Associated Influences on Information Uptake and Use
4.3.1. Being Informed about Coronavirus
A total of 90% of participants feel well-informed or very well-informed about coronavirus, irrespective of sex, age, education, or living with a child <18 years. However, participants with a lower income and participants living in the federal states of former East Germany were found to feel significantly less informed. This points to the presence of an economic as well as a regional gradient. While our data does not provide further clues on reasons for these differences, a number of reasons are plausible. Regarding economic differences, it is possible that the information on coronavirus available is sufficient for the majority, but that low-income participants have specific information needs that are not fully satisfied by the information circulating in major news channels. For instance, specific questions might arise more frequently among low-income participants during the pandemic concerning topics such as unemployment, social benefits, and social security, or youth services. The regional differences, however, may result from differences in patterns and frequencies of internet use in eastern and western Germany. While the latest Eurostat data shows that 9 out of every 10 households have internet access [61
], there are differences in internet use between households in the east and west of Germany with people living in eastern regions using the internet significantly less than their counterparts [62
]. Another explanation could be that the outbreak of COVID-19 started in the west of Germany [63
], which is why at the time of this survey, people living in the east of Germany may have felt less concerned about coronavirus and less engaged with seeking information.
However, the overall level of feeling informed is high, which might, at least partly, be attributed to government-led public health and media campaigns early on in the pandemic using different channels including social media. This finding may also reflect how well German public health authorities and public broadcasting agencies have reacted to the pandemic and associated challenges for society and individuals.
4.3.2. Being Confused about Coronavirus
Despite the findings, more than half of the German population (56%) feels somewhat confused or very confused about the amount of coronavirus information. Although it was expected that confusion and feeling informed were diametrically opposed indicators, this was not always the case. Even though a negative correlation was found between those variables, it was only small in size (r = −0.224). This means that there are a number of participants that feel well-informed and, at the same time, confused by all the information. Hence, it is not only necessary to provide citizens with coronavirus information they need so that they feel well-informed, but also to help them make sense of and evaluate additional and potentially conflicting information to avoid confusion.
We found that women reported feeling confused by the amount of coronavirus information more often than men. While the virus and virus information are not gender-specific, it may be related to the unequal distribution of care responsibilities, which is still prevalent in Germany [64
]. As women in Germany are more often engaged in the upbringing of children, providing childcare and care of elderly family members [64
], and more likely than men to seek online health information [66
], it is possible that they are confronted more often with confusing coronavirus information.
Further, we found that younger people were more confused than older people by coronavirus information. This can partly be explained by a more prevalent media use among younger people and thus the higher exposure to a broad range of potentially conflicting information, but there are probably other reasons that are unaccounted for in our data, such as frequently using social media websites (e.g., Twitter, Facebook, and Instagram), reading more about coronavirus, having fear of COVID-19, being stressed by the pandemic, and feeling uncertain regarding their future and the many changes that arise from the impact of COVID-19 on their lives.
4.3.3. Health Literacy and Infodemic
Regarding the relationship between feeling informed or confused and health literacy, we found that people with “inadequate” and “problematic” health literacy were significantly less likely to feel well-informed and more likely to feel confused by the amount of coronavirus information. It has to be noted that assumptions about cause and effect are difficult here, as “feeling informed” could be regarded as equivalent to two main dimensions of health literacy as measured by the HLS-COVID-Q22, namely the perceived ease in accessing and understanding health information. The same could be assumed for “feeling confused” and another core dimension of health literacy, namely the perceived ease or difficulty in appraising health information.
It has been highlighted that health education and teaching health literacy are critical prevention and health promotion measures for mitigating the adverse effects of the COVID-19 infodemic [10
]. The benefits of fostering health literacy, the acquisition of competencies, and the confidence to handle the sheer amount of information in an infodemic are not restricted to infectious diseases. Moreover, they will also be helpful in different public health areas, such as increasing critical thinking, promoting healthy behavior and managing risk behavior, informed decision making, empowerment, and health outcomes, as well as improving health care interaction, health communication, and adherence.
In order to facilitate health information seeking and understanding, information suppliers and providers, such as official public health organizations and health authorities, have a critical role in providing the public with high-quality health information [59
]. When designing and providing coronavirus and COVID-19-related information, providers must ensure that information is based on health literacy principles such as being easy-to-access, easy-to-understand, easy-to-use, culturally appropriate, and relevant to various populations [20
]. Accessing valid and reliable health information on the internet is among the greatest challenges for internet users [60
]. Social media providers, who are important actors in the infodemic [4
], must be urged to act responsibly, support the provision of reliable health information, and inhibit the spread of dis- and misinformation on their websites [13
]. In this regard, government support is needed to implement policies that hold social media and tech giants accountable, e.g., by fact-checking and flagging false information.
The perpetual stream of information overload that is driven by actors across society amplifies the COVID-19 infodemic. The infodemic must be acknowledged as a meta-risk in its own right, which interferes with people’s health literacy and their health outcomes. Limited health literacy in individuals, populations, and systems can cause adverse effects for parts and even the whole of society. When people lack the competencies to critically appraise health information, suppliers cannot ensure the means to protect valid and meaningful information against the many sources that spread invalid information. This can cause panic, destabilize the effectiveness of information distribution and public health interventions, and even threaten social cohesion and the political landscape. As Eysenbach highlighted in the context of infodemiology studies, “an epidemic of fear may exhibit similar characteristics as a true epidemic” [73
], which is why investment in population and system-level health literacy must be considered absolutely critical for public health responses against toxic health threats, such as coronavirus and COVID-19. This is supported by a recent study on health literacy and fear of COVID-19 in medical students, showing that higher levels of health literacy may protect from fear during the pandemic [38
4.4. Feasibility of the Questionnaire
The HLS-COVID-Q22 is a novel tool in the HLS-family and the first to measure coronavirus-related health literacy. Indicators for internal consistency demonstrate satisfactory reliability (α = 0.940 and ρ = 0.891), suggesting the success of the adaptation process and the feasibility of the instrument. Internal consistency estimates found in this study are in line with what has been found in previous studies using the HLS-EU-Q16 in German language (α = 0.96) [45
], which served as the original blueprint for developing the HLS-COVID-Q22. Internal consistency for the four subscales was also satisfactory, which means that they can be used to assess subjective coronavirus-related health literacy related to the areas of accessing, understanding, appraising, and applying coronavirus-related health information.
Construct validity for the four subscales suggested that the model fit was only moderate. However, the model fit indices clearly did not indicate an excellent model fit, neither based on the moderate cut-off values [49
] that were applied nor based on strict cut-off values as suggested by Hooper and colleagues [53
]. An acceptable model fit was only achieved in some indicators after a slight modification. In this modification step, we allowed the residuals of items 1 and 2, 19 and 20, and 20 and 21 to correlate, expressing a common source of variance. These modifications were acceptable from a theoretical and content-related standpoint, as those item pairs are concerned with finding information on the internet (items 1 & 2), following or using information received from the doctor (items 19 & 20), and using information on how to handle an infection with coronavirus (items 20 & 21). After allowing the correlation between the item’s residuals using moderate model parameters [49
], the model fit indices NFI, TLI, IFI, CFI, RMSEA, and SRMR of Model 2 suggested a sufficient model fit, whereas χ2
/df did not. To reach an excellent model fit based on both the strict [53
] or the more moderate [49
] values employed here, further modifications or even a revision of the underlying model might be necessary. However, we chose not to apply further modifications, such as correlations of residuals across factors (as suggested by SPSS Amos), as this would have implied further, potentially atheoretical deviation from the model which suggests 4 distinct factors. However, the high correlations (r = 0.65–0.90) found between the subscales (and thus, factors) conflict with the assumption of four independent factors and point to a second-order common factor, i.e., content-wise similarities between the highly correlated factors. This might contribute to less-than-optimal model fit and warrant further investigation of the underlying factor structure.
Earlier studies with the HLS-EU-Q have reported similar problems regarding model fit for the four-factorial model, assuming a rather unidimensional model [74
]. Unfortunately, most studies using the HLS-EU-Q do not provide data on construct validity. In a Norwegian study, the model fit was reported for the HLS-EU-Q16 for a one- and three-dimensional model with values for TLI = 0.911 and 0.939, CFI = 0.923 and 0.949, SRMR = 0.080 and 0.070, and RMSEA = 0.118 and 0.103, respectively [74
]. In a study conducted in six Asian countries using the HLS-EU-Q47 [36
], the model fit indices were reported for a three-factorial model representing the areas of health promotion, disease prevention, and health care. In each domain, the four-factor model then included accessing, understanding, appraising, and applying health information. For all six countries, RMSEA values ranged from 0.05–0.10, CFI ranged from 0.90–0.97, IFI ranged from 0.90–0.97, NFI ranged from 0.87–0.96, χ2
/df ranged from 2.55–21.85. In a Japanese study using the HLS-EU-Q47, CFI and RMSEA were reported for the three domains with values for CFI 0.937 and RMSEA 0.075 (health care), 0.943 and 0.079 (disease prevention), and 0.934 and 0.078 (health promotion) [75
]. The model fit indices of the HLS-COVID-22 are comparable to earlier studies. This calls for a more detailed approach to operationalising the four action areas, better reflecting the actions taken to achieve them, i.e., deconstructing the steps necessary to access information.
This study was conducted during the first weeks of the pandemic in Germany while people were already overburdened and stressed with the many changes they had to cope with and apply in everyday life and at the workplace. Therefore, we decided to keep the questionnaire rather short and did not include further scales and variables, such as COVID-19-related health knowledge and behavioral aspects. Based on the HLS-EU-Q, the HLS-COVID-Q22 is a self-report measure and does not assess the performance-based health literacy capabilities of individuals [76
]. For a better understanding of the associations and possible relationships between health literacy on the one hand and knowledge as well as health behavior on the other, we plan to include additional questions to the follow-up surveys (waves 2 and 3 of this measurement) to allow comparisons between subjective health literacy and objective performance-based skills and abilities. These waves will be implemented in 2020 as a three-country survey in Germany, Austria, and Switzerland in order to analyze change over time (German Federal Ministry of Health, grant number: ZMVI1-2520COR009).