Objectives The way health literacy is understood (conceptualised) should be closely linked to how it is measured (operationalised). This study aimed to gain insights into how health literacy is defined and measured in current health literacy research and to examine the relationship between health literacy definitions and instruments.
Design Systematic review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.
Data sources The MEDLINE, PsycINFO, ERIC and CINAHL databases were searched for articles published during two randomly selected months (March and October) in 2019.
Eligibility criteria We included articles with a quantitative design that measured health literacy, were peer-reviewed and original, were published in the English language and included a study population older than 16 years.
Data extraction and synthesis Six researchers screened the articles for eligibility and extracted the data independently. All health literacy definitions and instruments were considered in relation to category 1 (describing basic reading and writing skills, disease-specific knowledge and practical skills) and category 2 (social health literacy competence and the ability to interpret and critically assess health information). The categories were inspired by Nutbeam’s descriptions of the different health literacy levels.
Results 120 articles were included in the review: 60 within public health and 60 within clinical health. The majority of the articles (n=77) used instruments from category 1. In total, 79 of the studies provided a health literacy definition; of these, 71 were in category 2 and 8 were in category 1. In almost half of the studies (n=38), health literacy was defined in a broad perspective (category 2) but measured with a more narrow focus (category 1).
Conclusion Due to the high degree of inconsistency between health literacy definitions and instruments in current health literacy research, there is a risk of missing important information about health literacy considered be important to the initial understanding of the concept recognised in the studies.
PROSPERO registration number CRD42020179699.
- public health
- general medicine (see internal medicine)
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
To our knowledge, this is the first systematic review to investigate connections between health literacy definitions and instruments used in current health literacy research.
The health literacy definitions and instruments were categorised based on a health literacy model described by Nutbeam.
The initial plan to assess health literacy definitions and instruments for three categories was changed to two due to difficulties in distinguishing between two of the categories.
Searches were limited to two randomly selected months.
Health literacy is usually understood as cognitive and social skills that determine the motivation and ability to understand and use health information,1 and adequate health literacy is seen as a prerequisite for healthy behaviours. Researchers have increasingly worked to identify challenges associated with health literacy and investigate the role they play in an individual’s ability to comprehend self-care information and its relationship to health outcomes.2 Empirical studies have reported that low health literacy is associated with poor health-related outcomes, such as high hospital admission rates,3 4 low participation in preventive activities,5–9 poor self-management of chronic conditions,10 poor disease outcomes and high mortality.11–13
The concept of health literacy emerged in the 1970s when health education was viewed as social policy.14 A topic-specific query in the PubMed tools reveals a recent exponential growth of articles about health literacy, with 129 references between 1986 and 1990 increasing to more than 8000 in the past 5 years. Today, health literacy is seen as a global goal for enhancing health promotion through improved education and communication strategies to improve health outcomes.15
Health literacy is defined in numerous ways.14 16–18 In a systematic review by Sorensen et al,14 17 different definitions and 12 conceptual models were reported on health literacy. Another systematic review found that 34 research articles between 2010 and 2015 had an explicit objective to define the concept of health literacy.17 Moreover, the literature has reported the use and development of more than 150 health literacy instruments over the last decade.19–21 Traditionally, health literacy approaches have focused on individual skill deficits and health education based on the communication of factual information regarding health risks as well as how to use the health system,22 with the majority of health literacy research having used instruments measuring reading and numerical skills.23 However, in recent years, more multidimensional perspectives and instruments measuring health literacy have been introduced, such as the Health Literacy Questionnaire24 and the Health Literacy Survey European Questionnaire 47.25
The current and sometimes confusing use of various and inconsistent interpretations of health literacy is a challenge for the development of valid and reliable measurements.16 In 2000, Nutbeam proposed a health literacy model that is now widely cited in the health literacy literature and is seen by many health literacy researchers as useful in analysing health literacy abilities required in various health situations.22 According to Nutbeam, health literacy contains three different levels, progressing from basic skills in reading and writing (functional health literacy), to the ability to derive meaning from different forms of communication and apply new information to changing situations (interactive health literacy) and to more advanced cognitive skills which, together with social skills, can be applied to critically analyse information and to achieve policy and organisational changes (critical health literacy).
Different understandings of health literacy and different measurement tools may be useful as they complement each other and provide different perspectives. However, the way health literacy is understood (conceptualised) should be closely linked to how it is measured (operationalised) in each study context.23 Nguyen et al has described this as a ‘conceptual stumbling block’ that needs to be resolved for the field to progress.23 A first useful step for addressing this might be to systematically explore how it appears in current health literacy research. Hence, by performing a systematic review, our aim was to gain insights into how health literacy is defined and measured in current health literacy research. In particular, we will examine the relationship between health literacy definitions and instruments. This review may increase our understanding of potential conceptual and methodological challenges or gaps that need to be addressed in future research.
This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement26 and registered in PROSPERO (https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=179699). The review was designed with a time frame limited to two randomly selected months in 2019 (March and October). Due to a high number of health literacy articles published every year, random selection was chosen in order to reflect current health literacy research. To ensure that March and October were not unique in terms of the number of articles published, we performed the same search strategy using the same databases for one other randomly selected month in 2019. This search yielded nearly the same number of articles.
Systematic literature searches were conducted in collaboration with a trained librarian (4 March 2020). The MEDLINE, PsycINFO, ERIC and CINAHL databases were searched for the term ‘health literacy’. For all databases except for ERIC (where this was not possible), the search was automatically restricted to two randomly selected months in 2019: March and October. Citations in ERIC were manually assessed for articles published in March 2019 and October 2019.
The search yielded a total number of 1038 citations. Endnote V.X9 was used to manage the generated research articles. After removing duplicates, 951 citations remained. All records not published for the first time in 2019 were removed, leaving 414 articles for screening (see online supplemental appendix 1 for the search history).
The inclusion and exclusion criteria were developed a priori. The 414 published articles were distributed among six researchers (KHU, AKW, MHA, CRhB, SH and MHL), who worked in pairs. The articles were included if they fulfilled the following inclusion criteria: (1) ‘health literacy’ was mentioned in the title or abstract; (2) the article was peer reviewed; (3) the research was original; (4) it used a quantitative design; (5) it was published in the English language; and (6) the study population included individuals older than 16 years of age.
To reduce selection bias, the researchers independently screened the articles for eligibility according to the criteria. They then met in pairs to compare their results, resolve any conflicting opinions and decide whether to include each article. Conflicting opinions in pairs were presented and solved by the whole team.
Data extraction (selection and coding)
To achieve consistency in the data extracted from the included articles, an Excel V. 2019 spreadsheet was created. The initial question in this spreadsheet was whether health literacy was measured. If the answer was no, the article was excluded. The data extracted from the articles included information about the study design and context, such as country of origin and whether the study was conducted in a public health or clinical health setting. For clinical studies that included participants with health problems, the different types of diseases were categorised according to the International Statistical Classification of Diseases and Related Health Problems.27 For the public health studies, we categorised type of study populations . Furthermore, the data included information about whether the instruments were generic or disease specific, whether it was used in combination with other health literacy instruments or if a reference was provided for the instrument.
The published articles were analysed with respect to which health literacy definitions and instruments were used. Before the data extraction, a pilot assessment of three studies was performed to determine the feasibility of the data extraction tool. The initial plan was to assess health literacy definitions and instruments for three categories: (1) functional, (2) interactive and (3) critical health literacy, inspired by Nutbeam’ s description of different health literacy levels.22 As a result of the pilot testing, the number of categories was changed due to difficulties in distinguishing between the two latter levels. All health literacy definitions and health literacy instruments were therefore considered in relation to two categories: category 1 included definitions and instruments describing basic reading and writing skills, disease-specific knowledge and practical skill competencies needed to function in everyday situations; and category 2 included definitions and instruments that also described health literacy as skills to communicate and interact with healthcare providers as well as the ability to interpret and critically analyse health information (online supplemental appendix 2 illustrates the coding for the data extractions).
The researchers first extracted data and considered independently the definitions and instruments related to the two categories. They then met in pairs to compare their results and resolve any conflicting opinions. The remaining conflicting opinions within pairs were presented and solved by the whole team in a group meeting.
Patient and public involvement
No patient was involved.
The search yielded 1038 articles in total (see the online supplemental appendix 1 for the search history). After removing duplicates, 951 articles remained. Of the 163 articles that were read in full text, 1 was a duplicate, and 39 were excluded because they did not measure health literacy. Ultimately, 120 articles were included (see figure 1).
Characteristics of included studies
Sixty articles described studies with a clinical health focus, and 60 were conducted in public health settings. The clinical studies included a great variety of diseases, but the most frequent diseases were those in the circulatory system (n=11) and endocrine diseases (n=9) followed by mental illness (n=6) and cancer diseases (n=6). In the public health studies, the most frequent study group was the general population (n=19), looking into aspects such as vaccination programmes, oral care and mental health. Furthermore, a major part of the public health studies included students (n=13), mainly represented by health professional students. Six studies focused on parents/caregivers, while five studies focused on the elderly population.(table 1)
The majority of the studies used a cross-sectional design (n=90). Nine studies used a randomised controlled design, and 10 studies were quasi-experimental. Four of the studies used a longitudinal design, while one was a case–control study. Figure 2 presents an overview of all study designs.
As figure 3 indicates, the majority of studies were conducted in North America (n=56) and Asia (n=31). The fewest studies were performed in Africa (n=4) and South America (n=3).
Health literacy definitions and instruments
Out of the 120 included studies, 88 used generic health literacy instruments, while 32 were context-specific. Eleven studies used a combincation of health literacy instruments. A total of 77 studies used instruments from category 1 (describing basic reading and writing skills, disease-specific knowledge and practical skills), whereas 43 instruments were from category 2 (describing communication and interaction skills and/or the ability to interpret and critically analyse health information). The most frequently used health literacy instruments in category 1 were Newest Vital Sign28 (n=19) and Test of Functional Health Literacy29 (n=13). The most frequently used health literacy instruments in category 2 were The European Health Literacy Survey Questionnaire30 and The eHealth Literacy Scale31 (n=10).
Of the 120 articles, 79 provided a health literacy definition. Of these, 46 were public health studies and 33 were clinical health studies. Only eight studies provided a definition addressed to category 1. Five out of these eight studies focused on mental health challenges.
The most frequently used reference in category 2 was Nutbeam.22 32 33 ,34 The most frequently used reference in category 1 was a definition provided by Jorm and colleagues.35 36 Table 2 gives an overview of the text in the most cited health literacy definitions. Words in italics are examples of key words considered important in the decision to add the definition to category 1 or category 2. For instance, definitions that described health literacy with terms such as ‘knowledge’, ‘skills’ and ‘attitude’ were referred to category 1, while definitions with terms like ‘appraise’ and ‘social skills’ were referred to category 2.
Regarding the connection between health literacy definitions and instruments in the 79 relevant studies, 41 articles used health literacy definitions and instruments characterised at the same level. Thirty-three of these were in category 2, and 8 were in category 1 (see figure 4 and table 3).
In the remaining 38 studies, there was a disconnect between levels of health literacy definitions and instruments. In all of these, health literacy definitions were from category 2, and all instruments were in category 1 (see table 4).
This systematic review aimed to gain insights into how current research defines and measures health literacy and, in particular, whether studies consistently used definitions and instruments. The high number of articles published in the defined time frame shows that health literacy is of high research interest in both public health and the clinical field in large parts of the world.
We found a large variety of instruments used, and the majority of the included studies (79 out of 120) presented a health literacy definition as part of the study’s theoretical background. However, there seems to be an inconsistency between the definitions and the instruments in a significant number of the studies. In nearly half of the studies, health literacy was defined in a broad perspective (including aspects such as social health literacy competence and the ability to process and appraise health information) while using instruments with a more narrow focus (measuring basic skills and knowledge). As a result, almost half of the articles in our review lacked data on the participants’ ability to critically appraise health information and their social health literacy competence despite the fact that the authors had stated such aspects to be health literacy. This concern has previously been addressed. Numerous systematic reviews have reported on the diversity of understandings of health literacy and the various use of instruments not aligned to the definitions in current research.16–21 However, the current study is, to the best of our knowledge, the first review to systematically investigate the relationship between health literacy definitions and instruments from the perspective of a health literacy model.
The use of instruments that focused on functional health literacy (addressed to category 1), such as Newest Vital Sign and Test of Functional Health Literacy, were predominant in our review.28 29 This finding corresponds with previous findings.19 20 The widespread use of Newest Vital Sign and Test of Functional Health Literacy is justified by the fact that they are screening tools—that is, they are quick, available in an ever-increasing number of languages and can be adapted to different settings.37 However, a broad range of skills and tasks covering functional, interactive and critical domains needs to be included in measures of health literacy in order to capture the health literacy definitions used.32 In recent years, an increasing number of multidimensional instruments have been developed. In our review, The European Health Literacy Survey Questionnaire and The eHealth Literacy Scale, both addressed to category 2, were the multidimensional instruments most often used.30 31 In light of the increased focus on a more multidimensional perspective of health literacy, and the fact that the current review only includes studies from 2019, one could expect a higher use of multidimensional instruments. Instead, they represented only 43 out of 120 instruments.
Clearly, our study highlights the discrepancy between definitions and measurements as well as the narrow focus of health literacy in a large amount of empirical research. An important question to raise is therefore how the findings from this specific study can guide future research strategies to overcome the identified inconsistency. In other words, as most health literacy researchers seem to base their research on a broad understanding of health literacy, what can be done to facilitate an increased use of broad measurements? A first step should be to make researchers aware about the existing mismatch in current research. Furthermore, it seems necessary to develop more instruments that can answer the research questions posed. Despite the high number of instruments, there still seems to be a need for questionnaires in the field of health literacy that capture more multidimensional dimensions besides the functional aspects.
Nutbeam’ s description of three levels of health literacy inspired the categorisation of definitions and instruments.22 Initially, we planned to distinguish among the three levels. However, the line between levels 2 and 3 was challenging to distinguish. Therefore, it was decided to merge the two latter categories. In a recent publication from 2020, Nutbeam has provided a more thorough description of the three levels with a more detailed explanation of how the levels should be understood.38 Perhaps, these descriptions would have contributed to a clearer guidance in our work with the categorisation. However, this material was not available at the time of our work and, in general, this situation illustrates the challenge of adapting a theoretical model into practise.
The interpretation of terms used in the definitions and instruments that guided the choice of category also represented some challenges. These were resolved through discussions both in pairs and as a research team. For instance, definitions describing health literacy using terms like knowledge and ‘beliefs’ were included in category 1, while definitions using terms such as appraise’ and ‘understand and process’ were included in category 2. An interesting finding is that, among the eight studies providing a health literacy definition from category 1, the majority (n=5) were in the context of mental health. The predominant reference in these studies was the definition provided by Jorm and colleagues, who, unlike others, defined health literacy as being linked to beliefs and ‘attitudes’.35 36 Whether these terms are more closely linked to mental health literacy challenges, compared with other more ‘physical’ health literacy issues, is not clear but would be interesting for further investigation.
Another aspect worth noting is that we discovered that many publications did not cite the primary source of the health literacy definitions but instead referred to secondary sources (other researchers presenting definitions of health literacy). Referring to the original sources should be the first choice and would perhaps make it easier for readers to recognise the definition’s affiliation.
The present study has some limitations. First, this study was designed to analyse and describe health literacy research in two randomly selected months. This period may not be representative of health literacy research in general. However, a large number of health literacy studies are published every year. A random selection can therefore give a good picture of health literacy research. Second, we did not conduct a quality assessment screening of the included studies. This was considered less relevant for the current study as the aim of the current study was to explore connections between health literacy definitions and instruments rather than to assess methodology. Furthermore, searches were limited to the English language only. It is possible that similar studies may have been published in languages other than English.
The current review included only quantitative measurements. However, qualitative approaches might provide valuable and more in-depth insights into the field. For future research, it would be interesting to also explore how qualitative research links health literacy definitions to the research questions posed.
There seems to be an inconsistency between the definitions and the instruments used in a significant part of current health literacy research. This situation raises the risk of missing information about health literacy that was considered to be important in the initial understanding of the concept recognised in the studies. This gap should be taken into consideration in future health literacy research. We hope our work contributes to making explicit where the problem might be rooted and that it can be useful in the discussion about strategies for moving forward to better align health literacy measurement with definitions of health literacy.
Data availability statement
Data are available upon reasonable request.
Patient consent for publication
This study does not involve human participants.
Contributors KHU, MHA and AKW initiated the project and wrote and revised the manuscript. KHU led the project. KHU, MAH, AKW, SH, CRB and MHL designed the study, selected the articles and extracted the data. All authors contributed to drafting the manuscript and met authorship criteria. KHU had full responsibility for the work and the conduct of the study, had access to the data, and controlled the decision to publish.
Funding The Centre for Advanced Study in Oslo, Norway, funded and hosted our research project, The Body in Translation: Historicising and Reinventing Medical Humanities and Knowledge Translation, during the 2019/2020 academic year.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.