METHODS AND METHODOLOGIES
doi: 10.1111/scs.12125
Measuring health literacy – the Swedish Functional Health Literacy scale Josefin M. W angdahl MSPH (PhD Student)1 and Lena I. M artensson PhD, OTR (Senior Lecturer)2 1
Department of Public Health and Caring Science, Social medicine, Uppsala University, Uppsala, Sweden and 2Institute of Neuroscience and Physiology/Occupational Therapy, G€ oteborg University, G€ oteborg, Sweden
Scand J Caring Sci; 2015; 29; 165–172 Measuring health literacy – the Swedish Functional Health Literacy scale
Background: The benefits of health promotion efforts vary due to a complexity of reasons. One possible reason for an absence of effects is the level of functional health literacy among the individuals that participate in the interventions. Thus, valid and reliable instruments that capture these kinds of skills are needed. The aim of this study was to develop a Swedish Functional Health Literacy scale to use in health promotion and health prevention by translating the Japanese Functional Health Literacy scale into Swedish and testing some aspects of its validity and test–retest reliability. Methods: The research project comprised six phases including translation and back translation, validity tests of the two first versions of the instrument based on interviews with professionals and representatives for the target group of the instrument, and a test–retest of the first version among target groups.
Introduction Much work aims at increasing individuals’ health conditions in the form of health promotion and primary or secondary preventive interventions (1), for example health examinations, vaccination programmes, dental hygiene, prescription of contraceptive pills for teenagers, screening for prostate cancer et cetera. The benefits of health promotion efforts vary for a complexity of reasons (2, 3). One possible reason for an absence of effects that has not yet been given much attention is the level of health literacy among individuals that participate in the health-promoting interventions (4). Correspondence to: Josefin W angdahl, Department of Public Health and Caring Sciences, Uppsala University, Uppsala Science Park, Box 564, 751 22 Uppsala, Sweden. E-mail: [email protected] © 2014 Nordic College of Caring Science
Results: The items in the first two versions were experienced as unclear, which led to improvements of the next version. The final version of the translated instrument (the Swedish Functional Literacy scale) showed evidence of content validity, and the test–retest confirmed that the instrument had stability over time with a percentage agreement for the items ranging from 63% to 92% (M = 77.2%). Conclusion: The items in the Swedish version of the scale are equivalent to the original Japanese scale in terms of language and contents and cover the major aspects of functional health literacy as it is defined in the literature. The translated instrument shows stability over time, that is, reliability, at least for a part of the Swedish population. More validity tests of the Swedish Functional Health Literacy scale based on a broader population are needed. Keywords: health literacy, health promotion, health education, prevention, measurement. Submitted 19 January 2014, Accepted 10 February 2014
Health literacy (HL) has a variety of definitions. It basically concerns the ability to read and understand the information needed to carry out actions that preserve or improve health (5–9). A more complex understanding of HL also includes skills for critical reflection and analysis to allow informed decisions in health issues, in combination with the social skills needed to communicate health information in the society (8). Based on its complexity, HL is often divided into three dimensions: functional, communicative and critical HL, ranging from basic forms of cognitive skills to more complex ones, such as the ability to appraise, choose and implement health information in ordinary life and to use new terminology and mediate health information (9, 10). Regardless of the dimension, health literacy is understood as a context-dependent phenomenon, thus varying according to the situation in which a person has a need for information or for obtaining information in health issues (6). In this study, we focus on functional health literacy (FHL), that is, the 165
166
J.M. W angdahl, L.I. M artensson
skills necessary for being compliant with medication regimens, following advice regarding eating habits and life habits in general, understanding information about risky products, being familiar with home principles for self-care et cetera. Low levels of FHL have been associated with being an immigrant, having an uncompleted education, being old, having more hospitalisations, having fewer mammography screenings and influenza vaccines and having a poorer ability to interpret labels and health messages et cetera (11–13). Research also shows that, besides poorer health, low FHL conveys high costs for both individuals and society (14, 15). There are thus reasons for professionals that work with health issues to gain a comprehensive understanding of health literacy and pay attention to people’s FHL in practice. In research, data on FHL have most often been collected using the Test of Functional Health Literacy in Adults (TOFHLA) (16) and the Rapid Estimate of Adult Literacy in Medicine (REALM) (17). The TOFHLA is a multiple choice test that measures the comprehension of health-related concepts described in a medical context, for example patient information (16). The REALM assesses the reading comprehension of 66 de-contextualised medical terms (17). Due to the medical orientation of these assessments, their feasibility in the public health context is doubtful. An instrument that is more general concerning context is the Japanese Functional Health Literacy scale (the JFHL scale) by Ishikawa (18). Like the other Japanese scale, that is, the Communicative and Critical Health Literacy scale (J C&C HL scale), the SFHL scale is based on Nutbeam0 s definition of HL including abilities needed both for health prevention and health promotion, which means that it is related to a greater extent to a public health context (8). The JFHL scale is a short, easily administrated assessment. It comprises five items about persons’ skills in reading and understanding health information. The items can be self-assessed on a five-point Likert scale (19) ranging from ‘never’ to ‘often’. All items start with ‘You have. . .’ and are followed by: (i) found that the print was too small to read; (ii) found characters and words you did not know; (iii) found that the content was too difficult; (iv) needed a long time to read and understand them; and (v) needed someone to help you read them. The concept of health literacy has only recently been introduced in Scandinavia (6), and there is at present a total lack of instruments for assessing FHL in the Scandinavian languages. Filling this gap would make it possible to increase knowledge of FHL levels in those countries and facilitate the planning of effective programmes and evaluations in health promotion, prevention and primary care. Translating an existing instrument is viewed as preferable to constructing a new one. It saves time, that is, as the operationalisation process has already been carried
out (20), and makes it possible to make cross-cultural comparison studies (21) later on. To be able to measure all dimensions of the complex phenomenon of HL, questionnaires for both communicative and critical HL and functional HL are needed. The Japanese Communicative and Critical Health Literacy (J C&C HL scale) scale was recently translated into Swedish (22), while a scale for the functional dimension is still missing. Thus, the aim of this study was to develop a Swedish FHL to use in health promotion and health prevention by translating the JFHL scale into Swedish and testing some aspects of its validity and test–retest reliability.
Method Study setting and participants The study was carried out in the south-east and southwest of Sweden. Translation and validity testing of the instrument were based on recommended criteria for cross-cultural adaption projects (21). A translator group consisting of four professional translators was used to translate the JFHL into Swedish; two were recruited from a university and two from translation agencies. All of them had experience of tasks in the health domain. A committee of professionals was recruited to examine the quality of the translation and the content. Purposive and convenience sampling (23) was used to gather a variation of professions in the areas of public health/primary health. The committee consisted of seven members: one dietitian, two medical doctors, three midwives, one occupational therapist and one physiotherapist. In addition to having knowledge in their basic professions, they had expert competence in social medicine and public health, nursing, psychotherapy and research. Five of the members had a PhD degree. A test–retest group for testing the first version of the translated scale was recruited by convenience sampling (23). To form a group corresponding to a general population, they were recruited from a choir, a study circle in painting and a university course for occupational therapists. The reason for recruiting from group activities was to facilitate data collection from the same persons two consecutive weeks. Inclusion criteria for participation in the test–retest group were that the person should be adult (>18 years) and have sufficient language skills to be able to read, understand and fill in the first version of the translated instrument. The authors visited the groups and informed both verbally and in writing about the project and the meaning of informed consent. In total, 25 persons fulfilled the criteria and participated in the data collection by filling in the instrument twice, 2 weeks in a row. Of those participants, 24 people also filled in a form about demographics. The majority of the participants © 2014 Nordic College of Caring Science
Swedish Functional Health Literacy scale Table 1 Demographics of the participants in the test–retest group that filled in the response process form (n = 24) n = 24 Age Mean years (range) Gender Men Women Highest education level Primary school High school University or higher education Other Country of birth Sweden Other Nordic country
51.8 (18–75) 6 18 0 1 23 0 23 1
were women and had graduated from university/higher education. The ages of the participants ranged from 18 to 75 years. More specific demographic information is given in table 1. A target group for interviews was recruited by convenience sampling (23). To form a group corresponding to a general population, participants were recruited from an orienteering club, a choir and a healthy-living centre. Inclusion criteria for participation in an interview were the same as for participation in the test–retest group. For recruitment of participants from the orienteering club and the choir, the first author informed potential participants by e-mail about the project and asked people to take contact if they wanted to participate. For recruitment of participants from the healthy-living centre, staff from the centre informed potential participants verbally about the project and set up appointments between those and the first author. In total, ten people fulfilled the criteria and showed interest in participating in the target group. Most of them had graduated from university/higher education, their ages ranged from 34 to 78 years, and six of the ten participants were women. Seven of the participants were born in Sweden, and the other three had lived in Sweden for more than 20 years. The research project was approved by the regional ethical review board in Stockholm, Sweden (DNB 2011/197-31 /5). Informed consent was retrieved verbally from all participants.
Process The research process was divided into six phases in which numerical data and data in the form of text were collected using quantitative and qualitative methods, respectively. An overview of the research process is presented in Fig. 1. © 2014 Nordic College of Caring Science
167
Phase 1. The process started with a translation of the items of the original JFHL scale into Swedish and a back translation into Japanese according to guidelines regarding cross-sectional adoption of instruments (21). Two separate processes were carried out by the translator group. One of the translators then evaluated the translations and gave feedback. By considering the feedback and comparing the items with definitions of FHL (8), the researchers estimated which translations were the most appropriate and adjusted some translations of the items to better fit the Swedish language. This resulted in a first version of the Swedish Functional Health Literacy scale (the SFHL scale), which included the same items and categories as the original JFHL scale (18). However, the introduction text was longer and more detailed than the one in the original JFHL scale and referred to more general health information than the original scale, that is, it did not refer to instructions or leaflets from hospitals or pharmacies. Phase 2. A content validity form was created in which the first version of the SFHL scale was presented. This form included WHO0 s definition of health literacy (24), operational definitions of health literacy, operational definitions found in the literature (8) and questions regarding the instrument0 s content, comprehensibility and usability. The form was sent to the professionals with a request for feedback. After about 30 days, the professionals were interviewed by phone regarding their reflections on and opinions about the scale based on the questions in the form. Phase 2 also included interviews with four people in the target group to examine the contents, comprehensibility and response of the scale. The first author met the participants individually and asked them to fill in the scale and simultaneously talk aloud about their thoughts and feelings during this task. After completion of the task, the participants were asked if they experienced any items as problematic or difficult to understand and, if so, why. They were also asked to give suggestions for improvements to the scale and alternative formulations for the items. The interviews lasted about 30–40 minutes. Notes were taken about the participants’ statements. Phase 3. The first version of the SFHL scale was distributed to the test–retest group for test–retest. The analysis of the test–retest was made according to Svensson’s model (25) for identifying and separately measuring the systematic and occasional components of observed disagreement between the test and retest assessments. The method is based on the conclusion that categorical ordinal data should not be treated as interval or quoted data (25). A two-way bivariate ranking procedure was carried out in which the paired observations were replaced by ranks and the ranks connected to the pairs of observations
168
J.M. W angdahl, L.I. M artensson
Figure 1 Overview of the research process. The different colours of the boxes show the activity of the groups involved in each step and phase.
(25). Svensson’s free software (26) was used to identify disagreement of the positions of the ranks, that is, the relative position (RP) and estimate systematic disagreement by calculating the relative concentration (RC). Possible values of RP and RC range from 1 to 1, and a zero value means a lack of systematic disagreement or, in other words, stable test–retest assessments (27). The relative rank variance (RV) (25) was calculated in order to measure the individual occasional disagreement. The range of RV is between 0 and 1, where higher RV indicates a larger individual occasional contribution to the test–retest disagreement (25). The percentage agreement (PA) in the test–retest assessments was also calculated. Similar studies have considered a PA over 80% to be high and a PA between 60% and 80% to be fairly
good agreement (28). The same thresholds were used in this study. In connection with the retest, the participants in the test–retest group were asked to comment on the instrument in a response process form. Simultaneously, some notes were taken about verbally given experiences of the data collection and about the scale. Phase 4. To get an overall understanding of the opinions about the items and the scale, the qualitative feedback from the test–retest group, the professionals, the target group and comparisons with HL definitions (8) was analysed by examining structural patterns in the opinions collected and by taking into consideration similarities and differences. On the basis of this analysis, the authors © 2014 Nordic College of Caring Science
Swedish Functional Health Literacy scale constructed a second version of the SFHL scale. The major adjustments were that the heading and the introduction text were condensed to correspond to the original JFHL scale and that the scale response of ‘always’ was added. Those adjustments were the basis for the second version of the SFHL scale. Phase 5. Six participants in the target group were interviewed regarding the second version of the SFHL scale, in the same way as the participants in the target group in phase 2; that is, interviews were held about the content and response process. Phase 6. The content and the response process of the instrument were examined by both authors using the feedback from the professionals, the target group and the test–retest group. This examination led to one adjustment concerning the text flow. Full sentences were written for each item instead of dividing them into subheadings and items, which was the case in the two test versions. With this, the final version of the SFHL scale was established.
Results The first version The evidence of the content and the response process. Small differences were found in syntax and grammar which were caused by differences in expressions in the Swedish and Japanese languages. The items were, however, semantically equivalent. On the whole, the professionals and the target group judged the translated scale as being easy to use and satisfactory concerning structure and content. Nevertheless, there were some comments regarding the structure, content and validity of the scale. Concerning the introduction text and the instructions for filling in the scale, the opinions among the professionals were contradictory, as some of them regarded it as too extensive and detailed, while others viewed it as instructive and clear. The heading including the wording ‘health literacy’ was questioned as this expression may be experienced as difficult and unfamiliar. The target group thought the introduction text was long, complicated and difficult to understand because the items were linked to this. A suggestion was made to write out the full question for each item. Both the professionals and the target group raised some validity issues about the meaning of the word ‘information’ used in the items. It was not experienced as clearly stated as to whether the items referred to both national and international information or whether the items should be answered subjectively or on a more general basis. The target group was concerned about the same concept, mentioning that the understanding of information could vary depending on the context of the origin, for example whether it was in the form of an academic © 2014 Nordic College of Caring Science
169
paper, a technical instruction or a flyer. In addition, it was not experienced as clear whether the items referred to verbal, written and/or online information. Furthermore, professionals were concerned about how data gathered by the scale would be related to other variables, for example the degree to which a respondent0 s personality, including for example the traits of self-efficacy and empowerment, would influence the self-assessments. This was the same in the target group, who mentioned that the comprehensiveness of the information is dependent on the degree of meaning it has for the individual. The professionals also had ethical considerations about the use of the scale among alphabets, migrants, older people and people with different disabilities, who may have low literacy skills or low HL. Concerning the responses on the scale, only the professionals had an objection; they mentioned that a fifth alternative of ‘always’ should be considered. There were no considerations or comments about the cultural validity of the translation. To improve the scale, the professionals offered suggestions for restructurings and rephrasings to make the instructions more comprehensible. The target group suggested that concrete examples of health information should be given in the items. Suggestions were given by the professionals on the use of the scale, for example to assess health literacy on an individual or population level, when developing and evaluate materials and methods in clinical and in health promotion settings. Evidence of the test–retest reliability. The items showed test– retest stability, as the RP, RC and RV values were close to zero. The RV values of individual disagreement were negligible for all items in the test–retest. The percentage agreement (PA) measures between the tests ranged from 63% to 92%, with a mean of 77.2%. Good agreement (PA >80%) (28) was found for the last item and fairly good agreement (PA 60–80%) for the other four. Table 2 shows the systematic and individual agreement for the five items.
The second version Items in the second version of the SFHL scale were experienced by the new participants as being more comprehensible than those in the first version. However, those participants also commented on the meaning of ‘information’ and on the comprehension of the information being dependent on the degree of meaning it has for the individual. Concerning the scale response categories, none of the participants seemed to have difficulty giving an answer.
The final version In comparison with the major aspects of functional HL mentioned in the literature (5–9), the final SFHL scale covered the issues that it intends to assess.
170
J.M. W angdahl, L.I. M artensson
Table 2 Results of the test–retest analysis. Figures in parentheses are 95% confidence intervals (CI) of the measures (n = 25)
Item
Percentage agreement PA% (n)
Relative position RP (CI)
1 2 3 4 5
76 63 75 80 92
0.06 0.03 0.06 0.06 0.08
(25) (25) (25) (25) (25)
(0.12 (0.22 (0.22 (0.20 (0.18
to to to to to
Discussion The aim of this study was to develop a Swedish FHL to use in health promotion and health prevention by translating the JFHL scale (18) into Swedish and testing some aspects of its validity and test–retest reliability. The study has a number of limitations. One is that the target group became more homogeneous than was intended as concerns gender and education level. People with a low education, men and people with limited skills in the Swedish language are underrepresented in the test–retest and the target groups. This limitation makes it difficult to draw conclusions about the validity of the instrument in a general population. However, as concerns the examination of the content and response of the scale, these may not have changed in a more heterogenic target group, as the results in the homogenous target group varied a great deal. The results in themselves give sufficient information, indicating that the items in the SFHL scale might be too general, showing that the scale needed to be adjusted (and that further validity tests should be added together with additional validity tests). The interviews included only questions about the content, response process and usefulness; thus, there were no questions regarding participants0 experiences of situations related to each of the items, which may have resulted in a deeper understanding of the interpretations of the scale. The study also has some strengths. The translation and validity tests were carried out in a systematic way (29) and followed highly recommended guidelines (21, 23, 30). Involving both professionals and potential users, which is recommended when evaluating the content of an assessment (31), further strengthens the objectivity and the quality of the results. In the test of the Swedish FHL scale, the categorical ordinal data are treated in an adequate way (25) in comparison with the test of the original instrument (18, 22, 25). This may mean an improvement in quality in the original version as well, as the translations were made verbatim. Reliability testing based on small samples supports the statistical method used in the study (27, 32). However, the results would have been more robust (27) with a larger test group, that is, one or few discordance observations would affect the outcome less.
0.13) 0.15) 0.09) 0.08) 0.02)
Relative concentration RC (CI)
Relative rank variance RV (CI)
0.1 (0.26 to 0.06) 0.03 (0.25 to 0.18) 0.02 (0.12 to 0.16) 0.04 (0.12 to 0.19) 0.00 (0.00 to 0.00)
0.02 0.03 0.01 0.05 0.00
(0.00 (0.00 (0.00 (0.00 (0.00
to to to to to
0.06) 0.07) 0.04) 0.15) 0.00)
According to the results, the translated SFHL scale can be considered equivalent to the original, which implies an accurate translation. Concerning the content and response of the SFHL scale, the results show that the items in the SFHL scale were experienced as being relevant according to the operational definition of FHL (8, 9) and the content of the original JFHL. No cultural problems regarding the items were mentioned by any participant. However, having made the items more general to fit the public health perspective of HL may be regarded as a cultural adaptation. In comparison with the major aspects of FHL mentioned in the literature (6, 8, 9), the final SFHL scale is found to cover the issues that it intends to assess. Both professionals and participants in the target group experienced the items in the first two versions of SFHL scale as vague, interpreted them differently and/or commented that they could be interpreted differently according to who is answering. Both professionals and participants in the target group were furthermore uncertain as to which type of health information, what sources of health information and the language in the items referred to. Similar results were seen in a study testing the validity of the Dutch version of the FHL scale (33), and blurry definitions of concepts and unspecific instruments seem to be rather common in other HL instruments and studies using those. This has been criticised in a large review article examining the validity of HL instruments (34). Regarding different interpretations of the concept of ‘health information’, it would be interesting to examine what factors affect people’s interpretations, for example to examine the possible differences in comprehension of the concept based on the level of HL. If such a difference exists, the SFHL scale may overestimate HL among people with low HL and underestimate HL among people with high HL. Comparing results from the SFHL scale with other FHL instruments such as TOFHLA, REALM or NVS (The Newest Vital Sign) (34) could improve the SFHL’s criterion validity. The problem of not knowing how each individual interprets each item was also raised in connection with the translation process in the test of the Dutch FHL scale (33), which indicates a validity problem. This, together with the fact that not all people have the capacity to fill in the scale, raises the question of whether it is possible to have a short self-assessed general HL scale to © 2014 Nordic College of Caring Science
Swedish Functional Health Literacy scale assess HL in health promotion and prevention. More specific HL scales focusing on specific health promotion or prevention areas, such as the short version of the European Health Literacy Project questionnaire (The HLS-EU Q16) (35), may be more adequate and useful when assessing FHL in specific areas, for example when screening for FHL in primary care or evaluating a specific health promotion programme. Referring to specific health information in the items increases the validity of the scale and reduces very different interpretations of the items. Overall, the participants in the target group seemed to judge the items and the scale in the first versions to be fairly easier to understand than did the professionals. A reason for this may be an attempt on the part of the professionals to see the items from the user’s perspective, including possible factors of low health literacy and education levels. The higher comprehension of the second version of the SFHL scale indicates that it could be easier to understand items consisting of full questions than items consisting of the second part of sentences, which links back to a first half of a sentence in the ingress. The test–retest reliability with a PA range of 63–92% is judged to be satisfactory, based on judgments in similar studies using the same statistical method (28, 36). According to the literature, 100% agreement in a test– retest is impossible, as multiple factors influence estimations in a test – not just the instrument itself (36). The test–retest for the scale indicates stability over time (1 week), that is, the reliability of the scale, at least for a part of the Swedish population. Test–retest over a longer time period is not appropriate as HL is a dynamic phenomenon that may change over time (6).
Implications The SFHL scale could be useful in part of the population at both the individual and population levels, for example when screening for low HL or when examining changes in HL brought about by interventions. However, if the SFHL scale is used in a broader population including people with low education, men and immigrants, more validity tests are needed. Consideration also needs be taken to context and target groups to avoid selection bias. Additional studies are also needed to examine how different target groups
References 1 Halpin H, Morales-Suarez-Varela M, Martin-Moreno J. Chronic disease prevention and the New Public Health. Public Health Rev 2010; 32: 120–54. 2 Merzel C, D’Afflitti J. Reconsidering community-based health promotion:
© 2014 Nordic College of Caring Science
171
define and interpret health information and health literacy when comparing and analysing FHL levels within and between subgroups. Based on the results of this study, additional verbal questions may be necessary to focus on which sources and languages users of the scale refer to when they interpret the items and how interpretations of items differ between different countries and cultures.
Conclusions The items in the SFHL are equivalent to the original JFHL in terms of language and content and cover the major aspects of functional health literacy as it is defined in the literature. The translated instrument shows stability over time, that is, reliability, at least for a part of the Swedish population. More validity tests of the SFHL scale based on a broader population are needed.
Acknowledgements We would like to thank Doctor Ishikawa for the permission to translate and test the FHL scale and to professionals and target group giving feedback and testing the SFHL scale. The study was supported by the Adlerbert assistance fund, Sweden.
Author contributions Both authors have been active and equally contributed in the study conception/design, drafting of manuscript and in the critical revisions for important intellectual content. Both authors also contributed in data collection and analysis, but here Josefin M. W angdahl did the main part of the work.
Ethical approval The study was approved by the regional ethical review board in Stockholm, Sweden (DNB 2011/197-31/5).
Funding The study was financed by the Adlerbert Assistance Fund, Sweden, and the European Refugee Fund (ERF).
promise, performance, and potential. Am J Public Health 2003; 93: 557–74. 3 Hernandez L, Landi S. Promoting Health Literacy to Encourage Prevention and Wellness: Workshop Summary. 2011, The National Academies Press, Washington, DC. 4 Song L, Mishel M, Bensen JT, Chen RC, Knafl GJ, Blackard B et al. How
does health literacy affect quality of life among men with newly diagnosed clinically localized prostate cancer? Findings from the North Carolina-Louisiana Prostate Cancer Project (PCaP). Cancer 2012; 1: 3842– 51. 5 Kickbusch IS. Health literacy: addressing the health and education
172
6
7
8
9
10
11
12
13
14
15
J.M. W angdahl, L.I. M artensson
divide. Health Promot Int 2001; 16: 289–97. M artensson L, Hensing G. Health literacy – a heterogeneous phenomenon: a literature review. Scand J Caring Sci 2012; 26: 151–60. Peerson A, Saunders M. Health literacy revisited: what do we mean and why does it matter? Health Promot Int 2009; 24: 285–96. Nutbeam D. Health literacy as a public health goal: a challenge for contemporary health education and communication strategies into the 21st century. Health Promot Int 2000; 15: 259–67. Sorensen K, Broucke SV, Fullam J, Doyle G, Pelikan J, Slonska Z et al. Health literacy and public health: a systematic review and integration of definitions and models. BMC Public Health 2012; 12: 80. Murata J, Arkida M, Shirai F. A concept analysis of ‘health literacy’ – focusing on participants in health education classes at health centers (Japanese). J Jpn Acad Nurs Sci 2006; 26: 84–92. Adams RJ, Stocks NP, Wilson DH, Hill CL, Gravier S, Kickbusch I et al. Health literacy – a new concept for general practice? Aust Fam Physician 2009; 38: 144–7. Bennett IM, Chen J, Soroui JS, White S. The contribution of health literacy to disparities in self-rated health status and preventive health behaviors in older adults. Ann Fam Med 2009; 7: 204–11. Berkman ND, Sheridan SL, Donahue KE, Halpern DJ, Crotty K. Low health literacy and health outcomes: an updated systematic review. Ann Intern Med 2011; 155: 97–107. Eichler K, Wieser S, Brugger U. The costs of limited health literacy: a systematic review. Int J Public Health 2009; 54: 313–24. Vernon AJ. Health Policy Brief the High Economic Cost of Low Health Literacy in Wisconsin. 2009 Department of Health Policy and Management, UNC Chapel Hill and the National
16
17
18
19
20
21
22
23
24 25
26
Bureau of Economic Research (NBER), Wisconsin. Baker DW, Williams MV, Parker RM, Gazmararian JA, Nurss J. Development of a brief test to measure functional health literacy. Patient Educ Couns 1999; 38: 33–42. Davis TC, Long SW, Jackson RH, Mayeaux EJ, George RB, Murphy PW et al. Rapid estimate of adult literacy in medicine: a shortened screening instrument. Fam Med 1993; 25: 391–5. Ishikawa H, Takeuchi T, Yano E. Measuring functional, communicative, and critical health literacy among diabetic patients. Diabetes Care 2008; 31: 874–9. Svensson E. Statistical methods for repeated qualitative assessments on scales. Int J Audiol 2003; 42(Suppl 1): S13–22. Ishikawa H, Kiuchi T. Health literacy and health communication. Biopsychosoc Med 2010; 4: 18. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993; 46: 1417–32. W angdahl JM, Martensson LI. The communicative and critical health literacy scale – Swedish version. Scand J Public Health 2014; 42: 25–31. Kline T. Psychological Testing: A Practical Approach to Design and Evaluation. 2005, Sage Publications, Thousand Oaks, Calif. WHO. Health Promotion Glossary. 1998, WHO, Geneva. Svensson E. What is the therapeutic effect if the patient gets better, but no one knows how much better? Statistical method for paired ordinal data. L€akartidningen 2007; 104: 596–601. Svensson E. Svensson’s method - freeware and documentation 2013, 2 July, www.oru.se/Institutioner/Handelsho gskolan-vid-Orebro-universitet/Kont akt-och-presentation/Personliga-sidor/ Statistik/Elisabeth-Svensson/Svensoo ns-metod/Svenssons-metod—fri-prog
27
28
29
30
31
32
33
34
35
36
ramvara-och-dokumentation/Svenss ons-method—english/ Svensson E. Analysis of Systematic and Random Differences Between Paired Ordinal Categorical Data. 1993, Gothenburg University, Gothenburg. Kazdin AE. Artifact, bias, and complexity of assessment: the ABCs of reliability. J Appl Behav Anal 1977; 10: 141–50. Cook DA, Beckman TJ. Current concepts in validity and reliability for psychometric instruments: theory and application. Am J Med 2006; 119: 166.e7–16. Kimberlin CL, Winterstein AG. Validity and reliability of measurement instruments used in research. Am J Health Syst Pharm 2008; 65: 2276–84. Lynn MR. Determination and quantification of content validity. Nurs Res 1986; 35: 382–5. Allvin R, Ehnfors M, Rawal N, Svensson E, Idvall E. Development of a questionnaire to measure patientreported postoperative recovery: content validity and intra-patient reliability. J Eval Clin Pract 2009; 15: 411–9. van der Vaart R, Drossaert CH, Taal E, ten Klooster PM, Hilderink-Koertshuis RT, Klaase JM et al. Validation of the Dutch functional, communicative and critical health literacy scales. Patient Educ Couns 2012; 89: 82–8. Jordan JE, Osborne RH, Buchbinder R. Critical appraisal of health literacy indices revealed variable underlying constructs, narrow content and psychometric weaknesses. J Clin Epidemiol 2011; 64: 366–79. Sorensen K, Van den Broucke S, Pelikan JM, Fullam J, Doyle G, Slonska Z et al. Measuring health literacy in populations: illuminating the design and development process of the European Health Literacy Survey Questionnaire (HLS-EU-Q). BMC Public Health 2013; 13: 948. Goldstein G, Hersen M. Handbook of Psychological Assessment, 2nd edn. 1990, Pergamon, New York. Pergamon general psychology series.
© 2014 Nordic College of Caring Science
doi: 10.1111/scs.12125
Measuring health literacy – the Swedish Functional Health Literacy scale Josefin M. W angdahl MSPH (PhD Student)1 and Lena I. M artensson PhD, OTR (Senior Lecturer)2 1
Department of Public Health and Caring Science, Social medicine, Uppsala University, Uppsala, Sweden and 2Institute of Neuroscience and Physiology/Occupational Therapy, G€ oteborg University, G€ oteborg, Sweden
Scand J Caring Sci; 2015; 29; 165–172 Measuring health literacy – the Swedish Functional Health Literacy scale
Background: The benefits of health promotion efforts vary due to a complexity of reasons. One possible reason for an absence of effects is the level of functional health literacy among the individuals that participate in the interventions. Thus, valid and reliable instruments that capture these kinds of skills are needed. The aim of this study was to develop a Swedish Functional Health Literacy scale to use in health promotion and health prevention by translating the Japanese Functional Health Literacy scale into Swedish and testing some aspects of its validity and test–retest reliability. Methods: The research project comprised six phases including translation and back translation, validity tests of the two first versions of the instrument based on interviews with professionals and representatives for the target group of the instrument, and a test–retest of the first version among target groups.
Introduction Much work aims at increasing individuals’ health conditions in the form of health promotion and primary or secondary preventive interventions (1), for example health examinations, vaccination programmes, dental hygiene, prescription of contraceptive pills for teenagers, screening for prostate cancer et cetera. The benefits of health promotion efforts vary for a complexity of reasons (2, 3). One possible reason for an absence of effects that has not yet been given much attention is the level of health literacy among individuals that participate in the health-promoting interventions (4). Correspondence to: Josefin W angdahl, Department of Public Health and Caring Sciences, Uppsala University, Uppsala Science Park, Box 564, 751 22 Uppsala, Sweden. E-mail: [email protected] © 2014 Nordic College of Caring Science
Results: The items in the first two versions were experienced as unclear, which led to improvements of the next version. The final version of the translated instrument (the Swedish Functional Literacy scale) showed evidence of content validity, and the test–retest confirmed that the instrument had stability over time with a percentage agreement for the items ranging from 63% to 92% (M = 77.2%). Conclusion: The items in the Swedish version of the scale are equivalent to the original Japanese scale in terms of language and contents and cover the major aspects of functional health literacy as it is defined in the literature. The translated instrument shows stability over time, that is, reliability, at least for a part of the Swedish population. More validity tests of the Swedish Functional Health Literacy scale based on a broader population are needed. Keywords: health literacy, health promotion, health education, prevention, measurement. Submitted 19 January 2014, Accepted 10 February 2014
Health literacy (HL) has a variety of definitions. It basically concerns the ability to read and understand the information needed to carry out actions that preserve or improve health (5–9). A more complex understanding of HL also includes skills for critical reflection and analysis to allow informed decisions in health issues, in combination with the social skills needed to communicate health information in the society (8). Based on its complexity, HL is often divided into three dimensions: functional, communicative and critical HL, ranging from basic forms of cognitive skills to more complex ones, such as the ability to appraise, choose and implement health information in ordinary life and to use new terminology and mediate health information (9, 10). Regardless of the dimension, health literacy is understood as a context-dependent phenomenon, thus varying according to the situation in which a person has a need for information or for obtaining information in health issues (6). In this study, we focus on functional health literacy (FHL), that is, the 165
166
J.M. W angdahl, L.I. M artensson
skills necessary for being compliant with medication regimens, following advice regarding eating habits and life habits in general, understanding information about risky products, being familiar with home principles for self-care et cetera. Low levels of FHL have been associated with being an immigrant, having an uncompleted education, being old, having more hospitalisations, having fewer mammography screenings and influenza vaccines and having a poorer ability to interpret labels and health messages et cetera (11–13). Research also shows that, besides poorer health, low FHL conveys high costs for both individuals and society (14, 15). There are thus reasons for professionals that work with health issues to gain a comprehensive understanding of health literacy and pay attention to people’s FHL in practice. In research, data on FHL have most often been collected using the Test of Functional Health Literacy in Adults (TOFHLA) (16) and the Rapid Estimate of Adult Literacy in Medicine (REALM) (17). The TOFHLA is a multiple choice test that measures the comprehension of health-related concepts described in a medical context, for example patient information (16). The REALM assesses the reading comprehension of 66 de-contextualised medical terms (17). Due to the medical orientation of these assessments, their feasibility in the public health context is doubtful. An instrument that is more general concerning context is the Japanese Functional Health Literacy scale (the JFHL scale) by Ishikawa (18). Like the other Japanese scale, that is, the Communicative and Critical Health Literacy scale (J C&C HL scale), the SFHL scale is based on Nutbeam0 s definition of HL including abilities needed both for health prevention and health promotion, which means that it is related to a greater extent to a public health context (8). The JFHL scale is a short, easily administrated assessment. It comprises five items about persons’ skills in reading and understanding health information. The items can be self-assessed on a five-point Likert scale (19) ranging from ‘never’ to ‘often’. All items start with ‘You have. . .’ and are followed by: (i) found that the print was too small to read; (ii) found characters and words you did not know; (iii) found that the content was too difficult; (iv) needed a long time to read and understand them; and (v) needed someone to help you read them. The concept of health literacy has only recently been introduced in Scandinavia (6), and there is at present a total lack of instruments for assessing FHL in the Scandinavian languages. Filling this gap would make it possible to increase knowledge of FHL levels in those countries and facilitate the planning of effective programmes and evaluations in health promotion, prevention and primary care. Translating an existing instrument is viewed as preferable to constructing a new one. It saves time, that is, as the operationalisation process has already been carried
out (20), and makes it possible to make cross-cultural comparison studies (21) later on. To be able to measure all dimensions of the complex phenomenon of HL, questionnaires for both communicative and critical HL and functional HL are needed. The Japanese Communicative and Critical Health Literacy (J C&C HL scale) scale was recently translated into Swedish (22), while a scale for the functional dimension is still missing. Thus, the aim of this study was to develop a Swedish FHL to use in health promotion and health prevention by translating the JFHL scale into Swedish and testing some aspects of its validity and test–retest reliability.
Method Study setting and participants The study was carried out in the south-east and southwest of Sweden. Translation and validity testing of the instrument were based on recommended criteria for cross-cultural adaption projects (21). A translator group consisting of four professional translators was used to translate the JFHL into Swedish; two were recruited from a university and two from translation agencies. All of them had experience of tasks in the health domain. A committee of professionals was recruited to examine the quality of the translation and the content. Purposive and convenience sampling (23) was used to gather a variation of professions in the areas of public health/primary health. The committee consisted of seven members: one dietitian, two medical doctors, three midwives, one occupational therapist and one physiotherapist. In addition to having knowledge in their basic professions, they had expert competence in social medicine and public health, nursing, psychotherapy and research. Five of the members had a PhD degree. A test–retest group for testing the first version of the translated scale was recruited by convenience sampling (23). To form a group corresponding to a general population, they were recruited from a choir, a study circle in painting and a university course for occupational therapists. The reason for recruiting from group activities was to facilitate data collection from the same persons two consecutive weeks. Inclusion criteria for participation in the test–retest group were that the person should be adult (>18 years) and have sufficient language skills to be able to read, understand and fill in the first version of the translated instrument. The authors visited the groups and informed both verbally and in writing about the project and the meaning of informed consent. In total, 25 persons fulfilled the criteria and participated in the data collection by filling in the instrument twice, 2 weeks in a row. Of those participants, 24 people also filled in a form about demographics. The majority of the participants © 2014 Nordic College of Caring Science
Swedish Functional Health Literacy scale Table 1 Demographics of the participants in the test–retest group that filled in the response process form (n = 24) n = 24 Age Mean years (range) Gender Men Women Highest education level Primary school High school University or higher education Other Country of birth Sweden Other Nordic country
51.8 (18–75) 6 18 0 1 23 0 23 1
were women and had graduated from university/higher education. The ages of the participants ranged from 18 to 75 years. More specific demographic information is given in table 1. A target group for interviews was recruited by convenience sampling (23). To form a group corresponding to a general population, participants were recruited from an orienteering club, a choir and a healthy-living centre. Inclusion criteria for participation in an interview were the same as for participation in the test–retest group. For recruitment of participants from the orienteering club and the choir, the first author informed potential participants by e-mail about the project and asked people to take contact if they wanted to participate. For recruitment of participants from the healthy-living centre, staff from the centre informed potential participants verbally about the project and set up appointments between those and the first author. In total, ten people fulfilled the criteria and showed interest in participating in the target group. Most of them had graduated from university/higher education, their ages ranged from 34 to 78 years, and six of the ten participants were women. Seven of the participants were born in Sweden, and the other three had lived in Sweden for more than 20 years. The research project was approved by the regional ethical review board in Stockholm, Sweden (DNB 2011/197-31 /5). Informed consent was retrieved verbally from all participants.
Process The research process was divided into six phases in which numerical data and data in the form of text were collected using quantitative and qualitative methods, respectively. An overview of the research process is presented in Fig. 1. © 2014 Nordic College of Caring Science
167
Phase 1. The process started with a translation of the items of the original JFHL scale into Swedish and a back translation into Japanese according to guidelines regarding cross-sectional adoption of instruments (21). Two separate processes were carried out by the translator group. One of the translators then evaluated the translations and gave feedback. By considering the feedback and comparing the items with definitions of FHL (8), the researchers estimated which translations were the most appropriate and adjusted some translations of the items to better fit the Swedish language. This resulted in a first version of the Swedish Functional Health Literacy scale (the SFHL scale), which included the same items and categories as the original JFHL scale (18). However, the introduction text was longer and more detailed than the one in the original JFHL scale and referred to more general health information than the original scale, that is, it did not refer to instructions or leaflets from hospitals or pharmacies. Phase 2. A content validity form was created in which the first version of the SFHL scale was presented. This form included WHO0 s definition of health literacy (24), operational definitions of health literacy, operational definitions found in the literature (8) and questions regarding the instrument0 s content, comprehensibility and usability. The form was sent to the professionals with a request for feedback. After about 30 days, the professionals were interviewed by phone regarding their reflections on and opinions about the scale based on the questions in the form. Phase 2 also included interviews with four people in the target group to examine the contents, comprehensibility and response of the scale. The first author met the participants individually and asked them to fill in the scale and simultaneously talk aloud about their thoughts and feelings during this task. After completion of the task, the participants were asked if they experienced any items as problematic or difficult to understand and, if so, why. They were also asked to give suggestions for improvements to the scale and alternative formulations for the items. The interviews lasted about 30–40 minutes. Notes were taken about the participants’ statements. Phase 3. The first version of the SFHL scale was distributed to the test–retest group for test–retest. The analysis of the test–retest was made according to Svensson’s model (25) for identifying and separately measuring the systematic and occasional components of observed disagreement between the test and retest assessments. The method is based on the conclusion that categorical ordinal data should not be treated as interval or quoted data (25). A two-way bivariate ranking procedure was carried out in which the paired observations were replaced by ranks and the ranks connected to the pairs of observations
168
J.M. W angdahl, L.I. M artensson
Figure 1 Overview of the research process. The different colours of the boxes show the activity of the groups involved in each step and phase.
(25). Svensson’s free software (26) was used to identify disagreement of the positions of the ranks, that is, the relative position (RP) and estimate systematic disagreement by calculating the relative concentration (RC). Possible values of RP and RC range from 1 to 1, and a zero value means a lack of systematic disagreement or, in other words, stable test–retest assessments (27). The relative rank variance (RV) (25) was calculated in order to measure the individual occasional disagreement. The range of RV is between 0 and 1, where higher RV indicates a larger individual occasional contribution to the test–retest disagreement (25). The percentage agreement (PA) in the test–retest assessments was also calculated. Similar studies have considered a PA over 80% to be high and a PA between 60% and 80% to be fairly
good agreement (28). The same thresholds were used in this study. In connection with the retest, the participants in the test–retest group were asked to comment on the instrument in a response process form. Simultaneously, some notes were taken about verbally given experiences of the data collection and about the scale. Phase 4. To get an overall understanding of the opinions about the items and the scale, the qualitative feedback from the test–retest group, the professionals, the target group and comparisons with HL definitions (8) was analysed by examining structural patterns in the opinions collected and by taking into consideration similarities and differences. On the basis of this analysis, the authors © 2014 Nordic College of Caring Science
Swedish Functional Health Literacy scale constructed a second version of the SFHL scale. The major adjustments were that the heading and the introduction text were condensed to correspond to the original JFHL scale and that the scale response of ‘always’ was added. Those adjustments were the basis for the second version of the SFHL scale. Phase 5. Six participants in the target group were interviewed regarding the second version of the SFHL scale, in the same way as the participants in the target group in phase 2; that is, interviews were held about the content and response process. Phase 6. The content and the response process of the instrument were examined by both authors using the feedback from the professionals, the target group and the test–retest group. This examination led to one adjustment concerning the text flow. Full sentences were written for each item instead of dividing them into subheadings and items, which was the case in the two test versions. With this, the final version of the SFHL scale was established.
Results The first version The evidence of the content and the response process. Small differences were found in syntax and grammar which were caused by differences in expressions in the Swedish and Japanese languages. The items were, however, semantically equivalent. On the whole, the professionals and the target group judged the translated scale as being easy to use and satisfactory concerning structure and content. Nevertheless, there were some comments regarding the structure, content and validity of the scale. Concerning the introduction text and the instructions for filling in the scale, the opinions among the professionals were contradictory, as some of them regarded it as too extensive and detailed, while others viewed it as instructive and clear. The heading including the wording ‘health literacy’ was questioned as this expression may be experienced as difficult and unfamiliar. The target group thought the introduction text was long, complicated and difficult to understand because the items were linked to this. A suggestion was made to write out the full question for each item. Both the professionals and the target group raised some validity issues about the meaning of the word ‘information’ used in the items. It was not experienced as clearly stated as to whether the items referred to both national and international information or whether the items should be answered subjectively or on a more general basis. The target group was concerned about the same concept, mentioning that the understanding of information could vary depending on the context of the origin, for example whether it was in the form of an academic © 2014 Nordic College of Caring Science
169
paper, a technical instruction or a flyer. In addition, it was not experienced as clear whether the items referred to verbal, written and/or online information. Furthermore, professionals were concerned about how data gathered by the scale would be related to other variables, for example the degree to which a respondent0 s personality, including for example the traits of self-efficacy and empowerment, would influence the self-assessments. This was the same in the target group, who mentioned that the comprehensiveness of the information is dependent on the degree of meaning it has for the individual. The professionals also had ethical considerations about the use of the scale among alphabets, migrants, older people and people with different disabilities, who may have low literacy skills or low HL. Concerning the responses on the scale, only the professionals had an objection; they mentioned that a fifth alternative of ‘always’ should be considered. There were no considerations or comments about the cultural validity of the translation. To improve the scale, the professionals offered suggestions for restructurings and rephrasings to make the instructions more comprehensible. The target group suggested that concrete examples of health information should be given in the items. Suggestions were given by the professionals on the use of the scale, for example to assess health literacy on an individual or population level, when developing and evaluate materials and methods in clinical and in health promotion settings. Evidence of the test–retest reliability. The items showed test– retest stability, as the RP, RC and RV values were close to zero. The RV values of individual disagreement were negligible for all items in the test–retest. The percentage agreement (PA) measures between the tests ranged from 63% to 92%, with a mean of 77.2%. Good agreement (PA >80%) (28) was found for the last item and fairly good agreement (PA 60–80%) for the other four. Table 2 shows the systematic and individual agreement for the five items.
The second version Items in the second version of the SFHL scale were experienced by the new participants as being more comprehensible than those in the first version. However, those participants also commented on the meaning of ‘information’ and on the comprehension of the information being dependent on the degree of meaning it has for the individual. Concerning the scale response categories, none of the participants seemed to have difficulty giving an answer.
The final version In comparison with the major aspects of functional HL mentioned in the literature (5–9), the final SFHL scale covered the issues that it intends to assess.
170
J.M. W angdahl, L.I. M artensson
Table 2 Results of the test–retest analysis. Figures in parentheses are 95% confidence intervals (CI) of the measures (n = 25)
Item
Percentage agreement PA% (n)
Relative position RP (CI)
1 2 3 4 5
76 63 75 80 92
0.06 0.03 0.06 0.06 0.08
(25) (25) (25) (25) (25)
(0.12 (0.22 (0.22 (0.20 (0.18
to to to to to
Discussion The aim of this study was to develop a Swedish FHL to use in health promotion and health prevention by translating the JFHL scale (18) into Swedish and testing some aspects of its validity and test–retest reliability. The study has a number of limitations. One is that the target group became more homogeneous than was intended as concerns gender and education level. People with a low education, men and people with limited skills in the Swedish language are underrepresented in the test–retest and the target groups. This limitation makes it difficult to draw conclusions about the validity of the instrument in a general population. However, as concerns the examination of the content and response of the scale, these may not have changed in a more heterogenic target group, as the results in the homogenous target group varied a great deal. The results in themselves give sufficient information, indicating that the items in the SFHL scale might be too general, showing that the scale needed to be adjusted (and that further validity tests should be added together with additional validity tests). The interviews included only questions about the content, response process and usefulness; thus, there were no questions regarding participants0 experiences of situations related to each of the items, which may have resulted in a deeper understanding of the interpretations of the scale. The study also has some strengths. The translation and validity tests were carried out in a systematic way (29) and followed highly recommended guidelines (21, 23, 30). Involving both professionals and potential users, which is recommended when evaluating the content of an assessment (31), further strengthens the objectivity and the quality of the results. In the test of the Swedish FHL scale, the categorical ordinal data are treated in an adequate way (25) in comparison with the test of the original instrument (18, 22, 25). This may mean an improvement in quality in the original version as well, as the translations were made verbatim. Reliability testing based on small samples supports the statistical method used in the study (27, 32). However, the results would have been more robust (27) with a larger test group, that is, one or few discordance observations would affect the outcome less.
0.13) 0.15) 0.09) 0.08) 0.02)
Relative concentration RC (CI)
Relative rank variance RV (CI)
0.1 (0.26 to 0.06) 0.03 (0.25 to 0.18) 0.02 (0.12 to 0.16) 0.04 (0.12 to 0.19) 0.00 (0.00 to 0.00)
0.02 0.03 0.01 0.05 0.00
(0.00 (0.00 (0.00 (0.00 (0.00
to to to to to
0.06) 0.07) 0.04) 0.15) 0.00)
According to the results, the translated SFHL scale can be considered equivalent to the original, which implies an accurate translation. Concerning the content and response of the SFHL scale, the results show that the items in the SFHL scale were experienced as being relevant according to the operational definition of FHL (8, 9) and the content of the original JFHL. No cultural problems regarding the items were mentioned by any participant. However, having made the items more general to fit the public health perspective of HL may be regarded as a cultural adaptation. In comparison with the major aspects of FHL mentioned in the literature (6, 8, 9), the final SFHL scale is found to cover the issues that it intends to assess. Both professionals and participants in the target group experienced the items in the first two versions of SFHL scale as vague, interpreted them differently and/or commented that they could be interpreted differently according to who is answering. Both professionals and participants in the target group were furthermore uncertain as to which type of health information, what sources of health information and the language in the items referred to. Similar results were seen in a study testing the validity of the Dutch version of the FHL scale (33), and blurry definitions of concepts and unspecific instruments seem to be rather common in other HL instruments and studies using those. This has been criticised in a large review article examining the validity of HL instruments (34). Regarding different interpretations of the concept of ‘health information’, it would be interesting to examine what factors affect people’s interpretations, for example to examine the possible differences in comprehension of the concept based on the level of HL. If such a difference exists, the SFHL scale may overestimate HL among people with low HL and underestimate HL among people with high HL. Comparing results from the SFHL scale with other FHL instruments such as TOFHLA, REALM or NVS (The Newest Vital Sign) (34) could improve the SFHL’s criterion validity. The problem of not knowing how each individual interprets each item was also raised in connection with the translation process in the test of the Dutch FHL scale (33), which indicates a validity problem. This, together with the fact that not all people have the capacity to fill in the scale, raises the question of whether it is possible to have a short self-assessed general HL scale to © 2014 Nordic College of Caring Science
Swedish Functional Health Literacy scale assess HL in health promotion and prevention. More specific HL scales focusing on specific health promotion or prevention areas, such as the short version of the European Health Literacy Project questionnaire (The HLS-EU Q16) (35), may be more adequate and useful when assessing FHL in specific areas, for example when screening for FHL in primary care or evaluating a specific health promotion programme. Referring to specific health information in the items increases the validity of the scale and reduces very different interpretations of the items. Overall, the participants in the target group seemed to judge the items and the scale in the first versions to be fairly easier to understand than did the professionals. A reason for this may be an attempt on the part of the professionals to see the items from the user’s perspective, including possible factors of low health literacy and education levels. The higher comprehension of the second version of the SFHL scale indicates that it could be easier to understand items consisting of full questions than items consisting of the second part of sentences, which links back to a first half of a sentence in the ingress. The test–retest reliability with a PA range of 63–92% is judged to be satisfactory, based on judgments in similar studies using the same statistical method (28, 36). According to the literature, 100% agreement in a test– retest is impossible, as multiple factors influence estimations in a test – not just the instrument itself (36). The test–retest for the scale indicates stability over time (1 week), that is, the reliability of the scale, at least for a part of the Swedish population. Test–retest over a longer time period is not appropriate as HL is a dynamic phenomenon that may change over time (6).
Implications The SFHL scale could be useful in part of the population at both the individual and population levels, for example when screening for low HL or when examining changes in HL brought about by interventions. However, if the SFHL scale is used in a broader population including people with low education, men and immigrants, more validity tests are needed. Consideration also needs be taken to context and target groups to avoid selection bias. Additional studies are also needed to examine how different target groups
References 1 Halpin H, Morales-Suarez-Varela M, Martin-Moreno J. Chronic disease prevention and the New Public Health. Public Health Rev 2010; 32: 120–54. 2 Merzel C, D’Afflitti J. Reconsidering community-based health promotion:
© 2014 Nordic College of Caring Science
171
define and interpret health information and health literacy when comparing and analysing FHL levels within and between subgroups. Based on the results of this study, additional verbal questions may be necessary to focus on which sources and languages users of the scale refer to when they interpret the items and how interpretations of items differ between different countries and cultures.
Conclusions The items in the SFHL are equivalent to the original JFHL in terms of language and content and cover the major aspects of functional health literacy as it is defined in the literature. The translated instrument shows stability over time, that is, reliability, at least for a part of the Swedish population. More validity tests of the SFHL scale based on a broader population are needed.
Acknowledgements We would like to thank Doctor Ishikawa for the permission to translate and test the FHL scale and to professionals and target group giving feedback and testing the SFHL scale. The study was supported by the Adlerbert assistance fund, Sweden.
Author contributions Both authors have been active and equally contributed in the study conception/design, drafting of manuscript and in the critical revisions for important intellectual content. Both authors also contributed in data collection and analysis, but here Josefin M. W angdahl did the main part of the work.
Ethical approval The study was approved by the regional ethical review board in Stockholm, Sweden (DNB 2011/197-31/5).
Funding The study was financed by the Adlerbert Assistance Fund, Sweden, and the European Refugee Fund (ERF).
promise, performance, and potential. Am J Public Health 2003; 93: 557–74. 3 Hernandez L, Landi S. Promoting Health Literacy to Encourage Prevention and Wellness: Workshop Summary. 2011, The National Academies Press, Washington, DC. 4 Song L, Mishel M, Bensen JT, Chen RC, Knafl GJ, Blackard B et al. How
does health literacy affect quality of life among men with newly diagnosed clinically localized prostate cancer? Findings from the North Carolina-Louisiana Prostate Cancer Project (PCaP). Cancer 2012; 1: 3842– 51. 5 Kickbusch IS. Health literacy: addressing the health and education
172
6
7
8
9
10
11
12
13
14
15
J.M. W angdahl, L.I. M artensson
divide. Health Promot Int 2001; 16: 289–97. M artensson L, Hensing G. Health literacy – a heterogeneous phenomenon: a literature review. Scand J Caring Sci 2012; 26: 151–60. Peerson A, Saunders M. Health literacy revisited: what do we mean and why does it matter? Health Promot Int 2009; 24: 285–96. Nutbeam D. Health literacy as a public health goal: a challenge for contemporary health education and communication strategies into the 21st century. Health Promot Int 2000; 15: 259–67. Sorensen K, Broucke SV, Fullam J, Doyle G, Pelikan J, Slonska Z et al. Health literacy and public health: a systematic review and integration of definitions and models. BMC Public Health 2012; 12: 80. Murata J, Arkida M, Shirai F. A concept analysis of ‘health literacy’ – focusing on participants in health education classes at health centers (Japanese). J Jpn Acad Nurs Sci 2006; 26: 84–92. Adams RJ, Stocks NP, Wilson DH, Hill CL, Gravier S, Kickbusch I et al. Health literacy – a new concept for general practice? Aust Fam Physician 2009; 38: 144–7. Bennett IM, Chen J, Soroui JS, White S. The contribution of health literacy to disparities in self-rated health status and preventive health behaviors in older adults. Ann Fam Med 2009; 7: 204–11. Berkman ND, Sheridan SL, Donahue KE, Halpern DJ, Crotty K. Low health literacy and health outcomes: an updated systematic review. Ann Intern Med 2011; 155: 97–107. Eichler K, Wieser S, Brugger U. The costs of limited health literacy: a systematic review. Int J Public Health 2009; 54: 313–24. Vernon AJ. Health Policy Brief the High Economic Cost of Low Health Literacy in Wisconsin. 2009 Department of Health Policy and Management, UNC Chapel Hill and the National
16
17
18
19
20
21
22
23
24 25
26
Bureau of Economic Research (NBER), Wisconsin. Baker DW, Williams MV, Parker RM, Gazmararian JA, Nurss J. Development of a brief test to measure functional health literacy. Patient Educ Couns 1999; 38: 33–42. Davis TC, Long SW, Jackson RH, Mayeaux EJ, George RB, Murphy PW et al. Rapid estimate of adult literacy in medicine: a shortened screening instrument. Fam Med 1993; 25: 391–5. Ishikawa H, Takeuchi T, Yano E. Measuring functional, communicative, and critical health literacy among diabetic patients. Diabetes Care 2008; 31: 874–9. Svensson E. Statistical methods for repeated qualitative assessments on scales. Int J Audiol 2003; 42(Suppl 1): S13–22. Ishikawa H, Kiuchi T. Health literacy and health communication. Biopsychosoc Med 2010; 4: 18. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol 1993; 46: 1417–32. W angdahl JM, Martensson LI. The communicative and critical health literacy scale – Swedish version. Scand J Public Health 2014; 42: 25–31. Kline T. Psychological Testing: A Practical Approach to Design and Evaluation. 2005, Sage Publications, Thousand Oaks, Calif. WHO. Health Promotion Glossary. 1998, WHO, Geneva. Svensson E. What is the therapeutic effect if the patient gets better, but no one knows how much better? Statistical method for paired ordinal data. L€akartidningen 2007; 104: 596–601. Svensson E. Svensson’s method - freeware and documentation 2013, 2 July, www.oru.se/Institutioner/Handelsho gskolan-vid-Orebro-universitet/Kont akt-och-presentation/Personliga-sidor/ Statistik/Elisabeth-Svensson/Svensoo ns-metod/Svenssons-metod—fri-prog
27
28
29
30
31
32
33
34
35
36
ramvara-och-dokumentation/Svenss ons-method—english/ Svensson E. Analysis of Systematic and Random Differences Between Paired Ordinal Categorical Data. 1993, Gothenburg University, Gothenburg. Kazdin AE. Artifact, bias, and complexity of assessment: the ABCs of reliability. J Appl Behav Anal 1977; 10: 141–50. Cook DA, Beckman TJ. Current concepts in validity and reliability for psychometric instruments: theory and application. Am J Med 2006; 119: 166.e7–16. Kimberlin CL, Winterstein AG. Validity and reliability of measurement instruments used in research. Am J Health Syst Pharm 2008; 65: 2276–84. Lynn MR. Determination and quantification of content validity. Nurs Res 1986; 35: 382–5. Allvin R, Ehnfors M, Rawal N, Svensson E, Idvall E. Development of a questionnaire to measure patientreported postoperative recovery: content validity and intra-patient reliability. J Eval Clin Pract 2009; 15: 411–9. van der Vaart R, Drossaert CH, Taal E, ten Klooster PM, Hilderink-Koertshuis RT, Klaase JM et al. Validation of the Dutch functional, communicative and critical health literacy scales. Patient Educ Couns 2012; 89: 82–8. Jordan JE, Osborne RH, Buchbinder R. Critical appraisal of health literacy indices revealed variable underlying constructs, narrow content and psychometric weaknesses. J Clin Epidemiol 2011; 64: 366–79. Sorensen K, Van den Broucke S, Pelikan JM, Fullam J, Doyle G, Slonska Z et al. Measuring health literacy in populations: illuminating the design and development process of the European Health Literacy Survey Questionnaire (HLS-EU-Q). BMC Public Health 2013; 13: 948. Goldstein G, Hersen M. Handbook of Psychological Assessment, 2nd edn. 1990, Pergamon, New York. Pergamon general psychology series.
© 2014 Nordic College of Caring Science
