Assessing Health Literacy in Iran: An Image in Two Frames


 Background: The aim of this study is to draw a comprehensive picture of patients’ functional and critical health literacy and its association with socio-demographic variables and self-efficacy. We investigated the referrals to Isfahan health centers, in particular the dimensions of reading ability, numeracy and eHealth literacy using two different measures, NVS for the functional and the eHEALS for critical aspects. Methods: For testing individuals with low health literacy, the NVS is a reliable and quick screening tool. eHEALS helps to measure patients’ capabilities in Internet searches. Both of tools are validated in Persian. The measures were administered to a random sample (N = 384), which resembled Iran residents in terms, educational attainment and self-efficacy but not age and gender. Pearson’s correlation coefficient, χ2-test, independent t-test and regression were also used to assess the association between health literacy measures, self-efficacy and predictor variables. Results: More than 60% of our respondents had an inadequate or marginal level of health literacy with NVS, and with eHEALS, more than 50% were classified accordingly. NVS and eHEALS scores showed an association with higher education one of the predictors of health literacy. Higher eHEALS scores were associated with higher self-efficacy. eHEALS provides a valid self-reported assessment of patients’ eHealth literacy, and NVS provides a useful appraisal of patients’ functional health literacy. Conclusions: In our sample, more than 50% of patients had a low level of literacy. eHEALS provides a valid self-reported assessment of patients’ eHealth literacy, and NVS provides a useful appraisal of patients’ functional health literacy. Determining health literacy in patients leads to a better understanding of their perception, application and decision-making in health problems.

Nutbeam suggested a three-tiered concept of health literacy, (1) basic or functional literacy, which implies skills of reading and writing, and understanding and using health information in everyday situations; (2) communicative or interactive literacy, which implies advanced skills that enable active participation in everyday activities, extracting and interpreting health information from different sources such as by the interaction with health care providers and (3) critical literacy, which implies more advanced skills for critically analyzing information obtained from the environment and the media (3). The World Health Organization (WHO) de nes health literacy as "cognitive and social skills that determine a person's motivation and ability to access, understand, and use information, leading to improved quality of health" (4 p. 10). The WHO de nition is an acknowledgment that health literacy is not only a function of reading and numeracy capabilities, but also re ects both communicative/interactive and critical aspects (3).
Overall, individuals with limited health literacy have poor health and are less likely to utilize screenings for prevention. They lack access to appropriate health care, are less willing to pursue treatment, adhere to medication regimes less often, seek help in emergency wards more often, stay in hospital longer, show increased mortality and morbidity and seek help from health care services more often, have di culties with understanding health information available in media or on the Internet and cost the community more money (5). Two developments have furthered the popularity of health literacy in the academic eld of health communication. First, we encounter a paradigm shift from the modern age of medicine that was, hospital-centered, focused on disease, and put patients in a passive role, to an information age system of health care, which is person-centered, focuses on wellbeing, and allows active and dynamic roles for patients (6). The "patient-centered" paradigm in the health system necessitates that patients have particular skills to access and understand health information. Without such access and understanding, patients' involvement in making decisions about their medical conditions would be a risky and expensive endeavor. The second, much younger development is the advent and triumph of the Internet.
Nowadays, the vast availability of the Internet has facilitated access to Online Health Information (OHI) previously only available from health professionals. Search engines and Social Media have become devices routinely employed by patients, enabling them to act autonomously in the health care system and in public health settings (7).
AlGhamdi and Moussa showed for Saudi-Arabian patients that by 2010, "before coming to the clinic, 44% of outpatients in hospitals had searched the Internet for health information; 73% discussed the information with their doctors, and 72% of those who did so believed that this positively affected their relationship with health providers" (8 p. 363). Murray et al. (2003) (9) showed that bringing information to physicians' consultation makes patients more con dent during the consultation, and that once they had discussed their online ndings, they felt more trust in their physician's diagnosis. Internet search behavior leads patients to experience more con dence in their own judgment and therefore, when they enter into a comprehensive negotiation with their specialist, have their own ideas about treatment options and rely less on their specialists. The Internet has turned into an additional resource that empowers patients "to do something" rather than "just being told what to do" by their specialist (10). For autonomous acting in encounters with the health care system, health care consumers need a certain level of health literacy to comprehend their condition and treatment options. Empowerment can be dangerous if not accompanied by an adequate level of health literacy (11). Therefore, measuring this ability and identifying patients with potential health literacy de cits is important and requires speci c measures.

Measurement of Health Literacy
There are many measures of health literacy, two of which we used in our study, and which shall be brie y introduced here. The Newest Vital Sign (NVS) is a quick screening (12) tool to testing individuals for low health literacy. Respondents are shown a food label from an ice cream package and were subsequently asked questions about it. Answering these questions needs reading comprehension and mathematical ability. These abilities are directly associated with understanding basic health information. This tool is available in Persian (13). It can be administered in 3 minutes (12). Participants look at the information on a specially designed ice cream nutrition label and are then asked to answer six questions interpreting the information about it. Health providers can assess the patient's health literacy level and adjust the way to communicate with them based on the number of correct answers. The NVS measures only functional concept of health literacy and it is not a comprehensive assessment.
The e-Health Literacy Scale (eHEALS) conceptualized health literacy in a model reminiscent of a lily-six overlapping petals feed the pistil, which in turn overlaps the petals and binds them together. The model comprises six core skills or literacies: (1) traditional literacy, (2) health literacy, (3) information literacy, (4) scienti c literacy, (5) media literacy, and (6) computer literacy. The eight questionnaire items of eHEALS were developed based on this model. To assess users' knowledge and perceived skills at nd, evaluate, and apply eHealth, this measure is useful (14). There are also two additional questions, not included in the total score, to assess participants' perception of the Internet. We can say, for measuring health literacy these two scales complement each other.
Each measure captures an essential aspect of health literacy, NVS the functional health literacy, the possession of literacy skills su cient to read, understand, acquire and act on information. eHEALS aims at the critical aspects of health literacy and measures patients' capabilities in Internet searches. Critical aspects refer to the appropriate and correct use of online health information. As such, consumers require knowledge about scienti c aspects of what they search and nd on the Internet, and ability to distinguish reliable from unreliable sources. People need eHealth literacy to apply proper techniques for nding health information and use it for making healthy decisions (15).
In addition to knowledge, which plays a key role in therapeutic decision-making, self-e cacy is another variable that assesses the patient's ability to cope with a variety of worries and unforeseen situations. Self-e cacy was de ned by Bandura (1997) as "beliefs in one's capability to organize and execute the courses of action required to manage and prospective situation" (16 e68). Individuals need to have enough information and self-con dence to make right decisions in encountering with health problems. Self-e cacy could ease communication with health providers and health care services. Self-e cacy, along with some other variables, was shown to be associated with a number of management and state of health variables (17).
When the two measures are correlated, persons who score low on one of the measures will tend to also score low on the other. The less correlated the measures are, however, the more meaningful not only the comparison but also the combination of both becomes. We therefore also look at the antecedents of combinations of the two measures.

Health literacy studies in Iran
Several studies have been conducted to assess health literacy in the Iranian population, often patients or health care providers in health centers, such as diabetes clinics or hospitals. In terms of location, most studies were conducted at Tehran and Isfahan universities of medical sciences (18). More than half of the sampling frames aim at healthy individuals, including staff, pregnant women, students, and adults. Most Iranian studies that measure the level of health literacy used the Test of Functional Health Literacy in Adults (TOFHLA). Numerous of these studies showed that an increasing level of education raises the level of health literacy, and health literacy in healthy individuals had a signi cant relationship with income. Inadequate health literacy is more prevalent among the low-income population, rural residents, laborers and retirees. There were signi cant positive relationships between having a job, access to information resources, and levels of health literacy (18). The levels of health literacy with regard to self-care behaviors and self-e cacy have been measured in cases such as following diet recommendations and dosage instruction for medication (19), experience of higher anxiety, stress and negative emotional states (5) and frequency of daily exercise (20).
Most research on students reported moderate health literacy levels measured with Health Literacy for Iranian Adults (HELIA) (21,22) and eHEALS (23). Mahmoodi and Taheri showed that students gained health information through the internet and interaction with friends. There was no signi cant association between health literacy and gender. PhD students' eHealth literacy was found to be higher than MSc Students'. In addition, they reported a signi cant positive relationship between information literacy and student's health literacy, as information literacy explained 40% of the variance in health literacy. Based on the American Library Association de nition, "Information literacy is a set of abilities requiring individuals to 'recognize when information is needed and have the ability to locate, evaluate, and use effectively the needed information." (24 p. 1).
With information literacy, an individual can identify and retrieve credible information, represent and use information effectively and nally share it with others. There are two validation studies for the eHealth literacy scale (eHEALS) in Iran. The rst one was conducted on 525 youths randomly chosen in Yazd (25). The internal consistency of the scale was su cient (Cronbach alpha = 0.88, p < 0.001). The other sampled 223 patients in a military hospital in Tehran (Cronbach alpha = .93, p < 0.001) and found a statistically signi cant weak-tomoderate correlation between the eHEALS scores and education, computer knowledge, Internet knowledge and the use of Internet for health-related purposes and also with smoking, gender, age and Internet usefulness in health decision making (26).
A study was conducted among adults in Isfahan (13) to validate the NVS. In this particular sample, 25.5% of study subjects had inadequate health literacy, 36% had borderline literacy and 38.5% had adequate health literacy (Cronbach alpha = .8, p < 0.001). There were statistically signi cant associations between the educational level, age, gender, economic status and health literacy status. Adults with inadequate health literacy were more likely to be older, less educated, woman and had low income. In another study that used the NVS (27), less than half of the students surveyed at the University of Medical Sciences had inadequate health literacy, and scores among medical students were higher than non-medical students'.
In the case of referrals to health centers in Iran, two studies on the health literacy of pregnant women show that 50% of them have inadequate health literacy (28,29). The Ministry of Health and Medical Education conducted a national study in 2015 to measure the health literacy of Iranian aged 18-65 years (30). The study surveyed 20571 citizens. The measurement tool used in this study was the Health Literacy for Iranian Adults (HELIA). It uses 5-digit scales, and the score ranges from zero to 100 with a higher score indicating higher literacy. According to the results, the average of health literacy scores of adult urban literate population of Iran were 68.32 out of 100 points. The results showed that about 44% of the study population had limited health literacy. The mean score of health literacy was 68 ± 15.13 for men and 69 ± 15.16 for women. Among the variables considered, years of education and age had the strongest impact on health literacy. Almost one in two Iranians has limited health literacy. The 55 years old and over had the lowest scores, and the age range of 35 to 44 years obtained the highest ones. According to the survey, the highest level of limited health literacy was observed in those with one to ve years of education, and the lowest level of limited health literacy in those with an education of 13 years or higher. The ndings also showed that 42% of the respondents had access to health information most often by "radio and TV", 41% by the physicians and health professionals and 33% by the Internet, respectively. Another important nding is that about half of the study population had limited health literacy, which makes some social groups more vulnerable than others, including those over 55 years of age, the less educated and the unemployed.
Results of a systematic review and meta-analysis (30)  As far we know this is the rst study which used NVS and eHEALS among Iranian patients. The aim of this study is to draw a comprehensive picture of functional and critical aspects health literacy in a sample representative of the general population that refers to public health centers. We investigated the health literacy of referrals to Isfahan medical science centers, in particular the dimensions of reading ability and numeracy and eHealth literacy using two different measures. Besides, we used self-e cacy that directly affects health behaviors to assess patients' self-management. This study seeks to evaluate health literacy among patients and its association with socio-demographic variables. The results of this study can help to determine the state of health literacy of patients referring to public health centers in Isfahan and leads to a better understanding of their problems in health decision-making.

Expectations
From the studies that used NVS (13) or eHEALS (25,26) in Iran, we can expect that men, the elderly and those who had less schooling will score lower on the respective scales than their counterparts do. That should generally affect both measures. For the analysis of the combination, one could argue that, if more women than men decide a family's diet, they should be more familiar than men with using the information on a food label. And if more women than man tend for family members when someone is sick, they should have a realistic perception of their own capabilities in health matters, including the nding and understanding of health information on the eHEALS. As gender roles and habits can be expected to increase women's scoring on both scales, we would especially expect to nd women in the group that combines participants high in both measures. If decreasing cognitive performance is among the health conditions that come with age, we might expect a lower score for the older group on the NVS measure, but not necessarily on the eHEALS measure. That, however, is likely to be affected by lower computer use and abilities among the older generation. That is to say, lower scores on both scales can be expected of older respondents, and particularly a high percentage of the elderly should be found in the group with respondents low on both measures. Similarly, there are factors that go along with longer schooling in both scales: higher numeracy and more computer and Internet use for serious purposes. Again, we expect a higher share of those scoring high on both scales among the better educated. As both self-e cacy and eHEALS demand to reveal a person's self-perception, we expect many high eHEALS participants, irrespective of their NVS score, in the highly self-e cacious group.

Health literacy Measure
As was mentioned, we applied two measures of health literacy, NVS and eHEALS. We used the General Selfe cacy Scale (GSE-10) to measure perceived self-e cacy of participants. This scale has been used in 14 cultures with 12,840 participants and its reliability and validity is generally reported in some studies with Cronbach's alphas between 0.76 and 0.90 (35,36). Nezami et al. in 1996 developed the Persian version of GSE-10 scale (36).

Study population
Data collection for the present study took place from July to September 2019. 384 patients were randomly selected from two general polyclinics. Both polyclinics are located in the province of Isfahan and tertiary referral centers with a major primary health care in and out patients department which serves all demographic levels in Najafabad and other nearby cities and villages.
Without restriction to a speci c catchment area, everyone in Isfahan province can use these services. Their clients represent a random sample of the Isfahan community and the county in general and referrals to polyclinics can be a considered representative of the patients in public primary care clinics in Iran. The patients in the two polyclinics had a variety of diseases such as nerve conditions, CVD, diabetes, MS, kidney and cancer or even simple colds. Physicians who treat these patients are all specialists; and one to three general practitioners are one shift for a 24-hour period. Our population contains all patients who referred to one of the two clinics. Inclusion criteria were an age of 18 years or older, ability to speak Persian, and being in a physical and mental health to be able to ll out the questionnaires correctly.
Participates were contacted in the clinic waiting area. We invited everybody who had su cient time, 15 minutes or more for lling out all of the questionnaires, before being called in to see the doctor. After informed consent was obtained written and verbal from each participant, the researchers administered the paper-pen questionnaires, and the participants in themselves lled in demographic questions. Demographics included age (patients had to write down their age), gender and educational level. After demographic questions, we asked two questions about respondents physical and mental health during the past four weeks on a 5-digit scale ("poor" to "excellent") and six other questions about the importance of the following factors determining their quality of life: "Work/school, How you fell about yourself, Your physical state of being, Your mental state of being, Your family/ friends and Your bank balance", in 5-digit scale ("not at all important" to "extremely important"). Scoring and Data Analyzing in the NVS, the number of correct answers ranging from 0 to 6 indicates the level of health literacy. Based on the summative index, participants were grouped with scores of 0-1 indicating limited, 2-3 borderline and 4-6 adequate health literacy. As to the NVS, before lling we explained to participants that the label was for one cup only, but that the whole container carried 4 cups of ice cream and they have to calculate only grams and calories. During the answering, patients were allowed to retain the copy of the label so they could refer to it. The average time to complete the NVS was 10 minutes. After responding, every sheet's scores were recorded.
The eHEALS consists of eight questions to be answered on a 5-point Likert scale (1-strongly disagree to 5-strongly agree). The sum score ranges from minimum 8 to maximum 40 scores. Higher scores indicate higher literacy (14). The General Self-e cacy Scale (GSE-10) has 10 items based on 4-point scale (not at all true [1], hardly true [2], moderately true [3], exactly true [4]). The response to the 10 items have to be summed up to yield the total score ranging from 10 to 40 (35). The correlation among the NVS, eHEALS, the GSE-10 and socio-demographic questions were examined with Pearson's r. Internal consistency of all instruments (face and content) was assessed by Cronbach's alpha. We calculated Pearson's coe cients to assess the association between educational level, gender and age, as known predictors for health literacy, with health literacy measures.
To investigate differences in health literacy scores between female and male participants, an independent t-test was conducted. The association between education levels and health literacy was assessed by ANOVA. A linear regression model was run to estimate the relationship between the socio-demographic predictors and health literacy. For the combination of both health literacy measures, the respective scale were split at the median and combined into four groups. Low on both scales, low on NVS and high on eHEALS, low on eHEALS and high NVS, and high on both. The signi cance level of 0.05 was applied as the threshold in all statistical tests. All data were analyzed using SPSS (version 25), and parametric procedures were applied.

Sample Characteristics
A total of 384 patients completed all measures noted above. Participants were, on average, 33.5 ± 11.8 years old. The majority of them were female (68.5%) and highly educated (58%) ( Table 1). Most of the highly educated respondents had a bachelor degree, which corresponds to o cial statistics in Iran (37). Note: "High" education indicates university degree (vocational/technical, bachelor, master and PhD), "Medium" indicates some high school or diploma and "Low" indicates 6 years or less of schooling.

Health literacy measures
As far as we know, no established GSE-10 measure has been conducted among Persian patients. The Kaiser-Meyer-Olkin measure was .8 and the Bartlett test of sphericity was signi cant (χ 2 45 = 586.617, P < .001) for sample adequacy on GSE-10, which indicates the sampling have been meritorious. The GSE-10 mean score was 3 (range 1-4, SD = .4). We used the median score of the scale (median = 3.1) to create two groups: those with a high (mean ≥ 3.1) and those with a low self-e cacy score (mean ≤ 3.1). 56.8% (n = 218) had low scores and 43.2% (n = 166) had high scores in GSE-10. In our sample 33% of the participants (n = 128) achieved inadequate, 37% marginal (n = 142) and 30% (n = 114) adequate health literacy level based on the NVS scoring. The NVS mean score was 2.49 (SD = 1.7). The internal consistency of the NVS instrument applied in our study was slightly lower compared to the validation study conducted previously in Iran (12). Yet, the internal consistency was acceptable. eHEALS scores were used for a median split of the sample into a high and a low eHealth group, with the median = 3 (range 1-5, SD = 0.7) and missing data were 8.6% (n = 33) and their distribution was completely at random. In the same way, low-and high-NVS groups were formed. Association of health literacy measures with socio-demographic variables The eHEALS correlated with educational level, state of mental and physical health, two factors participants believed were important for achieving quality of life (work/education and family/friends) and GSE-10. There was no correlation between age or gender and eHEALS. Pearson's correlation between NVS and gender or age was insigni cant, and for educational level and two factors that participants believed were important for achieving quality of life (work/education and mental health) it was acceptable. There was an association with GSE-10, state of mental and physical health, and four factors important for the quality of life (work/education, feel about yourself, mental and physical health) ( Table 3). One-way ANOVA was conducted for test the relationship between highest achieved education and the eHEALS mean score, F (6, 377) = 4.83, p < 0.001. In case of patients who indicated primary school as their highest educational level, Tukey's HSD post-hoc test showed signi cant differences between them and all other educational levels. An independent t-test was run to evaluate differences in gender regarding eHEALS mean, GSE-10 mean and NVS summative. The test was signi cant for GSE-10, t (382) = 1.894, p < 0.05, but not for eHEALS, t (382) = .484 and for NVS t (382) = 1.223. Note: Items 4 and 5 were measured on a 5 type agreement scale from 1 = "Poor" to 5 = "Excellent," and the questions were on How would the patients rate their mental or physical health over the past 4 weeks. Items 6 to 11 were measured on a 5-type agreement scale from 1= "Not at all important" to 5= "Extremely important" and the questions were on How important the following factors were in determining patients' quality of life.

Regression model
We ran regression models for the association between education, gender and age with eHEALS, GSE-10 and NVS.

Combined health literacy measures
The two health literacy measures were hardly related to one another (r = .059; p = .25) and could therefore not be turned into a single measure. Nevertheless, the combination of the two measures by distributing respondents in low/high groups as described allows some view into the matter. As Table 5 shows, the combination generates four groups of equal size, indicative of the non-existing correlation between the two measures. Gender and age did not have an in uence on the combined health literacy measures, while education showed the expected effect: among individuals with low education, the low/low health literacy is the most frequent combination, while high/high appears most often among the people with more than average education. Self-e cacy did not affect the combined health literacy classi cation.

Discussion
We investigated health literacy in a socio-demographically diverse population of patients, which referred to two polyclinics in Isfahan. To our knowledge, this is the rst study to measure health literacy simultaneously with both NVS and eHEALS in Iran. The NVS measures functional health literacy like math and reading skills and eHEALS measured critical aspects of health literacy (38). When health literacy was measured by NVS, more than 60% of our respondents had inadequate or marginal level in health literacy; when the eHEALS was used, more than 50% were classi ed accordingly. In this study established predictors for health literacy were age, educational level and gender. Females, older age, higher socioeconomic status, normal self-reported cognition, and lower levels of disability, associated with higher odds of having functional health literacy (39).
In our study both measures of health literacy and their combination just had signi cant relations with level of education. In prior studies, educational attainment was introduced as proxy measure of health literacy. Many Persian language studies support our nding in case of education (40). For example, Banihashemi (43).
The NVS was also applied in two Australian and Singaporean studies (44,45). Among 2824 Australian participants, 55% had adequate functional health literacy, which compared with the 29.7% adequate health literacy that we found, is higher. In the three polyclinics in Singapore, researchers found almost half of 302 respondents attained a score of 6 in the NVS test. These polyclinics were representative of the public primary care clinics in Singapore. Besides, in Dutch validation of NVS among 289 patients, Fransen et al. (2011) argued that NVS is an appropriate measure to assess patients' health literacy levels but its application in different cultures makes di culties. Like their ndings, in some cases our participants claimed that content of the label was incomprehensible and caused confusion, or helpings were calculated instead of gram. Moreover, Iranian food labels differ in layout and content from the NVS label, and daily consuming of ice cream in Iran is less than it is in the United States (46). Looking at the combined analysis, we nd that there is not only no signi cant association; neither is there a trend that just fails to reach signi cance. We did not nd signi cant correlations between eHEALS and gender or age. Obtaining and adequately using eHealth information requires eHealth literacy. The present study therefore assumed female patients to have higher eHealth literacy than male patients, but the results did not show such differences. This agrees with the results of previous studies (47,48,25). Studies conducted with students reported higher scores for males (27,49), as did Dashti et al. (2017) (50).
These studies suggested potential cultural variations. In addition, in this study, there was a positive correlation between the combination of both measures and education. Education level affects eHEALS via online search of health information (51). There are two Iranian studies that reported patients' eHealth literacy was related to their education level, but some other studies (52,53) failed to show this association. Likewise, in the validation of the Dutch version of eHEALS (54), the translated scale was applied to a sample of patients with rheumatic diseases and measured correlation with age, education, and quality of Internet use.
Their results showed correlation between eHEALS and education and age were not signi cant. Principles of the self-e cacy and social cognitive theory, and eHealth literacy's six core skills are to be considered as precursors of behavior change and skill development, therefore, eHEALS can be considered as a measure of self-e cacy in the online health context (16). In this study, there is a signi cant positive relationship between eHEALS and GES-10, which indicates that the patients with higher level of the eHealth literacy had higher self-e cacy. Other researchers also found the effect of health literacy was mediated by knowledge and self-e cacy (59,60).
On one hand, critical and communicative health literacy provides social cognitive knowledge and comprehension skills necessary for proper disease management. These skills increase patients' self-con dence, helping them communicate effectively with health providers and make better use of health information resources (61). On the other hand, Diviani et al. (2015) believes that health literacy might play a role in OHI evaluation. Patients with low health literacy are less aware of OHI quality, less critical about their ability to evaluate it or apply in dealing with health providers (62). Social media are key role to spread and access OHI in Iran and even all over the world. The point is, although eHEALS is considered as a valid tool for assessing competency with the Internet, it is thought of failing to capture the skill related to the use of social media.
Consequently, relevant qualities of sources that should be part of the concept measured are in fact excluded. In the present study, there was a signi cant positive correlation between eHEALS and patients' physical and mental health ratings. When we ask questions about "mental health", patients' answers refer to their mood or satisfaction of life, not to "knowledge and beliefs about mental disorders, which aid their recognition, management or prevention" (63 p. 182). These results may be interpreted that patients with higher health literacy are more con dent and satis ed and have a better estimate of their mental and physical health. Mitsutake

Conclusions
The aim of this study was to draw a comprehensive picture of functional and critical aspects health literacy of Iranian patients. We investigated the health literacy of referrals to Isfahan medical science centers, in particular the dimensions of reading ability and numeracy and eHealth literacy using two different measures (NVS and eHEALS). This study sought to assess health literacy among patients and its association with socio-demographic variables and self-e cacy. In our sample, more than 50% of patients had low functional and eHealth literacy. Based on o cial reports (70), until the March of 2018, Internet penetration rate in Iran was 90%. Health providers' efforts should aim to raise awareness on OHI resource quality especially among low eHealth literate patients. Determining health literacy in patients leads to a better understanding of their perception, application and decision-making in health problems. We propose that further research in needed to measure health literacy with these two tools in the other public hospitals and polyclinics and in other provinces in Iran, particularly in provinces with different population texture form Isfahan. We propose that further research is needed to measure health literacy with different population such as more prevalent chronic diseases in Iran like MS, Diabetes or Cancer patients.

Limitations
The results of collecting data with NVS and eHEALS cannot entail the complex construct of health literacy with its functional, interactive/communicational and critical dimensions. Given that our populations were patients, it can be assumed that they were not in their normal physical health circumstances and were worried about their health status. In case of eHEALS, many older people, especially when they had not had good education refused to ll out questionnaires because they were ashamed to be assessed or did not have up-to-date information or access to new digital technologies (71). Furthermore, eHEALS measure the eHealth literacy of patients based on their self-reports and not actual performance or record of Internet use. Therefore, more studies that measure their actual online information usage or performance are required. Many respondents did not really understand the ice cream label of NVS on their own, and asked help because it is not an everyday treat in Iranian life style. Ethics approval and consent to participate The consent approval was obtained written and verbal. At rst of randomly choosing and meeting the referrals to the polyclinics, we verbally asked them to participate, if they had enough time and tendency to being involved, because the polyclinic's waiting halls were overcrowded in 6 days of each week and therefore at the rst glance we had to ask them face to face. also, being over crowded means there were a variety of Iranian ethnics who do not speaking in criteria Farsi, so we had to evaluating their ability to speak in Persian.
We informed them that the name or other personal information except demographic won't be necessary and use in our research.
When the referrals accepted to participate, we gave the questionnaires to the he/she. At The rst paragraph of every single one of questionnaires there was a written consent which explained that lling these questionnaires won't length be more than 15