Assessing Adult Health Literacy in Urban Healthcare Settings

o

r

i

g

i

n

a

l

c

o

m

m

u

n

i

c

a

t

i

o

n

Assessing Adult Health Literacy in Urban Healthcare Settings La Vonne A. Downey, PhD and Leslie S. Zun, MD, MBA

Objective: To determine if health literacy is lower among those using the emergency department as compared to those using community health clinics. A comparison was done of the health literacy level of patients at a level-I adult and pediatric emergency department and 3 community health clinics. The second purpose of the study was to identify and assess predictors of low or adequate health literacy in order to better accommodate the communication needs of these patients. Methods: The study used a convenience, cross-sectional design. The settings for the study were varied. The emergency department was a level-I inner-city pediatric and adult trauma center with 45,000 annual visits. The 3 outpatient clinics were all members of a 44-clinic network of private, community health centers that serve 175,000 underserved patients annually. Overall, 536 patients were approached to complete an in-person survey that included a 15-item written questionnaire and the Short Test of Functional Health Literacy in Adults (S-TOFHLA), a English- and Spanish-validated health literacy measure. Based on language spoken with healthcare providers, patients were given either the Spanish or English S-TOFHLA. Results: The refusal rate was 25.7%, with 536 patients approached and 398 completing the survey. Three of the four sites had a refusal rate of 8% per site. The fourth site had a higher refusal rate at 38%. Overall, 20% of subjects had marginal or inadequate functional health literacy. There was a significant difference at p=18.42, df=3, p=0.001 among sites. Three of the sites had ≥78% of the participants scoring at the adequate level. The fourth site, however, only had 66% who scored at the adequate level, with 34% of the scores in the inadequate functioning level as compared with 14% for the emergency department and its adjacent clinic and 3% for the clinic located at the city’s edge. Using logistical regression, there was an association between literacy level and education (t=2.653, sig 0.008) and age (t=-6.451, sig 0.001). Conclusion: Less-than-adequate functional health literacy was seen at 20% of all sites. The location of healthcare access was not as predictive of low functional health literacy as were demographic indicators, such as age and education levels. Key words: emergency department n healthcare n access

1304 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION

© 2008. From Roosevelt University, School of Policy Studies, Chicago, IL (Downey, assistant professor, policy studies); and Department of Emergency Medicine, Finch University/Chicago Medical School, Mount Sinai Hospital, Chicago, IL (Zun, chairman and professor of emergency medicine). Send correspondence and reprint requests for J Natl Med Assoc. 2008;100:1304–1308 to: Dr. La Vonne Downey, Assistant Professor/Roosevelt University/School of Policy Studies, 430 Michigan Ave., Chicago, IL 60605; phone/fax: (847) 360-1003; e-mail: [email protected]

Introduction

T

he definition of health literacy is the capacity to obtain, process and understand basic health information and services needed to make appropriate health decisions.1 Healthcare systems increasingly rely on individuals to manage their own health; the ability to do so is a function of a patient’s health literacy. Low health literacy is associated with lower compliance with treatment regimens and lower self-rated health status, as well as decreased ability to share in healthcare decision-making.2-4 Healthcare systems exhibit increased hospitalization rates and greater inpatient spending in serving patients with low health literacy than those with adequate health literacy.5 The American Medical Association recognizes limited patient literacy as a barrier to effective medical diagnosis and treatment, and suggests further research to address the issue of health literacy.6 The importance of adequate communication between the healthcare providers and non-English-speaking patients cannot be overemphasized. The lack of health literacy has been correlated with providing of inferior healthcare and an additional burden on healthcare resources. Williams found that 40% of the emergency department asthma patients read at or below the sixth-grade level and that this inadequate literacy level was strongly correlated with poor knowledge of asthma and improper metered-dose inhaler use.7 Schillinger found that poor health literacy has been associated with poor diabetic control.8 Several studies have illustrated the need to improve health literacy in order to improve patient care: Morales and Cunningham showed that a lack of health literacy in non-native speakers results in patients feeling less satisfied with medical care and with the explanations and responses they receive from the doctors treating them.9 VOL. 100, NO. 11, NOVEMBER 2008

Adult Health Literacy in Urban Healthcare Settings

Purpose

The main purpose of this study was to assess the functional health literacy of patients at multiple points of service, including the emergency department and 3 community health clinics. The second purpose of the study was to identify and assess predictors of low or adequate health literacy in order to better accommodate the communication needs of these patients.

Methods

The study used a convenience, cross-sectional design. The settings for the study were varied. The emer-

gency department, site 1, was a level-I inner-city pediatric and adult trauma center with 45,000 annual visits. The 3 outpatient clinics were all members of a 44-clinic network of private community health centers that serve 175,000 underserved patients annually. A description of these clinic sites were as follows: 1 across from the emergency department, site 2; 1 that served a mixture of clients on the edge of the city, site 3; and site 4 located in and served a predominately Hispanic population. This project was approved by the institutional review board. Inclusion criterion was a consenting adult (≥18 years) patient in the emergency department or 1 of the

Table 1. Demographic breakdown Age Breakdown by Site Site 18–28 1 37 36.6% 2 18 18.0% 3 20 20.6% 4 51 51.0% Total 126 31.7%

Age Range 29–41 30 29.7% 19 19.0% 35 36.1% 27 27.0% 111 27.9%

42–52 19 18.8% 31 31.0% 29 29.9% 13 13.0% 92 23.1%

≥53 15 14.9% 32 32.0% 13 13.4% 9 9.0% 69 17.3%

Total 101 100.0% 100 100.0% 97 100.0% 100 100.0% 398 100.0%

* Categories were determined by clusters in data responses

Education Level by Site Site Education Range Total Less than High-School Diploma/ More than High-School High-School Diploma GED Diploma/GED 1 28 38 35 101 27.7% 37.6% 34.7% 100.0% 2 24 27 49 100 24.0% 27.0% 49.0% 100.0% 3 22 23 52 97 22.7% 23.7% 53.6% 100.0% 4 42 37 21 100 42.0% 37.0% 21.0% 100.0% Total 116 125 157 398 29.1% 31.4% 39.4% 100.0% Income Range by Site Site <$14,999 1 46 45.5% 2 35 35.0% 3 56 57.7% 4 34 34.0% Total 171 43.0%

Income Range Total $15,000–$24,999 >$25,000 Unknown 25 30 0 101 24.8% 29.7% 0.0% 100.0% 18 47 0 100 18.0% 47.0% 0.0% 100.0% 22 19 0 97 22.7% 19.6% 0.0% 100.0% 37 26 3 100 37.0% 26.0% 3.0% 100.0% 102 122 3 398 25.6% 30.7% 0.8% 100.0%

* Categories were determined by clusters in data responses

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION

VOL. 100, NO. 11, NOVEMBER 2008 1305

Adult Health Literacy in Urban Healthcare Settings

3 outpatient adult clinics. Exclusion criteria were anyone unable to consent or refuse to consent, inappropriate mental capacity to complete the study, patients aged <18 years, and patients who communicate with their physician in a language other than English or Spanish. In order to achieve a cross-site comparison the enrollment goal was set at 100 patients per clinic site. Data were entered into SPSS® 14.0 (SPSS Inc., Chicago, IL), and a Pearson’s chi was used to determine if there was a significant difference at the ≤0.05 level among and within subjects for health literacy level, clinic site, age, education, income, race, usual source of medical care and self-rated ability to understand medical information and overall health. A second analysis using a logistical regression was also done to find correlations among health literacy level, clinic site, age, education, income, race, usual source of medical care and self-rated ability to understand medical information and overall health. All patients were approached to determine initial interest and eligibility. Once informed consent was obtained, subjects were given a written demographic/ screening questionnaire. Based on the primary language spoken with healthcare workers, each subject was given the validated versions of either English- or Spanish-language Short Test of Functional Health Literacy in Adults (S-TOFHLA). Spanish-speaking investigators were used to approach Spanish-speaking patients in the predominately Spanish-speaking clinic site. The health literacy screening questionnaire consisted of 3 parts. The first contained basic demographic questions such as age, race, income and education. The second part had healthcare specific questions such as language preference, use of interpreter services, usual source of care (clinic, emergency department, etc.), selfrated health literacy and overall health. The third part was the Short Test of Functional Health Literacy in Adults (S-TOFHLA). This validated English and Spanish version tests consisted of a 36-item, 2-passage instrument designed to test patients’ abilities to understand and uti-

lize health-related information.10 The patients were asked to choose the appropriate answer to each item among the 4 choices listed. An overall score was determined by adding the number of correct responses given in 7 minutes. The scoring of the S-TOFHLA consists of these 3 categories based upon the patient’s overall score: 0–16: inadequate functional health literacy, 17–22: marginal functional health literacy, and 23–36: adequate functional health literacy.11

Results

Overall, 536 patients were approached in order to compile 100 completed surveys per each of the 3 clinic and 1 emergency department site. The overall refusal rate was 25.7%, with 3 of the sites having a 8% rate and the fourth site having a higher refusal rate of 38%. The most common reasons for refusal to participate given by the 136 were as follows: did not have glasses/vision not good enough to read forms (40%), did not want to sign anything (30%), did not feel well enough to participate (20%), no reason given/not interested (10%). There was a significant difference (P=234.4, df=3, p=0.001) in terms of the racial/ethnic composition of the 4 sites. Overall racial/ethnic composition was 47% African American, 39% Hispanic, 8% white and 2% other. The majority of patients at site 1 (66%), site 2 (63%) and site 3 (59%) were African Americans, with <15% being Hispanic. Site 4 had 71% Hispanic and no African Americans. There was also a difference in ages based on sites (P=55.04, df=9, p=0.001) (Table 1). The age range was 28% between the ages of 18–28, 29% were 29–41, 24% were 42–52 and 18% were ≥53. Out of the total sample population, only 6% were age >65. The 2 sites that served the youngest population, those within the 18–28 ranges were site 1 at 37% and site 4 at 42%. Each of those clinics also had the lowest percentage of those who were ≥53, with site 1 having 14% and site 4 at 11%. There were 50% males and 50% females. There was a significant difference (P=29.1, df=6, p=0.001) between

Table 2. Usual source of medical information by site Site Usual Source Total Public Hospital Hospital Urgent Some No Prefer Doctors Health Outpatient Emergency Care Other Kind Usual Not Not Office Clinic Center Room Center of Place Place Sure to Answer 1 54 10 5 25 1 1 3 1 1 101 53.5% 9.9% 5.0% 24.8% 1.0% 1.0% 3.0% 1.0% 1.0% 100.0% 2 82 8 2 8 0 0 0 0 0 100 82.0% 8.0% 2.0% 8.0% 0.0% 0.0% 0.0% 0.0% 0.0% 100.0% 3 49 35 1 4 0 1 6 0 1 97 50.5% 36.1% 1.0% 4.1% 0.0% 1.0% 6.2% 0.0% 1.0% 100.0% 4 42 48 3 5 1 0 0 0 1 100 42.0% 48.0% 3.0% 5.0% 1.0% 0.0% 0.0% 0.0% 1.0% 100.0% Total 227 101 11 42 2 2 9 1 3 398 57.0% 25.4% 2.8% 10.6% 0.5% 0.5% 2.3% 0.3% 0.8% 100.0%

1306 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION

VOL. 100, NO. 11, NOVEMBER 2008

Adult Health Literacy in Urban Healthcare Settings

the sites in educational levels (Table 1). The education levels varied with 29% having less than a high-school diploma, 31% having a high-school diploma/GED and 40% having some or a college degree. The location with the highest percentage (42%) of those with less than a high-school education and the lowest percentage (21%) with some or 4 years of college was site 4. There was also a difference between income ranges (P=38.3, df=9, p=0.000) at the various sites (Table 1). The majority of participants (43%) made <$14,999 a year, with much smaller numbers (25%) earning $15,000–$24,999 and 30% earning >$25,000 a year. Sites 1–3 (emergency department 45%, adjacent emergency department clinic 35% and urban edge clinic at 58%) had the largest numbers of those earning <$14,999 per year, as compared to site 4 with (34%) within the same range. There was a significant difference (P=16.9 df=3, p=0.001) in terms of language spoken in the home. The overwhelming majority (98–100%) at sites 1–3 spoke English at home versus as compared with 31% of those at site 4. At site 4, 66% spoke Spanish and the remaining 3% spoke both English and Spanish. There was no significant difference among the sites as to the main source of medical information (Table 2). The usual source of medical information was 60% from the doctor’s office, 24% from public health clinics and 10% from hospital emergency departments. There was no significant difference (P=10.4, df=3, p=0.106) on how the respondents from each location rated their overall health status. The respondents at each site rated their health at similar levels, with 12% stating they were in excellent health, 24% very good, 33% good, 22% fair and 7% poor.

Literacy Levels

Three out of the 4 locations used the English test form (97–100%). The fourth site, however, had 33% using the English form and 68% using the Spanish S-TOFHLA

test form. Literacy level overall at the sites was 20% (100) at marginal or inadequate health literacy level, and 80% (300) had adequate health literacy levels. The literacy levels are broken down in order to illustrate those with adequate versus less-than-adequate literacy levels. There was a significant difference at P=18.4, df=3, p=0.001, however, among sites (Table 3). Three out of the 4 sites had ≥78% of the participants scoring at the adequate level with site 1 at 78%, site 2 at 81% and site 3 at 90%. The fourth site, however, had only 66% who scored at the adequate level, with 34% of the scores in the inadequate functioning level as compared with 21% for site 1 and 2 and 9% for site 3. This same difference was seen at the marginally functioning level with 7% at site 1, 5% at site 2, 6% at site 3 and 13% at site 4.

Predictors of Less-than-Adequate Health Literacy Age proved to be a significant predictor (t=-6.45, p=0.00). Of those patients with inadequate/marginal health literacy, 37% (n=28) were age >53, and 28% (n =21) were between 43–52 years of age. Education was also predictive, as (t=2.65, p=0.008) those patients with inadequate/marginal health literacy 80% (n=61) had a high-school diploma/GED or less.

Discussion

Approximately 20% of all patients surveyed had inadequate or marginal health literacy. Health literacy level did differ among the sites. Age, education and the language spoken at home were all found to be significantly related to health literacy levels as determined by the S-TOFHLA. Educational levels as seen in the emergency department and urban clinic populations are often low with a significant percentage of patients having less than a high-school education level.11 This phenomenon was also seen within this study, especially for those in the inadequate to marginal levels of health literacy. Sentell et al., however, did point to the relationship between

Table 3. Literacy level by site Site Literacy Range Total Marginal or Inadequate Adequate Health Health Literacy Level Literacy Level 1 22 79 101 21.8% 78.2% 100.0% 2 19 81 100 19.0% 81.0% 100.0% 3 9 88 97 9.3% 90.7% 100.0% 4 34 66 100 34.0% 66.0% 100.0% Total 84 314 398 21.1% 78.9% 100.0% * Categories were broken down in order to show difference between those with adequate and those with nonadequate literacy ranges

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION

VOL. 100, NO. 11, NOVEMBER 2008 1307

Adult Health Literacy in Urban Healthcare Settings

education and literacy.12 They felt that this literacy inequality was related to health disparities that were previously thought to be explained by race and educational factors alone. Lo et al. also saw a relationship between health literacy and those with less than a high-school education as well as recent immigrants.13 This would be consistent with what Powers et al. saw when they found that 40% of the emergency department patients could not read at an eighth-grade level.14 The patients who spoke Spanish (27%) at home and or preferred (30%) it as the language used in medical settings also had the highest levels of inadequate or marginal health literacy. Those who spoke only English at home and preferred English as the language for receiving medical information had an 82% adequate health literacy.

Limitations

There were several limitations to this study, the first of which is selection bias. Those patients with low literacy and or language skills may have declined in greater numbers to take part in the study, thus creating a higher measured literacy level than was really present within these 4 locations. This would have led to an underestimation of those who are within the marginal/inadequate health literacy levels. The fourth site had the largest refusal rate, which could have impacted the literacy levels. This could be due to a higher level of undocumented immigrants served at this site who did not want to take part due to legal concerns. Patients completing the non-STOFHLA aspect of the surveys may not have always fully comprehended the questions or answered them in accordance with the instructions. Finally, some patients may have underestimated their income in order to be eligible for a sliding scale payment system of care within the clinic settings. This may partly explain the inverse relationship between educational levels and income. Future research using this approach should enhance the questionnaire to order to obtain more descriptive information with relationship to clinic choice, transportation needs, frequency of visit and medical complaints, as these factors may explain who comes to each clinic and possibly correlate to the differences of health literacy seen.

and community health centers to provide such methods might add to their costs. Many of these healthcare settings are already under financial constraints. But the effects of lower health literacy on the quality of care (hence the health outcomes) also impacts the cost of delivery care within these settings, with lower levels of patients using self-care and preventive measures and increasing hospitalization. Reimbursement for the time spent increasing health literacy with patient education and on multidisciplinary teams could be shown to be cost effective insofar as it reduces costs in the long run and improves both patient and community health.

References

1. Win K, Schillinger D. Understanding of warfarin therapy and stroke among ethnically diverse anticoagulation patients at a public hospital. Abstract. J Gen Intern Med. 2003;18(suppl 1):278-279. 2. Baker DW, Gazmararian JA, Williams MV, et al. Functional health literacy and the risk of hospital admission among Medicare managed care enrollees. Am J Public Health. 2002;92:278-283. 3. Kim SP, Knight SJ, Tomori C, et al. Health literacy and shared decision making for prostate cancer patients with low socioeconomic status. Cancer Invest. 2001;19:684-691. 4. Ratzan SC, Parker RM. Introduction. In: National Library of Medicine Current Bibliographies in Medicine: Health Literacy. NLM Pub. 2002:2000-2001. 5. Institute of Medicine. Health Literacy. Washington, DC: National Academic Press; 2004;7. 6. American Medical Association: Council on Scientific Affairs. Health Literacy. JAMA. 1999;281:552-557. 7. Williams MV, Baker DW, Honig EG, et al. Inadequate literacy is a barrier to asthma knowledge and self care. Chest. 1998;114:1008-1015. 8. Schillinger D, Grumbach D, Piette J., Association of health literacy with diabetic outcomes. JAMA. 2002;288:475-482. 9. Morales LS, Cunningham WE, Brown JA. et al. Are Latinos less satisfied with communication by health care providers. J Gen Intern Med. 1999;14:257-258. 10. Davis TC, Long SW, Jackson RH, et al: Rapid estimate of adult literacy in medicine: A shortened screening instrument. Fam Med. 1993;25:391-395. 11. Parker R, Baker D, Williams M, et al. The Test of functional health literacy in adults (STOFHLA); a new instrument for measuring patients’ literacy skills. J Gen Intern Med. 1995;10:537-541. 12. Sentell TL, Halpin HA. Importance of Adult Literacy in Understanding Health Disparities. J Gen Intern Med. 2006;21:862-866. 13. Lo S, Sharif I, Ozuah PO. Health Literacy Among English –Speaking Parents in a Poor Urban Setting. J Health Care Poor Underserved. 2006;17:504-511. 14. Powers RD. Emergency department patient literacy and readability of patient directed materials. Ann Emerg Med. 1988;17:124-126. n

Conclusion

At least 20% of patients seen in urban healthcare settings have marginal or inadequate functional health literacy. This level may be higher among patients who are older, those with lower income and education levels, and in those patients whose most preferred language is not English. Low and or limited health literacy also contributes to lower levels of patient self-care in such diseases as hypertension, diabetes and asthma. The knowledge of these potential indicators of lower health literacy level could help healthcare professionals develop methods to help improve patients’ understanding of their health situations. Asking emergency departments 1308 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION

To photocopy, e-mail, post on Internet or distribute this or any part of JNMA, please visit www.copyright.com.

VOL. 100, NO. 11, NOVEMBER 2008