Bass, Pat F. III MD; Wilson, John F. PhD; Griffith, Charles H. MD; Barnett, Don R. MD
Dr. Bass is assistant professor, Departments of Internal Medicine and Pediatrics, University of Louisville School of Medicine, Louisville, Kentucky. At the time of the study, he was a faculty development fellow, University of Kentucky College of Medicine, Lexington. Dr. Wilson is professor, Department of Behavioral Sciences, Dr. Griffith is associate professor, and Dr. Barnett is staff physician, Colmery—O'Neil VA Medical Center, Topeka, Kansas.
Correspondence and requests for reprints should be addressed to Dr. Bass, University of Louisville School of Medicine, Office of Curriculum Development & Evaluation, Louisville, KY 40292; telephone: (502) 852-1864; fax: (502) 852-2368; e-mail: 〈firstname.lastname@example.org〉.
Purpose. To determine whether residents could identify patients with poor literacy skills based on clinical interactions during a continuity clinic visit. The authors hypothesized residents would overestimate patients' literacy abilities and fail to recognize many patients at risk for poor literacy.
Method. The Rapid Estimate of Adult Literacy in Medicine—Revised (REALM-R) was administered to screen patients for potential literacy problems. Residents were asked “Do you feel this patient has a literacy problem?” and answered yes or no. Continuity adjusted chi-square was used to test for overestimation of literacy abilities by residents.
Results. REALM-R scores and residents' evaluations of literacy were available for 182 patients. The residents believed 10% of patients (18) had literacy problems based on their clinical interactions. Only three patients passing the literacy screen were incorrectly identified as at risk for literacy. Of the 90% of patients (164) the residents perceived to have no literacy problem, 36% (59) failed the literacy screen.
Conclusion. Resident physicians overestimated the literacy abilities of their patients. A significant portion of these residents' patients may not have the skills to effectively interact with the health care system and are at increased risk for adverse outcomes.
According to the 1992 National Adult Literacy Survey, approximately 25% of Americans are functionally illiterate and another fourth have equivocal literacy skills.1 Poor literacy skills have been associated with a lack of knowledge about the process of disease and poor self-management skills in patients with chronic disease.2 Marginal functional literacy has been associated with poorer physical health, poorer psychological health, and higher health care costs.3,4 Furthermore, consent forms signed by patients with low literacy skills may not indicate informed consent because most consent forms are written at a college reading level or higher.2 The implication for physicians is that at least one fourth of our patients may not be able to discern instructions on prescription bottles, understand patients' education materials, use written directions to get mammography, or give informed consent.
While poor literacy skills have been associated with poor health outcomes, few patients share information about their levels of literacy with health care providers. In fact, most patients with poor literacy skills describe themselves as reading and writing English well or very well.5 Previous research has demonstrated that merely asking patients “Can you read” or “How many years of school did you complete” will also not determine a patient's literacy level.6 Further, there is a substantial amount of shame associated with poor literacy skills. In a study of urban patients with low levels of literacy, 67% did not reveal their reading difficulties to their spouses, and 19% had never told any-one.5
If poor literacy is associated with poor health outcomes and if health care providers are unable to determine literacy based on easily obtainable demographics, how do health care providers identify patients at risk for poor health literacy? Some experts recommend aggregate testing, while others recommend testing individual patients.7 Because of the length of time needed to administer them, the instruments currently available for screening patients' health literacy are impractical in busy practice settings. Surprisingly little research has been performed to determine whether physicians can identify patients with poor literacy abilities based on their clinical interactions. The purpose of this study was to determine whether residents could identify patients with low levels of literacy based solely on their clinical interactions with the patients during a continuity clinic visit. Given patients' unwillingness to share awareness of their poor reading skills with physicians, we hypothesized that residents would overestimate their patients' literacy abilities and not identify many patients at risk for potential health problems due to poor literacy skills.
Over an eight-week period in June and July 2000, the General Internal Medicine Clinic at the University of Kentucky College of Medicine administered a brief literacy exam, a patients' satisfaction survey, and a residents' questionnaire concerning different aspects of their continuity clinic. The clinic is staffed by members of the Division of General Internal Medicine, with residents obtaining a longitudinal ambulatory experience required by the American Board of Internal Medicine. Forty-five categorical, primary care, and combined internal medicine and pediatrics physicians attend continuity clinic at this site, with a resident-to-attending ratio of four to one. A wide spectrum of patients are cared for in the clinic, from adolescents to geriatric patients and from the indigent to deans of university colleges, and the purposes of the visits vary from yearly checkups to consultations for complex medical problems.
The residents' questionnaire covered areas that included satisfaction with the continuity clinic in general, satisfaction with individual patient encounters, and their perceptions of their continuity clinic patients' literacy. Completing the residents' satisfaction questionnaire was encouraged, but voluntary. The residents were asked, “Do you feel this patient has a literacy problem?” and they answered with “yes” or “no.” Patients recruited to participate in the study were a convenience sample presenting to their regularly scheduled appointments or call-in visits. A research assistant approached each patient after he or she had been placed in an exam room and said, “We are trying to find out words our doctors use that patients are unfamiliar with. Would you mind looking at this list of words and reading them aloud to me?” If the patient agreed to participate the research assistant administered the Rapid Estimate of Adult Literacy in Medicine—Revised (REALM-R),8 a new instrument to quickly screen for potential literacy problems that is described below. After checking out at the conclusion of the visit, the patient was asked to complete a patients' satisfaction questionnaire. The research was approved by the university's institutional review board.
The REALM-R has been previously correlated with the Wide Range Achievement Test—Revised (WRAT-R) and the Rapid Estimate of Adult Literacy in Medicine (REALM),8 two well-validated instruments but ones that are impractical for large-scale screening in a busy clinic. The correlations of the REALM-R with the REALM and the WRAT-R were .72 and .78, respectively. Cronbach's alpha for the REALM-R was .91, indicating reliability and internal consistency.8 The REALM-R asks patients to read a series of eight medical words (fatigue, jaundice, directed, allergic, colitis, constipation, anemia, and osteoporosis), and a correct response is given for each correct pronunciation. In a previous study,8 a score of 6 or less on the eight-item REALM-R identified 26 of 30 patients reading at a sixth-grade level based on the WRAT-R. A sixth-grade reading level was chosen as a cutoff for this study because prior studies have demonstrated significant comprehension problems among people who read at this level, and it is likely that patients reading at or below this level will have difficulty understanding most patients' education materials or oral instructions.9 The chi-square test was used to assess whether residents' misperceptions of patients' literacy adequacy arose from chance.
Data were available from 182 patients who completed the REALM-R and whose residents also completed the literacy question. The patients ranged in age from 18 to 93 years, and 85% were Caucasian. The median level of education of the continuity patients was 12 years. All 45 residents estimated the literacy abilities of their patients during the study. Table 1 shows the REALM-R scores based on the residents' perceptions of their patients' literacy. The residents believed only 10% of the patients (18) had literacy problems based on their clinical interactions. The residents perceived 90% of the patients (164) to have no literacy problem, but 36% of these patients (59) scored 6 or below on the REALM-R. Conversely, the residents identified as having literacy problems three patients who scored higher than 6 on the REALM-R, while the other 15 all scored 6 or below. The continuity-adjusted chi-square [(1 df) = 13.18, p < .001] demonstrated a statistically significant overestimation of patients' literacy by residents.
Patients' satisfaction were not correlated and the relevant ratings are not reported here.
Despite reports describing the association of literacy and health, few studies have examined whether physicians can identify patients with poor literacy skills. Residents perceived many patients to have no literacy problem although these patients scored poorly on the REALM-R, suggesting potential literacy problems. Given that many patients will hide literacy problems from physicians and that poor literacy is associated with poor health outcomes, residents' overestimation of patients' literacy is concerning. On the other hand, these residents incorrectly identified only a few patients as having problems.
Identification of patients with inadequate literacy is important because merely supplying patients with appealing, low-level reading materials does not improve comprehension to an acceptable level.9 Furthermore, physicians check patients' understanding of instructions less than 2% of the time.10 In our study, 36% of patients that the residents thought had no literacy problem could not read six of eight common medical words. This suggests a significant number of these residents' patients may not have had the literacy skills to effectively interact with the health care system and would, therefore, have been at risk for adverse health outcomes associated with poor literacy skills. Yet, this risk was not recognized by the residents.
Recognizing this risk, the American Medical Association Foundation's Partnership in Health program is launching a major initiative, Improving the Patient Physician Relationship Through Health Literacy. The initiative will develop programs and interventions to allow better navigation of the health care system for patients with low levels of health literacy skills. Additionally, medical schools and residency programs should consider adding to their curricula literacy awareness and teaching skills to facilitate communication with patients who have low levels of literacy. While the optimal ways to teach this topic to students and for physicians to best communicate with such patients are yet to be determined, students and residents need to realize that poor literacy is an important medical issue.
The study has several limitations. First, the study's sample came from a single institution, was a convenience sample of patients, and was relatively small. Given the demographics of the patients in our clinic, other sites may have more or fewer patients with poor literacy skills. Second, it may be that physicians' lack of experience was the reason residents failed to identify such a large portion of patients who scored poorly on the REALM-R. Although this is a possibility, given that there is little training for physicians in literacy issues in medical school or residency and given that patients often hide their literacy problems, we believe it is unlikely that more experienced physicians would have been any more adept at identifying patients with poor health literacy. Last, while participation by residents was voluntary, all participated in the study.
We found that resident physicians substantially underestimate patients' literacy based solely on their clinical interactions. A brief screening instrument such as the REALM-R could help clinicians identify patients for whom low levels of literacy may complicate health care. However, such identification may not change the dynamics of the clinical encounter if physicians are poorly educated in the care of patients with literacy problems. Further research should examine whether health literacy education for medical students and physicians is associated with improved communication with patients or the prevention of literacy-related adverse health outcomes.
1. Kirsh I, Jungenlut A, Jenjins L. Adult Literacy in America: A First Look at the Findings of the National Adult Literacy Survey. Washington, DC: National Center for Education Statistics, U.S. Department of Education, 1993.
2. Ad Hoc Committee on Health Literacy for the Council on Scientific Affairs. Health Literacy. Chicago, IL: American Medical Association, 1999:552–7.
3. Baker DW, Parker RM, Clark WS, Nurss JR. The relationship of patient reading ability to self-reported health and use of health services. Am J Public Health. 1997;87:1027–30.
4. Grosse RN, Auffrey C. Literacy and health status in developing countries. Annu Rev Public Health. 1989;10:281–97.
5. Parikh NS, Parker RM, Nurss JR, Baker DW, Williams MV. Shame and health literacy: the unspoken connection. Patient Educ Counseling. 1996;27:33–9.
6. Parker RM. What is health care literacy? Paper presented at the Health Literacy: A National Conference, Washington, DC, June 3, 1997.
7. Weiss BD, Coyne C. Communicating with patients who cannot read. N Engl J Med. 1997;337:272–4.
8. Bass III PF, Wilson JF, Griffith CH. A shortened instrument for literacy screening. J Gen Intern Med. 2001;16:117.
9. Davis TC, Fredrickson DD, Arnold C, Murphy PW, Herbst M, Bocchini JA. A polio immunization pamphlet with increased appeal and simplified language does not improve comprehension to an acceptable level. Patient Educ Counseling. 1998;33:25–37.
10. Braddock C, Fihn S, Levinson W, Jonsen A, Pearlman R. How doctors and patients discuss routine clinical decisions: informed decision making in the outpatient setting. J Gen Intern Med. 1997;12:339–45.