|Home | About | Journals | Submit | Contact Us | Français|
Health literacy-related problems can interfere with effective doctor-patient communication and effective patient care. This study examined several health literacy-related markers for patients seeking treatment in hospital emergency departments and physician and dentist offices for dental problems and injuries. Participants consisted of low-income white, black, and Hispanic adults who had experienced a dental problem or injury during the previous twelve months and who visited a hospital emergency department, physician, or dentist for treatment. A stratified random sample of Maryland households participated in a cross-sectional telephone survey. Interviews were completed with 94.8 percent of 423 eligible individuals. Multivariable logistic regression analyses were performed. Only 10.0 percent of the respondents expressed a difficulty understanding what they were told by the health provider, while 4.9 percent expressed a difficulty understanding the dental or medical forms they were asked to complete and 6.9 percent reported that they had difficulty getting the health provider to understand their dental problem or injury. Logistic regression analysis found that males and Hispanics were significantly (p<0.05) more likely to experience health literacy-related problems. In general, respondents did not express health literacy-related problems. Additional research is needed to identify health literacy-related barriers to effective patient-provider communication.
Good doctor-patient communication improves patient health and health care.1 Unfortunately, approximately one in three Americans lack health literacy2,3 in that they do not have the ability to obtain, process, or comprehend basic health information required to make appropriate decisions about their health.4 This lack of health literacy impedes doctor-patient communication, which in turn affects the patient’s health since effective doctor-patient communication has been shown to improve adherence to medical recommendations.1 This is especially important because approximately 40 percent of patients do not follow physician recommendations.5
Although patients consider communication skills to be among the top three skills needed by physicians, patients often rate their own doctor’s skills as unsatisfactory.6 At the same time, providers’ perceptions of their communication skills are often at variance from those of their patients.7 Problems with health literacy are expensive, costing an estimated $29 to $69 billion annually.8 Costs are incurred through the overuse of diagnostic tests and other services, as well as inappropriate or unnecessary referrals.9,10 Similarly, limited health literacy has been found to be a barrier to patients taking prescribed medications.11 Health literacy-related problems are of particular concern in hospital emergency departments where a substantial percentage of patients have been shown to have limited health literacy.12
As might be expected, a patient’s ability to acquire and use information is dependent upon the specific health problem, the type of provider, and the health care setting.13 Furthermore, improving health literacy requires more than simplifying the language of health information.14 Although research into the impact of health literacy on oral health has been limited,15 policy reports have suggested that inadequate oral health literacy may contribute to poor oral health.16,17 Our research study examined several health literacy-related markers among patients seeking treatment in hospital emergency departments (EDs) and physician and dentist offices for dental problems and injuries. Our hypotheses were that low-income and ethnic/racial minority patients would experience a greater prevalence of health literacy-related problems than higher income and white patients and that patients receiving care at EDs and physician offices would be more likely to experience health literacy-related problems than patients receiving care from dentists. This research was part of a larger study that examined the use of physicians, EDs, and dentists for the treatment of dental problems and injuries.18
The target population comprised low-income non-Hispanic white, non-Hispanic black, and Hispanic Maryland households with adults twenty-one years of age and greater who had experienced a dental problem or injury during the previous twelve months and who visited a physician, ED, or dental office for treatment of that problem. Low income was defined as respondents with annual family income less than $25,000, which is approximately 150 percent of the federal poverty level for a family of three. Participants with higher income were included in the study for comparative purposes and to examine the relative impact of income levels on respondent service use. Dental problem or injury was self-defined by a positive response to this question:
Have you had a dental problem or injury at any time during the past twelve months? By dental problem or injury we mean things like toothaches, accidents and other trauma, gum infections, jaw or face pain, dry or burning mouth, tongue or lip problems, sores or ulcers in the mouth, bleeding anywhere in the mouth, and pain caused by dentures, crowns, or bridges, but not routine dental care like cleanings or check-ups.
We did not attempt to assess theoretical levels of literacy, but used markers of health literacy that were designed to assess actual health literacy-related barriers to effective patient care and treatment outcomes. The respondents were asked to respond to three health literacy-related markers:
Respondents who answered in the affirmative to any of the questions were then asked, “What did you do?” Because it was necessary to conduct telephone screenings and interviews with Hispanic respondents who have limited English-language ability, the screening and interview instruments were translated into colloquial Spanish that could be understood by Spanish-speaking persons with limited education. Trained bilingual interviewers conducted the interviews with the Spanish-speaking persons.
2000 U.S. census data were used to divide the 3,058 block groups in Maryland according to the percentages of low-income persons and persons of different races/ethnicities they contained. Five groupings were created based on income and race/ethnicity. A random sample totaling 27,002 Maryland households with listed telephone numbers was selected from within each of the five block groups identified by poverty income level and racial/ethnic composition with the objective of having approximately equal numbers of interviewed persons from each grouping.
Interviews were conducted using Computer-Assisted Telephone Interviewing (CATI) technology to screen for eligible adults and to interview only one eligible adult per household. Interviewers completed the screener and interview in either English or Spanish, based on respondent needs. All of the 27,002 listed telephone numbers in the sample were called. However, 6,758 (25.0 percent) did not meet our specification as working residential landline telephones, but were instead business phones, cell phones, pay phones, fax machines, or non-working numbers. Of the remaining 20,244 working residential numbers, contact was made with 13,136 (64.9 percent). Of those contacted, 4,357 (33.2 percent) households completed a screening interview. From these, we identified 1,387 households that contained one or more eligible persons. Where there was more than one eligible person in a household, the CATI program randomly selected one to interview. In order to attain some balance between the numbers of persons visiting an ED or a physician’s office and a dentist, the CATI program was programmed to select a random sample of approximately 20 percent of the large majority of eligible adults who reported visiting only a dentist for treatment of their dental problem or injury. Interviews were completed with 401 (94.8 percent) of the 423 randomly selected eligible respondents: 282 female (70.3 percent) and 119 male (29.7 percent); forty-one Hispanic (10.2 percent), 144 white (35.9 percent), 199 black (49.6 percent), and seventeen other race/ethnicity (4.2 percent). Only twelve of the selected eligible respondents who were contacted for an interview refused; however, there were ten additional individuals who had not been recontacted when the study ended.
The sample cases were weighted to represent the size of the target population for analysis: low income or minority, twenty-one years of age or older, who had a dental problem or injury in the previous twelve months and had visited a physician, ED, or dentist for treatment. The weighted sample yielded the following population distribution (n=80,203): 33,280 male (41.5 percent) and 46,923 female (58.5 percent); 2,928 Hispanic (3.7 percent), 64,928 white (81.0 percent), 9,472 black (11.8 percent), and 2,875 other race/ethnicity (3.6 percent). There were no statistically significant associations among the respondent’s age, gender, race/ethnicity, and income with the exception that a larger percentage of males were in the older age groups (chi-square 3.0; df 3; p=0.03) and black respondents were more likely to be in the lower income groups than white or Hispanic respondents (chi-square 3.9; df 6; p=0.001). The research protocol was reviewed by the University of Maryland at Baltimore Office for Research Studies and judged exempt from review; however, a verbal informed consent was obtained from all participants. Respondents were sent a $10 gift card for participating in the interview.
Interviewed cases were weighted for analysis to adjust for sampling design, probability of selection, and unlisted, nonresidential, and unanswered telephones, as well as for screening and interview nonrespondents. Weighting was necessary because the sample design was developed to achieve over-sampling of low-income minority (Hispanic and non-Hispanic black) households. Analysis weights were used to restore proper representation to the study groups by adjusting for differences in sampling and nonresponse rates. The analysis is thus based on an estimate of the number of Maryland adults who had a dental problem/injury in the past twelve months and sought care from a physician, ED, or dentist (n=80,203).
Weighted tabular analysis was conducted using chi-square tests of statistical significance that accounted for the sample groupings and differential response rates. All statistical tabular analyses used SUDAAN (Research Triangle Institute, Research Triangle Park, NC), an analytic package designed especially to analyze complex survey samples with clustered and weighted data. In addition, we used logistic regression analysis modeling19 to test whether there were statistically significant variables associated with respondents’ experiencing any of the three health literacy-related markers: understanding what the treatment provider was saying, understanding the medical forms that needed to be completed, or difficulty getting the treatment provider to understand their dental problem or injury. In particular, we examined whether a respondent was more likely to report any of these problems when they went to see a particular type of health professional (dentist, physician, or an ED). If respondents saw more than one provider, we questioned them about their visit to the first provider seen.
We also assessed the association between measures in the model that represented covariates that we thought could have an important impact on the type of provider selected or the likelihood of experiencing a problem. These variables included demographic characteristics such as age, gender, educational level, income level, and race/ethnicity, along with health care utilization and access measures such as having Medicaid, health insurance, dental insurance, a regular physician, a regular dentist, and annual preventive medical and dental visits. Logistic regression modeling yields pseudo-R squares that can be interpreted and used as an indicator of how much the variables in the model contribute to reducing errors in predicting the likelihood of specific outcomes.
Patients seeking care for dental problem or injuries from physicians, EDs, or dentists were asked if there was any time during their visit when they did not understand what the health provider/staff member was trying to tell them. They were instructed to think of times when someone at the facility was explaining their illness, describing their treatment choices, or explaining how to take their medications. Only one in ten (10.0 percent) of the respondents expressed a difficulty understanding what they were being told. These respondents were asked what they did (Table 1). Overwhelmingly, these patients either asked the provider/staff member questions about the information they did not understand (58.1 percent) or they did nothing, acting as if they understood what they were being told (52.7 percent). Respondents were next asked if they had any difficulty understanding the medical forms they had to sign (insurance, consent, HIPAA, privacy). Only 4.9 percent of the respondents expressed a difficulty understanding the forms. When asked what they did about this problem, nearly all (91.2 percent) indicated that they asked the provider/staff member to explain the information they did not understand (Table 1). Finally, the respondents were asked if they had any difficulty getting the health provider/staff member to understand their dental problem or injury. Only 6.9 percent of the respondents indicated that this was a problem. To address this problem, the largest percentage of these respondents asked a friend or family member to explain their condition to the provider/staff member (34.3 percent), followed by 23.0 percent who told the provider to ask questions about what they did not understand (Table 1).
We next wanted to examine if there were statistically significant variables associated with a respondent experiencing health literacy-related problems. Due to the relatively small number of respondents who experienced a problem with any one of the health literacy markers, all of the respondents who experienced a problem with any marker were combined into one group for further analysis. Specifically, we examined whether respondents were more likely to experience a problem when they first or only went to see a particular type of health professional (dentist, physician, or ED). We also examined the association between factors that we thought could have an important impact on the type of provider selected or the likelihood of experiencing a problem. These variables included demographic characteristics along with health care utilization and access measures.
The results of the logistic regression statistical analysis indicated that the model we tested had a pseudo-R squared of 10.9 percent. This can be interpreted as an indicator of how much the variables in the model contributed to reducing errors in predicting whether or not respondents had a problem understanding what they were told or the forms they had to complete or had difficulty getting the provider or staff member to understand their dental problem or injury. Our analysis indicated that there were no statistically significant differences in the odds of whether or not respondents had a problem with these health literacy markers between going to a physician or ED or visiting a dentist. However, our analysis also indicated that two of the other variables in the model were statistically significant: race/ethnicity and gender. Respondents who were male had 106 percent higher odds of experiencing a problem with one of the markers than those who were female (OR 2.06; 95 percent CI 1.04–4.08; p<0.05). In addition, being Hispanic increased the odds of respondents experiencing a problem by 534 percent over white respondents (OR 6.34; 95 percent CI 2.22–18.09; p<0.01).
The importance of health literacy in addressing health disparities has been gaining increasing recognition,20 and improving dental health literacy, especially among racial/ethnic minority groups, is an important goal.16,17 Our hypotheses that low-income and ethnic/racial minorities would experience a greater prevalence of health literacy-related problems and that patients receiving care at EDs and physician offices would experience more health literacy-related problems than patients receiving care from dentists were not generally supported. There were no differences between treatment sites in respondent reports of problems with the health literacy-related markers, nor were their differences associated with respondent income. Hispanic respondents, however, did voice more health literacy-related concerns than did white respondents.
It was expected that dental problem-related patient interactions with dentists would exhibit fewer health literacy-related difficulties than interactions involving physicians or EDs. That this was not the case may reflect the fact that respondents had higher expectations for their interactions with dentists than they did for those with physicians or EDs.21 Although other reports have found associations between problems with patient-provider communication and respondent age,22,23 race,24 and income,23,24 we found no such association. The reason for these differences was not readily apparent. Our finding that Hispanics experienced greater difficulty with health literacy was consistent with other reports.23,24 Respondent strategies for dealing with difficulties in communication were generally constructive. The primary exception was those individuals who had difficulty understanding what they were being told and who chose to act as if there was no problem.
This study did not attempt to establish levels of health literacy nor factors that might be associated with those levels. Rather, we attempted to quantify the percentage of respondents who reported experiencing health literacy-related problems. The most unexpected finding was the relatively small percentage of respondents, 10 percent or fewer, who reported experiencing problems. This contrasts with national estimates that as many as one-third of Americans lack health literacy.2,3 Similarly, the most recent national study of health literacy found that approximately 80 million adults experience difficulties with health literacy.24,25 These two studies are consistent with a recent report that found approximately 29 percent of adult patients seeking dental care suffered low levels of oral health literacy.15
The difference in our findings and general health literacy findings may be due at least in part to the more direct and less complicated nature of patent-provider interactions surrounding dental problems.13 Dental encounters generally involve less complicated diagnoses, diagnostic procedures, treatment, and subsequent requirements for patient participation in the management of their conditions (following post-treatment directions, taking medications, dietary restrictions, etc.) than do medical encounters. In addition, differences may be due to the manner in which health literacy was assessed. For example, the study that examined oral health literacy among adult patients seeking dental care utilized a word-recognition test that required subjects to read aloud a list of words representing dental conditions, their prevention, and treatment, arranged in increasing difficulty, with points assigned for each word correctly pronounced.15 This approach has the advantage of being useful in identifying patients in advance who may experience problems in communication resulting from poor health literacy, but does not establish that they actually experienced a problem as a result of poor oral health literacy. The approach used in our study, however, was more direct, asking patients specifically about health literacy-related communication problems they actually experienced. Therefore, it is possible that respondents might have experienced a breakdown in communication of which they were not aware. That is, they may have thought they understood what was being said but in reality did not.
All cross-sectional surveys share the limitation that they ask questions about the past that depend on the respondent’s memory. This challenge was mitigated in our study because dental problems/injuries are not likely to be quickly forgotten.26 Furthermore, the time frame was limited to the most recent twelve-month period. One potential drawback to telephone surveys is the amount of noncoverage of the target population. In our case, there is a risk that a disproportionate number of low-income households will not have a listed telephone. However, the 2000 U.S. census reported that only 1.6 percent of Maryland households did not have telephone service in 1999, and only 7.9 percent of households with incomes below the Federal Poverty Level did not have telephone service.27 More recent national data from the National Health Interview Survey (January–December 2007) indicate that nationally the percentage of households with only wireless service has increased to 14.7 percent.28 Nevertheless, research has shown this not to have been a significant source of bias. For example, estimates from the 2004 and 2005 National Health Interview Survey of the use of health care services for adults with landline telephones showed relatively small differences from those for all adults.29 Differences between face-to-face surveys and telephone surveys have generally found few statistically significant differences and even fewer differences of practical significance.30 Although noncoverage bias has not resulted in discontinuing the use of general population telephone surveys in helping guide public health policy and program decisions,29 the ever-increasing use of cell phones will pose an increasing problem for public health data collection.31 Finally, although the findings are representative of Maryland residents with recent dental problems/injuries who visited an ED, physician, or dentist, they should be generalized to other populations with caution.
Doctor-patient communication plays an important role in the delivery of health care services. A recent initiative by the Agency for Healthcare Research and Quality encourages patients to become more involved with their health care by asking appropriate questions of their health care providers.32 At the same time, the Centers for Disease Control and Prevention has developed a free Health Literacy for Public Health Professionals online training program designed to educate health professionals about limited health literacy.33 Communication problems may contribute to racial or ethnic disparities evident in the delivery of health care services and concomitant disparities in health.34 In our study, only a few respondents expressed health literacy-related problems. Further, the covariates we examined that we thought might influence whether respondents experienced health literacy-related communication problems resulted in only a modest pseudo-R squared, suggesting that the variables in the model added only modestly to our understanding of the factors that influence communication difficulties associated with health literacy. There is clearly opportunity to investigate what other factors play an important role. Perhaps most importantly, future studies are needed to confirm the connection between health literacy and communication difficulties in actual patient encounters and subsequent health outcomes.
Grant Number 5R01DE017685-02 from the National Institute of Dental and Craniofacial Research supported this research.
Dr. Leonard A. Cohen, Department of Health Promotion and Policy, University of Maryland Dental School.
Dr. Arthur J. Bonito, Health Services and Social Policy Research Division, Research Triangle Institute.
Ms Celia Eicheldinger, Health Services and Social Policy Research Division, Research Triangle Institute.
Dr. Richard J. Manski, Department of Health Promotion and Policy, University of Maryland Dental School;
Dr. Robert R. Edwards, Department of Psychiatry, Johns Hopkins University School of Medicine.
Dr. Niharika Khanna, Department of Family and Community Medicine and the Greenebaum Cancer Center, University of Maryland School of Medicine.