Health Literacy and Communication Quality in Health Care Organizations

Abstract

The relationship between limited health literacy and poor health may be due, in part, to poor communication quality within health care delivery organizations. We explored the relationship between health literacy status and receiving patient-centered communication in clinics and hospitals serving communication-vulnerable patient populations. Thirteen health care organizations nationwide distributed a survey to 5929 patients. All patients completed seven items assessing patient-centered communication. One third also completed three items assessing health literacy. The majority of patients had self-reported health literacy challenges, reporting problems learning about their medical condition because of difficulty understanding written information (53%), a lack of confidence in completing medical forms by themselves (61%), and needing someone to help them read hospital/clinic materials (57%). Logistic regression models showed that, after adjustment for patient demographic characteristics and health care organization type, patients with limited health literacy were 28% to 79% less likely than those with adequate health literacy to report their health care organization “always” provides patient-centered communication across seven communication items. Using a scaled composite of these items, limited health literacy remained associated with lower reported communication quality. These results suggest that improving communication quality in health care organizations might help to address the challenges facing patients with limited health literacy. They also highlight that efforts to address the needs of patients with limited health literacy should be sensitive to the range of communication challenges confronting these patients and their caregivers.

Health literacy refers to a person's ability to understand and act on health information (IOM, 2004). A growing body of evidence demonstrates that compared to individuals with adequate health literacy skills, those with limited health literacy are more likely to misunderstand health information (Friedman, Hoffman-Goetz, & Arocha, 2006); face difficulty following medical instructions (Davis et al., 2006); inappropriately or infrequently use health care services (Gazmararian et al., 1999; Sudore et al., 2006); have worse physical and mental health (Wolf, Gazmararian, & Baker, 2005); experience higher rates of hospitalization (Baker et al., 2002); and have a shorter life expectancy (Baker et al., 2007). Efforts to overcome limited health literacy have included developing plain language, patient-friendly education materials and navigation aids (Stableford & Mettger, 2007); educating healthcare professionals about health literacy issues (Riley, Cloonan, & Rogan, 2008); redesigning patient informed consent forms (Lorenzen, Melby, & Earles, 2008); and using established communication methods such as the “teach back” techniques when communicating with patients (Villaire & Mayer, 2007). While experts agree that implementing a range of system-wide strategies may be the most effective means of overcoming limited health literacy (Murphy-Knoll, 2007; O'Leary, Davis, & Cordell, 2007), system-wide change to address limited health literacy has been difficult to stimulate and slow to develop in most health care organizations (Stableford & Mettger, 2007).

Patient-Centered Communication

Communication is one of the foundations of health care. Every health care interaction depends on effective communication, from making an appointment and registering for a visit to describing symptoms, discussing risks and benefits of treatments, and understanding care instructions. Good communication is linked to improved patient satisfaction, adherence to medical recommendations, and health outcomes. Today, many health care professionals believe that communication is more effective when it is patient-centered, or responsive to each patient's individual needs, values, and preferences (Stewart, 2001). While patient-centered communication is often described only in terms of individual clinician-patient interactions (Mead & Bower, 2002), health care organizations have a responsibility to facilitate patient-centered communication in all forms, including patient education materials, organizational signage, patient forms, and training providers to be better communicators (Epstein & Street, 2007).

It is especially important that health care organizations use patient-centered strategies to reach populations that may not receive or understand oral and written health information that is often presented in English using complex medical jargon. These include patients with limited or no English proficiency, limited health literacy, or cultural backgrounds that are not well understood by hospital or health system staff. In addition, health and health care disparities are created when miscommunication disproportionately affects certain patient populations (IOM, 2001). As a result, understanding and improving communication may be a key to addressing disparities (IOM, 2002), which is an important national health policy goal (IOM, 2001).

Ethical Force® Program

The Ethical Force® program is a multistakeholder collaborative that includes national experts in health literacy, cultural competency, medical ethics, and health care organizational policy; and was created by the American Medical Association to develop valid and reliable measures of the ethical environment in health care organizations. The “ethical environment” of a health care organization comprises all ethics-related facets of the organization, including the presence or absence of specific organizational values, infrastructure and other resources that make ethical practices more or less likely. The Ethical Force® Oversight Body (EFOB) selects specific aspects of the ethical environment––such as patient-centered communication (Wynia & Matiasek, 2006), privacy (Wynia, Coughlin, Alpert, Cummins, & Emanuel, 2001) or insurance coverage decisions (Wynia et al., 2004)—and uses a consensus process to develop performance expectations in that area for health care organizations. It then aims to create valid and useful tools to measure, monitor and improve an organization's ethical environment. This work is related to efforts to measure and improve health care organizational culture, yet available instruments measuring organizational culture in health care have been sparse (Gershon, Stone, Bakken, & Larson, 2004), and limited in their scope, ease of use, and measurement properties (Casida, 2008; Scott, Mannion, Davies, & Marshall, 2003).

To help prompt health care organizational change in the area of addressing limited health literacy and other communication challenges, the Ethical Force® program developed and validated a set of organizational assessment tools for patient-centered communication: The Organizational Communication Climate Assessment Toolkit (OCCAT) (Wynia, Johnson, McCoy, Griffin, & Osborn,). The OCCAT tools gather data on a health care organization's performance according to the perspectives of patients, clinicians, non-clinical staff, and leaders, and was field tested in 2006–2007. For this study, we use data gathered from patients during validation studies to explore the complex relationship between patient health literacy status and patient perceptions of experiencing patient-centered communication during clinical encounters.

Method

An assessment of health care organizational patient-centeredness was developed through a multistakeholder consensus process. This process has been previously described (Levine et al., 2007; Wynia et al., 2001) and a detailed description of the assessment is available online at www.EthicalForce.org. The assessment was validated in a field test with 13 widely-varying health care organizations, including 6 hospitals and 7 clinics nationwide (Wynia, Johnson, McCoy, Griffin, & Osborn, 2010) (Table 1). Sites were selected in a competitive process by a thirteen member Expert Advisory Panel on Patient-Centered Communication (member list available at: www.EthicalForce.org), which aimed to represent all regions of the country and a broad array of patient populations. Table 2 shows the demographic characteristics of the patients at the field test sites. Site-specific assessment results are not presented in this report because sites were promised anonymity for the purpose of testing the instruments.

Table 1. Participating clinics and hospitals in the field testing of the communication assessment tools

Organization type	Clinics/Hospitals
Large/Urban	• Children's Hospital & Research Center, Oakland, California• George Washington University Hospital, Washington, D.C.• Hennepin County Medical Center, Minneapolis, Minnesota• University of Chicago Hospitals, Chicago, Illinois• University of Mississippi Medical Center, Jackson, Mississippi
Small/Rural	• Geisinger Medical Group, Mount Pocono, Pennsylvania• Louisville Oncology, Louisville, Kentucky• Sierra Kings District Hospital, Reedley, California• Windham Hospital, Willimantic, Connecticut
FQHC	• Community Health Center, several cities across Connecticut• Family HealthCare, Visalia, California• Golden Valley Health Center, Merced, California• Open Door Family Medical Center, Ossining, New York

Table 2. Characteristics of all patients and those who also completed the health literacy items

	All patients	Completed HL items
	N = 5928	N = 2116
Characteristics	n (%)	n (%)
Age
18–24 years	781 (13)	275 (13)
25–44 years	2569 (43)	913 (42)
45–64 years	1642 (28)	650 (30)
65+ years	531 (9)	195 (9)
N/A ^a	405 (7)	133 (6)
Gender
Male	1323 (22)	502 (23)
Female	3915 (66)	1420 (66)
N/A	690 (12)	244 (11)
Race/Ethnicity
Caucasian	1533 (26)	646 (30)
African American	747 (12)	449 (21)
Hispanic/Latino	2383 (40)	820 (38)
Other	683 (12)	239 (11)
N/A	464 (8)	12 (.6)
Language Spoken
English	3609 (61)	1553 (72)
Spanish	1740 (29)	571 (26)
Other	210 (5)	42 (2)
N/A	369 (6)	0 (0)
Frequency of Visits
1 time	1040 (17)	349 (16)
2–5 times	2726 (46)	1011 (47)
6–10 times	923 (16)	361 (17)
>10 times	700 (12)	254 (12)
N/A	539 (9)	191 (9)
Organization Type
Large/urban	2022 (34)	886 (41)
Small/rural	853 (14)	300 (14)
FQHC ^b	3053 (52)	980 (45)

Survey Distribution Procedures

Field testing was conducted by an interdisciplinary team, led by researchers at the Institute for Ethics at the American Medical Association (AMA), and took place between November 2006 and August 2007. Patient surveys were available in paper or web-based format, and were available in eleven languages (English, Spanish, Haitian Creole, Hmong, Bosnian, Chinese, French, Polish, Portuguese, Somali, and Vietnamese). An automated telephone version was also created and available in four languages (English, Spanish, Haitian Creole, Hmong). All sites used convenience samples for data collection. For patient surveys, sites either distributed surveys to patients directly in outpatient clinics or prior to discharge from the hospital or mailed the surveys to patients' homes following an office visit or inpatient admission. All procedures were reviewed and approved by the Western IRB, Olympia, WA, which is an independent IRB contracted by the AMA to review research protocols. Several field test sites also underwent local IRB review, depending on the specifics of additional local assessment protocols.

Measures

All patient surveys included questions to gather demographic information as well as items to assess patients' perspectives on the patient-centeredness of the health care organizations in which they received their care. One third of all patients surveyed were randomly assigned to complete questions about health literacy, another third of patients completed questions about trust in the health care system, and the final third responded to a question about self-reported health status. In this paper, we present data on patients who responded to the health literacy items.

Demographics

Patient demographic information collected included age, gender, race/ethnicity (Caucasian, African American, Hispanic/Latino ethnicity, or other ethnicity), and language spoken (English, Spanish, or other language). To assess the degree of patient exposure to the hospital or clinic, an item asked, “How many times have you visited the clinic (or hospital) in the last six months?” Response options were 1 = one time, 2 = two-five times, 3 = six-ten times, and 4 = more than ten times.

Health Literacy

We used a slightly modified version of the three-item health literacy screening questions of Chew et al. (2008). These questions ask patients to report problems due to reading, understanding, and filling out forms, not due to poor vision: (a) “How often do you have problems learning about your medical condition because of difficulty understanding written information?”; (b) “How confident are you filling out medical forms by yourself?”; (c) “How often do you have someone like a family member, friend, hospital or clinic worker or caregiver, help you read hospital or clinic materials?” These screening questions have been validated against widely used measures of health literacy (Chew, Bradley, & Boyko, 2004; Chew et al., 2008) across a variety of settings (Sarkar et al., 2008; Wallace et al., 2007; Wallace, Rogers, Roskos, Holiday, & Weiss, 2006). Screening questions are on a 5-point Likert-type scale ranging from 1 = always have problems to 5 = never have problems. In keeping with prior studies, we coded patients who reported 1–4 = ever having problems as “limited health literacy” (corresponding to inadequate, low or marginal literacy on the REALM or STOFHLA), and those scoring 5 = never having problems as having “adequate health literacy” (Sarkar et al., 2008). We considered “don't know” responses as missing. Each question was analyzed separately, as the developers of these items recommend (Chew et al., 2004; Chew et al., 2008).

Patient-Centered Communication

To assess perception of a health care organization's patient-centered communication, we used items from the Ethical Force® program's Organizational Communication Climate Assessment Toolkit (OCCAT) patient survey. Seven items were extracted from a larger eighteen-item set intended to measure multiple performance domains: organizational commitment, data collection, community engagement, individual engagement, addressing health literacy, meeting language needs, and cross-cultural communication. Domain subscales have demonstrated acceptable statistical reliability and construct validity in two rounds of field testing (Wynia, Johnson, McCoy, Griffin, & Osborn, 2010).

For the purposes of this study, only items specific to overt, verbal communication in a clinical encounter were used. Five items ask: “Did doctors and nurses…(a) explain things in a way you could understand, (b) listen to you, (c) make sure you understood their instructions, (d) show respect for what you had to say, (e) ask if you have any questions?” The remaining two items ask: (f) “Is it easy to ask questions at the clinic/hospital?” and (g) “Does the clinic/hospital communicate well with patients?” Response options are 1 = never, 2 = sometimes, 3 = always. These items were analyzed with a binary outcome using the “top box” method (0 = never/sometimes, 1 = always) that addresses the uneven distributions common to patient satisfaction surveys (Velanovich, 2007). An overall composite score was also created with the sum of all seven items. The composite score ranged from 0 to 7 and showed high internal consistency reliability, α = 0.86.

Statistical Analyses

Demographic data were analyzed using means, frequencies, and cross tabulations to calculate descriptive statistics. The main study hypothesis was tested using a series of binary logistic regression analyses. Each dichotomized patient-centered communication item served as a dependent variable. The models were built in three steps. In step one, gender, age, race/ethnicity, primary language spoken, and frequency of visiting the hospital/clinic were entered. In step two, health care organization type (academic medical center or large urban hospital, smaller/rural hospital or health center, federally qualified health center) was entered, with academic medical center as the referent group. In the final step, health literacy was entered, with adequate health literacy as the referent group for calculating odds ratios. This analytic strategy tested whether patient perceptions of a health care organization's patient-centered communication differed for patients with limited health literacy relative to patients with adequate health literacy after controlling for gender, age, race/ethnicity, primary language spoken, frequency of hospital/clinic visits, and type of health care organization. Lastly, three t-tests were performed to examine health literacy differences on the scaled composite of patient-centered communication. Three multiple linear regression models also tested whether health literacy was related to the scaled composite after controlling for demographic factors and health care organization type. All analyses were done using SPSS version 17.0.

Results

Participant Characteristics

Thirteen participating health care organizations (Table 1) collected surveys from 5,929 patients, of whom a random subsample of 2,116 completed items about health literacy (Table 2). Because each site used convenience sampling methods, we cannot calculate precise response rates. Most patients were between 25 and 44 years of age, were female, and were members of racial/ethnic-minorities. Approximately one third of patients surveyed had limited English proficiency (LEP) and most were receiving care from a federally qualified health center (FQHC) or large/urban medical center. Nearly half reported visiting this facility two to five times in the last 6 months. Most patients had health literacy challenges, reporting problems learning about their medical condition because of difficulty understanding written information (53%), a lack of confidence completing medical forms by themselves (61%), and needing someone to help them read hospital/clinic materials (57%).

Relationship Between Limited Health Literacy and Patient-Centered Communication

In logistic regression models, adjusting for demographic characteristics and health care organization type, patients with limited health literacy on each of the three items had between 28% and 79% lower odds of reporting their health care organization “always” provides patient-centered communication on each of the seven communication items compared to those with adequate health literacy (Table 3). For example, patients who reported difficulty learning because of trouble understanding written information had 37% lower odds of reporting that doctors and nurses listened to them (95% CI: 0.27–0.52), 32% lower odds of reporting that doctors and nurses made sure they understood their instructions (95% CI: 0.23–0.45), and 28% lower odds of reporting that it is easy to ask questions at the hospital or clinic (95% CI: 0.21–0.38), compared to patients with adequate health literacy on this measure.

Table 3. Logistic regression models predicting perceptions of patient-centered communication from health literacy status, organization type, and demographics (the latter is not shown ^a )

	Health literacy items
	1. How often do you have problems learning about your medical condition because of difficulty understanding written information?	2. How confident are you filling out medical forms by yourself?	3. How often do you have someone like a family member, friend, hospital or clinic worker or caregiver, help you read hospital or clinic materials?
				95% CI for EXP(B)			95% CI for EXP(B)			95% CI for EXP(B)
PCC items	Predictor	df	OR	Lower	Upper	OR	Lower	Upper	OR	Lower	Upper
1. Did doctors and nurses explain things in a way you could understand?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.94	0.56	1.58	0.94	0.57	1.55	0.92	0.56	1.53
	FQHC	1	0.93	0.65	1.33	0.86	0.60	1.23	0.91	0.64	1.30
	Limited HL (Adequate)	1	0.28	0.20	0.40	0.40	0.29	0.57	0.40	0.29	0.54
	Constant	1	20.46			19.20			14.53
2. Did doctors and nurses listen to you?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.88	0.53	1.48	0.95	0.57	1.58	0.96	0.57	1.60
	FQHC	1	0.95	0.66	1.37	0.91	0.63	1.31	0.99	0.69	1.43
	Limited HL (Adequate)	1	0.37	0.27	0.52	0.62	0.45	0.85	0.49	0.35	0.67
	Constant	1	19.49			15.29			13.33
3. Did doctors and nurses make sure you understood their instructions?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.87	0.52	1.45	0.93	0.57	1.52	0.94	0.57	1.55
	FQHC	1	1.03	0.71	1.47	1.00	0.70	1.43	1.05	0.73	1.51
	Limited HL (Adequate)	1	0.32	0.23	0.45	0.57	0.41	0.78	0.38	0.28	0.52
	Constant	1	12.84			9.88			9.52
4. Did doctors and nurses show respect for what you had to say?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.86	0.50	1.49	0.87	0.51	1.48	0.89	0.52	1.52
	FQHC	1	0.93	0.64	1.35	0.86	0.59	1.25	0.89	0.61	1.30
	Limited HL (Adequate)	1	0.36	0.26	0.51	0.47	0.33	0.66	0.43	0.31	0.60
	Constant	1	15.02			14.70			11.35
5. Did doctors and nurses ask if you had any questions?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.64	0.40	1.02	0.69	0.44	1.08	0.67	0.42	1.06
	FQHC	1	0.68	0.49	0.95	0.68	0.49	0.95	0.69	0.50	0.97
	Limited HL (Adequate)	1	0.47	0.35	0.62	0.79	0.60	1.05	0.53	0.40	0.71
	Constant	1	10.80			8.49			8.38
6. Is it easy to ask questions at the hospital/clinic?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.75	0.49	1.17	0.84	0.55	1.28	0.81	0.52	1.24
	FQHC	1	0.90	0.66	1.24	0.89	0.65	1.22	0.95	0.69	1.30
	Limited HL (Adequate)	1	0.28	0.21	0.38	0.49	0.37	0.65	0.39	0.30	0.52
	Constant	1	10.56			8.23			6.87
7. Does the hospital/clinic communicate well with patients?	Organization (Large/Urban) ^b	2
	Small/Rural	1	0.85	0.55	1.31	0.95	0.62	1.45	0.95	0.62	1.45
	FQHC	1	0.80	0.59	1.09	0.82	0.61	1.11	0.85	0.63	1.15
	Limited HL (Adequate)	1	0.32	0.25	0.43	0.54	0.41	0.70	0.54	0.41	0.70
	Constant	1	9.49			7.22			5.89

The effects of limited health literacy on patients' overall perception of communication quality were explored using the scaled composite of all seven patient-centered communication items. Using t-tests and multivariable models (not shown in a table), we found that patients with limited health literacy provide consistently lower ratings of patient-centered communication quality. In bivariate analyses examining each health literacy item, patients with problems understanding written information, scored lower (M = 5.2, SD = 2.3) than patients with adequate health literacy (M = 6.2, SD = 1.5) on the patient-centered communication scale (t = 11.43, p < .001). Patients who lack confidence completing medical forms, also scored lower (M = 5.4, SD = 2.2) than patients with adequate health literacy (M = 6.1, SD = 1.7) on the patient-centered communication scale (t = 7.00, p < .001). Finally, patients who need someone to help read hospital or clinic materials, also scored slightly lower (M = 5.6, SD = 2.1) than patients with adequate health literacy (M = 5.8, SD = 1.9) on the patient-centered communication scale (t = 2.93, p < .003).

In a series of multiple regression models that controlled for patient demographic factors, including language, and health care organization type, limited health literacy, defined as having problems learning about a medical condition because of difficulty understanding written information or as lacking confidence in completing medical forms, was significantly related to lower patient-centered communication scores (β = −0.25, p < .001 and β = −0.14, p < .001, respectively), though limited health literacy, defined as needing someone to help read hospital or clinic materials, was not (β = −0.04, ns).

Discussion

Health care organizations and professionals have a responsibility to present health information in ways that patients, including those with limited health literacy, can use and understand (Stableford & Mettger, 2007). We studied a diverse sample of patients recruited from 13 different health care organizations, with a high burden of limited health literacy. We found that patients with limited health literacy were less likely to report “always” receiving patient-centered communication at their health care organization compared to those with adequate health literacy. This relationship between limited health literacy and lower perceived communication quality was maintained even after adjusting for patient demographics, including race, age and language, and health care organization type.

Approximately one-third of patients in our study had low English Proficiency (LEP), and studies suggest LEP is associated with higher rates of low health literacy (Osborn & Wynia, October 25–29, 2008). However, even after adjusting for LEP, health literacy was an independent predictor of perceptions of an organization's patient-centered communication quality in this study (i.e., health literacy is associated with the outcome above and beyond LEP). While not the intent of this paper, future research should explore potential interactions between language barriers and health literacy on patient perceptions of communication quality and patient health outcomes.

While the relationship between limited health literacy and perceptions of communication quality in health care organizations is undoubtedly complex, our findings are the first to use direct patient reports to suggest that organizations might be able to mitigate the effects of limited health literacy by improving patient-centered communication. At the same time, our findings also suggest the possibility that unequal quality of patient-centered communication across health care organizations might be contributing to the widespread problem of poor health outcomes among limited health literacy patients (IOM, 2004). For both reasons, improving patient-centered communication in health care organizations could be an important strategy for addressing the problems of limited health literacy.

There are several reasonable pathways by which limited health literacy could negatively affect patient experiences of communication quality in a health care organization. For example, patients with limited health literacy might also have a limited sense of self-efficacy in regard to communicating with health care professionals, which might negatively affect their perceptions of interactions. Similarly, one can consider this issue from the point of view of the organization and hypothesize that when the communication climate is poor, more patients will report problems related to limited health literacy, such as difficulty understanding instructions and needing help with documents. Our cross-sectional study cannot disentangle these varying pathways, which deserve consideration in future work.

To address the problems related to limited health literacy, experts have suggested that individual practitioners can improve communication with all patients by employing the following techniques: slowing down; using plain, nonmedical language; limiting the amount of information and repeating it; using the teach back technique; and creating a shame free environment (Kountz, 2009). In their reports on health literacy, the American Medical Association (AMA), the National Work Group on Literacy and Health, and the Institute of Medicine each called for greater efforts to educate physicians and other health professionals about issues related to health literacy, including techniques to communicate more clearly (Ad Hoc Committee on Health Literacy for the Council of Scientific Affairs, American Medical Association, 1999; IOM, 2004; National Work Group on Literacy and Health, 1998). However, recent evidence suggests that even when providers know about health literacy and the need for enhanced communication techniques, they underutilize these strategies (Turner et al., 2009). Many effective health communication strategies have been studied by physicians, nurses, and pharmacists, but remain unincorporated into routine clinical practice (Schwartzberg, Cowett, VanGeest, & Wolf, 2007). This study examined patients' views on organizational more than individual communication, yet our results corroborate these earlier findings, since organizations scored lower on our communication scale when patients reported that staff there did not communicate clearly and ensure patient understanding. The fact that patients with low literacy had 28% to 79% lower odds of reporting good quality communication suggests that there remains much that both organizations and individual practitioners can do to help meet the communication needs of these patients.

Study strengths include a large, diverse population and detailed assessments of patient experiences of communication within various types of healthcare organizations. We also used a well validated instrument to assess patient challenges related to health literacy. Nevertheless, the issues under study are complicated and this research has several limitations. First, because this is a cross-sectional analysis, we cannot conclude that limited health literacy causes unfavorable perceptions of patient-centered communication; although reverse causation is improbable, there may be underlying factors affecting both. Second, while we did not measure health literacy with one of the widely used measures of the construct (e.g., Short Test of Functional Health Literacy and the Rapid Estimate of Adult Literacy in Medicine), the health literacy screening items we used have been validated against these instruments (Chew et al., 2004, 2008). Moreover, the degree to which these screening items correlate perfectly with other literacy measures may be less relevant in practice, insofar as these items have predictive validity and can identify subgroups at risk for clinically important adverse events. Third, we did not measure numeracy in this study, so we could not address the relative contribution of numeracy versus health literacy. Numeracy and health literacy have been shown to be highly correlated (Huizinga et al., 2008). Fourth, patient reports of health care quality, while important for several reasons, do not always correspond to quality using other performance measures (Hrisos et al., 2009), though sometimes they do correspond to important outcomes (Brousseau, Gorelick, Hoffmann, Flores, & Nattinger, 2009). Similarly, our assessment of patient-centered communication is from a self-reported measure that focuses on a limited subset of communication experiences. Finally, despite the large sample size for patients and staff, this was a cross-sectional survey of a non-random, small number of sites (13), which precludes assessments of causation and most analyses by health care organization type or to examine within-site effects. We hope future studies will be longitudinal and can include larger numbers of health care organizations, which might allow detection of performance variation over time, to assess cuasality, and also detect variations according to region, additional site demographics, specific policy environmental factors and so on.

Conclusion

This study demonstrates that patients with limited literacy are considerably less likely to report receiving patient-centered communication from the organizations where they seek care, across an array of communication quality indicators. At the same time, our findings provide hope that feasible interventions to promote effective communication in health care organizations can improve the experience of care and outcomes for patients with limited health literacy. Widely endorsed methods to address the needs of patients with limited health literacy (e.g., improved educational materials, clearer forms, staff training in clear communication techniques, teach back methods and reinforcement) should be sensitive to the range of communication challenges facing these patients and their caregivers.

This work was funded in part by the California Endowment, grants 20043158 and 20062217, the Commonwealth Fund, and the Connecticut Health Foundation. Dr. Osborn is supported by a Diversity Supplement Award (NIDDK P60 DK020593). The authors wish to extend special thanks to our program officer at the California Endowment, Ignatius Bau, for his support of this work. In addition, many researchers have been involved in the work of the Ethical Force program over the last 6 years. Special thanks for their work on this project are due to Jennifer Matiasek, Jeff Jarosch, Ololade Olakanmi, Joanne Schwartzberg, and Cynthia Hedges-Greising. Finally, we thank the Oversight Body of the Ethical Force program and the Expert Advisory Panel on Patient-Centered Communication (named at: www.EthicalForce.org) for their unstinting support.

Notes

Note: FQHC = federally qualified health center.

Note: ^a Missing.

^b FQHC = federally qualified health center.

Note: ^a Results from steps two and three are presented, and have been adjusted based on the demographics entered in step one: age, gender, race/ethnicity, primary language spoken, and frequency of hospital/clinic visits. Step one has been omitted from the above table for simplicity. PCC = patient-centered communication, CI = confidence interval, B = beta estimate, df = degrees of freedom, OR = odds ratio, FQHC = federally qualified health center.

^b Reference group used to calculate Odds ratio is in parentheses.