GAZMARARIAN, JULIE A. MPH, PhD; PARKER, RUTH M. MD; BAKER, DAVID W. MD, MPH
Approximately 15% of the general adult population, 16–44 years old, are estimated to have a low level of reading ability.1 The impact of an individual's reading ability on his or her health status and use of health services is significant. Studies have shown that individuals with low health literacy have higher rates of self-reported poor health,2 hospitalizations,2 and medication errors.3
Of all the literacy skills needed in health care, reading comprehension is the most important.4 Studies consistently have found that there is often a large discrepancy between the average patient's reading comprehension and the ability levels needed to read patient education materials.4–10 A review of all the patient education pamphlets developed by ACOG found that almost all of the pamphlets describing contraceptive choices, sexually transmitted diseases, vaginitis, chlamydia, genital warts, and pregnancy are written at a grade level (11th grade or higher) that exceeds the literacy level of many of the women who would benefit greatly from the information.7 Health care providers should be aware of critical factors, such as their patients' reading skills, that may influence the effectiveness of their educational efforts. This need is especially applicable to health providers who work in public clinics. Although patients often are asked about the number of years of school they completed, research has shown that years of schooling is not a reliable predictor of reading level.8,9,11
An individual's reading skills can particularly affect family planning decisions. Men and women using contraception are more than likely to encounter written instructions, and if obtaining information through a health care provider, be given written brochures. Moreover, packaging of contraceptives regulated by the US Food and Drug Administration is required to include a patient insert describing how to correctly use the contraception method and warning the user of possible side effects and adverse reactions from the method. For the patient package inserts to be effective in communicating this information, they must be read, understood, and remembered by patients.
Individuals with low reading skills may have difficulty understanding patient package inserts. For example, one study found that three out of four patients using two local family planning clinics in the Midwest needed assistance in comprehending information in the patient package insert for oral contraceptives.12 Other studies have found that the patient package inserts for condoms13,14 and oral contraceptives14,15 are at a reading level higher (at least 10th grade) than that of many consumers who need the information.
To date, no research has explicitly examined the relationship between an individual's reading level and use of contraception services.5 A recent review article examining the relationship between literacy and contraception hypothesizes that functional health literacy influences contraceptive knowledge, attitudes, and behaviors.5 The purpose of this study was to examine the relationship between health literacy and contraceptive use and family planning knowledge among women enrolled in TennCare in Memphis, Tennessee.
Materials and Methods
Women from 13 through 45 years of age enrolled in Prudential HealthCare Community Plan as of March 1, 1996, were eligible to participate in the study. From the list of 2197 age-eligible women, we randomly selected individuals until we completed 500 surveys. Extensive efforts were made to locate women with missing phone numbers. If a phone number was missing or incorrect from the enrollment records, the project coordinator and interviewers used several other sources to obtain a telephone number: directory assistance, the Prudential appointment scheduling system for women accessing the system, the cross-reference system (which provides a phone number for an address listing), and visits to the addresses listed for the women. From the original sample of age-eligible women (n = 2197), we located 1136 women. This response rate is comparable to another study in a Medicaid population in which valid phone numbers were available for only approximately 50% of Medicaid beneficiaries.16 After excluding women who refused to participate (n = 204) and those who did not meet age and enrollment criteria (n = 216), 716 women were invited to participate in the survey, and 506 interviews were completed between March 1996 and April 1997. We excluded women less than 18 years of age (n = 95) because it is likely that their family planning practices are different from those of adults.
Trained interviewers (eight were from the TennCare community) contacted eligible women by telephone to schedule an interview in the respondent's home or another convenient location. Respondents who completed the 1-hour survey received an incentive ($25.00 grocery store certificate). An informed consent was administered as part of the questionnaire and confidentiality was assured to all participants. The study protocol was approved by the Prudential Center for Health-Care Research Institutional Review Board.
The measures of health literacy, maternal characteristics, and family planning knowledge and practices were obtained through a 1-hour in-person survey conducted in the respondent's home. Health literacy was measured by an abbreviated version of the Test of Functional Health Literacy of Adults.17 Reading skills were assessed by responses to one passage of a Medicaid Rights and Responsibility form written at a 10th grade reading level. This passage was a 20-item test that used a modified Cloze procedure, ie, every fifth to seventh word in the passage was omitted and four multiple-choice options were provided. The readability levels of this passage on the Gunning-Fog index is grade 10.4.18 This passage had a correlation of 0.93 with the full Test of Functional Health Literacy of Adults in previous studies,17 and the Cronbach α for the 20-item passage was .786. Women who answered fewer than 80% of the questions correctly (16 out of 20 questions) were classified as having low reading skills.
Maternal characteristics that were examined included age, race, marital status, education, employment, and poverty. Poverty status was determined by household income and family size. Respondents were asked to select one of several income categories that best represented their total household income during the last year. Poverty status was determined by Census guidelines for 1996, which provided estimated poverty thresholds for various family sizes.19 For example, a respondent living in a four-person household earning less than $16,036 would be considered living below the poverty level. The midpoint value for each of the income categories was assigned. The total number of persons living in the household and assigned income level for respondents were compared with the Census poverty thresholds.20 Women were classified as poor (less than 100% of poverty level), near poor (100%–199% of poverty level), or nonpoor (more than 200% of poverty level).
Measures of family planning knowledge that were examined included interest in learning more about birth control methods and awareness of when a woman is most likely to become pregnant during her monthly cycle. Whether the respondents would like to know more about birth control was assessed by asking if they thought that they knew a lot about the different methods of birth control and how effective they are, or if they would like to know more. Women's knowledge about the fertility cycle was determined by asking them when during the monthly menstrual cycle is the average woman most likely to become pregnant. To give respondents the absolute amount of chance to have a correct answer, we coded women who responded “right before,” “1 week after” or “2 weeks after” as answering correctly, and women who responded as “during period” or “no difference” as having an incorrect answer.
Family planning practices that were examined included pregnancy intendedness and current use and methods of birth control. Pregnancy intendedness was determined by the question: “Thinking back to just before you were pregnant, how did you feel about becoming pregnant?” (referring to most recent/current pregnancy). Intendedness was categorized as unwanted (did not want pregnancy then or at any time in the future), mistimed (wanted pregnancy later), or intended (wanted a pregnancy at that time). Women who had never been pregnant (n = 49) were excluded from any analyses of this variable.
Since the majority of respondents were either currently or recently pregnant, we asked, “During the month before you became pregnant, were you or your partner using any kind of birth control?” Women who had never been pregnant were asked, “Are you currently using any form of birth control?” Use of birth control the month before they became pregnant the most recent time (for currently and recently pregnant women), or current use of birth control (for never pregnant women) was determined and coded as either “yes” or “no.” All women who indicated that they had had voluntary sexual intercourse were asked, “Have you or any of your partners ever used any of the following types of birth control” followed by a list of options from which they could select as many that applied. Women who stated they had never had voluntary sexual intercourse (n = 6) were excluded from any analyses of this variable.
Frequencies were compared between reading skills (low compared with good) and maternal characteristics, family planning knowledge, and practices. χ2 analyses were conducted to determine significant differences between reading skill group and characteristics of interest.21 Because of the strong correlation between years of school and literacy and because literacy is a better marker of educational attainment than years of school completed, years of school completed was not used in the multivariate analysis with literacy. Multiple logistic regression analysis was conducted to determine the relationship between reading skill and family planning knowledge and practices, adjusting for variables that were statistically significant between reading skill groups. All analyses were conducted using the Statistical Analysis Software package (SAS, Cary, NC).22
Overall, 9.6% of the respondents were classified as having low reading skills (less than 80% correct) (Table 1). The majority of TennCare enrolled women who completed the in-person interview were at least 25 years of age, black, single, high school educated, employed (full or part-time), and poor (Table 2). Only education was related to reading skill level with 15.4% of women with a high school education or less having low reading skills compared with 3.5% of women with more than a high school education. Blacks were also more likely than whites to have low reading ability (11.1% compared with 4.2%), although the difference was not statistically significant (P = .09).
Pregnancy intendedness and current use of contraception did not vary by reading levels of respondents (data not shown). However, type of birth control ever used in the past did vary according to reading ability. Women who had ever used an intrauterine device, douching, rhythm, or levonorgestrel implants as methods of birth control had higher rates of low reading skills than did women who used other methods of birth control (Table 3). Women who indicated they would like to know more about birth control had lower reading levels (13.8% for those who wanted to know more compared with 6.8% for those who did not, P = .02). Women who did not know when they were more likely to become pregnant during their monthly cycle had lower reading skill levels (18.5% for those who did not know when they were more likely to become pregnant compared with 4.9% for those who did know, P = .001).
Logistic regression analysis indicated that women who had low reading skills were 2.2 times more likely to want to know more about birth control methods and 4.4 times more likely to not know when they were likely to become pregnant (Table 4). These relationships were significant even after controlling for other factors.
Almost 10% of the respondents in this study had low reading skills, which is about what we would expect in a low-income population for this age group.1 The rate of contraceptive use in this population was low, and women who had ever used an intrauterine device, douching, rhythm, or levonorgestrel had higher rates of low reading skills than did women who used other methods of birth control. Compared with women with good reading skills, women with low reading skills were more likely to want to know more about birth control methods and have incorrect knowledge about when they were most likely to get pregnant. Because this is one of the first studies to empirically examine the relationship between reading ability and family planning knowledge and practices, there is little research to use for comparison. However, several observations can be made.
A small but clinically important group of enrollees had poor reading ability. The prevalence in our study is similar to that reported by Williams et al18 for 18 to 30-year-old emergency department patients. The brief test used in this study was not designed to identify people with marginal reading ability, but previous studies suggest that another 10% would probably have milder, but clinically important, reading problems. Previous studies have shown that patients with marginal reading abilities have worse knowledge of chronic disease self-management than patients with adequate reading skills.23 Thus, it is likely that about one in five of the women in this study had limited reading ability that could affect their health care.
Enrollees who had ever used an intrauterine device for contraception had the highest rates of low reading skills. Given the requirement for informed consent prior to using this practice, this raises additional questions of whether women understand the consent form or whether providers are inserting more intrauterine devices in women who have low reading skills. Clearly, these questions warrant further research. Our data also indicated that enrollees who had used douching for contraception, an ineffective technique, were more likely than others to have low reading skills.
Although we could not determine a relationship to past unintended pregnancies, these factors could lead to unintended pregnancies in the future. Other research has shown that women with incorrect knowledge of their fertile period have a greater likelihood of contraceptive risk taking, defined as an inconsistency between women's desire to avoid unintended pregnancy and their failure to use readily available contraception.24
Results from this study support the hypothesis that literacy is associated with contraceptive knowledge.5 Although enrollees with low literacy had poorer knowledge, they were also more likely to express a desire to learn more about family planning. We do not know if this desire is related to lack of literacy, or if women with low skills are less likely to access birth control information. Either way, this is encouraging because it suggests that innovative educational strategies that do not rely exclusively on written materials are likely to be welcome.
There are several limitations with this study. First, we do not have information about correct use of contraceptives. It seems likely that women with low reading skills could be more likely to incorrectly use contraceptives. A recent study indicated that inconsistent use of oral contraceptives was more likely among minority women and women of lower socioeconomic status,25 factors also associated with reading skills. Second, we were not able to locate a large proportion of the population. However, once we could locate an enrolled woman, our screening and interview rates were fairly high. This response rate is comparable to another study in a Medicaid population in which valid phone numbers were available only for approximately 50% of Medicaid beneficiaries.16
Correct use of most contraceptives requires document, prose, and numeracy skills. However, the necessary level of proficiency probably varies with different methods of contraception. For example, understanding a calendar is critical for women taking oral contraceptives but is not necessary for effective condom use.5 Additional research is needed to determine what knowledge is essential for the effective use of contraceptives. Once we know this information we can develop tailored educational messages based on the audience's reading level and cultural content. Results from this study also indicate that education, beyond patient package inserts, is needed, particularly for individuals with low reading ability.
Results from this study have broad implications for health providers and organizations. Providers and payers should be aware of the prevalence of low literacy in their target population and modify educational messages and health services delivery accordingly. Providers need to be aware than just handing patients written materials may not be useful. When written materials are used, they need to be written at a simpler level than they are now and should be tested as to whether they actually communicate the information in a comprehensible way. Oral explanations, visual clues, and demonstration of tasks may be more effective educational methods for patients with low reading skills. Moreover, individuals with low reading skills are likely to possess a constellation of problems; thus, if a provider suspects a reading disability and the patient indicates a desire for assistance, the provider should refer her to a program that serves adults with reading difficulties.25
1. Kirsch I, Jungeblut A, Jenkins I, Kolstad A. Adult literacy in America: A first look at the results of the national adult literacy survey. Washington, DC: National Center for Education Statistics, U.S. Department of Education, 1993.
2. Baker DW, Parker RM, Williams MV, Clark WS, Nurss J. The relationship of patient reading ability to self-reported health and use of health services. Am J Public Health 1997;87:1027–30.
3. Baker DW, Parker RM, Williams MV, Pitkin K, Parikh NS, Coates W, et al. The health care experience of patients with low literacy. Arch Fam Med 1996;5:329–34.
4. Davis TC, Crouch MA, Wills G, Miller S, Abdehou DM. The gap between patient reading comprehension and the readability of patient education materials. J Fam Pract 1990;31:533–8.
5. Parker RM, Williams MV, Baker DW, Nurss JR. Literacy and contraception: Exploring the link. Obstet Gynecol 1996;88:72S–7S.
6. Wells JA, Ruscavage D, Parker B, McArthur L. Literacy of women attending family planning clinics in Virginia and reading levels of brochures on HIV prevention. Fam Plann Perspect 1994;26:113–15,131.
7. Zion AB, Aiman J. Level of reading difficulty in the American College of Obstetricians and Gynecologists patient education pamphlets. Obstet Gynecol 1989;74:955–60.
8. Streiff LD. Can clients understand our instructions? Image J Nurs Sch 1986;18:48–52.
9. Meade CD, Byrd JC. Patient literacy and the readability of smoking education literature. Am J Public Health 1989;79:204–6.
10. Nicoll A, Harrison C. The readability of health-care literature. Dev Med Child Neurol 1994;26:596–600.
11. Fredrickson DD, Washington RL, Pham N, Jackson T, Wiltshire J, Jecha LD. Reading grade levels and health behaviors of parents at child clinics. Kans Med 1995;127–9.
12. Holcomb CA. The Cloze procedure and readability of patient-oriented drug information. J Drug Educ 1983;13:347–57.
13. Richwald GA, Walmsley MA, Coulson AH, Moriskey DE. Are condom instructions readable? Results of a readability study. Public Health Rep 1988;103:355–9.
14. Swanson JM, Forrest K, Ledbetter C, Hall S, Holstine EJ, Shafter MR. Readability of commercial and generic contraceptive instructions. Image J Nurs Sch 1990;22:96–100.
15. Williams-Deane M, Potter LS. Current oral contraceptive use instructions: An analysis of patient package inserts. Fam Plann Perspect 1992;24:111–5.
16. Sisk JE, Gorman SA, Reisinger AL, Glied SA, DuMouchel WH, Hynes MM. Evaluation of Medicaid managed care: Satisfaction, access and use. JAMA 1996;276:50–5.
17. Parker RM, Baker DW, Williams MV, Nurss JR. The test of functional health literacy in adults: A new instrument for measuring patients' literacy skills. J Gen Intern Med 1995;10:537–41.
18. Williams M, Parker R, Baker D, Parikh N, Coates W, Nurss J. Inadequate functional health literacy among patients at two public hospitals. JAMA 9195;274:1677–82.
19. Lamison-White L. U.S. Bureau of the Census, Current Population Reports, Series P60-198, Poverty in the United States: 1996, U.S. Government Printing Office, Washington, DC, 1997.
20. U.S. Bureau of the Census. Statistical abstract of the United States: 1996. 100th ed. Washington, DC: U.S. Bureau of the Census; 1996.
21. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research: Principles and quantitative methods. New York: Van Nostrand Reinhold, 1982.
22. SAS Institute Inc. SAS/STAT User's Guide, Version 6, Fourth Edition, Volume 1. Cary, NC: SAS Institute Inc, 1989.
23. Williams MV, Baker DW, Parker RM, Nurss JR. Relationship of functional health literacy to patients' knowledge of their chronic disease. Arch Intern Med 1998;158:166–72.
24. Radecki SE, Beckman LJ. Contraceptive risk-taking in a medically-underserved, low-income population. Women Health 1994;21:1–15.
25. Peterson LS, Oakley D, Potter LS, Darroch JE. Women's efforts to prevent pregnancy: Consistency of oral contraceptive use. Fam Plann Perspect 1998;30:19–23.