Association of Low Caregiver Health Literacy With Reported Use of Nonstandardized Dosing Instruments and Lack of Knowledge of Weight-Based Dosing

Transcription

1 Association of Low Caregiver Health Literacy With Reported Use of Nonstandardized Dosing Instruments and Lack of Knowledge of Weight-Based Dosing H. Shonna Yin, MD; Benard P. Dreyer, MD; George Foltin, MD; Linda van Schaick, MS Ed; Alan L. Mendelsohn, MD Objective. Caregivers of young children frequently measure doses of liquid medications incorrectly. Use of nonstandardized dosing instruments and lack of knowledge that dosing is weightbased contribute to dosing errors. We sought to assess whether low caregiver health literacy was associated with these outcomes. Methods. This was a cross-sectional analysis of caregivers presenting to an urban pediatric emergency room. Dependent variables were caregiver reported use of nonstandardized dosing tools and knowledge of weight-based dosing. The independent variable was caregiver health literacy (Test of Functional Health Literacy in Adults [TOFHLA]). Results. Two hundred ninety-two caregivers were assessed: 23.3% reported use of nonstandardized liquid dosing instruments, and 67.8% were unaware of weight-based dosing. Caregivers who were unaware of weight-based dosing were more likely to use nonstandardized dosing tools (28.3% vs 12.8%; P.003). In unadjusted analyses, overall health literacy, reading comprehension, and numeracy were all associated with both dependent variables. In analyses adjusting for child age, health care experiences, and caregiver acculturation and education, inadequate/marginal overall health literacy was associated with lack of knowledge of weight-based dosing (adjusted odds ratio [AOR] 2.3; P.03), whereas lower reading comprehension was associated with both lack of knowledge (AOR 2.0; P.03) and reported use of nonstandardized instrument (AOR 2.4; P.007). Conclusions. Low health literacy, in particular reading comprehension, was associated with reported use of nonstandardized dosing instruments and lack of knowledge regarding weightbased dosing. Both caregiver health literacy and sociodemographic factors should be considered in the design of interventions to prevent medication administration errors. KEY WORDS: health literacy; literacy; medical errors; pediatric medication errors Ambulatory Pediatrics 2007;7: From the Department of Pediatrics, New York University School of Medicine and Bellevue Hospital Center, New York, NY. Address correspondence to H. Shonna Yin, MD, Department of Pediatrics, New York University School of Medicine, 550 First Avenue, New York, NY ( yinh02@med.nyu.edu). Received for publication November 1, 2006; accepted April 18, There has been limited study of caregiver dosing errors related to outpatient medication use in children. 1,2 A high frequency of errors involving liquid medicines has been reported, with 50% 70% of caregivers found to measure doses incorrectly or to report dosing outside the recommended range. 3 5 Caregiver use of nonstandardized instruments such as kitchen teaspoons is one factor that has been associated with liquid dosing errors Despite American Academy of Pediatrics (AAP) recommendations against use of nonstandardized dosing instruments, 11 kitchen teaspoons and tablespoons continue to be used to give medications to children, with studies reporting that between 20% and 73% of caregivers use a nonstandardized spoon for administering liquid medications. 5,8 Another factor that has been shown to be related to increased dosing errors is caregiver knowledge that dosing is weight-based, especially with over-the-counter medications. 9 In a study by Li et al, 4 caregivers who knew that weight was the criteria for determining dose made errors less frequently than those who identified some other factor such as age, gender, or height. Low literacy and health literacy have been shown to be related to medication self-management skills, including understanding of warning labels and adherence However, few published studies have assessed associations between literacy and medication dosing in children. In addition, there has been limited study of the link between health literacy and risk factors for medication error, such as use of nonstandardized dosing instruments and lack of knowledge of weight-based dosing. Given the high prevalence of literacy-related problems in the United States, especially in low socioeconomic status (SES) individuals, 18,19 additional study of the relationship between literacy and medication dosing is indicated. In this study, we sought to assess whether caregiver health literacy was related to risk factors for liquid medication dosing errors, including reported use of nonstandardized dosing tools and lack of knowledge about weightbased dosing. We hypothesized that low caregiver health literacy would be associated with reported use of nonstandardized dosing tools and lack of caregiver awareness AMBULATORY PEDIATRICS Volume 7, Number 4 Copyright 2007 by Ambulatory Pediatric Association 292 July August 2007

2 AMBULATORY PEDIATRICS Yin et al 293 of weight-based dosing. Better understanding of the association between health literacy and risk factors for liquid medication dosing errors has the potential to inform effective interventions to prevent pediatric medication errors. METHODS Study Sample This was a cross-sectional analysis of parents and caregivers presenting to the pediatric emergency service at Bellevue Hospital Center, an urban public hospital in New York City. The Bellevue pediatric emergency service serves many families who have low education and low SES and has approximately visits per year. Enrollment took place from July through October During times when enrollment was taking place, research assistants systematically assessed families to determine eligibility. Inclusion criteria were parent or caregiver with a child aged between 30 days and 8 years, nonurgent visit, 20 presence of primary caregiver (parent or legal guardian) responsible for giving medications, caregiver s language either English or Spanish, child s medication generally given in liquid form, and visit not involving psychiatric problem or child protection issue. Written, informed consent was obtained. This study was approved by the New York University School of Medicine Institutional Review Board and Bellevue Research Committee. Data Obtained Data was obtained by interview of the child s primary caregiver, conducted in the language preferred by the caregiver. Research assistants collecting the data were trained by one of the authors (H.S.Y.), with periodic review and observation to ensure that data was collected in a reliable manner. Dependent Variables One outcome variable was reported caregiver use of a nonstandardized measurement tool as a primary dosing instrument. Caregivers were asked about their use of dosing instrument for liquid medications with the question What do you use most of the time at home to give your child his/her liquid medicine? Caregivers were asked to choose from a set of photographs of dosing instruments based on those used in other studies: kitchen teaspoon, kitchen tablespoon, dosing spoon, measuring spoon, dosing cup, dropper, and syringe. 3,5 Answers were dichotomized in analyses, with caregivers selecting standardized measuring instruments (dosing spoon, measuring spoon, dosing cup, dropper, or syringe) compared with caregivers selecting nonstandardized tools (kitchen teaspoons and tablespoons). The second outcome variable was caregiver lack of knowledge about weight-based dosing, given its relationship to increased dosing errors. 4 Caregivers were asked, From the following list, what is the most important characteristic of a child when deciding what dose of medicine to give, with choices being gender, age, height, weight, and how the child is feeling (adapted from a study by Li et al 4 ). Answers were dichotomized, with caregivers selecting weight compared with caregivers who selected any other answer. Independent Variables The independent variable for this study was caregiver health literacy. In addition, we collected information about potential confounders, including sociodemographics and child health care related experiences. Caregiver health literacy level was assessed with the Test of Functional Health Literacy in Adults (TOFHLA), 21 which is composed of a 50-item reading comprehension section and a 17-item numeracy section. In the reading comprehension section, participants read incomplete passages related to health knowledge and are asked to fill in blanks by selecting 1 of 4 choices. The numeracy section is administered by research assistants and involves answering questions about instructions from prescription bottles and other health-related materials. The 2 sections are equally weighted, with total TOFHLA scores ranging from A score of 0 59 is categorized as inadequate functional health literacy; is categorized as marginal functional health literacy; is categorized as adequate functional health literacy. In 12 cases, time constraints resulted in the caregiver being unable to be tested using the full TOFHLA; in those cases, the caregiver was tested using the short version of the TOFHLA, 22 which yields the same set of categories for overall literacy. The TOFHLA has good correlation with the Wide Range Achievement Test and the Rapid Estimate of Adult Literacy in Medicine (correlation coefficients of 0.74 and 0.84, respectively) 21 and is validated in English and Spanish. 12,21 There are currently no standardized assessments that measure the health literacy skill of individuals comprehensively, as health literacy refers not only to the ability to understand health information, but also the ability to obtain and process this information to make decisions. 12 The TOFHLA has been utilized in multiple studies as a proxy measure for health literacy. 15,23 25 Sociodemographic data included child s age and gender, caregiver relationship to the child, and caregiver s country of origin, ethnicity, language, education, and SES. Child s age was dichotomized as 1 year and 1 year, because children younger than 1 year are not typically developmentally ready to use spoons or cups. 26 Caregiver country of origin was categorized as US born or birth place outside the US. Ethnicity was categorized as Latino and non-latino, as Latinos make up a majority of patients seen at Bellevue and no other ethnic group in this sample comprised a large enough proportion to define a separate category. Language was characterized based on the caregiver s choice on the TOFHLA (English or Spanish). Education was categorized by whether the caregiver had completed high school. Socioeconomic status was assessed using the Hollingshead Four Factor Index of Social Status (A. Hollingshead, unpublished data, 1975), based on education and occupation. Socioeconomic status was

3 294 Health Literacy and Risk Factors for Liquid Medication Dosing Error AMBULATORY PEDIATRICS coded to a 5-point scale, with 1 representing the highest resource level and 5 representing the lowest resource level. Child health care related experiences were assessed by asking caregivers whether the child had a regular health care provider and whether they had ever received a dropper or syringe in the clinic or emergency room. Statistical Analysis We assessed the association between health literacy and reported use of nonstandardized dosing instrument and lack of knowledge of weight-based dosing. Analyses were performed for both overall health literacy on the TOFHLA (with caregivers scoring inadequate or marginal compared with those with adequate scores) and for the reading comprehension and numeracy subscales (with caregivers scoring below the median compared with those scoring at or above). Both simple and adjusted analyses were performed. For simple analyses, Fisher exact tests for 2 2 tables and chi-square for comparisons with more than 2 categories were used. Adjusted analyses were performed using multiple logistic regressions. In these models, we included all variables considered a priori to be potential confounders, including caregiver education, country of origin, language, and SES, as well as age of child, regular child health-care provider, and experience of ever having received a dosing instrument in a health-care setting. Given strong previously found associations between health literacy and variables related to caregiver acculturation and education, 19 we also performed restricted regression analyses, excluding confounders of health literacy. In addition, we performed subgroup analyses to look for whether the association between health literacy and the dependent variables differed by sociodemographic characteristics, including age, education, and language spoken. RESULTS Study Sample During the enrollment period, there were 1863 visits of children aged between 30 days and 8 years; 745 visits occurred during the daytime and early evening hours when research assistants were present. Of 640 families with children who met age and nonurgent triage criteria, 422 (65.9%) were assessed for additional inclusion criteria. Although research assistants endeavored to assess consecutive families, this was not always possible due to families being called to be seen by the physician or being enrolled in other studies; as a result, 218 families were not assessed for additional inclusion criteria. There was no statistically significant difference in those who were assessed and those not assessed regarding age and gender. Of the 422 assessed for additional inclusion criteria, 74 (17.5%) were determined to be ineligible based on 1 or more of the following: primary caregiver (or person who regularly gives medications to the child) not present (26), inability to communicate in English or Spanish (39), and visit for psychiatric or child protective services issue (15). Of the 348 eligible families, 307 (88.2%) were enrolled in the study, whereas 41 (11.8%) refused. Of 307 families enrolled, 1 family left after completing only a minimal portion of the survey; in addition, 14 families were excluded from the analysis because they had never given a liquid medication to their child. Two hundred ninety-two caregivers were therefore included in the analysis, with 258 mothers, 30 fathers, and 4 others interviewed. Mean (SD) child age was 3.5 (2.4) years. Descriptive data are shown in Table 1. Health Literacy Mean (SD) caregiver TOFHLA score was 81 (16.5). Test of Functional Health Literacy in Adults scores were lower for those administered in Spanish (16.7% inadequate, 20.4% marginal, and 63.0% adequate) compared with those administered in English (5.5% inadequate, 13.3% marginal, and 81.2% adequate), with P.001. Other sociodemographic factors associated with lower health literacy skills included birthplace outside of the United States (P.001), Spanish language (P.001), less than high school education (P.001), and being in a lower Hollingshead SES category (P.001). Reported Use of Standardized Dosing Instrument and Knowledge of Weight-Based Dosing Caregiver utilization of dosing instrument and knowledge of the basis of medication dosing is shown in Table 2; 23.3% of caregivers reported that the primary dosing tool used at home was a nonstandardized kitchen teaspoon or tablespoon, and 67.8% of caregivers were unaware that weight was the basis for medication dosing. Among caregivers unaware of weight-based dosing, 27.3% of caregivers reported using a nonstandardized tool, compared with 12.8% of caregivers who were aware of weight-based dosing (P.003). Association of Low Caregiver Health Literacy With Reported Use of Nonstandardized Dosing Tool and Lack of Knowledge of Weight-Based Dosing Unadjusted Analyses In unadjusted analyses, caregivers with inadequate or marginal overall health literacy were more likely to report use of a nonstandardized dosing instrument compared with caregivers with adequate health literacy (34.7% vs 19.2%; P.01). Increased frequency of reported use of nonstandardized instruments was also found for caregivers scoring below the median in reading comprehension (35.2% vs 15.1%; P.001) and numeracy (34.3% vs 18.7%; P.005). Similarly, caregivers with inadequate or marginal health literacy were more likely to lack knowledge about weight-based dosing compared with caregivers with adequate health literacy (85.3% vs 61.2%, P.001). Lack of knowledge about weight-based dosing was also found more frequently for caregivers scoring below the median in reading comprehension (79.2% vs 55.9%, P.001) and numeracy (75.8% vs 62.1%; P.02). In subgroup analyses, the association between health literacy and reported use of dosing instrument varied by age of child. Amongst caregivers with children aged 1 year, 40% with inadequate or marginal health literacy reported use of a nonstandardized dosing instrument, com-

4 AMBULATORY PEDIATRICS Yin et al 295 Table 1. Descriptive Data n % Caregiver health literacy TOFHLA* Inadequate Marginal Adequate Child characteristics Age 1 y Gender Male Caregiver and household characteristics Child health care related experiences Country of origin Born outside the United States Race/ethnicity Latino Black or African American Asian White Other Language English Spanish Education Less than high school graduate Hollingshead Regular MD Yes Ever received a dosing tool from the Yes clinic or emergency department *TOFHLA indicates Test of Functional Health Literacy in Adults. Data missing for 3 subjects, 1 of whom only completed the numeracy portion of the test. Language of TOFHLA administration. Missing for 2 subjects. Hollingshead Socioeconomic Status (A. Hollingshead, unpublished data, 1975), with lower number representing higher socioeconomic status and greater family resources. pared with 22.9% of caregivers with adequate health literacy (P.01); in contrast, no difference was seen amongst caregivers with children aged 1 year (0% vs 4.5%; P 1.0). Variations in the association between health literacy and reported use of dosing instrument were not found for caregiver education or for language spoken. In addition, we found that the association between health literacy and lack of knowledge of weight-based dosing varied by caregiver language. Amongst English speaking caregivers, 88.6% with inadequate or marginal health literacy were unaware of weight-based dosing compared Table 2. Descriptive Data for Dependent Variables n % Primary instrument used for dosing medications Kitchen teaspoon Kitchen tablespoon Dosing spoon Measuring spoon Dosing cup Dropper Syringe Nonstandardized tool Standardized tool Knowledge of basis for dosing medication Weight Age How child is feeling Other/don t know Correct Incorrect with 54.1% of caregivers with adequate health literacy, P.001; in contrast, no association with health literacy was seen for Spanish speaking caregivers (82.5 % vs 76.5%; P.6). Variations in the association between health literacy and lack of knowledge of weight-based dosing were not found for child age or caregiver education. Adjusted Analyses Table 3 shows the results of separate multiple logistic regression analyses assessing the associations between caregiver health literacy and reported use of nonstandardized dosing tool and lack of knowledge of weight-based dosing. The first column shows associations by using the full model in which all potential confounders were included; the second column shows a restricted model in which acculturation and education variables previously found to be strongly related to health literacy were excluded. In full models with reported use of nonstandardized dosing tool as the dependent variable, only caregiver reading comprehension score below the median was found to be significantly associated (adjusted odds ratio [AOR)] 2.4; P.007). In restricted models, inadequate/marginal overall literacy (AOR 1.9; P.04), reading comprehension below the median (AOR 3.1, P.001), and numeracy below the median (AOR 1.9; P.03) were all found to be significantly associated. Other characteristics associated with reported use of nonstandardized dosing tools (in the full model with overall TOFHLA) included caregiver birthplace outside the United States (AOR 2.5;

5 296 Health Literacy and Risk Factors for Liquid Medication Dosing Error AMBULATORY PEDIATRICS Table 3. Multiple Logistic Regression Models Model Without Inclusion of Full Model Primary Independent Variable Health Literacy Confounders AOR 95% CI P Value AOR 95% CI P Value Reported use of nonstandardized dosing instrument Caregiver overall TOFHLA : marginal/inadequate * Caregiver TOFHLA reading comprehension score below the median * * Caregiver TOFHLA numeracy score below the median * Lack of knowledge of weight-based dosing Caregiver overall TOFHLA: marginal/inadequate * * Caregiver TOFHLA reading comprehension score below the median * * Caregiver TOFHLA numeracy score below the median * *P.05. Independent associations of caregiver health literacy with caregiver reported use of nonstandardized dosing instruments and lack of knowledge of weight-based dosing. Adjusting for caregiver education, birthplace, language, socioeconomic status, child age, regular health provider for child, and history of receiving dosing instrument in clinic or emergency department. Adjusting for child s age, regular health provider for child, and history of receiving dosing instrument in clinic or emergency department. AOR indicates adjusted odds ratio based on logistic regression model, adjusting for confounders. CI indicates confidence interval. TOFHLA indicates Test of Functional Health Literacy in Adults. P.02) and age of child 1 year or above (AOR 10.2; P.002). In full models with lack of knowledge of weight-based dosing as the dependent variable, both inadequate/marginal overall literacy and caregiver reading comprehension score below the median were found to be significantly associated (AOR 2.3, P.03; AOR 2.0, P.03, respectively). In restricted models, inadequate/marginal overall literacy (AOR 3.4; P.001), reading comprehension below the median (AOR 2.9; P.001), and numeracy below the median (AOR 1.8; P.047), were all found to be significantly associated. Other characteristics associated with lack of knowledge of weight-based dosing (in the full model with overall TOFHLA) included caregiver education less than high school (AOR 3.5; P.001), caregiver birthplace outside the United States (AOR 2.7; P.004), and no history of receiving a dosing instrument in the clinic or emergency department (AOR 2.0; P.02). DISCUSSION In this study, we assessed the relationship between caregiver health literacy and risk factors associated with medication dosing errors. We found that 23.3% of caregivers reported use of nonstandardized liquid dosing tools, which in prior studies has been associated with medication administration errors in young children. 5,11 In addition, we found that 67.8% of caregivers were unaware that medication dosing is based upon weight, and that caregivers who were unaware of weight as the basis for dosing were less likely to choose a standardized dosing instrument. The main study variable, health literacy, was strongly related to both reported use of standardized instrument and knowledge of weight-based dosing in unadjusted analyses. Although overall health literacy was no longer significantly related to reported use of nonstandardized instrument in multiple regression analysis, this may have been due to overcontrol, as acculturation and education are known to be highly correlated with health literacy. 27 Indeed, significant associations were found in restricted regression models in which acculturation and education variables had been removed. Reading comprehension in the TOFHLA was consistently associated with both outcomes in all analyses, whereas significant associations for numeracy were found only for unadjusted analyses and restricted models. Although prior studies have linked numeracy to health-related attitudes and behaviors, 28,29 to our knowledge there has been limited study comparing reading comprehension to numeracy in predicting health-related attitudes and behaviors. Additional research is indicated to better understand the relative contributions of each of these components of health literacy. Our findings are consistent with prior studies assessing the link between low overall health literacy and medication self-management skills Low health literacy skills have been linked to difficulty identifying medications, 14 difficulty following steps for the correct use of medications, 17 incorrect interpretation of prescription medication warning labels, 13 and poor medication adherence. 15,16 However, few studies have assessed the relationship between caregiver health literacy and appropriate medication dosing. In addition, there has been limited study of risk factors for medication errors in children, including use of nonstandardized dosing instruments and lack of knowledge of weight-based dosing, and the relationship between limited health literacy skills and these risk factors. We found that the association between low caregiver health literacy and reported use of nonstandardized dosing instrument varied by child s age. Reported use of nonstandardized liquid dosing instruments and associations between health literacy and use of these instruments were observed primarily with caregivers who had children 1 year of age or older. This may have been due to limited choices of dosing implements for infants and toddlers who may not be developmentally ready to use spoons and

6 AMBULATORY PEDIATRICS Yin et al 297 cups. 7 Also, many medications for infants (eg, antipyretics) are dispensed with a standardized dropper, and caregivers with older children may be less restricted in choice of dosing instrument. In addition, the association between low caregiver health literacy and lack of knowledge of weight-based dosing varied by caregiver language, with associations seen for English speakers only. Additional study of factors related to use of standardized dosing instruments and knowledge of weight-based dosing is indicated in non-english speaking populations. Prior receipt of a dosing instrument was also found to be independently associated with knowledge of weight-based dosing. This is consistent with previous work showing improvements in dosing behaviors associated with receipt of a dosing syringe together with a brief demonstration, 5 and may have implications for the development of interventions to prevent dosing errors. The other measure of child health care experiences, having a regular health care provider, was not associated with outcome variables. To our knowledge, there has been limited study of the association between engagement in care and medication dosing knowledge and behaviors, and additional study is indicated. There are limitations to our results. Because this was a cross-sectional study, we could not prove causal relationships between our independent and dependent variables. In addition, although caregiver use of a nonstandardized dosing tool has been found to be associated with decreased dosing accuracy, 5 the frequency of actual dosing errors was not evaluated in this study. Our assessment of utilization of standardized dosing tools may have been subject to social desirability bias. For example, we may have underestimated the number of caregivers who use teaspoons and tablespoons at home because questions were asked in a heath care setting where standardized dosing instruments are often used and provided. Additionally, we recruited only English and Spanish speakers from a primarily urban Latino patient population and only recruited study subjects during the daytime and early evening hours, which may affect generalizability. Lower-literacy patients may also have been less likely to enroll in our study due to discomfort with the health care system. 12 Finally, a single research assistant administered both the TOFHLA and the outcome questions. However, the TOFHLA typically followed the survey and was not scored until after the caregiver had completed all assessments; therefore, it is unlikely that this could have led to bias in the results. In summary, we found that caregiver reported use of nonstandardized liquid dosing instruments and lack of knowledge about weight-based dosing were more frequent among caregivers with low health literacy, and in particular, were related to lower reading comprehension. We also found associations between these outcomes and sociodemographic risk factors, including less than high school education, birthplace outside the United States, and not speaking English. In addition, reported receipt of a standardized dosing instrument was associated with increased knowledge of weight-based dosing. We conclude that both caregiver health literacy and sociodemographic characteristics, as well as provision of standardized dosing instruments, should be considered in the design of interventions to prevent medication administration errors. ACKNOWLEDGMENTS This study was performed with support from the Centers for Disease Control and Prevention (grant/cooperative agreement number T01 CD000146; Drs H. Shonna Yin, George Foltin, and Alan L. Mendelsohn). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control. Support was also provided by the New York University Department of Pediatrics. We would like to thank Marc Gourevitch and Mark Schwartz, who direct the Centers for Disease Control/New York University Medicine and Public Health Research Fellowship Program. We would also like to acknowledge the members of our project team for their work related to this study, including senior research assistants Stephanie DeFiores, Cheryl Dinglas, Meyling Oei, and Neha Sathe, and research staff members Isabel Bazan, Carolyn Cutler, Michelle Diaz, Leydi Espinal, Giselle Jose, Eric Levy, Heidi Mock, Katherine Milligan, and Susanna Silverman, as well as the staff in Bellevue Hospital s Pediatric Emergency Service for their assistance with this study. REFERENCES 1. Walsh KE, Kaushal R, Chessare JB. How to avoid paediatric medication errors: a user s guide to the literature. Arch Dis Child. 2005;90: McPhillips HA, Stille CJ, Smith D, et al. Potential medication dosing errors in outpatient pediatrics. J Pediatr. 2005;147: Simon HK, Weinkle DA. Over-the-counter medications. Do parents give what they intend to give? Arch Pediatr Adolesc Med. 1997; 151: Li SF, Lacher B, Crain EF. Acetaminophen and ibuprofen dosing by parents. Pediatr Emerg Care. 2000;16: McMahon SR, Rimsza ME, Bay RC. Parents can dose liquid medication accurately. Pediatrics. 1997;100: Rheinstein PH. Avoiding problems with liquid medications and dosing devices. Am Fam Physician. 1994;50: McKenzie M. Administration of oral medications to infants and young children. US Pharm. 1981;6: Madlon-Kay DJ, Mosch FS. Liquid medication dosing errors. J Fam Pract. 2000;49: Hyam E, Brawer M, Herman J, Zvieli S. What s in a teaspoon? Underdosing with acetaminophen in family practice. Fam Pract. 1989; 6: Litovitz T. Implication of dispensing cups in dosing errors and pediatric poisonings: a report from the American Association of Poison Control Centers. Ann Pharmacother. 1992;26: Yaffe SJ, Bierman CW, Cann HM, et al. Inaccuracies in administering liquid medication. Pediatrics. 1975;56: Schwartzberg JG, Van Geest JB, Wang CC, eds. Understanding Health Literacy. Implications for Medicine and Public Health, Chicago, IL: AMA Press, Davis TC, Wolf MS, Bass PF, et al. Low literacy impairs comprehension of prescription drug warning labels. J Gen Intern Med. 2006;21: Kripalani S, Henderson LE, Chiu EY, et al. Predictors of medication self-management skill in a low-literacy population. J Gen Intern Med. 2006;21: Kalichman SC, Ramachandran B, Catz S. Adherence to combination antiretroviral therapies in HIV patients of low health literacy. J Gen Intern Med. 1999;14: Graham J, Bennett I, Holmes W, Gross R. Medication beliefs as mediators of the health literacy antiretroviral adherence relationship in HIV-infected individuals. AIDS Behav 2007;11: Williams MV, Baker DW, Honig EG, Lee TM, Nowlan A. Inadequate literacy is a barrier to asthma knowledge and self-care. Chest. 1998;114: Nielson-Bohlman L, Panzer A, Kindig D, Health Literacy: A Pre-

7 298 Health Literacy and Risk Factors for Liquid Medication Dosing Error AMBULATORY PEDIATRICS scription to End Confusion. Washington, DC: Institute of Medicine- National Academies Press; Kutner M, Greenberg E, Jin Y, Paulsen C, White S. The Health Literacy of America s Adults. Results from the 2003 National Assessment of Adult Literacy. Report #: NCES Hyattsville, Md: National Center for Education Statistics; Maldonado T, Avner JR. Triage of the pediatric patient in the emergency department: are we all in agreement? Pediatrics. 2004;114: 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: Aguirre AC, Ebrahim N, Shea JA. Performance of the English and Spanish S-TOFHLA among publicly insured Medicaid and Medicare patients. Patient Educ Couns. 2005;56: Gazmararian J, Baker D, Parker R, Blazer DG. A multivariate analysis of factors associated with depression: evaluating the role of health literacy as a potential contributor. Arch Intern Med. 2000;160: Schillinger D, Grumbach K, Piette J. Association of health literacy with diabetes outcomes. JAMA. 2002;288: Williams MV, Baker DW, Parker RM, Nurss JR. Relationship of functional health literacy to patients knowledge of their chronic disease. A study of patients with hypertension and diabetes. Arch Intern Med. 1998;158: Frankenburg W, Dodds J, Archer P, Shapiro H, Bresnick B. Denver-II Screening Manual. Denver, CO: Denver Developmental Materials, Inc; Leyva M, Sharif I, Ozuah PO. Health literacy among Spanishspeaking Latino parents with limited English proficiency. Ambul Pediatr. 2005;5: Estrada CA, Martin-Hryniewicz M, Peek BT, Collins C, Byrd JC. Literacy and numeracy skills and anticoagulation control. Am J Med Sci. 2004;328: Schwartz LM, Woloshin S, Black WC, Welch HG. The role of numeracy in understanding the benefit of screening mammography. Ann Intern Med. 1997;127: