|
|
||||||||
Correspondence: Address correspondence to Daniel G. Morrow, Institute of Aviation, Aviation Human Factors Division, Willard AirportOne Airport Road, Q5, MC-394, Savoy, IL 61874. E-mail: dgm{at}uiuc.edu
| Abstract |
|---|
|
|
|---|
Key Words: Medication instructions Adherence Memory Health literacy
The relationship between patients' medication knowledge and adherence is articulated by the multifactor model of adherence developed by Park (e.g., Park & Jones, 1997). According to this model, adherence depends on cognitive abilities (e.g., working memory), medication and disease variables that influence beliefs about illness and treatment, and external cues (e.g., social support, reminders). Age primarily influences adherence through cognitive function. Thus, comprehension and other components of adherence may be compromised by age-related differences in working memory (Park et al., 1996; Salthouse, 1991) and health-related literacy (Baker et al., 2000).
An important strategy for increasing medication knowledge is to improve instructions, especially because the federal government has mandated a private-sector program to improve the quantity and quality of printed medication information provided to patients at pharmacies (Public Law 104-180, August 6, 1996). However, more information will only improve knowledge if this information is easy to understand and remember, especially for older adults who experience declines in the cognitive resources necessary for comprehension. Unfortunately, medication instructions now available at many commercial pharmacies are difficult to understand and use because of small print, complex language, and poor organization (Klein & Isaacson, 2003; Svarstad & Bultman, 1999; Wright, 1999).
We developed patient-centered medication instructions as part of a pharmacy-based educational intervention to improve health-related outcomes and adherence among older adults with CHF (for a description of the intervention study, see Murray et al., 2004). To validate these instructions, we first tested whether they were better understood, compared with instructions available in a chain pharmacy, by older adults with CHF.
The instructions were patient centered because they were designed according to patients' needs and abilities and they provided the patients the information needed to develop a plan for taking medication (Michie, Miles, & Weinman, 2003). More specifically, they were designed to support patients' comprehension and memory by minimizing demands on sensory (e.g., visual acuity) and cognitive (e.g., working memory) abilities that decline with age, and by taking advantage of patients' knowledge about taking medication. We took a multileveled approach, focusing on the content, language, organization, and presentation medium of the instructions. We conveyed information necessary for creating an accurate adherence plan (e.g., medication dose and schedule) by use of large print and simple language (short words and simple syntax, reflected in readability scores appropriate for low grade levels), which may especially benefit low literacy patients (Ad Hoc Committee on Health Literacy, 1999). We took a patient-centered approach to the organization as well as the content and language of the instructions. With increasing experience, older adults may develop a medication-taking schema that guides expectations about the information that should be provided by instructions. According to this schema, general information such as purpose should be followed by information about how to take the medication (e.g., dose and schedule), and then possible outcomes of taking the medication such as side effects (Morrow, Leirer, Andrassy, Tanke, & Stine-Morrow, 1996). We organized the instructions to match this schema because the order of information followed the schematic order. Finally, the presentation format of the instructions should also improve comprehension because we reinforced the text by icons that conveyed medication name, dose, and schedule. Combined icontext formats improve older adults' memory for procedural instructions (Morrell & Park, 1993) and medication instructions (Morrow, Hier, Menard, & Leirer, 1998).
We have found that patient-centered instructions improve memory for medication information among educated, healthy older adults (for a review, see Morrow & Leirer, 1999). However, it is important to test whether they also improve medication knowledge among patients with CHF who vary in education, literacy, cognitive ability, and health status, as a first step to investigating the impact of such instructions on health behaviors in clinical settings.
Study Aims
We tested the hypothesis that older adults with CHF better understand and remember the patient-centered instructions than a standard format used in a chain pharmacy, which would suggest that the patient-centered format increases patients' medication knowledge. We found earlier that older adults generally preferred these instructions over standard instructions (Morrow et al., 2004), but preferences do not always predict objective measures such as recall of instructions (Davis, Holcombe, Berkel, Pramanik, & Divers, 1998).
We also examined whether older adults better understand and remember instructions for familiar (e.g., currently prescribed) medications. On one hand, prior knowledge about the medication and illness may provide a framework for organizing the encoding and retrieval of new information from instructions. On the other hand, learning new information is sometimes more difficult for familiar than for unfamiliar health topics, perhaps because new information can conflict with previously learned information (Brown & Park, 2003). Medication familiarity also may moderate the impact of the patient-centered instruction format. Participants already taking a medication also may be familiar with the standard format, which could mitigate potential benefits of the novel format.
Finally, we investigated whether patient characteristics predicted recall of medication information, which would have implications for tailoring instructions. We examined how recall related to patient age, education, and measures of health-related literacy, working memory, and speed of mental processing.
| Methods |
|---|
|
|
|---|
Participants
A volunteer sample of 32 community-dwelling older adults diagnosed with CHF for which they were prescribed a diuretic and an ACE inhibitor was recruited at Wishard Health Services, Indianapolis, IN. Although the participants were not involved in the intervention study described by Murray and colleagues (2004), they met the study inclusion criteria.
Instructions
We compared patient-centered and standard pharmacy formats for familiar (the diuretic and the ACE inhibitor) and unfamiliar (the beta blocker and digoxin) medications. Figure 1 presents the patient-centered version for the diuretic. Although the two formats conveyed similar information, the patient-centered instructions should be better understood than the standard instructions for several reasons. They had a larger font (1214 point vs 810 point for the body of the instructions) and better readability scores (mean grade level of 7.4 vs 9.3, FleschKincaid readability formula), and they were shorter (mean of 251.3 vs 557.8 words; standard instructions contained more information about potential drug interactions and side effects). We organized the patient-centered instructions to be consistent with patients' medication-taking schema, and they contained icons that reinforced the text by explicitly conveying medication name, dose, and time information (for more details, see Morrow et al., 2004). The instructions for the familiar and unfamiliar medications did not differ in mean grade level of the FleschKincaid readability formula; familiar = 8.2, unfamiliar = 8.7, t(6) < 1.0.
|
Following the question task, we removed the instruction; after participants performed a 30-s distracter task (canceling each letter e in a passage), we asked them to recall information from the instruction (free recall task) and then answer questions about specific information (same as the explicit comprehension questionscued recall task). We measured memory according to the percentage of correctly recalled information. We measured comprehension by time and accuracy to answer questions with instructions in view. We defined answer time as the interval from the end of the question to the beginning of the answer, and we measured it blind to condition from the audiotape. The question task may relate to the ability to initially read instructions to understand how to take medication, or to search for information in response to an event such as experiencing side effects. The recall tasks tap the ability to retrieve knowledge about how to take the medication.
We measured several cognitive abilities. We measured health-related literacy by using the Short Test of Functional Health Literacy in Adults (S-TOFHLA; see Baker, Williams, Parker, Gazmararian, & Nurss, 1999). This 7-min test assesses the ability to read and understand actual health-related passages with readability levels of 4.3 and 10.4 grade levels (GunningFog index). Scores on this version of the test range from 0 to 36 (016 = inadequate health literacy, 1722 = marginal literacy, and 2336 = adequate literacy) and are often lower for older adults and predict health utilization (Baker et al., 2000). We measured verbal working memory, or the ability to simultaneously store and manipulate verbal information, by using the Listening Span task. Participants answered questions about progressively larger sets of simple spoken sentences (one to six sentences) and then recalled the final word of each sentence in the set. The span score is the size of the sentence set for which participants can reliably recall all sentence-final words (for details on materials and scoring, see Salthouse & Babcock, 1990). We measured processing speed by using the Letter Comparison and Pattern Comparison tasks (Salthouse & Babcock, 1991). In these timed paper-and-pencil tasks, participants decided as rapidly as possible whether pairs of letter sets or line patterns were the same or different (maximum score = 25). The working memory and speed measures account for age-related differences in performance of several verbal reasoning and memory tasks (Salthouse & Babcock, 1991).
Plan of Analysis
We analyzed recall and comprehension measures by using a randomized block factorial analysis of variance (ANOVA) design with instruction format (patient-centered vs standard pharmacy) and medication familiarity (currently prescribed vs never prescribed) as repeated measures (Kirk, 1995). We included type of test (free vs cued recall) as a repeated measure for recall. These analyses collapsed over the order of instruction presentation variable because order did not influence recall, F < 1.0. We also investigated whether patient characteristics predicted recall. Because instruction format influenced recall, we conducted analyses separately for patient-centered and for standard instruction recall. In a set of hierarchical regressions, we first examined whether variance in recall was related to patient age (Model 1). We then examined how much age-related variance was explained by differences in education and health-related literacy (Model 2). Finally, we examined whether additional variance in recall was explained by the cognitive measures (processing speed and working memory; Model 3). There was no evidence for multicollinearity among variables entered together in these analyses.
| Results |
|---|
|
|
|---|
|
To explore reasons for the patient-centered advantage, we examined format effects for each type of information probed in the recall task. Table 2 shows that the patient-centered format advantage primarily occurred for medication name, dose, and time (the Format x Familiarity interaction for time information was produced because the patient-centered benefit occurred only in the unfamiliar medication condition). This analysis suggests that the icons improved recall because these three items were conveyed by icons as well as by text in these instructions. This analysis also showed that the patient-centered benefit for recall was greater for some adherence-critical information than for overall recall, with a 23% improvement for name and 16% for dose information. Finally, the familiarity advantage occurred only for general facts that patients were likely to know before the study (medication purpose, name).
|
Patient-centered instructions were understood more accurately for unfamiliar medications (PC = 81%, standard = 74%), F(1, 93) = 5.8, p <.05, but the standard instructions were more accurate for familiar medications (PC = 75%, standard = 82%), F(1, 93) = 5.3, p <.05; Format x Familiarity interaction, F(1, 93) = 11.2, p <.001. Main effects of format and familiarity were not significant, Fs < 1.0.
Predictors of Medication Information Recall
Age accounted for 30% of the variance in recall of standard instructions (Model 1, Table 3). Education and health-related literacy accounted for an additional 27% of the variance, and the literacy measure appeared to eliminate the impact of age on recall (Model 2). Finally, the working memory and processing speed measures accounted for an additional 11% of the variance. The impact of health literacy on recall was eliminated, primarily when working memory was controlled for (Model 3).
|
| Discussion |
|---|
|
|
|---|
We previously found that older adults with CHF tend to prefer patient-centered instructions over instructions for the same medications from a chain pharmacy (Morrow et al., 2004). The present study shows that older adults also better understand and remember these instructions. Although only immediate recall was measured in our study, there is evidence that level of immediate recall of medication information predicts later recall at a 1-month test (McGuire, 1996).
The patient-centered and standard instructions differ on several dimensions, making it difficult to pinpoint why the patient-centered instructions were better remembered. For example, larger print reduced input limitations on comprehension, and simpler language (reflected in better readability scores) addressed literacy limitations. The icons may also have reinforced information about medication name, dose, and time. Indeed, analysis of instruction format effects for types of information (Table 2) suggested that icon-relevant information was especially likely to be recalled from the patient-centered instructions. Icons may particularly help less literate patients (Houts et al., 1998).
Instructions for familiar medications were better recalled than those for unfamiliar medications. An analysis by type of information suggested this advantage occurred for general facts that patients likely knew before the study. Although it is possible that variable content across instructions contributed to familiarity effects, familiar and unfamiliar instructions did not differ in mean readability scores.
Patient-centered instructions were understood more accurately for the unfamiliar medications, whereas the standard instructions were better understood for the familiar medications. Patients may have been familiar with the standard pharmacy format for medications they were already taking, which could mitigate effectiveness of the novel patient-centered format. However, the standard format advantage did not persist to the recall tasks, suggesting that any difficulty understanding the novel format for familiar medications was transient. Nonetheless, the patient-centered format may be easier to introduce for newly prescribed medications.
We also explored whether patient characteristics predicted recall. Although findings from these exploratory regression analyses must be interpreted with caution because of the small sample size, several findings are consistent with previous literature. The finding that age differences in recall were associated with differences in health literacy fits with evidence that less literate older adults are at risk for poor knowledge of self-care activities (Baker et al., 2000). There was also evidence that the effects of health literacy on standard instruction recall reflected age differences in working memory capacity, suggesting that age differences in health literacy reflect age-related differences in more basic cognitive resources (Baker et al., 2000). Working memory explained less variance in recall of the patient-centered instructions than of the standard instructions, suggesting that a patient-centered approach may especially benefit older adults with reduced working memory. However, the health-literacy measure predicted recall of the patient-centered instructions even after working memory was controlled for, perhaps because these instructions were primarily text based despite the presence of icons.
Implications for Improving Health Communication
Patient-centered instructions improved memory for medication information of older adults diagnosed with CHF. This suggests a fruitful approach to improving the design of instructions for prescribed medications that are available in many pharmacies, so that the instructions are useful as well as accurate, and thus likely to be used by older adults. The finding that participants with lower levels of health literacy recalled less of the patient-centered as well as the standard instructions suggests that such patients need to be supported by a patient-centered approach to spoken as well as printed communication (Morrow, 1997). These patients may also benefit from multimedia instructions such as those delivered by the Internet (Wright, 1999). Implementing these recommendations would require working with administrators, particularly regarding the use of computers to facilitate the distribution of patient-centered instructions in pharmacies.
Implications for Future Research
Patient-centered instructions may help increase medication adherence by improving patients' medication knowledge, and perhaps by influencing patients' beliefs about CHF and its treatment, both of which should support the ability to create effective adherence plans. However, this provisional conclusion must be verified by studies with larger diverse samples of patients in order to more definitively investigate the impact of literacy on patient knowledge and adherence. We also note that only one pharmacy format was tested in our study, and these instructions vary in quality across pharmacies (Svarstad & Bultman, 1999). Thus, future studies must assess larger samples of instructional materials. Finally, knowledge about medications is only one of many factors influencing adherence, such as external memory aids and the cost of medications (Park et al., 1999). The impact of the patient-centered instructions on adherence is now being investigated in an ongoing study assessing the impact of a multifaceted pharmacist-based educational intervention on adherence to CHF medications.
| Footnotes |
|---|
1 Institute of Aviation and the Beckman Institute, University of Illinois, Champaign-Urbana. ![]()
2 Indiana University School of Medicine and Center for Aging Research, Indianapolis. ![]()
3 Wishard Health Services, Indianapolis, IN. ![]()
4 Department of Psychology, University of MissouriColumbia. ![]()
5 Regenstrief Institute, Indianapolis, IN. ![]()
6 Pharmaceutical Policy & Evaluative Sciences Division, University of North Carolina at Chapel Hill. ![]()
Decision Editor: David E. Biegel, PhD
Received for publication March 17, 2004. Accepted for publication November 1, 2004.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. Jansen, P. N. Butow, J. C.M. van Weert, S. van Dulmen, R. J. Devine, T. J. Heeren, J. M. Bensing, and M. H.N. Tattersall Does Age Really Matter? Recall of Information Presented to Newly Referred Patients With Cancer J. Clin. Oncol., November 20, 2008; 26(33): 5450 - 5457. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Jansen, J. van Weert, N. van der Meulen, S. van Dulmen, T. Heeren, and J. Bensing Recall in Older Cancer Patients: Measuring Memory for Medical Information Gerontologist, April 1, 2008; 48(2): 149 - 157. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. G. Morrow, M. Weiner, D. Steinley, J. Young, and M. D. Murray Patients' Health Literacy and Experience With Instructions: Influence Preferences for Heart Failure Medication Instructions J Aging Health, August 1, 2007; 19(4): 575 - 593. [Abstract] [PDF] |
||||
![]() |
M. D. Murray, J. Young, S. Hoke, W. Tu, M. Weiner, D. Morrow, K. T. Stroupe, J. Wu, D. Clark, F. Smith, et al. Pharmacist Intervention to Improve Medication Adherence in Heart Failure: A Randomized Trial Ann Intern Med, May 15, 2007; 146(10): 714 - 725. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. G. Katz, S. Kripalani, and B. D. Weiss Use of pictorial aids in medication instructions: A review of the literature Am. J. Health Syst. Pharm., December 1, 2006; 63(23): 2391 - 2397. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Morrow, D. Clark, W. Tu, J. Wu, M. Weiner, D. Steinley, and M. D. Murray Correlates of health literacy in patients with chronic heart failure. Gerontologist, October 1, 2006; 46(5): 669 - 676. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| All GSA journals | Journals of Gerontology Series A: Biological Sciences and Medical Sciences | Journals of Gerontology Series B: Psychological Sciences and Social Sciences |