Determinants of Self-Management Strategies to Reduce Out-of-Pocket Prescription Medication Expense in Homebound Older people

Joseph R. Sharkey, PhD, MPH, RD, w Marcia G. Ory, PhD, MPH, w and Barry A. Browne, PharmD z

OBJECTIVES: To determine the extent to which homebound older people adopt strategies to reduce out-of-pocket prescription medication cost and the factors associated with level of cost-related medication management. DESIGN: Cross-sectional study. SETTING: Home-delivered meals programs in four North Carolina counties. PARTICIPANTS: Random sample of 222 home-delivered meal recipients aged 60 and older. MEASUREMENTS: The use of six different management strategies to reduce medication expenses was reported at the in-home assessment. Associations between level of costrelated medication management and sample characteristics, drug coverage, behaviors to cope with out-of-pocket medication expense, and payment difculty were examined. RESULTS: Forty-ve (20.3%) participants used one or more behaviors that restricted medication use; another 47 (21.2%) used one or more strategies to reduce out-of-pocket medication cost. Using medication restriction to reduce medication expense was more likely in older people who had difculty paying for medications (odds ratio (OR) 5 8.2, 95% condence interval (CI) 5 1.450.3), or used a strategy to cope with out-of-pocket expenses (choose food or medications (OR 5 5.1, 95% CI 5 1.715.7) or borrowed money or had another person pay for medications (OR 5 5.5, 95% CI 5 2.611.6)). Income, drug coverage, and medication use (prescribed and over-theFrom the Department of Social and Behavioral Health, wProgram on Health Promotion and Aging, School of Rural Public Health, Texas A&M University System Health Science Center, College Station, TX; and zScott and White Memorial Hospital, Temple, Texas. At the time of data collection, Dr. Sharkey was supported by a Doctoral Fellowship in Healthcare and Aging Research from the National Institute on Aging (1T32 AG00272-01 A1). The data analysis for this paper was supported by Small Grant Program, The Health Services Research Program, Texas A&M University System Health Science Center School of Rural Public Health, the Scott and White Hospital and Clinic, and the Scott and White Health Plan. Preliminary results were presented as an oral paper presented at the 25th Annual Meeting and Scientific Sessions of the Society of Behavioral Medicine in Baltimore, Maryland, March 2004. Address correspondence to Dr. Joseph R. Sharkey, Texas A&M Health Science Center School of Rural Public Health, 1103 University Drive, Suite 203, College Station, TX 77840. E-mail: sharkey@tamu.edu

counter) increased the likelihood of having increased difculty paying for medications. CONCLUSION: Clinicians should attempt to identify patients who are at risk for medication restriction and develop strategies for minimizing any unintended consequences of cost-related medication management behaviors. Providerpatient communication should include discussion of medication cost and appropriate medication management strategies. J Am Geriatr Soc 53:666674, 2005. Key words: homebound elderly; medication use; cost-related medication restriction

espite the importance of medication therapy for disease management,1 out-of-pocket prescription medication cost can be a significant barrier to medication adherence.2,3 For many elderly individuals, use of multiple prescribed and over-the-counter (OTC) medications, increasing costs for new and existing medications, concomitant lack of or limitations on drug coverage, and competing demands for limited resources (e.g., escalating prices for food, transportation, utilities, medical care, food) exacerbate this problem.48 The consequences of medication nonadherence in older persons are serious, with older adults at especially high risk for medical complications, increased medical care usage due to preventable adverse drug events, and unfavorable health outcomes.1,4,913 Characteristically, homebound older people report a large number of comorbid medical conditions that require prescription medications; additionally, these individuals must often make choices on allocation of limited resources.1416 This issue is of great concern for the growing number of homebound older people who live in the community, many of whom are women, members of minority groups, and poor or near-poor,17 but when considering nancial constraints and out-of-pocket costs for prescription medications, little is known of the strategies that homebound older people use to manage out-of-pocket cost or restrict medication use to reduce medication expense.1,18 Additionally, busy clinicians do not have the tools to predict

JAGS 53:666674, 2005 r 2005 by the American Geriatrics Society

0002-8614/05/$15.00

JAGS

APRIL 2005VOL. 53, NO. 4

MEDICATION EXPENSE STRATEGIES

667

who might be at risk and thus would need additional clinical intervention. The goal of the present study was to identify the factors associated with cost-related medication management (the extent to which homebound older people report the adoption of behavioral strategies to reduce prescription medication cost).

METHODS Subjects The Medication Management Study (n 5 222), a substudy of the North Carolina Nutrition and Function Study (NAFS) of homebound older persons, provided baseline data for this analysis. All participants took at least one prescribed medication each day and completed a comprehensive in-home assessment (December 2000 to May 2001). Details of the NAFS, conducted as a 2-year collaborative project between the School of Public Health at the University of North Carolina at Chapel Hill and Older Americans Act Nutrition Programs home-delivered meals service providers in four North Carolina counties, have been described elsewhere.19 Briefly, NAFS study participants were recruited by telephone from a probability sample of 1,866 home-delivered meal recipients. Subjects were considered eligible if they scored more than 16 of 22 possible points on the telephone version of the Mini-Mental State Examination20 and were able to answer questions without use of a proxy respondent. After supplemental research funding was obtained, the medication management component of NAFS was added to the protocol after 33% (n 5 116) had completed the in-home assessment. The Medication Management substudy subjects did not differ from the overall NAFS study population in terms of sex, race, age, income, education, living arrangement, medication use, or drug coverage. Informed consent was obtained from all participants, and the University of North Carolina at Chapel Hill School of Public Health institutional review board approved all study procedures. Data Collection A modied version of Andersens Behavioral ModelFMedication Cost Self-Management ModelFwas used to categorize many of the issues known to inuence cost-related medication adherence and served as a guide for data collection and analysis.21 This model suggests that preexisting population and medical characteristics (e.g., sociodemographics, health conditions, and prescribed and OTC medication use), enabling resources (drug insurance status), moderating behaviors (e.g., strategies used to cope with existing medication expense), and payment difculty (i.e., out-of-pocket costs for medications) may inuence the use of behavioral strategies used to reduce medication expense. Sample Population and Medical Characteristics Sample population and medical characteristics included self-reported sex, race (all participants were self-identied as white or black), age, monthly income, education (completedo9 years or 9 years), marital status (married or not married), living arrangement (lived alone or lived with others), comorbidity (self-report of 4 of 10 specific medical conditions), and prescribed and OTC medication use (including vitamin, mineral, and herbal supplements). Each

participant provided the medication containers for any prescribed and OTC medication currently being taken. Each medication container was visually inspected, and the name was transcribed directly from the container. Medication use was described in terms of a count of the number of different prescription medications. Because the distribution of total medications was highly skewed (range 131), the number of prescribed medications was divided into tertiles (0 5 14 different medications, 1 5 57, and 2 5 8). There were two variables for OTC medications: OTC medications (e.g., analgesics, nonsteroidal antiinammatory drugs, antacids, antibiotics, antidiarrheal, antiemetics, antihistamines, H2 receptor antagonists, laxatives, and sedatives); and supplements (vitamin, mineral, and herbal). The number of OTC medications ranged from 0 to 6 and was divided into tertiles (0, 1, and 2). The number of vitamin, mineral, and herbal supplements (VMHS) ranged from 0 to 8 and were also divided into tertiles (0, 1, and 2).

Enabling Resources (Drug Coverage) Using three mutually exclusive categories, prescription drug coverage at the time of the in-home assessment was described as having: no drug coverage, government drug coverage (included 85 participants with Medicaid coverage, 3 with indigent medication assistance coverage, and 2 who received Department of Veterans Affairs pharmacy benets), or supplemental drug coverage. The participants were also asked to report their monthly out-of-pocket expenses for prescription medications, which included all direct spending for prescription drugs, such as copayments, coinsurance amounts, and amounts not covered by an insurer.7 More than 96% (n 5 214) of the sample reported out-ofpocket prescription medication expenses, with median expense by drug coverage as: no drug coverage (median 5 $150), government coverage (median 5 $6), and supplemental drug coverage (median 5 $100). Moderating Behaviors and Payment Difculty Participants were queried about two strategies to cope with out-of-pocket medication expense: (1) if they borrowed money from another person for medicine or had a friend or family member pay for some or all of their medications and (2) if they experienced days when they had to choose between buying food and paying for medications.14 Difculty paying was based on how much perceived difculty the participant had paying for prescription medications (no difculty, little or some difculty, or great deal of difculty). Levels of Cost-Related Medication Management Cost-related medication management was operationalized from six behavioral questions that were previously used to measure strategies that individuals used to manage their medications to reduce out-of-pocket medication expense.18 Two behaviors related to restricting medication use were taking less medication than was prescribed: Do you take less medicine to make it last? and Do you go without medicine because of money? The other four behaviors had to do with medication resource decisions to reduce out-ofpocket cost: Do you buy part of a prescription instead of all of it? Do you take medicine only when you need it, such as when you have pain? Do you buy only the most important medicine? and Do you ask a physician for free

668

SHARKEY ET AL.

APRIL 2005VOL. 53, NO. 4

JAGS

samples of medication? Three mutually exclusive categories of cost-related medication management were dened: no cost-related medication management strategies (NCM), responded no to all six medication management behaviors; strategies to manage out-of-pocket medication cost (MCM), responded yes to any of the four medication resource decisions and no to both of the medication restriction behaviors; and cost-related medication restriction (CMR), responded yes to either medication restriction behavior.

additional items (age, being married, comorbidity, no drug coverage, and little or some payment difculty) that were significant correlates of CMR (if consideration was not given for other comparisons) lost statistical significance when the conservative Bonferroni correction was used. Median out-of-pocket medication expense varied among the participants, according to level of cost-related medication management: $25 for NCM, $50 for MCM, and $100 for CMR.

Statistical Analysis All statistical analyses were performed using Stata statistical software, Release 7 (Stata Corp., College Station, TX). Descriptive statistics for population- and disease-related characteristics, enabling resources, moderating behaviors, and payment difculty were calculated. Unadjusted multinomial logistic regression models examined the association (odds ratios (OR)) between each of the independent variables and levels of cost-related medication management. Because of a potential statistical problem with multiple testing, a conservative Bonferroni correction (a rejection region/number of tests to be conducted) was used to reduce the type 1 error rate for each individual test from 0.05 to 0.003.22 Adjusted binary and multinomial logistic regression models estimated the association (OR) between all independent variables in the model and moderating behaviors, payment difculty, and level of cost-related medication management. RESULTS Level of Cost-Related Medication Management Forty-ve (20.3% of the study sample) homebound participants reported CMR, using one or more behaviors that restricted medication use; 35 (15.8%) took less medicine to make the prescription amount last for a longer period of time; 31 (14%) went without medicine because of money; and 24 (10.8%) reported both behaviors. Sixty-eight (30.6%) participants asked a physician for free samples of medications; 51 (23%) bought partial prescription lls instead of the full prescription; 43 (19.4%) took medicine only when they felt the medicine was needed; and 39 (17.6%) bought only what they considered to be the most important medicine. Forty-seven participants (21.2%) reported MCM level of medication management (responded yes to one of the four medication resource decision behaviors and no to both medication restriction behaviors). The remaining 58% did not report employing any of the strategies to manage medication costs. Sample Characteristics Table 1 shows the sample prevalence of items from the Medication Cost Self-Management Model and the unadjusted odds for employing MCM or for CMR, each compared with using NCM. After statistically adjusting for multiple tests, one of the moderating behaviors (choose between food and drugs) and great payment difculty were associated with MCM, and both moderating behaviors and great payment difculty were associated with CMR. Five

Adjusted Correlates for Moderating Behaviors and Payment Difculty Table 2 shows variables, using results from multiple logistic regression models that controlled for other possible explanatory variables, that were significantly correlated with using moderating behaviors and having difculty paying for medications. Items associated with individuals who borrowed money or had family or a friend pay for medications included being female and younger (compared with individuals aged 85), no drug coverage or supplemental drug coverage (compared with those who had government coverage), and choosing between spending money for food and spending it on medications. Individuals who reported that they had to choose between food and medications were more likely to be participants who had a monthly income of $750 to $1,000 (middle category), used more prescription medications, or had to borrow money or have another person pay for their medications. Participants who experienced greater difculty paying for medications were individuals who reported income of $750 to $1,000, used eight or more prescription medications or two or more OTC medications, did not have government drug coverage (had no drug coverage or had supplemental drug coverage), or used either of the moderating behaviors to cope with out-of-pocket drug expense. Adjusted Correlates with Level of Cost-Related Medication Management Variables that increased the likelihood of homebound participants using MCM or CMR (Table 2) included younger age, comorbidity, borrowing money or having a family or friend pay for medications, having to choose between buying food or purchasing medications, and having great difculty paying for medications. DISCUSSION This study contributes to prior research by examining the use of behavior strategies to reduce medication expense in a vulnerable and growing homebound older population that is known to be difcult to recruit and retain in research studies.19 Using self-reported responses to six medicationmanagement behavioral questions,18 the current study provides a novel approach that estimates a three-level indicator of using cost-related medication management strategies: NCM, MCM, and CMR. This extends the use of a dichotomous measure of CRM,1 which may mask the use of strategies to manage cost by individuals who do not restrict medication use. More than 20% of the study sample used one or more behaviors that restricted medication use. This observation is of great concern, given that CMR (7% in a nationally

JAGS

Table 1. Unadjusted Odds Ratio of Level of Cost-Related Medication Management in 222 Homebound Elders by Items from the Medication Cost Self-Management Model
Strategies to Management Out-ofPocket Medication Cost Versus No Cost-Related Medication Management Strategies Odds Ratio (95% Condence Interval) Cost-Related Medication Restriction Versus No Cost-Related Medication Management Strategies

Total Sample n %

APRIL 2005VOL. 53, NO. 4

Item

178 44 122 100 75 85 62 139 44 39 73 149 46 176 131 91 59.0 41.0 20.7 79.3 32.9 67.1 62.6 19.8 17.6 1.50 (0.564.02) 2.45 (0.807.49) 1.0 0.74 (0.351.58) 1.0 1.49 (0.643.46) 1.0 0.70 (0.361.38) 1.0 33.8 38.3 27.9 2.63 (1.056.55) 2.09 (0.865.05) 1.0 54.9 45.0 1.0 1.10 (0.562.16)

80.2 19.8

1.57 (0.633.88) 1.0

1.10 (0.472.55) 1.0 1.0 1.56 (0.793.08) 4.13 (1.5910.76)w 2.15 (0.815.69) 1.0 1.38 (0.543.50) 1.29 (0.404.12) 1.0 1.74 (0.873.48) 1.0 3.03 (1.406.58)w 1.0 0.54 (0.271.07) 1.0

108 114 79 74 69

48.6 51.3 35.6 33.3 31.1

1.0 1.37 (0.702.67) 1.0 0.61 (0.261.42) 1.06 (0.482.32)

1.0 2.41 (1.184.89)w 1.0 2.12 (0.895.04) 2.12 (0.865.20)

Population and characteristics Sex Female Male Race White Black Age 6074 7584 85 Monthly income, $ o750 7501,000 41,000 Education completed, years o9 9 Marital status Married Not married Living arrangement Lived alone Lived with others Medical characteristics Comorbidity, number of medical conditions o4 4 Prescription medications 14 57 8

MEDICATION EXPENSE STRATEGIES 669

670

Table 1. (Continued) Strategies to Management Out-ofPocket Medication Cost Versus No Cost-Related Medication Management Strategies Odds Ratio (95% Condence Interval) 1.0 1.44 (0.613.42) 0.89 (0.411.93) 1.0 0.65 (0.271.56) 1.15 (0.532.50) 1.0 0.94 (0.342.62) 1.53 (0.713.28) 1.0 0.95 (0.422.16) 1.01 (0.442.30) Cost-Related Medication Restriction Versus No Cost-Related Medication Management Strategies

SHARKEY ET AL.

Total Sample n 80 46 96 104 57 61 46.8 25.7 27.5 36.0 20.7 43.2 %

Item

69 90 63 56 66 115 64 43 51.8 28.8 19.4 25.2 29.7 2.72 (1.196.22)w 3.31 (1.537.14)z 1.0 1.44 (0.673.13) 4.67 (1.7212.66)z

31.1 40.5 28.4

1.31 (0.602.87) 1.0 0.48 (0.201.14)

2.96 (1.336.58)w 1.0 0.67 (0.261.74) 10.69 (4.8323.63)z 11.68 (5.3125.69)z 1.0 3.28 (1.298.34)w 23.85 (8.4966.99)z

Over-the-counter medications 0 1 2 Vitamin/mineral/herbal supplements 0 1 2 Enabling resources Prescription drug coverage None Government Supplemental Moderating behaviors Borrow money or have family/friend pay for medications Choose between food and drugs Payment difculty No difculty Little/some difculty Great deal of difculty

P  .05; w.01. z Statistically significant after Bonferroni adjustment for multiple comparisons (P  .003).

APRIL 2005VOL. 53, NO. 4 JAGS

JAGS

Table 2. Adjusted Odds Ratio of Using Moderating Behaviors, Having Payment Difculty, and Level of Cost-Related Medication Management in 222 Homebound Older Persons
Moderating Behaviors Strategies to Manage Out-ofPocket Medication Cost Versus No Cost-Related Medication Management Strategies Cost-Related Medication Management Behavior Strategies Cost-Related Medication Restriction Versus No CostRelated Medication Management Strategies

Borrowed Money or Had Family/Friend Pay for Medications Had to Choose Between Food and Medications Increased Difculty Paying for Medications Adjusted Odds Ratio (95% Condence Interval) 1.65 (0.564.84) 1.74 (0.763.99) 1.54 (0.643.71) 0.84 (0.381.88) 2.38 (0.737.74) 0.58 (0.241.41)

APRIL 2005VOL. 53, NO. 4

Variable

3.74 (1.1612.06) 2.38 (0.975.79)

0.65 (0.182.31) 0.54 (0.181.64)

3.95 (1.3311.71) 3.25 (1.208.83) 1.09 (0.412.91) 0.50 (0.191.30) 1.97 (0.784.98) 3.58 (1.508.54)

1.96 (0.636.08) 2.38 (0.846.72)

5.99w (1.3726.23) 2.86 (0.6811.91)

1.58 (0.475.30) 0.55 (0.152.02)

1.31 (0.335.22) 3.95 (1.0514.89)

1.91 (0.705.27) 5.13 (1.7814.77)

1.05 (0.264.17) 1.77 (0.437.22)

0.58 (0.122.78) 0.32 (0.052.08)

1.52 (0.683.38)

1.50 (0.663.40)

1.14 (0.542.41)

0.66 (0.271.63)

1.69 (0.594.84)

0.91 (0.302.81)

1.53 (0.514.54)

2.51 (0.966.55)

2.44 (0.708.43)

3.19 (0.7613.41)

1.17 (0.482.83)

0.60 (0.251.43)

0.80 (0.361.77)

0.94 (0.382.35)

1.51 (0.455.04)

1.11 (0.472.62)

0.91 (0.412.02)

1.18 (0.572.44)

1.42 (0.593.39)

3.46 (1.1010.85)

Female (male 5 reference) Black (white 5 reference) Age 85 (reference) 6074 7584 Monthly income, $ 1,000 (reference) o750 7501,000 Education Completed 9 years (reference) o9 years Marital status Not married (reference) Married Living arrangement Lived with others (reference Lived alone Comorbidity o4 medical conditions (reference) 4 medical conditions Prescription medication use, number of drugs 14 (reference) 57 8 2.90 (1.137.42) 6.33 (2.2417.87) 1.11 (0.482.55) 3.14 (1.257.93) 0.43 (0.161.15) 0.50 (0.171.51)

MEDICATION EXPENSE STRATEGIES

0.72 (0.281.81) 0.52 (0.171.57)

1.11 (0.333.68) 0.33 (0.071.45)

671

672

Table 2. (Continued)
SHARKEY ET AL.

Moderating Behaviors Strategies to Manage Out-ofPocket Medication Cost Versus No Cost-Related Medication Management Strategies

Cost-Related Medication Management Behavior Strategies Cost-Related Medication Restriction Versus No CostRelated Medication Management Strategies

Borrowed Money or Had Family/Friend Pay for Medications Had to Choose Between Food and Medications Increased Difculty Paying for Medications Adjusted Odds Ratio (95% Condence Interval)

Variable

2.22 (0.816.08) 1.29 (0.553.02)

0.35 (0.111.11) 1.17 (0.532.59)

1.53 (0.603.89) 2.99 (1.386.50)

1.69 (0.614.67) 0.73 (0.291.85)

1.40 (0.355.66) 0.87 (0.292.66)

1.58 (0.673.77) 1.77 (0.694.52)

1.17 (0.482.84) 1.38 (0.553.45)

0.95 (0.422.18) 1.05 (0.472.32)

0.60 (0.221.63) 1.40 (0.553.54)

1.01 (0.323.17) 1.08 (0.333.52)

4.55 (1.6712.40) 5.43 (1.4919.78) 2.04 (0.785.35) 0.67 (0.192.37) 4.92 (2.1711.15) 5.46 (2.4712.09)

21.21 (7.9156.83) 5.58 (1.9216.23) 5.63 (2.5712.31)

0.63 (0.182.23) 0.22 (0.050.88) 1.99 (0.715.58) 2.98 (1.127.89)

0.82 (0.183.78) 0.32 (0.061.80) 5.91 (2.0017.47) 5.12 (1.6615.75)

Number of over-the-counter medications 0 (reference) 1 2 Vitamin, mineral, and herbal supplements 0 (reference) 1 2 Prescription drug coverage Government (reference) None Supplemental Borrow money or have friend/ family pay No (reference) Choose food or medications [not have to choose] Difculty paying None (reference) Little/some Great deal 1.32 (0.496.54) 2.04 (0.439.72) w2 5 52.7 df 5 19 Po.001 0.21 w2 5 67.8 df 5 19 Po.001 0.25 w2 5 171.0 df 5 20 Po.001 0.38

5.27 (2.3311.89)

2.47 (0.679.11) 8.26 (1.3550.32) w2 5 115.2 df 5 44 Po.001 0.27

Likelihood ratio test for model

APRIL 2005VOL. 53, NO. 4

Pseudo coefcient of determination of model

P.05; w.01; z.001. w2 5 chi-square; df 5 degrees of freedom.

JAGS

JAGS

APRIL 2005VOL. 53, NO. 4

MEDICATION EXPENSE STRATEGIES

673

representative sample) in older people has been shown to increase the risk for subsequent decline in health.1 These ndings show that homebound older people who have coexisting medical conditions, or who are using strategies to cope with existing medication expense (moderating behaviors) or experience great difculty paying medication cost, restrict prescription medications because of cost. The study did not nd that prescription or OTC medication use or prescription drug coverage was directly associated with medication restriction, but these variables were associated with intervening variablesFmoderating behaviors or payment difcultyFthat preceded MCM or CMR. For example, participants who reported great difculty paying for medications were individuals who used a greater number of prescription and OTC medications, reported a monthly income of $750 to $1,000, or had no drug coverage or supplemental drug coverage. These variables (monthly income, prescription medication use, and drug coverage) are important because, in addition to payment difculty, individuals who reported any of these characteristics also reported using moderating behaviors (strategies to cope with medication expense). This observation suggests greater vulnerability in individuals who reported middle level of income; few in this income category would qualify for government drug coverage, and even fewer would be able to afford the premiums for supplemental drug coverage.23 Thus, older people in this income group would reasonably report a high out-of-pocket medication expense that may constitute a greater proportion of monthly income. These data suggest that perceived difculty paying for medications may be a critical topic for communication between healthcare providers and patients and a target for interventions that address the number of medications used, out-of-pocket drug expense, or competing needs. The least likely in this sample to directly experience great difculty paying for medications and hence indirectly resorting to strategies to reduce medication expense were older persons who had government drug coverage and subsequently the lowest out-of-pocket expenses. This observation reects the effect of government coverage in the lowestincome group. For this group, the effect of any reduced government coverage is potentially of great concern, given several factors: the serious difculty that would occur if coverage were discontinued, the potential negative effects of restrictions on copayments and coverage limits,4,5 and the fact that these individuals have the lowest level of nancial resources. These factors suggest that even modest changes to government drug coverage programsFmuch less program discontinuationFcould result in increased use of medication restriction behaviors and place additional older people at great risk of adverse outcomes from medication nonadherence. Although this study of homebound older people has a number of strengths, several limitations should be acknowledged. The measure of medication expense-reduction strategies did not identify which drugs or how often restricting behaviors occurred.24 An additional limitation was lack of information on the inuence of other factors on management strategies, such as perceived severity of symptoms, effectiveness of treatments, perceived benet of medication, medication side effects, or needs of others (e.g., children).25 Finally,

additional limitations include the lack of information regarding scheduled versus as-needed medications and maintenance/consistency of drug coverage throughout the year. In conclusion, economic factors play an important role in medication decisions.26 Given the cost burden of pharmaceuticals on elderly persons with chronic diseases,27 this study extends knowledge by identifying the vulnerability of homebound older persons to adopting strategies to manage out-of-pocket medication costs or to restrict medication use because of cost. Given the strong association between costrelated medication restriction and worsened health outcomes,1 clinicians should attempt to identify patients who are at risk and develop strategies for minimizing need forFor consequences ofFcost-related medication-management behaviors. Information on medication cost and appropriate medication management strategies needs to be included in provider-patient communications.1,2,24 Such an approach will aid the clinician and supporting care team to determine which medication regimens carry the lowest possible out-of-pocket cost without compromising treatment effectiveness. As part of this dialogue, patient education within the medical encounter as well as referral to other patient support sources needs to include understandable information on the importance of each prescribed medication, information on sources of low-cost drugs, and linkage with coverage programs. Although many intervention strategies manage patient-directed reminders,28 disease management and targeted interventions must take cost-related factors affecting medication adherence, including coping strategies, into account.29 Although these recommendations may result in additional clinical time in the already-pressed medical encounter, the importance of restructuring clinical interactions and referral patterns to attend to factors that affect barriers to medication adherence and preventable adverse events and medical complications cannot be neglected.

ACKNOWLEDGMENTS The authors acknowledge the hard work and dedication to this project by our community partners, who actively participated in the study as in-home coordinators: Amy Walls, RN, and the Chatham County Council on Aging, Turquious Byrd and Senior Resources of Guilford/Mobile Meals, Minnie McBurnett and the Johnston County Council on Aging, and Viki Baker and Meals on Wheels of Wake County, Inc. REFERENCES
1. Heisler M, Langa KM, Eby EL et al. The health effects of restricting prescription medication use because of cost. Med Care 2004;42:626634. 2. Salzman C. Medication compliance in the elderly. J Clin Psychiatry 1995;56:1822. 3. DiMatteo MR. Patient adherence to pharmacotherapy. The importance of effective communication. Formulary 1995;30:596605. 4. Piette JD, Wagner TH, Potter MB et al. Health insurance status, cost-related medication underuse, and outcomes among diabetes patients in three systems of care. Med Care 2004;42:102109. 5. Fortess EE, Soumerai SB, McLaughlin TJ et al. Utilization of essential medications by vulnerable older people after a drug benet cap: Importance of mental disorders, chronic pain, and practice setting. J Am Geriatr Soc 2001;49:793797. 6. Kennedy J, Erb C. Prescription noncompliance due to cost among adults with disabilities in the United States. Am J Public Health 2002;92:11201124. 7. Kreling DH, Mott DA, Wiederholt JB et al. Prescription Drug Trends: A Chartbook Update, 2001. The Kaiser Family Foundation [on-line]. Available

674

SHARKEY ET AL.

APRIL 2005VOL. 53, NO. 4

JAGS

8. 9.

10. 11.

12.

13.

14.

15.

16. 17.

at www.kff.org/rxdrugs/loader.cfm?url=/commonspot/security/getle.cfm& PageID=13796 Accessed April 24, 2004. Hwang W, Weller W, Ireys H et al. Out-of-pocket medical spending for care of chronic conditions. Health Aff 2001;20:267278. Gurwitz JH, Field TS, Harrold LR et al. Incidence and preventability of adverse drug events among older persons in the ambulatory setting. JAMA 2003;289:11071116. Neal WW. Reducing costs and improving compliance. Am J Cardiol 1989; 63:17B20B. Col N, Fanale JE, Kronholm P. The role of medication noncompliance and adverse drug reactions in hospitalizations of the elderly. Arch Intern Med 1990;150:841845. Malhotra S, Karan R, Pandhi P et al. Drug related medical emergencies in the elderly: Role of adverse drug reactions and non-compliance. Postgrad Med J 2001;77:703707. Kuo Y-F, Raji MA, Markides KS et al. Inconsistent use of diabetes medications, diabetes complications, and mortality in older Mexican Americans over a 7-year period. Diabetes Care 2003;26:30543060. Sharkey JR. Risk and presence of food insufciency are associated with low nutrient intakes and multimorbidity among homebound older women who receive home-delivered meals. J Nutr 2003;133:34853491. Millen BE, Silliman RA, Cantey-Kiser J et al. Nutritional risk in an urban homebound older population. The nutrition and healthy aging project. J Nutr Health Aging 2001;5:269277. Golden AG, Preston RA, Barnett SD et al. Inappropriate medication prescribing in homebound older adults. J Am Geriatr Soc 1999;47:948953. Ponza M, Ohls JC, Millen BE. Serving Elders at Risk: The Older Americans Act Nutrition Programs. National Evaluation of the Elderly Nutrition Program. 199395. Princeton, NJ: Mathematica Policy Research, Inc., 1996.

18. Mitchell J, Mathews HF, Hunt LM et al. Mismanaging prescription medications among rural elders: The effects of socioeconomic status, health status, and medical prole indicators. Gerontologist 2001;41:348356. 19. Sharkey JR. Responding to the challenges of recruiting and retaining homebound older men and women in a study of nutrition and function: A universitycommunity collaborative strategy. J Nutr Elder 2002;22:1536. 20. Roccaforte WH, Burke WJ, Bayer BL et al. Validation of a telephone version of the Mini-Mental State Examination. J Am Geriatr Soc 1992;40:697702. 21. Andersen RM. Revisiting the behavioral model and access to medical care: Does it matter? J Health Soc Behav 1995;36:110. 22. Kleinbaum DG, Kupper LW, Muller KE et al. Applied Regression Analysis and Other Multivariate Models. Pacific Grove, CA: Duxbury Press, 1998. 23. Poisal JA, Chulis GS. Medicare beneciaries and drug coverage. Health Aff 2000;19:248256. 24. Piette JD, Heisler M, Wagner TH. Problems paying out-of-pocket medication costs among older adults with diabetes. Diabetes Care 2004;27: 384391. 25. Schafheutle EI, Hassell K, Noyce PR et al. Access to medicines: Cost as an inuence on the views and behaviour of patients. Health Soc Care Community 2002;10:187195. 26. Stuart B, Grana J. Ability to pay and the decision to medicate. Med Care 1998;36:202211. 27. Steinberg EP, Gutierrez B, Momani A et al. Beyond survey data: A claimsbased analysis of drug use and spending by the elderly. Health Aff 2000; 19:198211. 28. Bennett JW, Glasziou PP. Computerised reminders and feedback in medication management: A systematic review of randomised controlled trials. Med J Aust 2003;178:217222. 29. Balkrishnan R. Predictors of medication adherence in the elderly. Clin Ther 1998;20:764771.