Does Packaging with a Calendar Feature Improve Adherence to Self-Administered Medication for Long-Term Use? A Systematic Review

Does Packaging with a Calendar Feature Improve Adherence to Self-Administered Medication for Long-Term Use? A Systematic Review

Clinical Therapeutics/Volume 33, Number 1, 2011 Does Packaging with a Calendar Feature Improve Adherence to Self-Administered Medication for Long-Ter...

307KB Sizes 0 Downloads 6 Views

Clinical Therapeutics/Volume 33, Number 1, 2011

Does Packaging with a Calendar Feature Improve Adherence to Self-Administered Medication for Long-Term Use? A Systematic Review Barbara K. Zedler, MD1; Priyanka Kakad, MS2; Susan Colilla, PhD, MPH1,3; Lenn Murrelle, MSPH, PhD1; and Nirav R. Shah, MD, MPH4 1

Venebio, LLC, Richmond, Virginia; 2Department of Pharmacotherapy and Outcomes Science, School of Pharmacy, Virginia Commonwealth University, Richmond, Virginia; 3Integrative Epidemiology, LLC, Cherry Hill, New Jersey; and 4Division of General Internal Medicine, New York University School of Medicine, New York, New York, and Geisinger Center for Health Research, Danville, Pennsylvania ABSTRACT Background: The therapeutic benefit of self-administered medications for long-term use is limited by an average 50% nonadherence rate. Patient forgetfulness is a common factor in unintentional nonadherence. Unit-of-use packaging that incorporates a simple day or date feature (calendar packaging) is designed to improve adherence by prompting patients to maintain the prescribed dosing schedule. Objective: To review systematically, in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement, randomized controlled trial evidence of the adherence benefits and harms of calendar blister packaging (CBP) and calendar pill organizers (CPO) for self-administered, long-term medication use. Methods: Data sources included the MEDLINE and Web of Science and Cochrane Library databases from their inception to September 2010 and communication with researchers in the field. Key search terms included blister-calendar pack, blister pack, drug packaging, medication adherence, medication compliance, medication compliance devices, medication containers, medication organizers, multicompartment compliance aid, persistence, pill-box organizers, prescription refill, randomized controlled trials, and refill compliance. Selected studies had an English-language title; a randomized controlled design; medication packaged in CBP or CPO; a requirement of solid, oral medication self-administered daily for longer than 1 month in community-dwelling adults; and at least 1 quantitative outcome measure of adherence. Two reviewers extracted data independently on study design, sample size, type of intervention and control, and outcomes. Results: Ten trials with a total of 1045 subjects met the inclusion criteria, and 9 also examined clin-

62

ical outcomes (seizures, blood pressure, psychiatric symptoms) or health care resource utilization. Substantial heterogeneity among trials precluded metaanalysis. In 3 studies, calendar packaging was part of a multicomponent adherence intervention. Six of 10 trials reported higher adherence, but it was associated with clinically significant improvement in only 1 study: 50% decreased seizure frequency with a CPO-based, multicomponent intervention. No study reported sufficient information to examine conclusively potential harms related to calendar packaging. Limitations: All trials had significant methodological limitations, such as inadequate randomization or blinding, or reported insufficient information regarding enrolled subjects and attrition, which resulted in a moderate-to-high risk of bias and, in 2 studies, unevaluable outcome data. Trials were generally short and sample sizes small, with heterogeneous adherence outcome measures. Conclusions: Calendar packaging, especially in combination with education and other reminder strategies, may improve medication adherence. Methodological limitations preclude definitive conclusions about the effect size of adherence and clinical benefits or harms associated with CBP and CPO. High-quality trials of adequate size and duration are needed to assess the clinical effectiveness of such interventions. (Clin Ther. 2011;33:62–73) © 2011 Elsevier HS Journals, Inc. All rights reserved. Accepted for publication February 1, 2011. doi:10.1016/j.clinthera.2011.02.003 0149-2918/$ - see front matter © 2011 Elsevier HS Journals, Inc. All rights reserved.

Volume 33 Number 1

B.K. Zedler et al. Key words: calendar packaging, drug, medication adherence, medication error, systematic review.

BACKGROUND Medication taking is a complex, multidimensional behavior. The most common reasons given by patients for not taking their medications are forgetfulness (30%), other priorities (16%), deciding to omit a dose (11%), lack of information (9%), and emotional reasons (7%); 27% give no reason.1 Factors that are associated negatively with adherence include increased complexity or duration of a medication regimen, side effects, very old age, extreme poverty, social isolation, and psychiatric diagnoses, especially paranoia.2 The risk of nonadherence is especially high when multiple predisposing factors converge, such as cognitive impairment and the use of numerous medications for multiple chronic conditions in the elderly. The patient’s education, intelligence, and knowledge about their disease; race or ethnicity; (most) age and income levels; actual seriousness of the disease; and efficacy of the treatment do not predict levels of adherence consistently, whereas individual beliefs and perceptions about disease and treatment efficacy and social influences do.2–5 Complex interactions among determinants of this multifaceted behavior are likely. Adherence is defined as how well one takes a medication in relation to its prescribed dosing regimen (dose, interval, and duration). The International Society for Pharmacoeconomics and Outcomes Research (ISPOR)6 considers “adherence” and “compliance” to be synonyms, defined as the extent to which a patient acts in accordance with the prescribed interval and dose of a dosing regimen. Adherence to a medication’s prescribed dosing regimen is a key determinant of its effectiveness. The World Health Organization estimates that adherence to long-term therapy for chronic illnesses in developed countries averages 50%.7 Nonadherence to self-administered medicine (ie, underuse, overuse, or other failure to use as prescribed) represents the most common type of medication error studied in self-care settings.8 Medication nonadherence leads to unnecessary disease progression and complications, avoidable hospitalizations, premature disability, a lower quality of life, and even death.9 The use of low-cost interventions designed to improve adherence has resulted in significant increases in the effectiveness of health interventions and savings in health

January 2011

care utilization and overall health care costs.7 Improving adherence may be the most effective way to address chronic medical conditions.7,10 Innovative pharmaceutical packaging has the potential to leverage the therapeutic benefit of self-administered drugs substantially by reducing nonadherence. For example, medication packaging that incorporates a simple day or date calendar feature can prompt the patient to maintain the prescribed dosing schedule. Calendar pill organizers (CPO) are plastic trays or boxes with separate compartments labeled with the days of the week, which generally are filled by the patient from pharmacy-dispensed vials containing loose pills. Calendar blister packaging (CBP) is unit-of-use (generally 1 month) packaging in which each pill’s blister is labeled with the day of the week or date of the month, providing a visual record of when the patient last took their medicine. Some corticosteroids and antibiotics for the short-term treatment of limited conditions are packaged in CBP, and CBP has been used in the United States since the 1960s for oral contraceptives and, more recently, for female hormone replacement therapy.4,11 Generally, it has not been used for other classes of daily dosed medications for long-term use, except to package the initial titration phase of a few, such as memantine. Our objective was to perform a systematic review of the available scientific evidence from randomized controlled trials regarding the impact of calendar packaging (CBP, CPO) on adherence to self-administered medications for long-term use.

METHODS A systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement12 and guided by the Cochrane Handbook for Systematic Reviews of Interventions.13

Study Selection Criteria Studies included in the systematic review met the following criteria: (1) they had an English-language title; (2) they were randomized controlled trials of medication packaged in blister packaging or a pill organizer that incorporated a reminder system for the day of the week as part of the packaging; (3) they examined solid, oral medication prescribed and selfadministered daily for any duration longer than 1 month; (4) their populations included community-

63

Clinical Therapeutics dwelling adults aged 18 years or older; and (5) they included at least 1 objective (non–self-reported), quantitative outcome measure of medication adherence.

Search Methods and Sources We used validated search methods14 and the assistance of a professional librarian to identify clinical studies and review articles in 4 databases: Medline, Web of Science, the Cochrane Library, and Current Controlled Trials (http://controlled-trials.com/). Each electronic database was searched from its initial inclusion date to September 30, 2010. The main search terms included randomized controlled trial AND (medication OR drug) AND (compliance OR adherence OR persistence OR patient compliance OR medication adherence) AND (drug packaging OR pack$ OR compliance ADJ device/device$ OR medication/ medic$ ADJ organizer/container/box$ and refill compliance, prescription refill, multicompartment compliance aid, blister pack, and blister-calendar pack. We examined reference lists from retrieved reviews and sentinel trials recursively and communicated with researchers in the field to identify any additional trials.

Data Collection Two reviewers (BKZ and PK) independently reviewed all titles and studies identified in the search. The full text of the citation was retrieved for study reports with no abstract available and for studies with unclear methodology. Attempts were made to contact authors of published studies to provide or clarify key trial details or methods. Data items were dual-abstracted independently using a standardized evidence table, with discrepancies resolved by consensus. If consensus could not be achieved, the disagreement was resolved by discussion with a third reviewer (NS).

Assessment of Bias A Jadad score15 was calculated to assess the quality of each included study as reported. Ranging from 0 (very poor) to 5 points (rigorous), this score is a validated measure of the methodological quality and likelihood of unbiased results based on the reported description of randomization, double blinding, and subjects who withdrew or dropped out of the trial before completion.

Qualitative Data Synthesis Data synthesis was limited to a narrative summary of the included evidence. Quantitative data pooling

64

was not possible because of marked heterogeneity among the identified studies in patient populations, medical conditions, calendar packaging specifications, and measures of adherence and clinical outcomes. We were unable to perform subgroup or sensitivity analyses because of the small number of randomized trials that met all inclusion criteria and their general poor methodological quality, including inadequate reporting of key data elements and methodological features.

RESULTS Trial Flow and Study Selection The literature search resulted in a sample of 2145 titles (869 Medline, 569 Web of Science, 707 Cochrane). After excluding duplicates, we identified 1663 potential controlled trials regarding the impact of calendar packaging on medication adherence. We excluded 1585 after reviewing titles and abstracts and performed a full-text review of 78 articles. Sixty-eight retrieved studies were excluded for the following reasons (some studies had more than 1 reason): nonrandomized, controlled design (36); lack of calendar packaging (25); medication prescribed for less than 1 month (5); use of nonsolid, oral formulation (2); self-reported adherence only (2); and nonadult subjects (1). Ten studies met all inclusion criteria and provided data for analysis.

Study and Subject Characteristics Characteristics of the studies that were included in the systematic review are listed in Table I. Three of the 10 included studies evaluated CBP, 11,16,17 and 7 evaluated CPO.18 –24 All studies used a prospective, parallel, randomized trial design except for a trial of vitamin supplementation in CPO, which used a 2 ⫻ 2 factorial design.23 The sample size for analysis in the intervention arm ranged from 1324 to 89,23 and the duration of the packaging intervention ranged from 223 to 12 months.11,16,17,19,20,24 Patients were taking prescription medication for hypertension,11,17,20,21 type 2 diabetes,19 epilepsy,22 or serious mental illness,16,24 except for 1 trial of vitamin supplementation to prevent disease.23 One study enrolled elderly subjects with a variety of medical conditions and an episode of medication mismanagement,18 and 1 studied seriously mentally ill patients with a history of nonadherence.16 Two limited enrollment to elderly subjects.11,18

Volume 33 Number 1

January 2011

Table I. Methodology of studies included in the review, by condition treated. Intervention

Study*

Condition Treated

Study Medication

Study Duration (mo)

n

Medication Packaging

Control

n

Outcomes

Medication Packaging

Adherence

Clinical

Hypertension

Lisinopril

12

47

CBP, pharmacy-filled

38

Standard vial of loose pills

MPR; proportion with on-time refills (⫾5 d); self-reported adherence

DBP, SBP at 6, 12 mo; proportion with decreased DBP, SBP at 6, 12 mo; morbidity and health care resource utilization

Becker et al17 (1986) USA

Poorly controlled diastolic hypertension

Any antihypertensive

12

86

CBP

85

Standard vial of loose pills

Pill count; proportion with pill count ⱖ95%

DBP; proportion with DBP ⬍90 mm Hg

Skaer et al20 (1993) USA

Hypertension

Verapamil

12

85

CPO, pharmacy-filled

78

Standard vial of loose pills

MPR

Health care resource utilization (prescriptions, lab, MD visits, hospitalizations)

Rehder et al21 (1980) USA

Hypertension

Any antihypertensive

3

UNK

CPO, pharmacy-filled

UNK

Standard vial of loose pills

Pill count; proportion with pill count ⱖ95%

DBP

Valenstein et al16 (2009) USA

Serious mental illness and MPR ⬍80%

Antipsychotic and all general medications

12

54

CBP, pharmacy-filled, plus mailed refill reminders, phone call if miss refill, medication and packaging education

58

Standard vial of loose pills

MPR, 0–6 mo and 6–12 mo; composite adherence measure based on MPR ⱖ0.8, selfreported almost/always took antipsychotic, and detectable blood level

PANSS, QWB, CSQ-8

Azrin24 (1998) USA

Severe mental Illness

Psychotropics

12

13

CPO (“Mediset”) plus medication and adherence education

13

Standard vial of loose pills

Pill count; self-reported adherence

Psychiatric symptoms (SCL-90-R)

Peterson et al22 (1984) Australia

Epilepsy

Any anticonvulsant

6

26

CPO (“Dosett box”) plus mailed refill reminders, customized dosing schedule, nightly CPO checks by patient

24

Standard vial of loose pills

Proportion with on-time refills (⫾7 d); Proportion with subtherapeutic anticonvulsant plasma level

Seizure occurrence

Skaer et al19 (1993) USA

Type 2 diabetes

Glyburide

12

53

CPO

78

Standard vial of loose pills

MPR

Health care resource utilization (prescriptions, lab, MD visits, hospitalizations)

WinlandBrown18 (2000) USA

Any chronic condition with medication mismanagement

Any long-term medication

6

16

CPO, pharmacy-filled

24

Standard vial of loose pills

Pill count (missed doses)

Health care resource utilization (MD visits, hospitalizations, home health visits)

Huang et al23 (2000) USA

Disease prevention

Vitamins C and E

2

89

CPO, patient-filled

94

Standard vial of loose pills

Pill count; proportion with pill count ⬎90%; serum vitamin concentration; self-reported adherence

None

65

CBP, calendar blister package; CPO, calendar pill organizer; CSQ-8, Client Satisfaction Questionnaires; DBP, diastolic blood pressure; decr, decreased; MD, physician; MPR, medication possession ratio; PANSS, Positive and Negative Symptom Scales; QWB, Quality of Well-being Scales; SBP, systolic blood pressure; SCL-90-R, Symptoms Checklist 90-R; UNK, unknown. *Study by first author, publication year, and country where conducted.

B.K. Zedler et al.

Schneider et al11 (2008) USA

Clinical Therapeutics

Types of Calendar Packaging All CBP interventions were pharmacy-filled. Some CPO were filled by the pharmacy,11,18,20,21 some by the patient from vials of loose pills,23,24 and some studies did not identify the filler.19,22 Importantly, the calendar packaging was the sole adherence intervention employed in 7 trials11,17–21,23 but was combined with other adherence-promoting strategies in 1 CBP16 and 2 CPO22,24 studies. Other components of a multicomponent intervention included 1-time medication education,16,22,24 adherence education,22,24 and “packaging education”16 by the pharmacist, mailed refill reminders,16,22 phone or mail contact if refills were missed,16,22 and customized dosing schedules.22

Adherence Outcomes Eight of the 10 studies used indirect objective measures of adherence, including pill counts (to calculate the number of missed doses18 or the pills taken as a percentage of the total number prescribed and dispensed17, 21,23,24) and prescription refill rates (to calculate the proportion of on-time refills11,22 or the medication possession ratio [MPR; total days of medication supplied by the pharmacy as a percentage of the days in the study period16,19,20 or the days between the first and last fills11) (Table I). The other 2 studies assessed adherence more directly by measuring blood concentrations of vitamin supplements23 or anticonvulsants.22 Adherence outcomes of the 10 included studies are shown in Table II. Five of the 10 studies had significant methodological flaws in assessing adherence that limited the interpretation of the results,11,17,19,20,23 and the adherence data in 2 other studies18,21 were unevaluable because of substantial reporting gaps for outcome data and key study conduct details (Table III). Of the 8 evaluable studies, 6 reported an increase in adherence.

Clinical Outcomes Clinical outcomes were assessed in most studies and included blood pressure,11,17,21 seizures,22 psychiatric symptoms,16,24 well-being and satisfaction scales,16 and health care resource utilization11,18 and associated costs.19,20 Methodological flaws in several studies limited the interpretation of the clinical outcome,11,17,19,20,23 and substantial gaps in the reported information for 2 studies18,21 rendered their clinical data unevaluable (Table III). Of the 9 trials that as-

66

sessed a clinical outcome, only one22 reported improvement in a clinically important treatment outcome associated with improved adherence. This CPO-based, multicomponent intervention study of persons with epilepsy22 found that an increase in medication adherence was associated with a greater than 50% reduction in seizure occurrence over 6 months.

Harms Only 2 of the included studies reported systematically collecting study-related adverse event (AE) data.11,16 No trials reported, or appeared to solicit, packaging-related AEs except for comments in some reports regarding patient difficulty in opening the CBP, which was sometimes attributed to the required childproof feature. In summary, the 10 included studies did not provide evidence for the presence or absence of harm from CBP or CPO. The insufficient information about harm was owing to a failure to assess it systematically or to report it in a standardized fashion.25 However, the reported comments suggest (1) a need to specifically assess new packaging designs for usability and (2) the potential for toxicity resulting from increased medication adherence16 or dosing confusion.

Assessment of Risk of Bias in Included Studies Although the studies on which this review is based met prespecified inclusion criteria, almost all had important methodological flaws that limit the validity of the results. Table III summarizes methodological quality criteria from the included studies. The Jadad scores ranged from 0 to 3 on a 5-point scale, indicating very high to moderate risk of bias, respectively. Gaps were often due to insufficient information provided in the published study reports about key quality criteria, including randomization; blinding; numbers of subjects enrolled, withdrawn, and dropped out and reasons for early termination; and statistical handling of missing data, such as that caused by subject attrition. We were unable to contact authors of some included trials for relevant key data missing from publications. Intentionto-treat (ITT) analysis of adherence and clinical outcome data was clearly not performed in 7 studies,* and its use was unclear in three.18,21,23 Analysis in accordance with ITT principles, whereby all patients randomly assigned to a treatment are analyzed together, regardless of whether they received or completed that *References 11, 16, 17, 19, 20, 22, 24.

Volume 33 Number 1

B.K. Zedler et al. treatment, is critically important in trials of interventions for a complex behavior such as medication adherence. Use of ITT analysis prevents overestimating the potential benefits and underestimating the potential harms of an intervention, that is, subjects who are least adherent or experience an AE from the intervention may terminate early.

DISCUSSION Summary of the Evidence Calendar packaging of medications for long-term use is intuitively attractive as a simple adherence strategy, but only a paucity of clinical trials have assessed its adherence benefits systematically, and essentially none has evaluated its potential for harm. Six of the 10 studies11,16,19,20,22,24 reviewed here reported an increase in adherence. The improvement in adherence was associated with a clinically meaningful benefit in 1 study: reduced seizure frequency in an epilepsy trial that used a CPO-based, multicomponent intervention.22 Importantly, the 3 studies using multicomponent interventions16,22,24 reported improved adherence. This result is consistent with the concept of medication taking as a complex multidimensional behavior and previous reviews that found single interventions to be ineffective10,26 –28 In fact, 2 studies of hypertensives20 or diabetics19 using a CPO alone or a mailed refill reminder alone reported a nonsignificant trend to reduced total health care expenditures associated with an increase in medication adherence. However, combining both interventions was associated with a significantly greater improvement in adherence (MPR), which was reflected in 2-fold to 3-fold higher mean annual prescription costs per patient than for either intervention alone, plus a significant 2-fold to 4-fold reduction in mean physician, hospital and total health care expenditures. It has been argued that increased adherence alone is inadequate and that improvement in a clinically relevant health outcome must also be demonstrated to validate an intervention’s effectiveness.10,29 Calculation of the cost effectiveness of a medication adherence intervention, moreover, ultimately requires consideration of a number of direct and indirect factors. These include the actual cost of providing the intervention, the increase in professional dispensing fees due to increased prescription fills, and efficiencies in the drug distribution chain, such as the time savings and reduction in dispensing errors associated with unit-of-use packages30 that can enable a shift in pharmacist ser-

January 2011

vices from dispensing to greater patient interaction, including counseling and education. The results reported from the trials included in this review should be interpreted with caution, in light of their moderate to very high risk of bias due to methodological shortcomings. Adherence and clinical outcomes data were inevaluable, in fact, in 2 studies because of substantial gaps in the information published.18,21 Of concern are comments in some reports that suggest the potential for harm from calendar packaging (eg, difficulties using the packaging design, toxicity resulting from increased drug exposure, dosing confusion). In a recent analysis of Dutch patient reports regarding the practical aspects of using blister-packaged marketed medications, 46% of those who reported problems cited manual dexterity difficulties, confusion about opening the package, or confusion related to changes in the package’s appearance.31

Limitations of the Evidence The published literature on medication adherence interventions is a relatively small body that is not well indexed. Much of the published research on medication adherence interventions is of poor methodological quality with a high risk of bias.10,26 Some gaps may reflect inadequate reporting of key study elements and results rather than inherent design or conduct flaws. More recent studies were generally reported in a more robust fashion,11,16 and compliance with published global standards such as the CONSORT (Consolidated Standards of Reporting Trials) Statements25,32 will go far in facilitating the assessment of possible bias in individual trials. Additional study design gaps limited the validity of the results from the included studies. First, the adherence outcome measures were heterogeneous and often inadequately defined in the published reports. There is no gold standard for measuring adherence behavior. ISPOR has provided guidance to standardize the terminology and operationalize definitions for the purpose of conducting research on the effect of adherence interventions.6 Furthermore, a value for adherence outcome measures that universally, accurately, and reliably dichotomizes adherence or nonadherence does not exist. Evidence-based threshold values have been established for a few specific conditions, however, in which correlation with a health outcome has been documented. For example, use of at least 95% of highly active antiretroviral therapy for human immunodefi-

67

Study* Schneider et al11 (2008) USA

Condition Treated Hypertension

Adherence Outcomes MPR: 93% vs 87% at 12 mo Proportion with on-time refills: 80.4% vs 66.1% at 12 mo

Becker et al17 (1986) USA

Poorly controlled diastolic hypertension

Pill count ⬎80%: 84.0% vs 75.3% at 3 mo

Skaer, et al.20 (1993) USA

Hypertension

MPR: 67% vs 56% at 12 mo

Rehder et al21 (1980) USA

Hypertension

Pill count: unevaluable owing to large amount of missing data

Valenstein et al16 (2009) USA

Serious mental illness and MPR ⬍80%

MPR: 91% vs 64% at 6 mo 86% vs 62% at 12 mo Proportion with positive composite adherence measure†: 50% vs 17% at 6 mo 35% vs 18% at 12 mo

Volume 33 Number 1

Azrin24 (1998) USA

Severe mental illness

Peterson et al22 (1984) Australia

Epilepsy

Mean pill count: 92% vs 74% at 2 mo Proportion with on-time refills: 88% vs 50% at 6 mo Change in proportion with subtherapeutic anticonvulsant blood level baseline to 6 mo: 54% to 12% vs 45% to 52%

P 0.039 0.012

NS ⬍0.05

Clinical Outcomes Mean DBP: 73.2 vs 77.7 (10.2) at 6 mo 72.0 vs 75.2 at 12 mo Mean SBP: 132.7 vs 138.2 at 6 mo 130.9 vs 136.5 at 12 mo Proportion with decr DBP: 47% vs 37% at 6 mo 48% vs 18% at 12 mo Proportion with decr SBP: 49% vs 63% at 6 mo 46% vs 41% at 12 mo Decr in adjusted DBP over 12 mo “significantly greater” in CBP group Decr in adjusted SBP over 12 mo “consistently lower but NS” MD visits, ED visits, hospitalizations, angina, MI, stroke, serum creatinine Change in mean DBP baseline to 3 mo‡: –3.2 vs –2.8 mm Hg Adjusted mean annual prescription costs per person: incr $48.17 Adjusted mean annual total health care costs per person: decr $13.66

P

0.0367 NS 0.21 NS 0.39 0.0313 0.21 0.31 0.0104 NS NS NS

⬍0.05 NS

Unevaluable owing to insufficient data reported

⬍0.0001 ⬍0.0001

0.003 0.06

Mean total PANSS: 58.8 vs 61.7 at 6 mo 55.3 vs 57.1 at 12 mo Mean QWB: 0.09 vs 0.12 at 6 mo 0.060 vs 0.061 at 12 mo Mean CSQ-8: 4.3 vs 5.0 at 6 mo 5.7 vs 5.2 at 12 mo

0.29 0.46 0.27 0.24 0.96 0.51

⬍0.01

Psychiatric symptoms by SCL-90-R at 2 mo: data not reported

NS

⬍0.01

Median 6-mo seizure rate: decr 6 to 2.5 vs no change

⬍0.01

⬍0.005

Clinical Therapeutics

68 Table II. Results of studies included in the review.

January 2011 Table II (continued). Study*

Condition Treated

Adherence Outcomes

Skaer et al19 (1993) USA

Type 2 diabetes

Winland-Brown18 (2000) USA

Any chronic condition with medication mismanagement

Number of missed doses by pill count: insufficient data and details reported to calculate

Huang et al23 (2000) USA

Disease prevention

Median pill count at 2 mo: vitamin C: 100% vs 99% vitamin E: 100% vs 99% Pill count ⱖ90% at 2 mo: 91% vs 94% Change in mean serum vitamin level at 2 mo: vitamin C: ⫹3.6 vs ⫹3.9 mg/dL vitamin E: ⫹5.8 vs ⫹8.1 mg/dL

MPR: 71% vs 57% at 12 mo

P

⬍0.05

Clinical Outcomes Adjusted mean annual prescription costs per person: incr $48.17 Adjusted mean annual total health care costs per person: decr $13.66

P

⬍0.05 NS

Unevaluable owing to insufficient data reported

None 0.42 0.34 Not reported

0.47 0.06 for interaction term

CSQ-8 ⫽ Client Satisfaction Questionnaires; DBP ⫽ diastolic blood pressure; decr ⫽ decreased; ED ⫽ emergency department; incr ⫽ increased; MD ⫽ physician; MI ⫽ myocardial infarction; MPR ⫽ medication possession ratio; NS ⫽ not significant; PANSS ⫽ Positive and Negative Symptom Scales; QWB ⫽ Quality of Well-being Scales; SBP ⫽ systolic blood pressure; SCL-90-R ⫽ Symptoms Checklist 90-R. *Study by first author, publication year, and country where conducted. † Significant predictors of 6- and 12-month adherence by the composite adherence measure in a logistic regression analysis were medication packaging (smallest odds ratio, 12-month OR 5.41; 95% CI, 1.61, 18.12) and baseline MPR (12-month OR 2.51; 95% CI, 1.49, 4.21).16 ‡ The majority of the average 3 mm Hg improvement in DBP in both groups actually occurred before the packaging assignment and may reflect the nonspecific effects of increased attention paid to all study subjects before and after randomization (aggressive follow-up with visit reminders by mail and phone, free medication).17

B.K. Zedler et al.

69

Assessor Blind to Packaging Group Allocation†

Study*

Randomization

Allocation Concealment

Adherence Outcome

Clinical Outcome

Description of Dropouts and Withdrawals

Intention-to-Treat Analysis

Jadad Score‡

Other Limitations

Volume 33 Number 1

Schneider et al11 (2008)

By biostatistics department-generated randomization log

Not reported

No

Yes

Incomplete

No

3

No baseline assessment of adherence Small sample size Dropout rate: 22% CBP and 27% control groups

Becker et al17 (1986)

Method not stated

Not reported

Unclear

Yes

Incomplete

No

2

No baseline assessment of adherence Unintended adherence co-interventions used (eg, visit reminders) Cost-free medication limits generalizability Dropout rate: 9% overall.

Skaer et al20 (1993)

Method not stated

Not reported

Unclear

Unclear

No

No

1

No baseline assessment of adherence or clinical outcomes Small sample size Dropout rate: not reported Cost-free medication limits generalizability

Rehder et al21 (1980)

Method not stated

Not reported

No

No

No

Unclear

0

No baseline assessment of adherence or clinical outcomes Dropout rate: not reported High rate of missing data for adherence outcome Duration too short to determine adherence to long-term treatment

Valenstein et al16 (2009)

By central coordinating center

Not reported

No

No

Incomplete

No

3

CBP assessed as part of multicomponent intervention. Dropout rate: 22% CBP and 22% control groups

Azrin24 (1998)

Method not stated

Not reported

Unclear

Unclear

Incomplete

No

2

CPO assessed as part of multicomponent intervention Very small sample size Dropout rate: 8% CBP and 15% control groups

Peterson et al22 (1984)

By coin toss

Not reported

Unclear

Yes

Incomplete

No

3

CPO assessed as part of multicomponent intervention Small sample size Dropout rate: 4% CPO and 8% control groups

Skaer et al19 (1993)

Method not stated

Not reported

Unclear

Unclear

No

No

1

No baseline assessment of adherence or clinical outcomes Small sample size Dropout rate: not reported Cost-free medication limits generalizability

Clinical Therapeutics

70 Table III. Methodological quality of studies included in the review.

Clinical Therapeutics/Volume 33, Number 1, 2011

Does Packaging with a Calendar Feature Improve Adherence to Self-Administered Medication for Long-Term Use? A Systematic Review Barbara K. Zedler, MD1; Priyanka Kakad, MS2; Susan Colilla, PhD, MPH1,3; Lenn Murrelle, MSPH, PhD1; and Nirav R. Shah, MD, MPH4 1

Venebio, LLC, Richmond, Virginia; 2Department of Pharmacotherapy and Outcomes Science, School of Pharmacy, Virginia Commonwealth University, Richmond, Virginia; 3Integrative Epidemiology, LLC, Cherry Hill, New Jersey; and 4Division of General Internal Medicine, New York University School of Medicine, New York, New York, and Geisinger Center for Health Research, Danville, Pennsylvania ABSTRACT Background: The therapeutic benefit of self-administered medications for long-term use is limited by an average 50% nonadherence rate. Patient forgetfulness is a common factor in unintentional nonadherence. Unit-of-use packaging that incorporates a simple day or date feature (calendar packaging) is designed to improve adherence by prompting patients to maintain the prescribed dosing schedule. Objective: To review systematically, in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement, randomized controlled trial evidence of the adherence benefits and harms of calendar blister packaging (CBP) and calendar pill organizers (CPO) for self-administered, long-term medication use. Methods: Data sources included the MEDLINE and Web of Science and Cochrane Library databases from their inception to September 2010 and communication with researchers in the field. Key search terms included blister-calendar pack, blister pack, drug packaging, medication adherence, medication compliance, medication compliance devices, medication containers, medication organizers, multicompartment compliance aid, persistence, pill-box organizers, prescription refill, randomized controlled trials, and refill compliance. Selected studies had an English-language title; a randomized controlled design; medication packaged in CBP or CPO; a requirement of solid, oral medication self-administered daily for longer than 1 month in community-dwelling adults; and at least 1 quantitative outcome measure of adherence. Two reviewers extracted data independently on study design, sample size, type of intervention and control, and outcomes. Results: Ten trials with a total of 1045 subjects met the inclusion criteria, and 9 also examined clin-

62

ical outcomes (seizures, blood pressure, psychiatric symptoms) or health care resource utilization. Substantial heterogeneity among trials precluded metaanalysis. In 3 studies, calendar packaging was part of a multicomponent adherence intervention. Six of 10 trials reported higher adherence, but it was associated with clinically significant improvement in only 1 study: 50% decreased seizure frequency with a CPO-based, multicomponent intervention. No study reported sufficient information to examine conclusively potential harms related to calendar packaging. Limitations: All trials had significant methodological limitations, such as inadequate randomization or blinding, or reported insufficient information regarding enrolled subjects and attrition, which resulted in a moderate-to-high risk of bias and, in 2 studies, unevaluable outcome data. Trials were generally short and sample sizes small, with heterogeneous adherence outcome measures. Conclusions: Calendar packaging, especially in combination with education and other reminder strategies, may improve medication adherence. Methodological limitations preclude definitive conclusions about the effect size of adherence and clinical benefits or harms associated with CBP and CPO. High-quality trials of adequate size and duration are needed to assess the clinical effectiveness of such interventions. (Clin Ther. 2011;33:62–73) © 2011 Elsevier HS Journals, Inc. All rights reserved. Accepted for publication February 1, 2011. doi:10.1016/j.clinthera.2011.02.003 0149-2918/$ - see front matter © 2011 Elsevier HS Journals, Inc. All rights reserved.

Volume 33 Number 1

Clinical Therapeutics from changes in adherence, general and packagingspecific adverse events, usability,37,38 and patient-reported outcomes39 should be assessed over the study period after obtaining robust baseline outcome measurements. The duration of an adherence intervention for medications intended for long-term use should be at least 6 months to adequately capture initial adherence and subsequent persistence. Intermediate treatment outcomes, such as blood pressure, may serve as initial endpoints, but single or combination interventions that demonstrate a positive benefits-harms balance should ultimately be assessed for their impact on major morbidity endpoints, such as stroke and myocardial infarction, as well as health economics. The reporting of controlled clinical trials should conform to current standards.25 Assessing the effects of various daily dosing schedules and numbers of calendar-packaged daily medications would be useful. Medication packaging designs found to be adherence-enhancing in carefully designed and executed controlled clinical trials should then be studied in combination with other adherence interventions, such as education and behavioral strategies, to optimize the likelihood of having a significant impact on adherence and associated clinically meaningful health outcomes. Conversely, components of effective combination interventions can be assessed in factorial studies. Effect size should be calculated for comparison purposes. An adherence strategy of even small effect size that is broadly implemented on a population level could significantly leverage therapeutic effect and achieve significant public health benefit.

CONCLUSIONS Calendar packaging, especially in combination with education and other reminder strategies, may improve medication adherence. Methodological limitations preclude definitive conclusions about the effect size of adherence and clinical benefits or harms associated with CBP and CPO. High-quality trials of adequate size and duration are needed to assess the clinical effectiveness of such interventions.

ACKNOWLEDGMENTS We are grateful to J. Lyle Bootman, PhD, ScD; Joyce Cramer, BS; Gbenga Ogedegbe, MD, MPH; Spencer E. Harpe, PharmD, PhD, MPH; Richard Levy, PhD; and Patricia Slattum, PharmD, PhD, for their thoughtful review of and comments on this report. This work was

72

funded by MWV Healthcare, which had no role in the design or conduct of the systematic review or in the preparation, review, or approval of the manuscript. NS has received unrestricted research funding from AstraZeneca, Roche, and Pfizer. The authors have indicated that they have no other potential conflicts of interest with regard to the content of this article.

REFERENCES 1. Cramer J, Spilker B. Patient Compliance in Medical Practice and Clinical Trials. New York: Raven Press; 1991:414. 2. Haynes RB. A critical review of the determinants of compliance. In: Sackett DL, Haynes RB, eds. Compliance with Therapeutic Regimens. Baltimore, Md: Johns Hopkins University Press; 1976. 3. Steiner JF. Can we identify clinical predictors of medication adherence. . . and should we? Med Care. 2010;48:193–195. 4. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005;353:487– 497. 5. Steiner JF, Ho PM, Beaty BL, et al. Sociodemographic and clinical characteristics are not clinically useful predictors of refill adherence in patients with hypertension. Circ Cardiovasc Qual Outcomes. 2009;2:451– 457. 6. Cramer JA, Roy A, Burrell A, et al. Medication compliance and persistence: Terminology and definitions. Value Health. 2008;11:44 – 47. 7. Sabate E, ed. Adherence to Long-Term Therapies: Evidence for Action. Geneva: World Health Organization; 2003. 8. Aspden P, Wolcott JA, Bootman JL, Cronenwett LR, eds. Preventing Medication Errors. Washington, DC: National Academies Press; 2007. 9. National Council on Patient Information and Education. Enhancing prescription medicine adherence: A national actionplan.2007.http://www.talkaboutrx.org/documents/ enhancing_prescription_medicine_adherence.pdf. Accessed November 5, 2010. 10. Haynes RB, Ackloo E, Sahota N, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2008:CD000011. 11. Schneider PJ, Murphy JE, Pedersen CA. Impact of medication packaging on adherence and treatment outcomes in older ambulatory patients. J Am Pharm Assoc. 2008;48:58 – 63. 12. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Ann Intern Med. 2009;151:264 –269, W64. 13. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2 [updated September 2009] ed: The Cochrane Collaboration; 2009. Available from www.cochrane-handbook.org. 14. Dickersin K, Manheimer E, Wieland S, et al. Development of the Cochrane Collaboration’s CENTRAL Register of controlled clinical trials. Eval Health Prof. 2002;25:38 – 64.

Volume 33 Number 1

B.K. Zedler et al. 15. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: Is blinding necessary? Control Clin Trials. 1996;17:1–12. 16. Valenstein M, Kavanagh J, Lee T, et al. Using a pharmacy-based intervention to improve antipsychotic adherence among patients with serious mental illness. Schizophr Bull. 2009. Nov 21. [Epub ahead of print] 17. Becker LA, Glanz K, Sobel E, et al. A randomized trial of special packaging of antihypertensive medications. J Fam Pract. 1986;22:357–361. 18. Winland-Brown JE, Valiante J. Effectiveness of different medication management approaches on elders’ medication adherence. Outcomes Manag Nurs Pract. 2000;4:172–176. 19. Skaer TL, Sclar DA, Markowski DJ, et al. Effect of value-added utilities on prescription refill compliance and Medicaid health care expenditures—a study of patients with non–insulin-dependent diabetes mellitus. J Clin Pharm Ther. 1993;18: 295–299. 20. Skaer TL, Sclar DA, Markowski DJ, Won JK. Effect of value-added utilities on prescription refill compliance and health care expenditures for hypertension. J Hum Hypertens. 1993; 7:515–518. 21. Rehder TL, McCoy LK, Blackwell B, et al. Improving medication compliance by counseling and special prescription container. Am J Hosp Pharm. 1980;37:379 –385. 22. Peterson GM, McLean S, Millingen KS. A randomised trial of strategies to improve patient compliance with anticonvulsant therapy. Epilepsia. 1984;25:412– 417. 23. Huang HY, Maguire MG, Miller ER 3rd, Appel LJ. Impact of pill organizers and blister packs on adherence to pill taking in two vitamin supplementation trials. Am J Epidemiol. 2000;152:780 –787. 24. Azrin NH, Teichner G. Evaluation of an instructional program for improving medication compliance for

January 2011

25.

26.

27.

28.

29.

30.

31.

32.

chronically mentally ill outpatients. Behav Res Ther. 1998;36:849 – 861. Ioannidis JP, Evans SJ, Gotzsche PC, et al. Better reporting of harms in randomized trials: An extension of the CONSORT statement. Ann Intern Med. 2004;141:781–788. Heneghan CJ, Glasziou P, Perera R. Reminder packaging for improving adherence to self-administered longterm medications. Cochrane Database Syst Rev. 2006:CD005025. Schroeder K, Fahey T, Ebrahim S. How can we improve adherence to blood pressure-lowering medication in ambulatory care? Systematic review of randomized controlled trials. Arch Intern Med. 2004;164:722– 732. Roter DL, Hall JA, Merisca R, et al. Effectiveness of interventions to improve patient compliance: A metaanalysis. Med Care. 1998;36:1138 – 1161. McDonald HP, Garg AX, Haynes RB. Interventions to enhance patient adherence to medication prescriptions: Scientific review. JAMA. 2002;288:2868 –2879. Lipowski EE, Campbell DE, Brushwood DB, Wilson D. Time savings associated with dispensing unit-ofuse packages. J Am Pharm Assoc (Wash). 2002;42:577–581. van Geffen EC, Meuwese E, Philbert D, Bouvy ML. Problems with medicine packages: Experiences reported to a Dutch medicine reporting system. Ann Pharmacother. 2010;44: 1104 –1109. Schulz KF, Altman DG, Moher D; Consort Group. CONSORT 2010 statement: Updated guidelines for reporting parallel group randomized trials. Ann Intern Med. 2010; 152:726 –732.

33. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med. 2000;133:21–30. 34. Sackett DL, Haynes RB, Gibson ES, et al. Randomised clinical trial of strategies for improving medication compliance in primary hypertension. Lancet. 1975;1:1205–1207. 35. Wu JR, Moser DK, De Jong MJ, et al. Defining an evidence-based cutpoint for medication adherence in heart failure. Am Heart J. 2009;157: 285–291. 36. McCarney R, Warner J, Iliffe S, et al. The Hawthorne Effect: A randomised, controlled trial. BMC Med Res Methodol. 2007. doi: 1471-22887-30 [pii] 10.1186/1471-2288-730. 37. DeRosier J, Stalhandske E, Bagian JP, Nudell T. Using health care Failure Mode and Effect Analysis: The VA National Center for Patient Safety’s prospective risk analysis system. Jt Comm J Qual Improv. 2002;28:248 – 267. 38. US Food and Drug Administration. Guidance for Industry. Medical device use-safety: Incorporating human factors engineering into risk management. 2000. http://medical. cms.itri.org.tw/pdf/u09.pdf. Accessed October 28, 2010. 39. US Food and Drug Administration. Guidance for Industry: Patient-reported outcome measures: Use in medical product development to support labeling claims 2009. http://www.fda. gov/downloads/Drugs/Guidance ComplianceRegulatoryInformation/ Guidances/UCM193282.pdf. Accessed November 5, 2010.

Address correspondence to: Barbara K. Zedler, MD, Venebio, LLC, 7400 Beaufont Springs Drive, Suite 300, Richmond, VA 23225. E-mail: [email protected]

73