Problem Statement: Low adherence to medication regimens is common and contributes to reduced effectiveness of treatment, increased secondary health care costs, medication wastage, and increased antimicrobial resistance in communicable diseases. Poor adherence is multifactorial and difficult to address, but the complexity of the medication regimen has been identified as an important contributor, and is amenable to passive intervention. Objective: To quantify the advantages of fixed-dose combination pills and unit-of- use packaging compared with medications in their usual presentation, in terms of adherence to treatment and improved outcomes. Design: Systematic review of randomized and quasi-randomized controlled trials. Setting and Population: Trials were included if they involved adults taking more than one oral self-administered medication, an intervention consisting of a combination pill or unit-of-use packaging system compared with separate pills in usual containers, and at least one outcome measure relating to adherence, the pharmacological goal of medication, or the cost of therapy. Outcome Measures: 1. Adherence to the regimens under study; by pill counting, self-report, or biological markers. 2. Clinical or intermediate end-point relevant to condition being treated; blood pressure, HbA1c, HIV viral load, CD4+, TB sputum conversion. Results: Fourteen trials met the inclusion criteria, three using fixed-dose combination pills, and 11 with unit-of-use packaging. The trials involved treatments for communicable diseases (n=5), combinations of blood pressure–lowering medications (n=3), diabetic patients with multiple medications (n=1), and management of multiple medications by the elderly (n=5). There were trends to improved adherence and/or clinical outcomes in all but two of the trials, which reached statistical significance in 4 of 7 trials reporting a clinically relevant or intermediate endpoint, and 7 of 12 studies reporting medication adherence. However, outcome measures and settings were heterogeneous. Methodological issues, particularly trial size and length and losses to follow-up, further limited interpretation. Conclusions: Combination pills and unit-of-use packaging are likely to improve adherence in a range of clinical conditions and settings. However, almost all the trials have been too small or have other limitations; important uncertainty about the size of these benefits remains. Large simple trials are needed to provide quantitative information so that the cost-benefit of these interventions can be determined. Do fixed-dose pills or unit-of-use packaging improve adherence? A systematic review Connor J, Rafter NJ, Rodgers A. School of Population Health, University of Auckland

Poor adherence to medication… Is extremely common even in affluent countries (estimated at 50% for long term therapy) Leads to reduced effectiveness or treatment failure Increases secondary healthcare costs Contributes to antimicrobial resistance in communicable disease Is multifactorial: the complexity of the medication regimen is an important contributor Five interacting dimensions affecting adherence Social and economic factors Health care team/system factors Condition-related factors Therapy-related factors Patient-related factors (from “Adherence to long-term therapies: evidence for action” WHO, 2003)

FDCs and unit-of-use packaging Fixed dose combinations (FDC) combine two or more medications for the same indication in a single pill Unit-of-use packaging includes methods of pre- packaging multiple medications intended to be taken together, and includes blister packing and medication organisers, with or without calendar labeling Both reduce the complexity of medication dosing FDCs also reduce the number of pills taken Blister packaging allows for calendar labeling to prompt the user For example oral contraceptive pills –combined advantages of FDC (single pill) and blister calendar pack (reminder system). Also portability, preservation, storage benefits

Study aims To quantify the advantages of fixed dose combination pills and unit-of-use packaging, compared with medications in their usual presentation, in terms of adherence to treatment and improved outcomes

Methods A systematic review of randomised and quasi- randomised controlled trials Inclusion criteria: Adult participants taking more than one oral self-administered medication Intervention consisting of FDC or unit-of-use packaging Comparison group taking same medications in usual presentation (free combinations) At least one outcome measure related to adherence, the pharmacological goal of medication (e.g. BP, HIV viral load), or cost of therapy Search: electronic databases, bibliographies of relevant publications, websites of conferences and institutions involved in research, policy and regulation relating to pharmaceuticals Review: 2 independent reviewers judged eligibilty for inclusion, extracted data, and assessed methods No quantitative combination of findings (meta- analysis) due to substantial variation in settings, conditions studied, interventions, adherence and outcome measures, and study quality

Results: studies included 14 trials met inclusion criteria, conducted between 1980 and 2002 Interventions: 3 FDC, 11 packaging Conditions: Tuberculosis x 2, HIV, Leprosy, Malaria, Hypertension x 3, Diabetes x 1, Multiple medications in geriatric patients x 5 Outcome measures: 7 reported clinically relevant endpoints, 12 reported at least one adherence measure, one reported cost Adherence measures: pill counting, self-report, diary cards, patient medication adherence questionnaire (PMAQ), urine testing, composite measures

Slides 7 and 8 (landscape)

Of 12 studies measuring adherence, 7 reported a statistically significant improvement, but measures of adherence varied Four out of seven studies measuring clinical outcomes found a significant improvement: sputum conversion in TB, reduction in blood pressure x3, and improved diabetic control Study quality was generally poor: many trials were too small, too short to assess long term adherence, had high losses to follow-up, or did not report intention-to-treat analysis. Methods of measurement of adherence were usually unvalidated and sometimes unclear The two trials using the most robust methods both showed significant improvements with the intervention: Eron et al (2000) and Simmons et al (2000) Yeboah-Antwi et al (2001) showed that prepackaging of unit doses anti-malarial drugs not only improved adherence, but also reduced cost and waiting time by 50% Very few trials have been conducted in developing countries. Results: evidence of effect on adherence and treatment

Conclusions Combination pills and unit-of-use packaging are likely to improve adherence to medication and treatment outcomes in a range of clinical conditions and settings. However, direct evidence of the size of any benefit is weak as few trials have been carried out,and most have been too small or have other significant limitations. The small number of good quality trials show benefits, but important uncertainty remains.

Implications FDCs and co-blistering of medications have important applications in both communicable and non-communicable disease. In particular, the prevention of cardiovascular disease and the treatment of HIV, tuberculosis and malaria all employ multiple medications and are hampered by poor adherence to therapy Evidence of the size of adherence benefits and therefore cost-effectiveness of this approach is poor and large trials are required. Important questions remain about the most appropriate ways to measure adherence to medication The effect of using FDCs and co-blistering will be modified by many other factors including: –Appropriate training for health workers –Simple written/pictorial instructions for patients to accompany medication –Cost-effective and culturally appropriate modes of communication with families and communities

Research needs Trials Knowing the comparative benefits of different modes of delivery is important for allocative efficiency There is an ethical responsibility to evaluate effectiveness, adverse events, and cost-benefit when interventions are introduced Trials can be conducted in an appropriate context as interventions are being introduced. This opportunity is lost once the interventions are established. New trials must Be big enough for sufficient power Be long enough to determine long term effects Use reliable adherence and outcome measures Use intention to treat analysis New trials can Compare FDCs, co-blistering and free combinations, with no placebo arm Also measure patient safety and development of resistance in communicable disease