Clinical and economic consequences of non-adherence Mickaël Hiligsmann Maastricht University, CAPHRI Research Institute, the Netherlands, Department of Public Health Sciences, Belgium ESPACOMP 15 th Annual Meeting 2012 1
The problem Background Poor adherence is common in chronic conditions Drugs don't work in patients who don't take them Poor adherence clinical benefit of therapy has an impact on cost C. Everett Koop May therefore compromise the clinical and economic effects of drug therapies 2
The problem Poor adherence Step 4 Economic value of improving adherence Step 1 Clinical effects Step 2 Societal effects Step 3 Economic effects 3
The problem Why should we assess the economic value? Rising demand of health care Budget constraints Rapid development of medical technological possibilities Choices have to be made Efficiently allocate health care resources 4
Health economic evaluation - Background Annual rise in the number of published studies Increased use of economic data in decisions about the reimbursement or use of health technologies Formal use of economic evaluations in health care decisionmaking (e.g. drug reimbursement) The fourth hurdle: efficacy, safety, quality and costeffectiveness 5
Full economic evaluation «Comparative analysis between two or more health technologies in terms of costs and effects» Cost A Cost B Intervention A Intervention B Outcome A Outcome B Differences in costs? Differences in outcome? Relationship? 6
Incremental cost-effectiveness ratio ICER = (C A C B ) / (E A E B ) = C/ E = The additional cost per Quality-Adjusted Life-Years (QALY) gained from the comparator treatment Intervention adopted if ICER < λ (= willingness to pay per effectiveness unit) Commonly accepted thresholds varied between 20,000 and 80,000 per QALY gained 7
Plan de coût-efficacité Cost-effectiveness plane Cost difference Reject REJECT? Adoption B Effect difference ICER = slope of the line between the estimate and the origin Cost-effectiveness threshold value? ADOPTION 8
The problem The burden of osteoporosis in Europe 1 in 3 women and 1 in 5 men aged 50 years will have an osteoporotic fracture 22 millions of women and 5.6 millions of men have osteoporosis 3.5 millions new fractures every year (10,000 per day) Costs of osteoporosis: 37 billions 43,000 men and women death as a consequence of fractures IOF-EFPIA, Kanis, IOF-ECCEO Congress 2012 9
Compliance data Poor adherence with osteoporosis medications Proportion of patients with adequate adherence (MPR 80%) at 12 months for oral bisphosphonates Rabenda et al. Expert Opin Pharmacother, 2009;10:2303-15 10
The problem Poor adherence Step 4 Economic value of improving adherence Step 1 Clinical effects Step 2 Societal effects Step 3 Economic effects 11
Increased risk of osteoporotic fractures Increased risk of osteoporotic fractures Measurement Impact of non-adherence on osteoporotic fracture risk 1,5 Meta-analysis 113,376 patients +28% (18%-38%) 1,5 Meta-analysis 171,063 patients +46% (34%-60%) 1 1 0,5 0,5 0 Adherent (MPR>80%) Non-adherent (MPR<80%) 0 Adherent (MPR >80%) Non-adherent (MPR<80%) Ross et al. Value Health 2011;14:571-81 Imaz et al. Osteoporos Int 2010;21:1943-51 12
Increased risk of hip fractures Measurement Impact of non-adherence on hip fracture risk 1,5 +35% (17%-56%) 1 0,5 Belgian Social Security Database Osteoporotic women 45 Case-control study: 901 Hip Fx - 4505 controls Adjusted for age and duration of follow-up Daily or Weekly ALD 0 Adherent (MPR>80%) Non-adherent (MPR<80%) Rabenda et al. Osteoporos Int 2008;19:811-18 13
Increased risk of osteoporotic fractures Increased risk of hip fractures Measurement Impact of non-persistence on osteoporotic fracture risk 2,5 Meta-analysis 57,534 patients +40% (29%-52%) 2,5 Belgian women +148% (118%-180%) 2 2 1,5 1,5 1 1 0,5 0,5 0 Persistent Non-persistent 0 Persistent Non-persistent Ross et al. Value Health 2011;14:571-81 Rabenda et al. OI 2008;19:811-18 14
Healthy adherer effect = Adherence to drug treatment may be a surrogate marker for overall healthy behavior High adherence to placebo fracture risk by 50% (33%-78%) Curtis et al. Med Care 2011;49:427-35 No evidence of healthy adherer bias in a frail cohort of seniors Cadarette et al. Osteoporos Int 2011;22:943-54 Observational study limited role of healthy adherer effect Curtis et al. Arthritis Care Res 2012 [Epub Ahead of Print] 15
The problem Poor adherence Step 4 Economic value of improving adherence Step 1 Clinical effects Step 2 Societal effects Step 3 Economic effects 16
The problem Efficacy clinical effectiveness SIMULATION MODEL (e.g. Markov model) To estimate outcomes (fractures, Quality-Adjusted Life-Years) Scenarios: 1. No treatment 2. Real-world adherence 3. Full adherence 17
The problem Incorporating adherence in modeling PERSISTENCE At risk of discontinuation within 3 years (6, 12, 18, 24, 30 and 36 months) Refill gap period IMPLEMENTATION In the subgroup of persistent patients High implementation (MPR>=0.8) Low (MPR < 0.8) Lower efficacy for the low group Drug costs adjusted by mean MPR in the group RE-INITIATION RATES One year after stopping therapy 18
Incremental effectiveness per patient Measurement The societal burden of poor adherence in Ireland 0,05 0,04 Real-world adherence Full adherence 0,03 0,02-43% -44% Burden of persistence: + 90% Efficacy 0,01 Non-adherence 0 Number of fractures prevented QALY gained Clinical effectiveness Hiligsmann et al. Value Health 2012;15:604-12 19
The societal burden of poor adherence Belgium (1) Sweden (2) -61% -69% Adherence / persistence Refill gap period Treatment re-initiation (1) Hiligsmann et al. Health Policy 2010;96:170-77 (2) Landfeldt et al. Bone 2011;48:380-88 20
The problem Poor adherence Step 4 Economic value of improving adherence Step 1 Clinical effects Step 2 Societal effects Step 3 Economic effects 21
Impact of non-adherence on cost-effectiveness Impact on effectiveness Impact on costs Non-adherence Clinical effectiveness Impact on cost-effectiveness Non-adherence Therapy cost Fracture-related costs? Total healthcare costs 22
Incremental effectiveness per patient Costs, in Measurement Impact of poor adherence on effectiveness and costs 0,04 1500 0,03 0,02-43% -44% 1000 500 0 0,01-500 -1000 0 Number of fractures prevented Real-world adherence QALY gained Full adherence -1500 Therapy costs Disease costs Total costs Real-world adherence Full adherence Hiligsmann et al. Value Health 2012;15:604-12 23
Impact of poor adherence on cost-effectiveness IRELAND Budget of 20.000 Real-World: 1.68 QALYs Full Ad: 3.15 QALYs Hiligsmann et al. Value Health 2012;15:604-12 Cost-effectiveness plane. The incremental cost-effectiveness ratio is represented by the slope of the line from the origin 24
Incremental cost (in ) Measurement Impact of poor adherence on cost-effectiveness 150 /QALY 10,279 /QALY 3,909 BELGIUM 100 50 0 Real-world adherence Full adherence 0 0,01 0,02 0,03 0,04 Incremental effectiveness (in QALY) Budget of 20.000 Real-World: 1.95 QALYs Full Ad: 5.12 QALYs Hiligsmann et al. Health Policy 2010;96:170-77 25
Impact of medication adherence on the costeffectiveness of bisphosphonates vs. no treatment Hiligsmann et al. Calcif Tissue Int 2010;86:202-210 26
The problem Poor adherence Step 4 Economic value of improving adherence Step 1 Clinical effects Step 2 Societal effects Step 3 Economic effects 27
Adherence intervention and cost-effectiveness Adherence intervention Cost of adherence intervention Improved adherence More doses taken Fewer fractures More side effects Increased drug & medical costs Increased drug & medical costs Less QALYs Reduced fracture costs More QALYs Cost- Effectiveness Adapted from Strom O, ISPOR 2009 28
ICERs for adherence interventions vs usual care No specific interventions No studies have examined the feasibility and acceptability of a specific adherence-enhancing intervention Hypothetical interventions ICERs for a variety of hypothetical interventions Costs: marginal (e.g. monitoring) and one-time costs (e.g. education program) Effectiveness: improvements between 10% and 50% (adherence and/or persistence) 29
Measurement The economic value of improving medication adherence 150 Annual cost (in ) of adherenceenhancing intervention 100 50 0 140 85 50 WTP = 50,000 per QALY 0 25000 50000 75000 100000 Cost (in ) per QALY gained Cost (in ) per QALY gained of hypothetical adherence-enhancing interventions according to their cost and effect on adherence Adherence i mprovement of 10% Adherence i mprovement of 25% Adherence i mprovement of 50% Hiligsmann et al. Value Health 2012;15:604-12 30
Measurement The economic value of improving medication adherence United States A hypothetical intervention with a one-time cost of $250 reducing discontinuation by 30% ICER of $29,571 per QALY gained Sweden ICER threshold of 60,000 10% improvement 225 30% improvement 676 50% improvement 1130 Belgium ICER threshold of 45,000 10% improvement 73 25% improvement 149 50% improvement 239 Patrick et al. JCEM 2011 Landfledt et al. Bone 2011 Hiligsmann et al. HealthPolicy 2012 31
Sensitivity analyses on the effects of an intervention to improve economic outcomes Patrick et al. JCEM 2011;96:2762-70 32
Measurement Discussion Key findings - Approximately 50% of the benefits of osteoporosis medications are lost due to poor adherence and persistence - Poor adherence with osteoporosis medications results in a doubling of the cost per QALY gained from these medications - Programs to improve adherence have the potential to be an attractive approach to improve the allocation of resources 33
Discussion Economic - Adherence = important determinant of cost-effectiveness analyses Persistence and adherence should be an integral part of pharmacoeconomic analyses (1,2) - Lack of inclusion could bias the results and lead to suboptimal allocation of resources (3) - Importance of estimating the economic value of adherenceenhancing interventions (1) Hughes et al. Value in Health 2007:10:498-509 (2) Hiligsmann et al. Expert Review Pharmaco Out Res 2012;12:159-166 (3) Hiligsmann et al. Pharmacoeconomics 2011;29:895-911 34
Measurement Conclusion Non-adherence with medications clinical benefits of drugs economic value of drugs = critical hurdle to disease management Improving adherence is urgently needed but is a complex task The development of effective and cost-effective interventions to support adherence should be a priority for patients, healthcare providers and the pharmaceutical company 35
Thank you for your attention m.hiligsmann@maastrichtuniversity.nl 36