Medication adherence in America: the $300 billion problem hiding in plain sight¶
Half of all Americans with chronic conditions do not take their medications as prescribed, a figure that has remained stubbornly unchanged since the WHO first documented it in 2003. This non-adherence costs the U.S. healthcare system an estimated $100–$300 billion annually, contributes to roughly 125,000 preventable deaths per year, and drives up to 25% of all hospitalizations. For seniors on Medicare — who fill more prescriptions, see more providers, and face more complex regimens than any other group — the problem is especially acute. CMS has made medication adherence a cornerstone of its Star Ratings system, tying billions in bonus payments to plan performance on adherence measures. Yet national average PDC scores still hover around 86–89%, and only 7 out of 521 Medicare Advantage contracts earned 5-star ratings in 2025. The gap between current performance and optimal adherence represents both a massive public health challenge and a significant market opportunity for technology-based solutions — particularly wearables targeting seniors, a category where no dominant product yet exists.
Half of chronic disease patients never follow through¶
The foundational statistic has not changed: approximately 50% of patients with chronic conditions do not take medications as prescribed (WHO, 2003; Osterberg & Blaschke, NEJM, 2005; CDC MMWR, 2017). This figure holds across measurement methods, countries, and decades. Disease-specific rates vary but paint a consistently troubling picture.
Diabetes medication adherence is particularly uneven. Meta-analyses show only 67.9% of patients on oral antidiabetics achieve a medication possession ratio (MPR) above 80%. Adherence to specific drug classes ranges from 36.7% (thiazolidinediones) to 47.3% (DPP-4 inhibitors) in large claims databases. Insulin adherence hovers around 62–64% for established users, with self-reported non-adherence at roughly 29%. Primary non-adherence — never filling the first prescription — runs as high as 31.4% for diabetes medications (Fischer et al., Journal of General Internal Medicine, 2010).
Hypertension non-adherence ranges from 27% to 45% depending on measurement method. A landmark 2022 meta-analysis in the Journal of the American Heart Association (Lee et al.) covering 27 million patients across 161 studies found non-adherence rates of 27% by biochemical assay, 28% by refill records, and 40% by patient questionnaire. Roughly 50% of patients discontinue antihypertensive therapy within the first year.
Statin therapy sees similar attrition: about 50% of patients discontinue within one year of initiation. High adherence (PDC ≥ 80%) was found in only 35–38% of primary prevention patients and 58–64% of post-heart-attack patients in an AHA study of U.S. data from 2007–2014 (Muntner et al., JAHA, 2018).
Heart failure medications fare worst. One 2023 study of 427 heart failure patients found 92.5% had low-to-moderate adherence. Long-term adherence to ACE inhibitors and beta-blockers hovers around 50–60%.
Prescriptions that never get filled represent a distinct and underappreciated category. A meta-analysis of 33 studies covering 539,156 patients (Cheen et al., 2019) found an overall primary non-adherence rate of 17%, with rates highest for osteoporosis and dyslipidemia medications (25% each). At the pharmacy counter, IQVIA reported that 98 million new therapy prescriptions were abandoned in 2023 across the U.S. — and 55% of prescriptions costing over $250 out-of-pocket are never picked up.
The staggering economic toll of not taking pills¶
The most frequently cited cost figure comes from the New England Healthcare Institute (NEHI, 2009), which estimated $289 billion in annual avoidable medical spending from non-adherence and related medication problems. The CDC and FDA have subsequently referenced a $100–$300 billion range. A critical distinction must be made with the Watanabe et al. (2018) meta-analysis in the Annals of Pharmacotherapy, which estimated $528.4 billion (2016 dollars) in costs from prescription drug-related morbidity and mortality. This broader figure encompasses all drug-related problems — adverse reactions, dosing errors, drug interactions, and non-adherence — not non-adherence alone.
Per-patient costs are substantial. Cutler et al. (2018, BMJ Open) conducted a systematic review of 79 studies across 14 disease groups and found disease-specific annual costs of non-adherence ranging from $949 to $44,190 per person, with all-cause non-adherence costs between $5,271 and $52,341 per person per year.
The human cost is equally stark. An estimated 125,000 deaths annually in the U.S. are attributed to medication non-adherence — a figure consistently cited across the literature since Bosworth et al. (2011, American Heart Journal) and reinforced by the CDC, though the original estimate dates to 1989. Between 33% and 69% of all medication-related hospital admissions are caused by poor adherence (Osterberg & Blaschke, NEJM, 2005), and non-adherence accounts for an estimated 10–25% of all U.S. hospitalizations. CMS's own National Impact Assessment found that improving medication adherence among Medicare beneficiaries could save $11.6 billion (statins), $12.4–$15.7 billion (hypertension), and $0.5–$1.8 billion (diabetes) annually.
Seniors face a perfect storm of adherence barriers¶
Medicare beneficiaries are the population most affected by non-adherence, driven by three compounding factors: high disease burden, polypharmacy, and age-related cognitive and physical decline.
About 90% of adults aged 65 and older take at least one prescription medication. The median number of chronic medications for seniors doubled from 2 to 4 between 1988 and 2010 (NHANES data), and current estimates place the average at 5 or more. Polypharmacy — defined as taking 5 or more medications — affects 42–44% of community-dwelling seniors, a figure that has nearly tripled from 12.8% in 1988. Hyperpolypharmacy (10+ medications) affects 18–20% of the 65+ population (Lown Institute; NHANES trend analysis, 2023). Nursing home residents average 7 prescription medications.
The clinical consequences are severe. Older adults visit emergency departments more than 600,000 times each year for adverse drug events — more than twice the rate of younger populations (CDC Medication Safety data). A landmark NEJM study (Budnitz et al., 2011) found that just four drug categories — warfarin, insulins, oral antiplatelet agents, and oral hypoglycemics — accounted for two-thirds of emergency hospitalizations among seniors. Adults 65+ are nearly 7 times as likely as younger patients to be hospitalized for adverse drug events.
Hospital readmissions compound the problem. Approximately 20% of Medicare patients are readmitted within 30 days, and studies estimate that 20–40% of readmissions in older adults are medication-related, with 70% of these deemed preventable (Glans et al., PLOS ONE, 2020). A 2024 meta-analysis in Drugs & Aging found a pooled prevalence of 9% for drug-related readmissions across studies of adults 65+, with 15–22% deemed preventable.
Cost-related non-adherence is a growing concern. By 2022, approximately 1 in 5 adults aged 65+ reported not taking medications as prescribed due to cost (JAMA Network Open, 2023), up from 14.4% in 2016 (Medicare Current Beneficiary Survey). Nearly 40% of older adults say they or someone they know has skipped filling a prescription due to cost (AARP, 2024).
PDC scores and Star Ratings drive billions in plan revenue¶
CMS measures medication adherence through Proportion of Days Covered (PDC) across three drug categories, and these measures carry enormous financial weight in the Star Ratings system.
National average PDC scores for 2025 (performance year 2023, per Pharmacy Quality Solutions) show Medicare Advantage plans averaging 86% for diabetes medications (3.2 stars), 89% for RAS antagonists/hypertension (3.3 stars), and 88% for statins (3.3 stars). Standalone Part D plans perform notably worse, averaging just 2.4–2.9 stars on adherence measures.
The cut points to reach top ratings are demanding and rising annually. For 2025, MAPD plans need a PDC of 87% for 4-star diabetes adherence, 90% for 4-star hypertension, and 89% for 4-star statin adherence. Five-star thresholds now sit at 91–93% — leaving vanishingly thin margins. Notably, today's 2-star cut point now exceeds the 5-star cut point from 2012, reflecting the relentless upward creep of performance benchmarks.
The financial stakes are enormous. Plans that achieve an overall rating of 4.0 stars or higher qualify for a 5% quality bonus payment (QBP) applied to their benchmark. In 2025, total Medicare Advantage quality bonus payments are estimated at $12.7 billion (KFF), up from $3 billion in 2015 — a fourfold increase over a decade. The average QBP per enrollee is approximately $372, meaning a plan with 100,000 members stands to gain roughly $37 million annually from clearing the 4-star threshold. Higher star ratings also unlock larger rebate percentages: 70% at 4.5+ stars versus 50% below 3.5 stars.
Each of the three adherence measures is triple-weighted in the Star Ratings formula, giving them a combined weight of 9 out of 81 total weighted stars (11.1%) in 2026. When related medication measures are included, adherence-adjacent metrics account for roughly 21% of the total weighted score. A decade-long analysis found that 70–74% of plans achieving 4+ stars on adherence also achieved 4+ stars overall. Yet performance at the top remains rare: only 7 of 521 MA-PD contracts (1.3%) earned 5 stars in 2025, a dramatic collapse from 74 contracts in 2022.
Forgetfulness, not defiance, is the primary culprit¶
Understanding why patients don't take their medications is essential to designing effective interventions. Research consistently shows that unintentional non-adherence is 2–3 times more prevalent than intentional non-adherence.
Forgetfulness is the single most-cited barrier, reported by 39–62% of non-adherent patients across studies (Express Scripts survey of 600,000 patients; Harris Interactive panel of 24,017). Other unintentional factors include running out of medication (37%), being too busy, and carelessness about timing. Intentional non-adherence — deliberate decisions to skip or alter doses — accounts for a smaller but significant share: 6–28% of patients across studies. Common intentional reasons include feeling better or believing the medication is unnecessary (20–30%), fear of side effects (21–26%), trying to save money (8–22%), and distrust of medications.
Regimen complexity has a profound, well-quantified effect. A meta-regression of 51 studies (Claxton & Coleman, JMCP, 2012) found that compared to once-daily dosing, adherence declines by 6.7 percentage points for twice-daily, 13.5 points for three-times-daily, and 19.2 points for four-times-daily regimens. For timing adherence — taking the right dose at the right time — the penalties are even steeper: 27%, 39%, and 54% drops respectively. This is perhaps the strongest evidence base in all of adherence research and directly supports simplification-focused interventions.
Low health literacy independently increases the risk: patients with limited literacy are 2.6 times more likely to be unintentionally non-adherent and experience 68% more prescription misinterpretations (Cureus, 2024 systematic review). Cost barriers escalate sharply with out-of-pocket expense — IQVIA data show prescription abandonment under 5% at zero cost but jumping to 45% above $125 and 60% above $500.
What actually works, and what the evidence really says¶
The Cochrane Collaboration's definitive review of 182 RCTs on adherence interventions (Nieuwlaat et al., 2014) delivered a sobering conclusion: current methods are "mostly complex and not very effective." Only 5 of 17 low-risk-of-bias trials improved both adherence and clinical outcomes. That said, several intervention categories show meaningful evidence of benefit.
Pharmacist counseling and MTM have the strongest effect sizes. A 2023 meta-analysis (Kelly et al.) found pharmacist counseling was associated with 4.41 times greater odds of adherence (95% CI: 2.46–7.91). Among older adults specifically, a meta-analysis of 40 RCTs (Marcum et al., 2021, Journal of the American Geriatrics Society) showed a Cohen's d of 0.57 at 3 months, declining to 0.22 at 12 months — suggesting the need for sustained engagement. A 2025 PDC-based meta-analysis of 29 studies found pharmacist-led interventions improved mean PDC by +0.08 (8 percentage points) and adherence rates by a relative 9%.
Medication synchronization programs show strong real-world evidence. A Health Affairs study of approximately 23,000 matched Medicare beneficiaries found med-sync patients achieved a mean PDC of 0.87 versus 0.84 for controls — a 3-point gain that was 3 times larger for patients with low baseline adherence. Hospitalization rates were 9% lower and ED visits 3% lower in the synchronized group. A meta-analysis of 9 studies found med-sync associated with 2.29 times higher odds of adherence. In single-pharmacy studies, control-group PDC ranged from 58–63% versus 80–87% in intervention groups, with patients showing 3.4 to 6.1 times greater odds of adherence.
Smart pill bottles show promise but inconsistent standalone evidence. In a multiple myeloma RCT, smart bottles with pharmacist follow-up produced a 12.6 percentage-point median adherence improvement. A breast cancer study found smart bottles nearly doubled the odds of being adherent at 12 months. A Korean RCT of 61 breast cancer survivors using a Pillsy smart bottle found statistically significant adherence improvement (P=.004). However, an integrative review noted that only 1 of several studies showed significant improvement when the smart bottle was the sole intervention — the technology works best when paired with human follow-up.
Text message reminders produce mixed results. A meta-analysis by Thakkar et al. found they approximately double the odds of adherence, and a CHD-specific meta-analysis of 6 RCTs (n=1,678) showed 2.85 times greater adherence with text reminders. However, a rigorous multi-center RCT (TEXTMEDS, Circulation, 2022) found no significant improvement in self-reported medication adherence at 6 or 12 months. A 2024 Cochrane review rated the evidence as "very uncertain," with effects tending to be short-lived.
Wearable devices remain the least-studied category with no robust RCT evidence yet demonstrating medication adherence improvement. However, an Evidation Health/Humana study of 8,500 patients found that those who used activity trackers (Fitbit, Garmin, Apple Watch) had significantly higher medication adherence across diabetes, hypertension, and dyslipidemia — a strong signal that engagement with wearable technology correlates with better health behaviors. A 2026 discrete choice experiment found 84.2% of older adults preferred simple, low-burden interactions for medication management, while only 15.8% valued complex wearable coordination — a critical design insight.
Pharmacists are the most accessible but underutilized clinicians¶
A 2020 study in JAMA Network Open (Berenbrok et al.) using Medicare claims for 681,456 beneficiaries found that the median number of community pharmacy visits was 13 per year versus 7 primary care encounters — nearly double. The gap widened in rural areas to 14 versus 5. More than 10% of commercially insured adults visited a pharmacy but had no physician visit over an entire year. This positions pharmacists as the healthcare professionals with the most frequent, regular patient contact.
Yet this access is underutilized. The economic case for pharmacist-led interventions is compelling: the APhA Foundation estimates that every $1 spent on pharmacist patient care services yields $16.70 in healthcare savings. The Outcomes network (Cardinal Health) reported $738 million in healthcare savings in 2020 through 4.5 million adherence interventions. AdhereHealth reports an ROI of $2.28 per dollar spent and Star Rating improvements of 1–1.5 points per adherence measure in the first year.
The medication adherence technology market reflects growing investment in this space. Valued at approximately $4–5 billion in 2025, it is projected to reach $9–16 billion by 2030–2032, growing at a 11–15% CAGR. A scoping review in JMIR Aging identified 114 smart medication adherence products on the market — predominantly stationary devices (dispensers, pill boxes) and phone-dependent apps. Tracxn identifies 815 companies globally in the adherence space, but the market remains fragmented with no clear winner, particularly in wearable form factors — a notable gap for a company developing a senior-focused wearable device.
Conclusion: where the opportunity lies¶
The medication adherence problem is vast, well-documented, and stubbornly persistent. The core statistics — 50% non-adherence, $100–300 billion in costs, 125,000 deaths — have barely moved in two decades. What has changed is the financial infrastructure around the problem: CMS now channels $12.7 billion annually in quality bonuses tied partly to adherence metrics, creating a direct revenue incentive for Medicare Advantage plans to invest in solutions that move PDC scores even a few percentage points.
The evidence points to three critical design principles for an effective senior adherence wearable. First, simplicity is non-negotiable — adherence declines 6.7 percentage points with each additional daily dose, and 84% of older adults demand low-burden interactions. Second, technology alone is insufficient — the strongest outcomes come from pairing devices with human touchpoints, particularly pharmacist follow-up, which increases adherence odds by 4.4 times. Third, the pharmacy is the right distribution channel — seniors visit pharmacies nearly twice as often as their doctors, and med-sync programs already demonstrate that pharmacy-based engagement can lift PDC by 3–26 percentage points.
The absence of a dominant wearable product in a $5-billion, 14%-CAGR market with 815 competitors suggests the category is ripe for a well-designed entrant. The winning product will likely not be the cleverest technology but rather the one that most seamlessly bridges the gap between the 13 annual pharmacy visits and the 352 days in between when seniors are managing complex regimens on their own.