Patient Compliance with Prescription Drug Regimens

Medication adherence is critical to treatment efficacy and safety over time, and noncompliance can undermine patients’ health and increase costs to the healthcare system. This analysis of a nationally representative sample of U.S. adults (n= 23,325) who were prescribed medications within a 12 months time frame revealed that: 

  • Around 30 million patients asked their doctor if there was a lower cost alternative medication
  • Around 18 million patients skipped dose, reduced dosing, delayed filling/refilling prescriptions, or used alternative therapies in order to save money
  • Patients in poor health are less compliant than their healthier counterparts
  • Noncompliant actions vary by type of insurance coverage and chronic conditions

Patient Rx Noncompliance Due to Cost.jpg

Prevalence of Prescription Drug Noncompliance. Among all adult patients who were prescribed medications in 2014, an estimated…

  • 31 million patients asked their doctor if there was a lower cost medication
  • 13 million patients delayed filling a prescription in order to save money
  • 10 million patients intentionally skipped doses of their medication to save money
  • 10 million patients intentionally took less medication to save money
  • 6 million patients turned to alternative therapies to save money
  • 3 million patients bought their medication from another country where it was less expensive

Above estimates are most likely underestimates because the survey questions focused on actions taken to save money, and does not include patients who had simply forgotten or intentionally refused to take their medication. 

Strictly speaking, only 4 of these 6 actions imply noncompliance to treatment regimen, because when patients ask their doctors for a cheaper medication or buy their medications from another country, they may still take their medications as instructed. In contrast, patients are not adhering to their prescribed dosing regimen when they skip dose, reduce dose, delay filling or refilling their prescriptions, or use alternative therapies instead of prescription drugs. Weighted projections of these 4 actions combined revealed that around 18 million adults took these actions in order to save money.

Rx Noncompliance by Patient Health Status.jpg

Differences by Personal Health Status. Compared to patients in better health, patients who believe they are in poor or fair health are significantly less likely to comply with their treatment regimen. For example, 20% of patients in poor health delayed filling a prescription due to cost, compared to only 4% of patients in excellent health. This relationship is probably circular - that is, if patients who are in good health tend to comply with their treatment regimen, this adherence to medication could then result in better health maintenance over time. The bigger implication here is that patients who report being in poor health are clearly more assailed by cost concerns than their healthier counterparts.

Rx Noncompliance by Insurance Coverage.jpg

Principal component analysis (PCA) of the insurance market space revealed that patients with Medicaid coverage only behave in similar ways to patients without any insurance coverage at all. These two groups are most likely to skip doses, lower doses, delay refills or not fill a prescription due to cost.

Patients with Medicare coverage are most likely to buy medications from abroad, and to ask their doctor if there was a lower cost medication. Clearly, Medicare beneficiaries feel more empowered to seek cheaper alternatives than do Medicaid beneficiaries with no other coverage.
 
Medicare beneficiaries who have supplementary coverage from private plans are least likely to take any noncompliant action compared to other patients, including Medicare beneficiaries who have no other coverage. 

Rx Noncompliance by Chronic Condition.jpg

Adherence of prescription drug treatment regimens vary across patients with different chronic conditions, namely:

  • Hypertension
  • High Cholesterol
  • Heart Conditions (includes coronary heart disease, angina pectoris, heart attack, etc.)
  • Stroke
  • Arthritis
  • COPD / Asthma
  • Diabetes

Patients with COPD/Asthma were most likely to skip doses, lower doses, delay refills, or not fill a prescription due to cost. Patients with Arthritis were most likely to ask their doctors if there was a lower cost medications or try alternative treatments. Patients with hyperlipidemia (or high cholesterol) were the least likely to skip doses, lower doses, delay refills, or not fill a prescription; but they were the most likely to buy their medications from abroad in order to save money. These odds ratios are computed for each condition while controlling for all other conditions in attempt to extract the unique effect for each condition despite comorbidities.

However, the NHIS does not measure treatment adherence for each chronic condition separately. The survey measures whether patient was diagnosed with various conditions, and also whether they took any action that did not comply with the treatment regimens of their prescribed medications. So we cannot know for sure if the noncompliant actions pertain to which chronic condition if they have comorbidities, or to some other acute but short term ailment.


Data Source

The National Health Interview Survey (NHIS) conducted by the U.S. National Center for Health Statistics (NCHS) is an outstanding public resource on population health data. The sample adult module collects data via in-person interviews with close to 37,000 adults, and permits projections to the population of around 240 million civilian noninstitutionalized adults residing in the United States. Sampling and interviewing for the NHIS is conducted continuously year-round. The sample design follows a multistage area probability design that maximized representative sampling of residential households and noninstitutional group quarters, such as college dorms. All estimates in this analysis were weighted by the final annual sample adult weight (WTFA_SA) that includes design, ratio, non-response and post-stratification adjustments for 2014 NHIS sample adults.

In short, the data for this analysis is NOT from web respondents who have opted in to participate in market research. Instead, the respondents were randomly sampled from the U.S. adult population, and interviewed in their homes in person by highly trained field representatives employed by the U.S. Census Bureau. The full sample contained close to 37,000 adult respondents, of which over 23,325 were prescribed medications within the past 12 months. This analysis focus on data from this subsample of 23,325 respondents, weighted represent the population of 147 million U.S. adults who were prescribed medications within a 12 months time frame. More methodology details can be found here.