What question did this study set out to answer?

To determine whether the observed relationship between poor antihypertensive adherence and worse clinical outcomes is driven by unmeasured healthy user bias rather than drug efficacy.

May 17, 2011Heart

Overestimation of the effects of adherence on outcomes: a case study in healthy user bias and hypertension

View Full Paper

Ask AI

Bookmark

Why the study?

Does poor adherence to antihypertensive medications independently increase the risk of adverse outcomes when controlling for blood pressure, or is it confounded by healthy user bias?

Population

Veterans with hypertension starting antihypertensive treatment

Comparison

Poor adherence to antihypertensive medications vs Good adherence (implied by the <80% threshold)

Design

Cohort

Authors

BSBrian C. SauerUniversity of Utah JNJonathan R. NebekerUniversity of Utah

Discussion

Member takes

Implication

Healthy user bias likely inflates adherence benefits in observational data; leaves open the true causal effect of antihypertensive adherence independent of BP.

Key Points

To determine whether the observed relationship between poor antihypertensive adherence and worse clinical outcomes is driven by unmeasured healthy user bias rather than drug efficacy.
Analyzed data from veterans with hypertension initiating antihypertensive therapy.
Developed causal models predicting hospitalization, myocardial infarction (MI), and death for poor adherence (<80%), comparing models adjusted for baseline covariates against models additionally adjusting for time-varying blood pressure (BP).
Models without time-varying BP adjustment showed that poor adherence was associated with increased risks of 3.7% for hospitalization, 28.1% for MI, and 23.3% for death.
Controlling for time-varying BP yielded nearly identical risk increases (3.4% for hospitalization, 27.7% for MI, and 23.4% for death), indicating that adherence-related risk reductions occur independently of blood pressure control.

Structured PICO

Does poor adherence to antihypertensive medications independently increase the risk of adverse outcomes when controlling for blood pressure, or is it confounded by healthy user bias?

Population

Veterans with hypertension starting antihypertensive treatment

Intervention

Poor adherence (<80%) to antihypertensive medications

Comparator

Good adherence (implied by the <80% threshold)

Outcome

Risks of hospitalisation, myocardial infarction (MI) and deathhard clinical

Observational studies of medication adherence are likely heavily confounded by the healthy user bias, overestimating the true magnitude of adherence effects on clinical outcomes.

Limitations

Observational adherence studies using rich clinical data are impossibly confounded

Cite This Study

Sauer et al. (2011) studied this question.

synapsesocial.com/papers/6a7e8ae40af838441d0d1fc5 https://doi.org/10.1136/hrt.2011.223289

Topics

Hypertension management

View Full Paper

Ask AI

Bookmark

Implication

Healthy user bias likely inflates adherence benefits in observational data; leaves open the true causal effect of antihypertensive adherence independent of BP.

Overestimation of the effects of adherence on outcomes: a case study in healthy user bias and hypertension

Discussion

Member takes

Implication

Key Points

Structured PICO

Limitations

Cite This Study

Topics

Overestimation of the effects of adherence on outcomes: a case study in healthy user bias and hypertension

Discussion

Member takes

Implication

Key Points

Structured PICO

Limitations

Cite This Study

Also Consider

Topics

Also Consider