Evaluating the Consequences of a Hypertension Management Incentive
A hypertension control bonus increased repeat blood pressure checks but may have harmed borderline patients.
*Quasi-experimental difference-in-differences study; Level 3 (OCEBM).
Citation
Boone C, Robicsek A. Evaluating the Consequences of a Hypertension Management Incentive. JAMA Intern Med. 2026;186(8):965-973. doi:10.1001/jamainternmed.2026.1637
Background
Financial rewards tied to quality measures are common, but they can change documentation more than health. This study evaluated a physician bonus for documenting blood pressure below 140/90 mm Hg.
Patients
334,364 adults aged 18 to 85 years with diagnosed hypertension and primary care encounters in 103 practices. Pregnant patients were excluded.
Intervention
Physician contract bonus tied to hypertension control performance.
Control
Usual primary care practices without the hypertension control bonus.
Outcome
Blood pressure remeasurement, documented control, medication dose changes, and hospitalization for stroke or acute coronary syndrome.
Follow-up Period
Primary analysis: 2021 to 2022; hospitalizations assessed up to 1 year.
Results
| Outcome | Population | Effect | NNT or NNH |
|---|---|---|---|
| Repeat blood pressure measurement | All patients | +1.90 percentage points (95% CI 0.70 to 3.10) | NNT 53 |
| Documented blood pressure control | Initial systolic 140-145 mm Hg | +4.07 percentage points (95% CI 2.11 to 6.02) | NNT 25 |
| Repeat blood pressure measurement | Initial systolic 140-145 mm Hg | +5.61 percentage points (95% CI 2.90 to 8.33) | NNT 18 |
| Existing medication dose increased | Initial systolic 140-145 mm Hg | −1.11 percentage points (95% CI −1.95 to −0.27) | — |
| Stroke or acute coronary syndrome hospitalization | Initial systolic 140-145 mm Hg, 1 year | +0.52 percentage points (95% CI 0.17 to 0.87) | NNH 192 |
The main overall hypertension population had no significant improvement in blood pressure control, medication changes, or cardiovascular hospitalizations. The control was current usual care without the incentive.
Limitations
Nonrandomized design, possible practice-level confounding, COVID-era data, one health system, possible missed hospitalizations outside the network, and unclear generalizability.
Funding
Québec public research fund; no reported conflicts.
Clinical Application
Do not equate better metric performance with better care; avoid incentive designs that reward “repeat until normal” blood pressure documentation.
Discussion
Sign in to join the discussion.
In this quasi-experimental difference-in-differences study, the physician-facing hypertension-control financial incentive was associated among patients with initial SBP 140-145 mm Hg with more documented control (+4.07 pp) and higher 1-year stroke/ACS hospitalization (+0.52 pp); given the nonrandomized design, should this change quality-incentive policy or clinic practice? The authors used a difference-in-differences design to estimate that practices adopting the hypertension-control incentive had a 0.52 percentage point higher 1-year stroke/ACS hospitalization risk among patients with initial SBP 140-145 mm Hg. Which assumption is most important for interpreting this estimate as the effect of the incentive?