How do minor changes in the definition of blood pressure control affect the reported success of hypertension treatment?

OBJECTIVE: In 1999, the first reporting year of the Health Plan Employer Data and Information Set, our organization noted a significant difference in the proportion of patients with controlled blood pressure (BP) when the target was changed from < 140/90 mm Hg to < or = 140/90 mm Hg. We compared these data with a second larger dataset to determine if these findings were an isolated or common phenomenon and to determine what factors influenced this difference. STUDY DESIGN AND METHODS: Subjects were drug-treated hypertensive patients. Blood pressure measurements were taken from outpatient medical records.
RESULTS: The percent of hypertensive patients with controlled BP levels was significantly greater (P <.001) when the cutoff of < or = 140/90 mm Hg was used rather than <140/90 mm Hg. When a single BP measurement was used to estimate control, the percent of patients with controlled hypertension increased 12.7% simply by changing the definition to include the level of systolic BP = 140 mm Hg or diastolic BP = 90. When multiple BP readings were used to estimate control, the difference was much less pronounced (2.7%).
CONCLUSIONS: When evaluating the success of hypertension treatment, defining BP control as a BP of < or = 140 mm Hg systolic and < or = 90 diastolic mm Hg will result in significantly more people having controlled BP than when <140 mm Hg systolic and <90 mm Hg diastolic are used. This difference is most likely the result of end-digit preference. The use of multiple measures reduces this effect. Policy makers and guideline authors should consider these aspects when setting BP goals.