Table 3.

  Sensitivity, specificity, and likelihood ratios for a positive and negative test of different measures in predicting ambulatory pressure treatment thresholds

	Systolic pressure					Diastolic pressure
	Sensitivity (%)	Specificity (%)	Likelihood ratio (positive)	Likelihood ratio (negative)		Sensitivity (%)	Specificity (%)	Likelihood ratio (positive)	Likelihood ratio (negative)
Doctor	83.7	50.7	1.7	0.32		79.5	63.0	2.1	0.33
Nurse 1—mercury	67.9	67.1	2.1	0.48		75.0	33.6	2.2	0.38
Nurse 1—oscillometric	69.9	73.2	2.6	0.41		68.1	69.6	2.2	0.46
Nurse 2—mercury	66.7	75.3	2.7	0.44		79.2	74.5	3.1	0.28
Nurse 2—oscillometric	58.5	83.3	3.5	0.50		80.9	76.9	3.5	0.25
Self measurement in surgery 1	77.1	58.3	1.8	0.39		100	65.1	2.9	–
Self measurement in surgery 2	80.0	90.9	8.8	0.22		92.3	74.4	3.6	0.10
Home measurement	84.1	68.7	2.7	0.23		80.4	70.6	2.7	0.28
Last three clinic measurements	85.8	43.4	1.5	0.33		84.8	58.8	2.1	0.26

Thresholds at which drug treatment for high blood pressure would be started or changed: ambulatory readings—previous recommendations suggested that in patients with newly diagnosed or borderline hypertension ambulatory readings or home readings of >145/95 mm Hg would indicate a need for treatment,⁵ and at >135/85 mm Hg poor control would warrant changing treatment for established hypertension^2 5; clinic readings (by doctor, nurse, or self measurement in the surgery)—in patients with newly diagnosed hypertension drug treatment would be started at >160/100 mm Hg,⁴ and poor control would warrant a change in treatment for established hypertension at >140/90 mm Hg.³