Table 1.

Characteristics of included studies

First author and location	Year	Participants Total (male:female), Mean (SD) age	Measurement method	Randomised order of measurement	Blinding of doctors and nurses measurements	Blinding of outcome assessment	Summary risk of bias judgment	Notes
Bo³⁸ Turin, Italy	2008	224, (96:128) subjects attending lipid clinic, age 59.8 (13.2) years. 110 on antihypertensive treatment	Standard mercury sphygmomanometer and appropriately sized cuff. Mean of three consecutive readings after 5 minutes rest	No	Unclear	No	High	Same protocol for doctors and nurses
Bo³⁹ Turin, Italy	2009	273 (109:164) patients age 72.6 (9.7) years. 84.2 on antihypertensive treatment, referred to hospital Alzheimer Evaluation Unit	Nurse readings obtained 15 minutes before and 10 minutes after physician readings using same protocol	No	Unclear	No	High	Nurse before doctor data extracted, nurse after doctor data not significantly different
Culleton²⁹ Calgary, Canada	2006	107 (50:57) hypertensives age 57 (12) years referred for ABPM	Nurse measured BP twice, 5 minutes apart, after 5minutes rest, with mercury sphygmomanometer. Patients were seen within 3 hours by own family physician for a usual BP measurement	No	Yes	No	High	Probably different protocols: Doctors’ ‘usual measurement’ protocol not described
de Blok³⁰ Amsterdam, Netherlands	1991	31 (13:17) patients with suspected hypertension age 43 years (range 17–63) referred to specialist hypertension clinic	Three readings collected after 5 minutes resting seated with random zero sphygmomanometer BP measurement by doctor and nurse in random order	Yes	Unclear	Yes	Low	Same protocol for doctors and nurses. Conference abstract only. Sex data do not = 100%; reported as stated in paper
Gerin³¹ New York US	2001	27 (9:18) patients age 57.6 (11.6) years attending hypertension clinic	Hawkesly random zero sphygmomanometer used by physicians and nurses to measure BP after 5 minutes rest seated. Mean of three measurements recorded at two visits 1 week apart	Unclear	Unclear	Yes	Low	Same protocol for doctors and nurses. Order of measurement was varied at visits; randomisation not described
Gil²⁸ Alicante Spain	1994	50 (13:37) hypertensives in hospital clinic age 59.3 (7.9) years and 50 (23:27) years randomly selected non-hypertensive clinic attendees age 47.2 (10.1) years	BP measured by mercury sphygmomanometer after 5 minutes rest. Doctors measured BP first then nurses	No	Unclear	No	High	Same protocol for doctors and nurses. Spanish text English abstract: translated by CEC
Kumpusalo²⁵ Kuopio, Finland	2002	148 (64:84) hypertensives age 59.2 years visiting one of six primary healthcare centres	BP measured twice with mercury sphygmomanometer and twice with Omrom HEM 705 (graded B/A by BHS). Means of all readings recorded	Yes	Unclear	Yes	Low	Same protocol for doctors and nurses Only doctor and nurse means from automated recordings extracted for analyses
La Batide-Alanore³² Broussais Hospital, Paris	2000	1062 (525:537) patients referred to hypertension clinic (2/3 on treatment). Age 51.8 (13.7) years	Nurses measured three BPs with 5-minute intervals using oscillometric device (Press Mate BP 880 Colin; graded B/B by BHS), patients lying supine. Then physicians measured BP with mercury sphygmomanometer after a mean of 15 minutes lying horizontally	No	Yes	No	High	Different measurement protocols for nurse and doctor
Labinson et al.²⁶ Connecticut, US	2008	65 (29:36) African-American hypertensives age 53.6 (12.7) years recruited from primary care centre	Clinic BP measured x3 with mercury sphygmomanometer by white male study physician and white female study nurse. All wore white coats. Mean of three readings for each observer	Yes	Unclear	Unclear	Low	Same protocol for doctors and nurses. BP also measured by African–American male study physician (data not extracted), to assess effect of race of doctor
Mansoor et al.³³ University of Connecticut, US	1996	64 hypertensives (23:41), mean age 65 (12) years; 59% on treatment referred to health centre for ABPM	Two BP readings with mercury sphygmomanometer, sitting after 10 minutes rest	Yes	Yes	Unclear	Low	Same protocol for doctors and nurses
Richardson and Robinson³⁷ London, England	1971	372 men attending clinic for BUPA health screening	Nurses then doctors recorded supine BP within 1 hour with mercury sphygmomanometers	No	Unclear	No	High	Probably the same protocol for doctors and nurses
Salvador et al.³⁵ Toulouse, France	1990	28 hypertensives (19:9), mean age 46 (12) years enrolled in study of β blockade on alerting reaction	Nurse measured BP with automated monitor seated after 10 minutes rest. Patients spent 12 minutes on an automated oscillometric monitor (data not extracted) then physician measured BP with mercury sphygmomanometer after 5 minutes rest	No	Unclear	No	High	Different measurement protocols for nurse and doctor. Original article in French; translated by CEC Model of automated sphygmomanometer not stated
Sokolovic et al.³⁶ Sarajevo, Bosnia Herzegovina	2012	120 patients (35:85), mean age 47.2 years (No SD) recruited to a prospective clinical trial. Normotensive and hypertensive subgroups	BP first measured by physician then 15 minutes later by nurse	No	Unclear	Unclear	High	Poster abstract only, no further detail available. Authors contacted but no reply. SD imputed from matched studies
Veerman and van Montfrans²⁷ Amsterdam, Netherlands	1992	110 consecutive patients with hypertension on treatment (55:55) mean age 46 (range 22–80) referred to specialist clinic for ABPM.	Patients seated quietly for 10 minutes had three readings made within 5 minutes with a Hawkesly random zero sphygmomanometer	Yes	Unclear	Yes	Low	Same protocol for doctors and nurses
Yoon et al.³⁴ Korea	2010	1019 treated hypertensives (488:531) mean age 56 (10) from nine hospital clinics	BP measured with Omron HEM 747, mean of two measurements after 2 minutes rest seated	Unclear	Unclear	Yes	Low	Same protocol doctors and nurses. Hypertensives had no change of drugs in previous 6 months

ABPM = Ambulatory blood pressure monitoring. BP = blood pressure.