MEDLINE, CINAHL, Embase, and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched from their respective start dates to 1 September 2011 using the text words ‘white coat’. Additional citations were identified from reference lists of full texts included in the review. Searches of individual journal collections (American Journal of Hypertension, Blood Pressure Monitoring, Hypertension, Journal of Human Hypertension, Journal of Hypertension), conference abstracts, and personal reference archives were also undertaken. No language restrictions were applied. Search updates for new publications were run until 30 September 2013.
How this fits in
Blood pressure measurements are elevated in the presence of a health professional — the ‘white coat effect’ — and this appears to be greater for doctors than for nurses. There is evidence for lower blood pressure outcomes in nurse-led clinics for hypertension compared with usual care by doctors, but the white coat effect is not taken into account in many such studies. This is the first systematic review to compare blood pressure measurements made by doctors and nurses; there is evidence that blood pressures recorded by doctors are systematically higher than those recorded by nurses within the same setting. The findings suggest that studies comparing doctor and nurse interventions require independent outcome measures to avoid bias, and that doctor-measured blood pressures may introduce bias into clinical decision making in hypertension.
Studies were sought reporting blood pressures measured by doctors and nurses during the same clinic visit, for adults aged ≥18 years with or without hypertension. Primary outcome measures were mean systolic and diastolic blood pressures measured by nurses and by doctors or differences between them, and proportions of patients diagnosed with white coat hypertension (defined as clinic blood pressure >20/10 mmHg higher than ambulatory or home-measured blood pressure), according to blood pressure measurements by doctors and nurses.
One author screened retrieved citations and abstracts for potentially relevant studies and a second author checked the selections. Studies assessed as definite or uncertain for inclusion were retrieved in full.