Automated blood pressure devices improve outcome  

Professor Mark Nelson presented the following paper at the 2007 Annual Scientific Meeting of the HBPRCA.

RESULTS OF CRAB: A CLUSTER RANDOMISED CONTROLLED TRIAL OF AN AUTOMATED VERSUS MANUAL DEVICE FOR BLOOD PRESSURE MANAGEMENT
MR Nelson, TM Winzenberg, Menzies Research Institute, University of Tasmania, Hobart

It is likely that automated devices will replace mercury sphygmomanometers. This has the potential to improve blood pressure measuring practices in general practice. CRAB is a cluster randomisation trial of automated (OMRON office digital blood pressure HEM-907 monitor) versus manual devices for BP management. Objectives were to determine the effect of such devices over a 1 week period on digital preference, on measurement of BP, antihypertensive drug prescribing and mean systolic and diastolic BP measurements. We had power in excess of 0.70 for all our objectives. Practices were recruited in southern Tasmania and randomly allocated to intervention (supplied with automated monitors in all clinical rooms and other BP measuring devices were removed) or control (practices did not have such devices). Records of all adult patients who attended the practice in the study week were audited for BP recordings and management. Analysis: unpaired t-tests for effects on mean BP otherwise c2 test. Results: 24 practices enrolled (3 withdrew - 2 control, 1 intervention), 3355 records were reviewed with 768 BP recordings and 1333 hypertensive individuals. For all individuals in intervention vs. control practices the percentage of BP recordings ending in ‘0’ was significantly lower (SBP 20% vs. 72%, DBP 22% vs. 70% p<0.00001), and more multiple BP recordings were made (8% vs. 3% p<0.001). For hypertensive individuals in the intervention vs. control practices there were more recordings made (38% vs. 31% p<0.01), a greater proportion with multiple recordings (5% vs. 1% p < 0.01) and an increased number of antihypertensive drugs prescribed (7% vs. 4% p = 0.03). The mean SBP recordings, but not DBP, was significantly higher in intervention practices [138.2 (sd 17.7) vs. 132.0 (sd 20.8) mmHg p<0.0001].

Conclusions: It seems that automated machines induced more measurements, reduced digit preference, forced up BP recordings, and pushed GPs into action. It seems likely that such machines will have a beneficial effect on BP management in general practice.