Computerized reporting improves the clinical use of ambulatory blood pressure measurement

Ambulatory blood pressure measurement (ABPM) is being used increasingly in clinical practice. One previous study has shown that there can be considerable variance between expert observers in the interpretation of ABPM data. The purpose of this study was to show whether computer-generated reports with the dablABPM system would provide more consistency in the interpretation of data than reports from expert observers.

Twenty-six international experts in hypertension were invited to participate and 17 agreed to do so. Twelve ABPMs generated by the Spacelabs device that were considered representative of the patterns likely to be seen in practice were sent to each participant for reporting. The corresponding dabl reports with an automatic interpretation were generated according to the European Society of Hypertension guideline for comparison with the observer reports. Each of the observer-interpreted Spacelabs reports for the 12 ABPM patterns were coded, analysed and compared with the automatically interpreted dablABPM reports. Both sets of data were analysed for interobserver variability, observer v dablABPM consistency and the time taken for observer reportage. The main analysis determined issues of definite disagreement, namely the presence or absence of nocturnal dipping. Further analysis determined the presence or absence of white-coat phenomena and the severity of hypertension.

Incorrect diagnoses were made in 13 instances. White-coat hypertension and white-coat effect, although obvious in many instances, were not identified in five ABPMs; the severity of hypertension was not reported in four ABPMs; the severity of nocturnal hypertension was not diagnosed in one ABPM by nine experts and isolated diastolic hypertension was not identified by six experts in two ABPMs.

This study provides evidence to show that observer variance in reporting ABPMs is common even among experts and that computer-generated interpretative reports of ABPM data improve the diagnostic decisions based on the data generated by 24-h blood pressure recording.