Can ambulatory blood pressure machines improve postural blood pressure data collection for geriatric patients admitted to a geriatric ward? : a pilot study

![CDATA[Aims: Are Ambulatory Blood Pressure Monitors (ABPMs) equivalent to nursing measurements of blood pressure (BP) &postural blood pressure? Can ABPMs improve data collection rates for postural BP measurement, improving clinical diagnosis? Methods: Over a combined period of 12 months across a two year period from 2013, 97 geriatric patients were recruited to the study from admissions to St Vincent’s Hospital Sydney. Standard nursing observations were supplemented with readings obtained from an OSCAR-II 24-hour ambulatory blood pressure monitor every 30 minutes during the day and hourly when asleep. Postural blood pressure measurements were collected by nursing staff & manually by the research team using the ABPM. Student’s t test was used to assess differences in blood pressure data which was normally distributed. A Bland-Altman plot (validated to compare two methods of collecting similar clinical data) was used to investigate differences in blood pressure data on patients who had complete set of ABPM & nursing systolic & diastolic, lying & postural readings. Results: Of the 97 patients enrolled in the study, only 29 patients had a complete set of nursing readings & 84 had a complete set of ABPM readings. Of these, 22 patients had a complete set of ABPM & nursing data, which were used for the Bland-Altman analysis. There were no statistically significant differences in blood pressure data collected between ABPM and nursing measurements. The Bland-Altman plot demonstrated considerable variability with the differences in postural drops observed for the two methods, with no unusual trends. Conclusions: This study demonstrates that ABPMs can improve the rates of collection of postural blood pressure readings, aiding clinicians with diagnosis. However our analysis demonstrates variability between ABPM postural measurements and ‘standard nursing readings’. This may be explained by small sample size. A larger study with improved methodology could demonstrate whether there is adequate correlation between the two collection methods.]]