This cross-sectional study was conducted in a teaching hospital in Chennai, Tamil Nadu, India using 240 subjects who attended the medicine outpatient department.

The participants included were in age group of 20-80 years of both the genders were included in the study. Patients with normal BP and willing to participate in the study were included in the study. Patients who were critically ill, known hypertensives, neurological diseases, renal and endocrine disorders were excluded from the study.

The study was explained to the subjects in their native tongue and consent was obtained. Subjects of both the gender of age group 20-80 years without any neurological, renal or endocrine disorders were selected for the study.

The study was carried out during the months of April to August 2014 in the General Medicine outpatient department of a teaching hospital in Chennai. Approval from the institutional ethical committee was obtained before the beginning of the study.

BP was measured in the right arm of all the subjects after five minutes rest. DIAMOND Sphygmomanometer (ideal surgical company, India) with Littmann classic II stethoscope (3M, USA), DIAMOND aneroid meter (Ideal surgical company, India) with Littmann classic II stethoscope (3M, USA), and OMRON digital BP monitors (Omron Healthcare Manufacturing Vietnam Company, Singapore) were used to record the resting BP. Five minutes of interval was given between consecutive readings. The subjects were seated comfortably with back supported and right arm bared, with no constrictive clothing and legs uncrossed. The arm of the subjects was supported at the level of the heart and the cuff was chosen so that it covered 80% of the arm circumference. The mercury column was deflated at 2 to 3mm/sec and care was taken that neither the subject nor the observer talk during the measurement. Three trained observers (Two trained second year MBBS students and one physiology assistant professor) recorded the BP independently to avoid observer bias. Each observer recorded the BP thrice with gap of five minutes and the best of the three measurements were documented.

The data were analysed using SPSS version 21.0. The mean and standard deviation of the BP measured using sphygmomanometer was compared with the aneroid measurements and digital measurements using paired t-test. A p-value of <0.05 was taken to be statistically significant. The mean and standard deviation of difference between the sphygmomanometer and aneroid readings and sphygmomanometer and digital readings were used to test the AAMI criteria. The cumulative percentage of absolute difference between the sphygmomanometer and aneroid readings and sphygmomanometer and digital readings were calculated in Microsoft Excel 2010 and tested for the BHS criteria. The Bland-Altman scatter plots were prepared using Microsoft Excel 2010.

Comparison of measurements of both SBP and DBP using sphygmomanometer and aneroid meter showed that the mean difference between the two methods for SBP was 0.675±7.12 mmHg and for DBP it was 1.75±6.3 mmHg [Table/Fig-1]. Thus the measurements done using aneroid meter satisfies the AAMI criteria which specify a mean difference between devices of <5 mmHg with a standard deviation of ≤8 mmHg [10]. While comparing the SBP and DBP measurements using sphygmomanometer and digital methods, the mean difference for SBP was found to be 1.65±11.08 mmHg and for DBP, 2.76±8.8 mmHg [Table/Fig-2]. Thus the measurements using digital method do not satisfy the AAMI criteria for both SBP and DBP where the standard deviation remains higher than the required 8 mmHg.

While focusing on the BHS criteria [Table/Fig-3], the absolute difference in the measurements using sphygmomanometer and aneroid meter falls under grade C for both SBP and DBP which is considered to be poor validation [Table/Fig-4] and in case of digital measurements, it falls under grade D for SBP and DBP which is considered to be very poor validation [Table/Fig-5]. The grades represent the cumulative percentage of the absolute difference between the readings falling within 5, 10, 15 mmHg of the standard sphygmomanometer readings. Grade A and B are acceptable for the usage of the devices while grades C and D are not recommended [8].

To evaluate the agreement between the manual and automated devices, Bland-Altman scatter plots were used, where the mean of the readings taken by the two methods is plotted against the difference in the readings [11]. The Limits Of Agreement (LOA) are calculated using the formula:,

LOA=Mean±1.96xStandard deviation.

[Table/Fig-6] shows the Bland-Altman scatter plots where mean of the SBP readings from sphygmomanometer and aneroid are plotted against difference in the readings between the two. The bias (average variation in the readings from standard) was 0.675 with the limits of agreement from -13.28 to 14.63. The bias in DBP readings between sphygmomanometer and aneroid was 1.75 (-10.6-14.1) [Table/Fig-7]. Bland Altman plot of sphygmomanometer versus digital readings show a bias of -1.41 with the limits of agreement from -21.66 to 18.84 for SBP [Table/Fig-8] and 1.008 (-15.26 – 17.276) for DBP [Table/Fig-9]. Most of the readings in all the plots lie with the limits of agreement.

[Table/Fig-6]:

Bland-Altman plot of SBP measured by manual and aneroid method of BP measurement.

Correlation R=0.04514 (p=0.48); Slope=0.0213 (p=0.487); Intercept=-1.844 (p=0.61); Bias=0.675 (-13.28-14.63); The upper and lower limits of agreement are given in boxes

[Table/Fig-7]:

Bland-Altman plot of DBP measured by manual and aneroid method of BP measurement.

Correlation R=0.021 (p=0.74); Slope=0.015 (0.74); Intercept=0.607 (p=0.86); Bias=1.75 (-10.6-14.1); The upper and lower limits of agreement are given in boxes

[Table/Fig-8]:

Bland-Altman plot of SBP measured by manual and digital method of BP measurement.

Correlation R=0.0817 (p=0.21); Slope=-0.12 (p=0.20); Intercept=119.17 (p<0.00***); Bias=-1.41 (-21.66-18.84); The upper and lower limits of agreement are given in boxes

[Table/Fig-9]:

Bland-Altman plot of DBP measured by manual and digital method of BP measurement.

Correlation R=0.249 (p <0.00***); Slope=-0.213 (p <0.00***); Intercept=17.052 (p<0.00***); Bias=1.008 (-15.26-17.276); The upper and lower limits of agreement are given in boxes

This study was conducted to assess the reliability and accuracy of the aneroid meter and digital BP monitors which are off late used in an increasing level in clinical practice to measure BP. The readings produced by aneroid and digital monitors are compared with that of the manual mercury manometer technique and their validity is tested against three well recognised criteria for the usage of medical devices viz., AAMI criteria, BHS protocol and Bland Altman plots. Aneroid meter satisfied the AAMI criteria while digital did not and both aneroid and digital showed minimal bias with a majority of reading lying between the limits of agreement in the Bland Altman plots. But on focussing on the BHS protocol both the aneroid and digital methods got poor validation. The devices used were pre-validated before the beginning of the study despite which the aneroid meter showed poor reliability according to one of the criteria and the digital monitor showed poor reliability according to two out of the three criteria.

For more than two centuries, mercury sphygmomanometer was considered synonymous to BP measurement. But mercury manometers pose a threat where they are the main source of mercury pollution and WHO considers it to be a serious occupational hazard [12]. Mercury is a neurotoxin and produces severe health hazards. It is prohibited in some parts of the world and the European Union is suggesting a more widespread ban [13]. Many countries have completely moved on from manual mercury manometers to record BP and rely on automated ones [14]. The automated devices can be recorded by anyone without proper training and were also proven to have less of white coat effect on patients [8,15].

Separate protocols and criteria are available to validate the automated devices. There are studies with evidence to limit the usage of automated devices on normotensives and not on hypertensives [15,16]. In our study we found that digital monitor and aneroid too either over read or under read the BP and this tendency of the automated devices was reported by a few studies, irrespective of whether readings were taken from healthy children [17], children with diabetes [18], or adult patients attending a hospital emergency department [19] or day surgery clinic [16]. In case of trauma or in patients with deteriorating conditions, inaccurate reading from automated BP monitors could have grave consequences.

Aneroid meter showed far more reliability than digital BP monitors but nevertheless there were discrepancies. But as mentioned earlier they need periodic calibration requiring sphygmomanometers and their preference over sphygmomanometer taking in to account of the mercury hazards overweighs its relative diminished reliability. Since both sphygmomanometers and aneroid meters need trained professionals to record BP, their white coat effect is also comparatively more than digital monitors [8, 15]. With the increased usage of automated devices to measure BP, physician’s awareness of the variations in their readings is vital.

Limitations of the study are the age group included were of wider range. Recording were done at different times rather than being simultaneous.

Automated BP measurements were found to produce discrepancies in reading in clinical settings. The reliability and accuracy of the automated devices are questionable. Though, digital BP monitors have more convenience based advantages over manual methods, their usage should be done with increased caution. Mentioning of the type of instrument which was used to measure the BP should be included in patient’s data.