Cardiovascular Journal of Africa: Vol 23 No 8 (September 2012) - page 43

CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 8, September 2012 AFRICA 457
of Family Medicine, Lagos State University Hospital (LASUTH),
Lagos, Nigeria. The general out-patients’ clinic is the first contact
clinic for all patients presenting at the hospital with non-urgent
conditions. The study was approved by the Institution’s research
ethics committee (LHREC/08/054).
Participants were consecutively recruited in the waiting hall
of the clinic, where patients are seated and take their turn to see
the nurse for evaluation of vital signs. The study was introduced
to each patient by the clinic nurse, who offered each adult the
option of having automated blood pressure measurement with
the use of a device that could obtain six consecutive readings
in a designated office. Associated time required and attendant
discomfort from repeated cuff inflation and deflation were
explained to prospective participants. Signed informed consent
was obtained from all individuals who agreed to participate.
Inclusion criteria were: willingness to voluntarily participate
and the absence of any acutely distressful condition such as
fever, breathlessness and pain. Patients excluded were those with
irregular pulse rhythm or a mid-upper arm circumference greater
than 42 cm (oscillometric devices are unreliable in persons with
arrhythmias or mid-upper arm circumference > 42 cm), and all
patients who had smoked cigarettes or taken coffee on the day
of examination.4
Blood pressure measurements were conducted in an office
devoid of noise or vibrations, offering optimum comfort to the
participants. Patientswere instructedbya trained researchassistant
to relax and avoid arm movement during the measurements.
The research assistant witnessed and documented the reading
obtained from the first measurement, and then left the room.
Participants were trained and instructed to remove the cuff and
press a door bell, calling for the return of the research assistant
after the completion of the six measurements. All measurements
were taken with an automated oscillometric blood pressure
machine, BpTRU BPM-100 (VSM Medical Technologies,
Canada) with strict adherence to the American Heart Association
recommendations for clinical BP measurement.4
The BpTRU machine uses an oscillometric algorithm for the
determination of systolic and diastolic BP. It is designed to take
six BP readings with a programmable rest interval between each
measurement (resting time between measurements can be set at
one, two or three minutes). The interval between measurements
has been shown to have no effect on the readings obtained by
this device. 8 For this study, the resting time was set at one minute
to keep the time committed to the study by each volunteer to the
lowest possible, in order to minimise the interference of the study
with their primary aim of clinic attendance.
The BpTRU device automatically discards the first reading
and computes the average systolic and diastolic BP from the
average of the last five readings. All six readings (including
the first reading) as well as the computed average are digitally
displayed. The device has been validated with the British
Hypertension Society (BHS) protocol and passed with an A/A
grade. 9
All the BP readings displayed by the machine were accurately
documented. The average readings computed by the device were
also recorded. Participants’ biodata, history of hypertension/use
of antihypertensive medications were also noted. Data collection
was concluded on the 186th participant (after six months) due to
resource (time, space and personnel) constraints. At this point,
it was judged that the sample would suffice for the purpose
of the study because adequate representation of hypertensive
and non-hypertensive participants had been attained to ensure
adequate statistical power for sub-population analysis. The
final study sample was separated into two sub-populations,
namely, a hypertensive sub-population (individuals with a
previous diagnosis of hypertension) and a mixed sub-population
comprising normotensive and yet undiagnosed possibly
hypertensive individuals.
Three BP measurement models, namely single, triple
(average of three consecutive readings) and quintuple (average
of five consecutive readings) BP readings were created from
the database. We evaluated the effect of inclusion or exclusion
of FBPR on: (1) distribution of participants’ BP readings in a
JNC-7 classification model for the mixed sub-population and
a modified JNC-7 classification model (in which the optimal
BP and pre-hypertensive domains were merged and relabeled
stage 0) for the hypertensive sub-population, (2) consideration
of a diagnosis of hypertension among the mixed sub-population
of the study sample, and (3) clinical assessment of BP control
among the hypertensive sub-population of the participants. The
correlations between the readings that included FBPR and those
excluding FBPR were also evaluated.
Lastly, differences between the compared readings in each
measurement model were evaluated to determine and compare
the proportion of differences among the hypertensive and
mixed sub-populations respectively, those = 0 mmHg and those
≥ 5 mmHg. (Differences ≥ 5 mmHg are considered clinically
significantly different in the British Hypertension Society’s
protocol for validating BP devices).10
Variables used in data analysis included:
• first BP reading (SYS-1 and DIA-1)
• second BP reading (SYS-2 and DIA-2)
• average of three readings, including FBPR: SYS1-3 and DIA1-3
• average of three readings, including FBPR: SYS2-4 and DIA2-4
• average of five readings, including FBPR: SYS1-5 and DIA1-5
• average of five readings, including FBPR: SYS2-6 and DIA2-6.
Data redesignation became imperative because of the need
for joint consideration of systolic and diastolic values for
JNC classification and relativity to the threshold of 140/90
mmHg. SYS1-3/DIA1-3, SYS2-4/DIA2-4 , SYS 1-5/DIA1-5 and SYS2-6/
DIA2-6 were designated as AVE1-3, AVE2-4, AVE1-5 and AVE2-6 ,
respectively. Likewise, SYS-1/DIA-1 and SYS-2/DIA-2 were
designated as FBPR and SBPR, respectively. Cases in which
systolic and diastolic readings fell under different JNC-7 stages
were treated by classifying the BP reading under the higher
category, as recommended in the JNC-7 report. This principle
was extended to the classification of the BP readings into
< 140/90 or ≥ 140/90 mmHg.
Statistical analysis
Data analysis was performed using GraphPad Prisms version 5
for Windows. All study data were first evaluated with descriptive
statistics. In addition to the final study sample, sub-groups that
included the hypertensive sub-population (those who had been
diagnosed as having hypertension) and a mixed sub-population
(comprising true normotensives and undiagnosed hypertensives)
were identified, analysed and compared.
Correlation statistics and comparison of mean values were
performed after evaluation of data for normality. Comparisons