CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 2, March 2012
AFRICA
85
End digit preference in blood pressure measurement in
a hypertension specialty clinic in southwest Nigeria
OE AYODELE, EO SANYA, OO OKUNOLA, AA AKINTUNDE
Abstract
Background:
One of the observer errors associated with
blood pressure (BP) measurement using a mercury sphyg-
momanometer is end digit preference (EDP) which refers
to the occurrence of a particular end digit more frequently
than would be expected by chance alone. Published reports,
mainly from outside Africa, have shown a high prevalence
ranging from 22 to 90% of end digit zero in BP readings
taken by healthcare workers (HCWs). This study examined
the prevalence of EDP and patients’ and physicians’ charac-
teristics influencing the occurrence of EDP.
Methods:
A retrospective review was undertaken of BP read-
ings of 114 patients seen over a two-month period at our
hypertension specialty clinic.
Results:
Nurses and physicians displayed a high frequency of
preference for end digit zero in systolic blood pressure (SBP)
and diastolic blood pressure (DBP) readings. The prefer-
ence for end digit zero was, however, higher for nurses than
for physicians (SBP: 98.5 vs 51.2%,
p
<
0.001; DBP: 98.5 vs
64.3%,
p
<
0.001). Among the physicians, the consultant staff
displayed the least preference for end digit zero compared to
resident doctors. There was no statistically significant differ-
ence in gender, age, weight, height and BMI of those with
BP readings with end digit zero compared with those with
non-zero end digits.
Conclusion:
The high prevalence of EDP for zero argues for
the training, retraining and certification of HCWs in BP
measurement and the institution of a regular monitoring and
feedback system on EDP in order to minimise this observer
error.
Keywords:
end digit preference, blood pressure measurement,
hypertension specialty clinic, southwest Nigeria
Submitted 7/1/11, accepted 30/8/11
Cardiovasc J Afr
2012;
23
: 85–89
DOI: 10.5830/CVJA-2011-045
The toxicity of mercury notwithstanding, the mercury sphyg-
momanometer remains the most widely used apparatus for blood
pressure (BP) measurement in Nigeria.
1
Although protocols for
measuring BP are well described and standardised,
2-5
inaccuracies
of measurement do occur from the use of faulty or malfunction-
ing equipment, improper technique, or observer errors or bias.
5-9
Observer biases described in clinical practice and trials include
duplication of previous BP recordings, rounding off to or below
pre-set cut-off values for the diagnosis or control of hypertension
and rounding off BP values to a particular end digit.
5-20
End digit preference (EDP) refers to the occurrence of a
particular end digit more frequently than would be expected by
chance alone and it is a widely accepted indicator of low-quality
BP measurement.
5-8
In view of the fact that mercury sphygmoma-
nometers are calibrated in increments of 2 mmHg, individual
readings should only end in 0, 2, 4, 6 and 8.
If BP measurements are done strictly according to guidelines,
the expected prevalence of each of the digits should be 20%.
However, various studies have shown that there is an increased
prevalence of end digit 0 (zero), ranging from 22 to 90%,
depending on the clinical setting (primary healthcare, specialty
clinic, drug-trial setting), the skill or qualification of the health-
care worker who took the BP, and the presence of feedback-
monitoring systems for EDP.
10-20
End digit preference leads to over- or underestimation of
actual BP. Underestimation of BP could mean missing the diag-
nosis of hypertension in a patient, which can result in significant
morbidity and mortality due to lack of treatment.
2-5
On the other
hand, overestimation of BP could result in inappropriate diagno-
sis of hypertension, inappropriate labelling, lifetime subjection
to antihypertensive treatment with its attendant side effects, and
reduction in quality of life and financial status due to loss of
work or hospitalisation.
2-5,21-23
Furthermore, EDP had been associated with difficulty in
assessing associations between blood pressure and other poten-
tial cardiovascular risk factors by reducing the power of statisti-
cal tests.
10
Therefore the diagnosis of hypertension, the eligibil-
ity for treatment, the assessment of adequacy of BP control,
the recruitment for clinical trials on blood pressure and other
cardiovascular risk factors and, by implication, the validity of
the findings of such trials all depend on proper measurement and
recording of BP.
There is a dearth of publications on EDP in BP measurement
from Africa, despite the fact that one of the earliest publications
on this clinical entity was from South Africa.
10
We therefore
conducted this study to determine the frequency of EDP in
systolic (SBP) and diastolic blood pressure (DBP) readings taken
by nurses and attending physicians in our hypertension speciality
clinic. We also determined patients’ and physicians’ characteris-
tics influencing the occurrence of EDP.
Methods
We retrospectively reviewed the medical records of patients
with hypertension attending the Hypertension Clinic of Ladoke
Akintola University of Technology Teaching Hospital, Osogbo,
Department of Medicine, Ladoke Akintola University of
Technology, Osogbo, Osun State, Nigeria
OE AYODELE, MB BS, FWACP
OO OKUNOLA, MB, BS, FMCP
AA AKINTUNDE, MB ChB, FWACP
Department of Medicine, University of Ilorin, Ilorin, Kwara
State, Nigeria
EO SANYA, MB ChB, FWACP
