German Hypertension League DHL® German Hypertension Society

---

German version Last Update: May, 2003

Clinical Evaluation of blood pressure measurement devices
for the attainment of the test seal
of the German Hypertension Society

1 Preface / Introduction 

The clinical test must be executed according to DIN EN 540. The following test program makes this possible. However since it is to be checked whether a device can be awarded a special test seal of the "Deutsche Liga zur Bekämpfung des hohen Blutdrucks" (German league against Hypertension) it exceeds the DIN regulation mentioned with respect to individual requirements. This concerns among other things the selection and instruction of the examiners, the composition of the test subject group and the evaluation of the measuring results. Apart from the average systolic and diastolic differences between device measurements and comparison measurements the correspondence between systolic and diastolic errors of measurement is evaluated.
For reasons of equal treatment of all devices independent from the application place (upper arm, wrist, finger) simultaneous comparative measurements are not performed during the clinical test which exclusively uses indirect, non-invasive measurements. All measuring comparisons are always carried out sequentially.
The aforementioned DIN regulation applies to questions not regulated in this protocol e.g. concerning special device types, such as for instance for long-term measurement.

2 Test procudure 

2.1 General information 

Due to the meaning of the clinical test the measurement of the blood pressure must be executed by the examiners (at least three) particularly carefully and precisely. Mercury or aneroid manometers with a scale division value of 2 mm Hg are suitable as excess pressure reference measuring instruments. The calibrated excess pressure reference measuring instruments must comply with the requirements of DIN EN 1060-1 bis E DIN EN 1060-3, but may only show margins of error of 1 mm Hg for falling cuff pressure prior to the start of the clinical test¹.

The auscultatory blood pressure measurement by the examiners must be performed according to the recommendations of the World Health Organisation [1] The recommendations mean among other things that the cuff should be positioned such that the lower edge of the cuff is located about 2.5 cm above the elbow, the outflow rate of the cuff pressure amounts from 2 mm Hg/s to 3 mm Hg/s at the most, that the first and last audible noise (Phase I or Phase V, defined in [2]) should be used to determine the systolic and diastolic blood pressure. If the blood pressure is taken under physical stress Phase IV can additionally be used to determine the diastolic blood pressure value.

The examiners can be qualified medical personnel but also non-professionals who were instructed in regard to how to take the blood pressure. In particular laymen must be trained with respect to their measuring accuracy as examiners prior to the start of the test. For this, comparative measurements with experienced examiners as well as videos and films with measuring sequences are suitable. For tests, only those examiners who produce differences of over 10 mm Hg for only 5% and differences of over 5 mm Hg for 10% of the comparative measurements during simultaneously conducted comparative measurements with experienced examiners are used.

The auscultatory reference blood pressure readings described in the following should always be performed by two examiners using a twin stethoscope for whom the display of the non-invasive blood pressure measuring device to be tested is not visible. The auscultated blood pressure values which were determined separately by the examiners must be documented independently. The third examiner operates the non-invasive blood pressure measuring device to be tested. The examiners alternate according to the procedure described in the appendix. The auscultatory reference value is the average value of the two values auscultated by the examiners.

When taking readings the examiners must avoid parallax errors. If aneroid manometers are used for reference measurements, each of the examiners must use his own for the reading in order to avoid parallax errors. For mercury manometers the reading can possibly be taken by both examiners using one device.

NOTE: Reading the values on the excess pressure reference measuring instrument should be carried out as precisely as possible. Roundings to zero and five affect the test result negatively.

2.2 Test procedure - Sequential comaprison blood pressure measurement at the same arm

2.2.1 Application of the test procedure

The testing procedure is among other things suitable for blood pressure measuring devices which operate

• with continuous decrease of pressure (decrease of pressure rate > mm Hg/s) and/or
• with a pulse-controlled decrease of pressure (decrease of pressure rate  > 3 mm Hg per heart beat) and/or
• with gradual decrease of pressure,
• or measure at the wrist or finger.

This testing procedure is also suitable for blood pressure measuring equipment which measures during the inflation phase.

2.2.2 Testing device 

2.2.3 Selection of test subjects 

The selection of test subjects as well as their number depends on the intended purpose of the device to be tested. For the clinical test at least 96 test subjects are required who must fulfil the following requirements:

Table 1: Distribution of the 96 test subjects (minimum numbers)

The criterion for the selection of the blood pressure level is the average value of the last two measurements of in total three random values, even if the values fall into another group during the test measurements.

The test subjects specified in Table 1 do not have to fulfil both (systolic, diastolic) criteria concerning their blood pressure. With a blood pressure of p_sys / p_dia = 138/92 mm Hg a test subject of the age of 25 can for example be placed in two different groups but the required minimum number of test subjects must nevertheless be reached. The average blood pressure level during the test should correspond to the blood pressure criterion of the respective 8 subgroups for at least 75% of the test subjects.

Blood pressure measuring devices which measure on the wrist or on the finger must be additionally tested on 20 test subjects with manifest diabetes mellitus. These must fulfil the following requirements:

Additionally, 30 females subjects can be included in the test procedure, among them at least ten who are pregnant in the second or third trimester.

A test seal for a blood pressure measuring device tested in the subgroup of pregnant women can only be awarded when the device has passed sucessfully the complete validation procedure.

Exceptionally, tests of a blood pressure measuring device on the pregnant subgroup can be performed before or after the rest of the tests; however, passing of the test procedure for the subgroup of pregnant women does not imply the use of the test seal.

The costs to test the subgroups accrue in addition to the costs of the main test and are assessed separately.

2.2.4 Cuffs

The cuff of the blood pressure measuring device to be tested must be selected and applied according to the manufacturer's recommendations (see 9.2 of DIN EN 1060-1:1995). If the manufacturer's recommendations do not match the recommendations of "Deutsche Liga zur Bekämpfung des hohen Blutdrucks" (German League against Hypertension) the "Test seal" League work group decides on the further procedure.

If cuffs for upper arm circumferences over 33 cm are offered for the non-invasive blood pressure measuring equipment then measurements must be carried out on 20 additional test subjects.

These must fulfil the following requirements:

2.2.5 Test execution 

Figure 1: Comparison blood pressure measurement on the same arm.

The test must be performed under conditions which are adapted to the intended purpose designated by the manufacturer, see Fig. 1.

At least three trained examiners must execute more or less the same number of measurements according to the log in Appendix II. A recovery period of the test subject of at least 1 min must be observed between the measurements. The temporal sequence of the measurements should change between auscultation and test subject in accordance with the log in Appendix II.

As the comparison of the values measured by the blood pressure measuring device with the blood pressure values determined auscultatorily (average value of the two blood pressure values auscultated using a twin stethoscope) cannot be effected by a simultaneous measurement on the same arm, but only sequentially one after the other on the same arm, it must be observed that the two measurements take place under approximately the same conditions.

The measurements of each test subject must be executed within approximately 40 min.

2.2.6 Data analysis

The maximum difference between the two examiners who execute the auscultatory blood pressure measurement using a twin stethoscope and manometer may amount to 4 mm Hg at the most. A comparative measurement is only included in the analysis if the 4 mm Hg criterion is valid for both the auscultated systolic and the diastolic blood pressure values. Only test subjects for whom at least 3 comparative measurements comply with the 4 mm Hg criterion are included in the analysis.

As the measurements are executed in temporal succession it is important that stable hemodynamic conditions are present. A test subject can only be included in the analysis if the reference values, valid in accordance with the 4 mm Hg criterion, of the entire series of measurements deviate systolically by no more than 16 mm Hg and diastolically by no more than 12 mm Hg from each other.

Obvious runaway values (or data transfer errors) are removed from the evaluation. To do this, systolic and diastolic differences between the unit under test and control measurement are eliminated which deviate from the mean difference by more than two standard deviations (2 sd-criterion).

When an evaluation is made according to the DIN 58 130 - protocol in addition to the evaluation of the test seal protocol the first 85 of 96 (resp. 120) test persons of the test seal examination are to be selected that meet the other criteria of the DIN protocol.

The analysis is to be executed as follows:

First the auscultated reference value must be determined:

Subsequently the following differences must be determined:

From the differences of all values of the test subjects the mathematical average values

and the appropriate empirical standard deviations

must be calculated.

The results of the analysis of all measurements must be within the limits determined in 7.9 by E DIN EN 1060-3:1995-11.

In addition points are assigned for each comparative measurement, e.g. four, if both, the systolic and diastolic reading difference, i.e. dp_SYS and dp_DIA, are equal to or smaller than 5 mm Hg; three, if one of the two reading differences is equal to or smaller than 5 and the other one is equal to or smaller than 10 mm Hg; two points, if both reading differences are equal to or smaller than 10 mm Hg. For all other comparative measurements the device receives zero points.

The calculation of the mean difference between test and control measurement and it's standard deviation according to the DIN 58 130 - protocol is made on the basis of the first group of three contiguous measured values, which meet the remaining criteria (4 mm Hg and 8/12 mm Hg) of the test protocol.

2.2.7 Evaluation

The test is passed if the aforementioned limits are adhered to. At present they amount to:

    5 mm Hg,

    5 mm Hg,

s(dp_SYS)    8 mm Hg,

s(dp_DIA)    8 mm Hg,

In addition a device must have reached half (50%) of the maximum attainable points score.

2.2.8 Documentation 

For each test subject the following details must be recorded for each series of measurements:

For each series of measurements the following details of the test subject must be recorded:

---

Literature:

---

Appendix I: Abbreviations

---

Appendix II: Log of the clinical test

Test subject-No. ...
 Page 1

Log of the clinical test

Table 1: Details of the test subject

Test subject-No.
Test subject identification (e.g. initials)
Date of measurement
Sex	male / female
Age (years)
Size (cm)
Weight (kg)
Measuring point (Test subject)	right / left upper arm / wrist / finger
circumference at the measuring points of the test subject (cm)	right:                  /left:
Cirumference of right    /left upper arm (cm)	right:                  /left:
Medication
Comments

---

Test subject-No. ...
Page 2

Table 2: Test subject

Details as for Test subject-No.
Test device
Manufacturer
type name
Serial number
Cuff applied	type name:          size (cm):
Software version (if specified)
Other

Table 3: Examiner

Details as for Test subject-No.
Examiner A
Examiner B
Examiner C
Hospital Address: Telephone Fax

Test subject-No. ...
Page 3

Tab. 4: Devices for the reference measurement

Details as for Test subject-No.
Reference measuring devices
Reference measuring device I (error of display max. 1 mm Hg)
Design	with mercury / aneroid manometer
manufacturer
type name
Serial number
Date of the last instrumentation check (Calibration)
Insufflator size of the cuff (cm)	............................x
Reference measuring device II (error of display max. 1 mm Hg), if required
Design	with mercury / aneroid manometer
Manufacturer
Type name
Serial number
Date of the last instrumentation check (calibration)
Insufflator size of the cuff (cm)	............................x
Twin stethoscope
Model name

Test subject-No. ...
Page 4

Tab. 5: Sequential comparative measurements (the calculation of the average value can occur seperately);
recovery period between the measurements at least 1 min.