CORDIS Archive

Non-invasive sphygmomanometers (blood pressure devices)

1. CONFORMITY WITH THE WORK PROGRAMME

This topic falls under the Competitive and Sustainable Growth Programme, generic activity Measurement and Testing. Specifically, it is related to Objective GROW-2000-6.2.1 Methodologies to support standardisation and Community policies for which expressions of interest have been called.

2. KEYWORDS

medical devices, test procedures, sphygmomanometer, accuracy, simulators, blood pressure

3. SUMMARY OF JUSTIFICATION

The Council Directive 93/42/EEC of 14 June 1993 concerning medical devices states in Annex I, Essential Requirements, clause 10.1 and 13.6 (p) that the accuracy of devices with a measuring function has to be indicated. The determination of the overall system accuracy of automated non-invasive sphygmomanometers according to the harmonised standard EN1060 is difficult, since clinical trials are required because technical tests with appropriate simulators are not available at present. None of the currently available simulators are able to substitute subjects, because the generated signals are ideal, not reflecting the special human characteristics and its variability.

4. BACKGROUND

On its last meeting (September 28/29, 1999 in Stockholm) CEN TC 205, "Non-active medical devices" has accepted a new work item titled "Test procedures to determine the overall system accuracy of automated non-invasive sphygmomanometers" proposed by CEN TC 205 WG 10 "Non-invasive sphygmomanometer". At present only clinical trials, which compare the measurement of the device with the measurement of a physician are suitable to determine the overall system accuracy, since currently available simulators are not able to replay realistic human signals [1]. Although national standards (in Germany, UK, USA) require 85 subjects to perform the clinical trial, this number is – due to time and costs – already the absolute minimum. Both, manufactures and physicians are interested in reducing the number of subjects, therefore additional tests with simulators substituting human subjects must be introduced.

5. ECONOMIC AND SOCIAL BENEFITS

About 15 % of the adult European population suffer from high blood pressure. Every year in Germany alone more than 100,000 automated non-invasive sphygmomanometer for professional use (medical areas) and more than 1,000,000 home use devices are sold, only clinical thermometers are sold in greater numbers. The conformity tests for the CE mark of automated non-invasive sphygmomanometers require tests of the overall system accuracy. So far, only clinical trials are suitable to determine these data. These trials are time and cost consuming (ca. 3 – 6 month, ca. 5000 – 20000 €) and still limited to the most probable cases (diseases, diagnostic aspects). Tests with adequate simulators should allow to identify the limits of application of the sphygmomanometers and should be a helpful tool for the manufacturer to improve the accuracy and the performance in general. Higher performance would imply an increase in quality for diagnosis and clinical treatment of patients, reducing costs due to mistreatment and additional measurements and tests.
Automated non-invasive sphygmomanometers for professional use are manufactured by several European companies (Siemens, Datex-Ohmeda, Kontron etc.). They are sharing the internal market to high percentage. About 20 years ago European manufacturers also had high market shares for home use devices, since that time Japanese and recently other Asian manufacturers (Taiwan, Korea, China) share about 90% of the internal market. The competition at present is only for the lowest price, not for the best quality or performance. The European manufacturers are hardly able to compete for the price but for quality, therefore an increase in this field by more objective test procedures with simulators would imply quality in diagnosis.

6. SCIENTIFIC AND TECHNOLOGICAL OBJECTIVES

The simulator shall be able to test non-invasive oscillometric sphygmomanometers. It shall generate pressure oscillations identical to those obtained on the human upper arm or wrist. These oscillations are small pressure changes appearing in the cuff during the measurement cycle of the sphygmomanometer, starting at cuff pressures higher than the systolic blood pressure and disappearing at cuff pressures lower than the diastolic blood pressure. These oscillations are evaluated by the software of the sphygmomanometer and the systolic and diastolic blood pressures are determined and displayed. The simulator shall be able to generate these raw data of such a high quality, that the sphygmomanometer under test accepts them as "human" data. The raw data generated by the simulator shall be recorded from human subjects reflecting typical subjects and measuring conditions. They have to be validated by physicians, stating for each raw data the systolic and diastolic blood pressures. Since modern sphygmomanometer sometimes take into account also other parameter such as ecg etc. when evaluating the oscillations, the recorded data should also include these data too, if they are available.
A digital database shall represent typical oscillations taken for representative groups of subjects, due to age, sex, disease, measuring conditions (rest, movement, ventilated etc.), artefacts (from the patient or the environment). The raw data recorded from each subject (oscillations, cuff pressure, optional: Korotkoff sounds, invasive blood pressure, ecg) shall be validated by physicians before becoming reference values for the simulation. Also additional tests with simulator are required to proof, that it is able to substitute human subjects.

The database shall include:
a) upper arm simulation: ca. 150 subjects (low/normal/high blood pressure: 20/20/40; male/female equally distributed; age: up to 20 years 30 people; 20-40 years 30 people; 41-70 years 30 people; over 71 years 30 people; high pulse rate: 30 people; artefacts: 30 people;
b)wrist simulation: ca. 150 people (distribution equal upper arm simulation);
c)neonates: ca. 30 people;
d)special groups: 150 people (physical load: 50 people; anaesthesia: 50 people (ventilated: 20, non-ventilated: 30); various: 50 people (arrhythmia: 20, …). To get these approximately 500 representative simulation data at least 1500 data have to be recorded and validated.

Further proposals should consider the improvement of the simulation system, in particular the hardware to generate the oscillations. A revision of the operating software for the new designed hardware can also be expected.

7. TIME SCALE

The results will be exploited by amendment to standards and their update. As a consequence the project has to be finished within 3 years.

References