Close

Search

Abstract

Background

We conducted a prospective study to test the validity of a new continuous non-invasive blood pressure (NIBP) monitor (CNAP) (CNAP™ Monitor 500).

Methods

One hundred patients undergoing elective surgery under general anaesthesia were included in the study after informed written consent. The CNAP finger cuffs were placed on the fingers of one arm, an arterial catheter was inserted into the same arm and data were recorded simultaneously. Agreement between invasive arterial pressure (IAP) and blood pressure obtained by CNAP was compared using the Bland–Altman method for repeated measurements. The data from the first 50 patients (software V3.0) were used to improve the software of the CNAP (software V3.5), which was then evaluated in another 50 patients. We defined a clinically acceptable agreement according to the standards of the American Association for the Advancement of Medical Instrumentation for NIBP measurements [limits of agreement (LOA) ±15 mm Hg].

Results

We analysed 524 878 paired measurements in 100 patients. The mean bias of the mean arterial pressure in the first 50 patients was −2.9 mm Hg (sd 10.6 mm Hg, LOA −23.7 to 17.9 mm Hg), and in the consecutive 50 patients (using software V3.5) the bias was −3.1 mm Hg (sd 9.5 mm Hg, LOA −21.6 to 15.4 mm Hg).

Conclusions

The new CNAP monitor showed an agreement with the IAP that is promising but did not match our predefined criteria.

  • The CNAP™ Monitor 500 is a new non-invasive device for continuous arterial pressure measurement

  • This study compared the CNAP with direct arterial pressure measurements, using an adjusted software algorithm

  • Mean bias between CNAP and direct measurements was reasonable, but limits of agreement were wide.

Haemodynamic monitoring is a requirement for every patient in emergency and intensive care medicine, in the operating theatre, during the postoperative period, and during interventions with distinct haemodynamic changes. 1 While oxygen saturation and heart rate are measured continuously, blood pressure is usually assessed non-invasively by intermittent oscillometric measurement every 3–10 min.

For several years, monitors for non-invasive continuous blood measurements have been in development, but have not yet reached clinical practice in anaesthesia. The first devices for the measurement of the arterial pressure waveform at the finger, based on the principle of arterial wall unloading developed by Penaz, 2 were introduced in the early 1980s. 3 In 2007, a new device for continuous non-invasive blood pressure (NIBP) measurement, which can be connected to a standard Draeger monitor, was released (Infinity™ CNAP™ SmartPod™, CNSystems Medizintechnik AG, Graz, Austria). Recently, a study evaluated the CNAP™ pod (Software version 2.9.14) connected to a Draeger Infinity Delta™ monitor (Draeger Medical Systems, Inc., Danvers, MA, USA) with promising results. 4 A stand-alone monitor (CNAP™ Monitor 500) with improved software (V3.0) is also now available. In this study, we tested whether the agreement between invasive arterial pressure (IAP) measurement and the new CNAP has improved, as a result of technical advancements 5 in 50 patients. Based on these data the software was adapted, and evaluated in another 50 patients (V3.5 software).

Methods

After obtaining approval from the local ethics committee board (ref.: 709/2008) and written informed consent, we enrolled 100 patients undergoing major orthopaedic surgery. The inclusion criteria were: the need for intraoperative invasive arterial pressure monitoring, a body weight of >40 kg and age ≥18 yr. Exclusion criteria were contraindications or inability to measure NIBP on the fingers of the arm with the arterial catheter (Table1).

Table 1

Patient characteristics presented as mean (range). Number of patients (n). *Beta blockers, ACE inhibitors, calcium antagonists, angiotensin receptor antagonists

V3.0 V3.5
Number 50 50
Age (range) 65.1 (33–87) 62.9 (30–96)
Sex (male/female) (n) 27/23 30/20
Weight (kg) 76 (48–125) 73 (48–102)
Height (cm) 170 (150–185) 173 (158–190)
BMI (kg m−2) 26 (17.4–46.7) 24 (16.9–38.4)
ASA class I, II (n) 28 27
ASA class III, IV (n) 22 23
Arterial hypertension (n) 33 24
Atrial fibrillation (n) 2 4
No oral vasoactive medication* (n) 18 23
Vascular disease (n) 10 8
Diabetes mellitus (n) 11 9
V3.0 V3.5
Number 50 50
Age (range) 65.1 (33–87) 62.9 (30–96)
Sex (male/female) (n) 27/23 30/20
Weight (kg) 76 (48–125) 73 (48–102)
Height (cm) 170 (150–185) 173 (158–190)
BMI (kg m−2) 26 (17.4–46.7) 24 (16.9–38.4)
ASA class I, II (n) 28 27
ASA class III, IV (n) 22 23
Arterial hypertension (n) 33 24
Atrial fibrillation (n) 2 4
No oral vasoactive medication* (n) 18 23
Vascular disease (n) 10 8
Diabetes mellitus (n) 11 9

Table 1

Patient characteristics presented as mean (range). Number of patients (n). *Beta blockers, ACE inhibitors, calcium antagonists, angiotensin receptor antagonists

V3.0 V3.5
Number 50 50
Age (range) 65.1 (33–87) 62.9 (30–96)
Sex (male/female) (n) 27/23 30/20
Weight (kg) 76 (48–125) 73 (48–102)
Height (cm) 170 (150–185) 173 (158–190)
BMI (kg m−2) 26 (17.4–46.7) 24 (16.9–38.4)
ASA class I, II (n) 28 27
ASA class III, IV (n) 22 23
Arterial hypertension (n) 33 24
Atrial fibrillation (n) 2 4
No oral vasoactive medication* (n) 18 23
Vascular disease (n) 10 8
Diabetes mellitus (n) 11 9
V3.0 V3.5
Number 50 50
Age (range) 65.1 (33–87) 62.9 (30–96)
Sex (male/female) (n) 27/23 30/20
Weight (kg) 76 (48–125) 73 (48–102)
Height (cm) 170 (150–185) 173 (158–190)
BMI (kg m−2) 26 (17.4–46.7) 24 (16.9–38.4)
ASA class I, II (n) 28 27
ASA class III, IV (n) 22 23
Arterial hypertension (n) 33 24
Atrial fibrillation (n) 2 4
No oral vasoactive medication* (n) 18 23
Vascular disease (n) 10 8
Diabetes mellitus (n) 11 9

The CNAP finger cuffs were placed on the fingers of one hand and a radial arterial catheter (20-gauge, 45 mm long and connected to a transducer by a 150 cm long line with low compliance) was sited in the same arm. The pressure transducer was placed at the mid-thoracic point (zero point) and the system flushed. A standard flush test was performed to determine the natural frequency and the damping coefficient of the invasive catheter.

The CNAP™ 500 Monitor System consists of a sensor placed on the finger (second and third digit), an NIBP cuff for calibration and the CNAP monitor. The continuous measurement is performed on one finger at a time and changes to the other finger after a defined period (default setting of 30 min) which allows long-term measurement of up to 24 h on the one hand.

The finger is illuminated with infrared light. A part of the light is absorbed by the pulsatile blood volume such that the transmitted light through the finger is a measure of the pulsatile blood volume. The finger cuff is inflated to keep the blood volume and accordingly the transmission of light constant. 6 The primary signal used for blood pressure measurement is the pressure in the finger cuff; this is calibrated every 15 min with the standard NIBP cuff on the same arm (Table2).

Table 2

Systolic, mean, and diastolic pressures recorded by the arterial catheter and the CNAPTM 500 using V 3.0 (n=50 patients) and V 3.5 software (n=50), expressed as mean (sd). Bias, mean of the differences (arterial catheter−CNAP); sd, standard deviation of the differences corrected for repeated measurements; LOA, level of agreement, calculated as bias (sd) of the differences multiplied by 1.96

Blood pressure (mm Hg) Arterial catheter CNAP Bias sd LOA
Lower limits Upper limits
Software V3.0. (249 508 beats analysed Systolic 104.5 (17) 103.6 (17.4) 3.4 16 −27.9 34.8
Diastolic 56.4 (8.8) 59.7 (10.6) −4.4 10.8 −25.5 16.8
Mean 73.3 (8.6) 76.4 (12.8) −2.9 10.6 −23.7 17.9
Software V3.5 237 562 beats analysed Systolic 103.5 (16.3) 102.6 (15.8) 0.9 13.2 −24.9 26.8
Diastolic 56.6 (8.6) 59.4 (10) −2.8 8.6 −19.7 14.1
Mean 73 (11.2) 75.81 (11.7) −3.1 9.45 −21.6 15.4
Blood pressure (mm Hg) Arterial catheter CNAP Bias sd LOA
Lower limits Upper limits
Software V3.0. (249 508 beats analysed Systolic 104.5 (17) 103.6 (17.4) 3.4 16 −27.9 34.8
Diastolic 56.4 (8.8) 59.7 (10.6) −4.4 10.8 −25.5 16.8
Mean 73.3 (8.6) 76.4 (12.8) −2.9 10.6 −23.7 17.9
Software V3.5 237 562 beats analysed Systolic 103.5 (16.3) 102.6 (15.8) 0.9 13.2 −24.9 26.8
Diastolic 56.6 (8.6) 59.4 (10) −2.8 8.6 −19.7 14.1
Mean 73 (11.2) 75.81 (11.7) −3.1 9.45 −21.6 15.4

Table 2

Systolic, mean, and diastolic pressures recorded by the arterial catheter and the CNAPTM 500 using V 3.0 (n=50 patients) and V 3.5 software (n=50), expressed as mean (sd). Bias, mean of the differences (arterial catheter−CNAP); sd, standard deviation of the differences corrected for repeated measurements; LOA, level of agreement, calculated as bias (sd) of the differences multiplied by 1.96

Blood pressure (mm Hg) Arterial catheter CNAP Bias sd LOA
Lower limits Upper limits
Software V3.0. (249 508 beats analysed Systolic 104.5 (17) 103.6 (17.4) 3.4 16 −27.9 34.8
Diastolic 56.4 (8.8) 59.7 (10.6) −4.4 10.8 −25.5 16.8
Mean 73.3 (8.6) 76.4 (12.8) −2.9 10.6 −23.7 17.9
Software V3.5 237 562 beats analysed Systolic 103.5 (16.3) 102.6 (15.8) 0.9 13.2 −24.9 26.8
Diastolic 56.6 (8.6) 59.4 (10) −2.8 8.6 −19.7 14.1
Mean 73 (11.2) 75.81 (11.7) −3.1 9.45 −21.6 15.4
Blood pressure (mm Hg) Arterial catheter CNAP Bias sd LOA
Lower limits Upper limits
Software V3.0. (249 508 beats analysed Systolic 104.5 (17) 103.6 (17.4) 3.4 16 −27.9 34.8
Diastolic 56.4 (8.8) 59.7 (10.6) −4.4 10.8 −25.5 16.8
Mean 73.3 (8.6) 76.4 (12.8) −2.9 10.6 −23.7 17.9
Software V3.5 237 562 beats analysed Systolic 103.5 (16.3) 102.6 (15.8) 0.9 13.2 −24.9 26.8
Diastolic 56.6 (8.6) 59.4 (10) −2.8 8.6 −19.7 14.1
Mean 73 (11.2) 75.81 (11.7) −3.1 9.45 −21.6 15.4

In contrast to previous devices using the vascular unloading technique, CNAP uses a special algorithm to prevent signal drift. Arterial pressure fluctuations within a frequency band <0.02 Hz (cycle duration of ∼50 s) are detected and corrected for vasomotor changes. 6 The values of the CNAP system were compared with the invasive arterial blood pressure assessed on the same arm on a beat-to-beat basis. Using the data from the first 50 patients, the algorithm for the CNAP monitor was adjusted (software version 3.0 to 3.5) and the next 50 patients were analysed as described previously. The main difference between the software versions 3.0 and 3.5 is the calibration logic. In version 3.0, the uncalibrated finger values were calibrated to mean arterial pressure of the upper arm cuff, and the systolic and diastolic values were adjusted. In version 3.5, the finger values are calibrated to systolic and diastolic upper arm pressure, and the mean pressure is adjusted according to proprietary calibration logic.

Secondly, in version 3.5, a quality indicator of the signal is calculated during the finger search period, and the device automatically switches to the other finger if the signal is too weak. This prevents unreliable measuring phases.

Data collection

Each patient received a radial arterial catheter, a standard adult oscillatory blood pressure cuff and the CNAP. The CNAP was connected to the patient monitor (Draeger Medical Systems, Inc., Danvers, MA, USA) and all synchronized signals were exported from the patient monitor automatically using data collection software. The invasive and non-IAP curves were sampled at a frequency of 100 Hz. The raw data of the analogue curve were processed using an open source beat detector 7,8 for the definition of every single cardiac cycle. The filter algorithms used to automatically reject non-physiological values from the statistical analysis, such as arterial damping and technical errors (for example zeroing, drawing of blood samples) were:

  • Invasive and CNAP pulse pressure had to be between 10 and 150 mm Hg

  • Systolic pressure had to be between 20 and 300 mm Hg for both methods

  • Diastolic pressure had to be between 15 and 250 mm Hg for both methods

Sample size

The European Society for Hypertension suggested that, for the validation of new NIBP devices, 33 subjects and 99 paired measurements are required. 5 To cope with non-predictable technical problems, we decided to include 50 patients in both parts of the study (software versions V3.0 and V3.5).

Statistical analysis and normative considerations

There are two different relevant guidelines for the validation of blood pressure measuring devices: one from the Association for the Advancement of Medical Instrumentation 9(AAMI) and the other from the European Working Group for the Validation of Blood Pressure Measuring Devices. 5 The AAMI standard clearly states that it does not cover finger blood pressure monitors, as a greater bias has to be expected. Both guidelines have been developed for the comparison of automated NIBP devices with the auscultatory (Riva-Rocci) method and both describe the problems of comparing IAP with automated NIBP devices. There are no direct recommendations for validation of a continuous non-IAP device. Although this guideline may not be directly applicable, we decided to compare our results with the AAMI standards. 5 A device is considered acceptable if its estimated probability of a tolerable error is at least 85%. This leads to acceptable standard deviations from 4.81 to 6.95 mm Hg depending on the sample mean error. A device is automatically rejected if the sample mean error is greater than ±5 mm Hg. Our predefined aim was to have a sample mean error of <5 mm Hg and limits of agreement (LOA) according to AAMI standards. The agreement was described using the Bland-Altman method {mean of the differences and LOA [mean of the differences (sd) of the differences × 1.96] of paired measurements}. 10 Data were analysed using the SPSS statistical program, Version 16.0 (SPSS Inc.,  Chicago, IL, USA).

To correct for the effect of repeated measurements, a random effects model was fitted to the data of the arterial pressure for both methods and their difference. 11 The variances have been estimated based on 100 randomly chosen samples from each subject.

Results

For evaluation of the CNAP (software V3.0), we analysed 271 025 paired measurements from 50 patients. Of these, 21 517 (7.9%) beats were excluded. The algorithm of the monitor was then adapted and data from another 50 patients including 253 853 beats were analysed (V3.5 software). Of these, 16 291 beats (6.8%) were excluded. The average number of beats excluded per patient was 21.8% (range 6.5–49.7%).

The CNAP software V3.0 overestimated the mean arterial blood pressure by 2.9 (sd 10.6 mm Hg, LOA −23.7 to 17.9 mm Hg). The mean bias for the CNAP software V3.0 (IAP–CNAP) was −2.9 mm Hg (sd 10.6 mm Hg, LOA −23.7 to 17.9 mm Hg).

The CNAP software V3.5 overestimated the mean arterial blood pressure by 3.1 mm Hg (sd 9.5 mm Hg, LOA −21.6 to 15.4 mm Hg) compared with the IAP on the same arm. The bias for the CNAP software V3.5 (IAP–CNAP) was −3.1 mm Hg (sd 9.5 mm Hg, LOA −21.6 to 15.4 mm Hg).

Discussion

This is the first report of the CNAP™ Monitor 500 (V3.5), a continuous NIBP monitor which has real-time beat-to-beat measurement.

We found a clinically acceptable agreement between mean arterial pressure recorded by the CNAP Monitor 500 (V3.0) and direct intra-arterial measurement. After adjustment of the software (from V3.0 to V3.5), based on the data of 50 patients, the agreement improved. The correlation and agreement between the invasive and the non-invasive continuous mean and diastolic pressure measurements are promising, but do not comply with the requirements of the AAMI standard using V3.5 software (Fig.1).

Fig 1

Bland–Altman plot showing the agreement of the mean arterial pressure. Horizontal lines indicate mean of the differences (IAP–CNAP) and LOA [bias (sd) of the differences × 1.96].

Bland–Altman plot showing the agreement of the mean arterial pressure. Horizontal lines indicate mean of the differences (IAP–CNAP) and LOA [bias (sd) of the differences × 1.96].

Bias was greater for systolic pressure indicating a lower agreement between CNAP and the invasive method, consistent with previous data. 12

In recent years, a number of devices have become available for measuring blood pressure non-invasively and continuously. The Finapres monitor (Ohmeda, Denver, CO, USA) was the best known in anaesthesia. The basic principle behind finger blood pressure devices such as the Finapres or the CNAP device is the vascular unloading technique introduced by Penaz. 13,14 A difference between the CNAP and the Finapres monitor is the calibration to NIBP cuff (adjustable from 5 to 60 min), which results in real beat-to-beat values and allows verification of the continuous signal between defined intervals.

In contrast to other devices, CNAP uses digital control loops and mathematical algorithms that deal with online detection of vasoactivity. The CNAP has been especially designed for vasomotor changes occurring during anaesthesia. 6 Several conditions such as vasoconstriction, hypothermia, or vascular disease may interfere with the measurement of the plethysmogram although our study was not designed to detect such interference. We included patients with a wide variety of co-morbidities, such as diabetes mellitus, arterial fibrillation, or atherosclerosis. Moreover, 22% of the patients needed vasopressor therapy during the study period. Our data therefore represent a typical patient sample of daily anaesthetic practice. In addition, it is important to use the right-sized finger cuff. The CNAP Monitor provides three different sizes of finger cuffs to guarantee a good quality of measurement of the CNAP signal. The main advantages of the CNAP compared with IAP are that it is faster [mean (sd) 150 (30) compared with 310 (60)] until blood pressure measurement, it carries no risk of needle stick injury, and its usage is pain free for the patient. Although arterial catheter related complications are rare, 15–17 they can be safely avoided, if non-invasive monitoring is used.

The age of our patients is continuously rising, and the number of patients with cardiac and cerebro-vascular disease is increasing. Intraoperative hypotension has been identified as an independent predictor even for 1 yr mortality 18 and intraoperative cardiac arrest is often preceded by hypotension. 19 The CNAP™ Monitor 500 offers the possibility of continuous haemodynamic measurement compared with the standard NIBP where a 3–5 min interval between recordings is used.

It is well known that arterial pressure measured in different body regions differs significantly. 5 Additionally, different methods for measuring blood pressure at the same site lead to different values. 12 This variability of methods and sites has to be taken into account when analysing a new method. The relevant guidelines for the validation of blood pressure measuring devices have been developed for comparison of automated NIAP devices with the auscultatory (Riva-Rocci) method. Although these guidelines may not be applicable to the comparison of IAP with automated NIBP devices, we decided to compare our results with the strict AAMI standards. Therefore, the LOA between CNAP and IAP may fail our predefined criteria, but the results are still valuable in daily clinical practice.

Further studies are necessary to evaluate the advantages and limitations of this technology in patients with common conditions such as severe atherosclerosis, aortic stenosis, or chronic heart failure. The first results of the CNAP™ Monitor 500 (V3.5) in 50 patients are promising but did not reach our predefined criteria.

Declaration of interest

None declared.

Funding

Equipment used in this study was provided by CNSystems, Graz, Austria. We received no additional funding.

Acknowledgement

Equipment used in this study was provided by CNSystems, Graz, Austria.

References

1

,  ,  , et al.

Continuous non-invasive arterial pressure technique improves patient monitoring during interventional endoscopy

,

Int J Med Sci

,

2009

, vol.

6

 (pg.

37

-

42

)

2

.

[A portable finger plethysmograph.]

,

Scr Med (Brno)

,

1954

, vol.

27

 (pg.

213

-

34

)

3

,  ,  ,  .

Fifteen years experience with finger arterial pressure monitoring: assessment of the technology

,

Cardiovasc Res

,

1998

, vol.

38

 (pg.

605

-

16

)

4

,  ,  , et al.

Precision and accuracy of a new device (CNAP™) for continuous non-invasive arterial pressure monitoring: assessment during general anaesthesia

,

Br J Anaesth

,

2010

, vol.

105

 (pg.

264

-

72

)

5

,  ,  , et al.

Working group on blood pressure monitoring of the European Society of Hypertension International Protocol for validation of blood pressure measuring devices in adults

,

Blood Press Monit

,

2002

, vol.

7

 (pg.

3

-

17

)

6

,  ,  , et al.

Continuous non-invasive blood pressure monitoring using concentrically interlocking control loops

,

Comput Biol Med

,

2006

, vol.

36

 (pg.

941

-

57

)

7

,  ,  .

PhysioNet: a research resource for studies of complex physiologic and biomedical signals

,

Comput Cardiol

,

2000

, vol.

27

 (pg.

179

-

82

)

8

,  ,  .

BioSig: the free and open source software library for biomedical signal processing

,

Comput Intell Neurosci

,

2011

, vol.

2011

 pg.

935364

 

9

ANSI/AAMI

,

American National Standard for Manual, Electronic, or Automated Sphygmomanometers.

,

2002

Arlington, VA: Association for the Advancement of Medical Instrumentation

10

,  .

Measuring agreement in method comparison studies

,

Stat Methods Med Res

,

1999

, vol.

8

 (pg.

135

-

60

)

11

,  .

Using the Bland–Altman method to measure agreement with repeated measures

,

Br J Anaesth

,

2007

, vol.

99

 (pg.

309

-

11

)

12

,  .,

Direct Measurement of Blood Pressure. Handbook of Hypertension

,

1991

Amsterdam: Elsevier

(pg.

315

-

33

)

13

.

[Current photoelectric recording of blood flow through the finger]

,

Cesk Fysiol

,

1975

, vol.

24

 (pg.

349

-

52

)

14

,  ,  .

Beitrag zur fortlaufenden indirekten Blutdruckmessung

,

Z innere Med

,

1976

, vol.

31

 (pg.

1030

-

3

)

15

,  ,  ,  ,  .

Arterial catheter-related infection of 2,949 catheters

,

Crit Care

,

2006

, vol.

10

 pg.

R83

 

16

,  ,  .

Clinical review: complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine

,

Crit Care

,

2002

, vol.

6

 (pg.

199

-

204

)

17

,  ,  , et al.

Incidence and outcome of radial artery infections following cardiac surgery

,

Ann Thorac Surg

,

2003

, vol.

76

 (pg.

801

-

4

)

18

,  ,  ,  .

Anesthetic management and one-year mortality after noncardiac surgery

,

Anesth Analg

,

2005

, vol.

100

 (pg.

4

-

10

)

19

,  ,  , et al.

Predictors of survival following cardiac arrest in patients undergoing noncardiac surgery: a study of 518 294 patients at a tertiary referral center

,

Anesthesiology

,

2003

, vol.

99

 (pg.

259

-

69

)

© The Author [2012]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please email: journals.permissions@oup.com

© The Author [2012]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please email: journals.permissions@oup.com