Patient characteristics
From 18th March 2022 to 30th November 2022, 272 patients were eligible. Among them, 50 patients (15.4%) were included in the study. No statistical difference in baseline characteristics was found between included and excluded patients. Excluded patients were more frequently admitted in operating room for thrombectomy (19.8% vs. 0%, p < 0.001) or diagnostic arteriography (58.6% vs. 6%, p < 0.001). Included patients were more frequently admitted for the treatment of SAH (12.2% vs. 38%, p < 0.001). An additional file illustrate the comparison between included and excluded patients [see Additional file 1]. Most patients were excluded due to the absence of invasive blood pressure monitoring during the procedure. Another subject was also excluded due to the impossibility of obtaining a valuable ABP curve on the Nexfin device. An additional file show the flow diagram of the study [see Additional file 2]. Enrolled patients were mainly women (62%), with a mean age of 58 (± 12) years and were mostly ASA 2 (28%) or ASA 3 (60%). Half of patients had hypertension or other cardiovascular comorbidities. Nearly all patients were admitted in the operating room for a cerebral embolization (94%) under general anaesthesia (96%). Nineteen patients (38%) needed an emergency procedure for SAH. Baseline characteristics are summarized in Table 1.
Table 1
| Overall |
| n = 50 |
Age (years) | 58 (12) |
Gender (men/women) | 19/31 |
Weight (kg) | 71.1 (16.5) |
BMI (kg/m2) | 25.2 (5.2) |
Core temperature (°C) | 36.5 (0.6) |
SAPS II | 23.8 (8.3) |
Comorbidities | |
Cardio-vascular | 25 (50) |
Myocardial infarction | 4 (8) |
Arrhythmia | 1 (2) |
Hypertension | 25 (50) |
Arteritis | 1 (2) |
COPD | 3 (6) |
Diabetes | 1 (2) |
Chronic kidney disease | 5 (10) |
Medication | |
Anti-hypertensive agents | 20 (40) |
Beta-blockers | 2 (4) |
ASA score | |
1–2 | 31 (63.2) |
3–5 | 19 (36.8) |
Emergency procedure for SAH | 19 (38) |
Fisher score | |
1–2 | 2 (10.6) |
3 | 5 (26.3) |
4 | 12 (63.2) |
WFNS score | |
1 | 14 (73.7) |
2 | 1 (5.3) |
4 | 2 (10.5) |
5 | 2 (10.5) |
Type of procedure | |
Embolization | 47 (94) |
Other | 3 (6) |
General anaesthesia | 48 (96) |
Norepinephrine infusion | 17 (34) |
Data are given as n (%) or mean ± SD. ASA: American Society of Anaesthesiologists; BMI: Body Mass Index; COPD: Chronic Obstructive Pulmonary Disease; SAH: Sub-Arachnoid Haemorrhage; SAPS: Simplified Acute Physiology Score; SD: Standard Deviation; WFNS: World Federation of Neurologic Surgeons.
Accuracy of ABP measurements
In our dataset, we recorded 380 different paired ABP measurements with the Nexfin and the arterial catheter. Compared to invasive measurements, the Nexfin device showed a good correlation for measures of SAP (r2 = 0.78, p < 0.001) and MAP (r2 = 0.82, p < 0.001) (Fig. 1A and 1D). Between the two devices (arterial catheter and the Nexfin), Bland-Altman analysis showed a mean bias of 9.6 mmHg (-15.6 to 34.8 mmHg) for SAP and − 0.8 mmHg (-17.2 to 15.6 mmHg) for MAP (Fig. 2A and 2D). Three hundred and seventy one different paired ABP measurements between intermittent arm cuff and arterial catheter were also recorded. Compared to invasive measurements, the Arm Cuff showed a good correlation for SAP (r2 = 0.80, p < 0.001) and MAP (r2 = 0.74, p < 0.001) (Fig. 1B and 1E). Between the two devices (arterial catheter and Arm Cuff), Bland-Altman analysis showed a mean bias of 5,8 mmHg (-30.4 to 22.9 mmHg) for SAP and − 1,4 mmHg (-17.3 to 14.4 mmHg) for MAP (Fig. 2B and 2E). Three hundred and seventy two different pairs of ABP measurement between intermittent Arm Cuff and the Nexfin device were recorded. Compared to Arm Cuff, the Nexfin showed a good correlation for SAP (r2 = 0.71, p < 0.001) and MAP (r2 = 0.63, p < 0.001) (Fig. 1C and 1F). Between the two devices (Arm Cuff and the Nexfin), Bland-Altman analysis showed a mean bias of -3.8 mmHg (-30.4 to 22.9 mmHg) for SAP and − 0.6 (-22 to 20.8 mmHg) for MAP (Fig. 2C and 2F). For DAP measurements, the three devices showed a moderate to good correlation. An additional file show the relationship between absolute values of DAP on invasive and non-invasive devices [see Additional file 3]. DAP measurements obtained with arterial catheter and the Nexfin showed the best correlation (r2 = 0.72, p < 0.001). Compared to arterial catheter, Bland-Altman analysis showed a mean bias of -3.7 (-19.9 to 12.5 mmHg) and − 7 (-24.4 to 10.4 mmHg) respectively, for the Nexfin and the Arm Cuff. An additional file show the Bland-Altman analysis of DAP obtained with invasive and non-invasive devices [see Additional file 4]. Compare to the Arm Cuff, a mean bias of -3.3 (-24 to 17.4 mmHg) was recorded for the Nexfin device’s measurements.
Analysis of ABP changes during norepinephrine infusion
Compared to the ABP changes with the radial artery catheter (before and after the beginning of norepinephrine), the four-quadrant plot analysis showed a concordance rate of 92% and 70% respectively for the ABP changes measured with the Nexfin and the Arm Cuff (Fig. 3A and 3B). With the same analysis, ABP changes measured with the Nexfin and the arm Cuff were concordant in 76% (Fig. 3C).
Clinical concordance
Compared to the arterial catheter, the error grid analysis revealed that 91.3% of SAP were in zone A, 8.1% were in zone B and, 0.5% in zone C for the Nexfin (Fig. 4A). Considering the MAP, 86.1% of the measurements pairs were in zone A and 13.9% in zone B (Fig. 4B). In the same analysis, 95.4% of SAP were in zone A with intermittent Arm Cuff, 3.7% were in zone B and 0.8% in zone C (Fig. 4C). MAP measurement with intermittent arm cuff were in zone A for 77.1%, in zone B 22.5% and 0.2% in zone C (Fig. 4D). The comparison between Nexfin and the arm cuff pressure measurements showed 89% of the reading in zone A, 10.4% in zone B and 0.05% in zone C for SAP, whereas for MAP there were 82.3% in zone A, 16.6% in zone B, 1.1% in zone C (Fig. 4E and 4F).
Evolution of ABP in each device
In comparison to arm cuff, SAP measures with invasive method were significantly higher (p < 0.001) and SAP measures with Nexfin were significantly lower (p < 0.001). In comparison to arm cuff, MAP measures with invasive method were significantly lower (p < 0.01). In comparison to arm cuff, DAP measures with invasive method were significantly lower (p < 0.001) and DAP measures with Nexfin were significantly lower (p < 0.001). An additional file illustrate the trend in each ABP parameter [see Additional file 5].