Forty-nine ICU patients and 11 healthy volunteers were included. Of the healthy volunteers there were 4 women and 7 men. The mean age was 24.5 years. In the ICU patients there were 41 missing values (6.3%), including 34 values above the detection limit. In the healthy volunteers, there were 10 missing values (2.5%), including 4 values above the detection limit.
Table 1: Demographic data of ICU patients.
Gender
|
N (%)
|
Male
|
28 (57.0 %)
|
Female
|
16 (33.0 %)
|
NA
|
5 (10.0 %)
|
Mean age
|
62.0 ± 17.1 years
|
Diagnosis
|
|
Sepsis
|
15 (30.6 %)
|
Hypovolemic shock
|
4 (8.2 %)
|
Trauma
|
4 (8.2 %)
|
Respiratory failure
|
3 (6.1 %)
|
Severe burn injury
|
3 (6.1 %)
|
Cardiogenic shock
|
4 (8.2 %)
|
Other
|
11 (22.4 %)
|
NA
|
5 (10.2 %)
|
Instrument comparisons
In the ICU group (n=49), we found no significant difference in measured lactate between the LP2 and the ABL using arterial blood (Table 2, Figure 1). We found significantly higher values with LP2 than ABL using central venous blood (Table 2, Figure 2).
In the healthy volunteer group (n=11), we found no significant difference between LP2 and ABL using arterial blood but significantly higher values for LP2 when using peripheral venous blood (Table 2). We found no statistically significant difference between LP2.2 and LP2.1 in both arterial and venous blood. In a mixed effects model where the instrument discrepancy could vary during follow-up, LP2 measured lower values in arterial blood compared to ABL at “rest” (data not shown), but we found no significant difference between LP2 and ABL at the other time points.
Table 2: Results of instrument comparisons with arterial and venous blood in ICU patients and healthy volunteers.
ICU patients
|
Estimate (Ratio)
|
95 % CI
|
p-value
|
Arterial
|
|
|
|
LP2.1 vs ABL
|
1.03
|
0.99 to 1.08
|
0.140
|
LP2.2 vs ABL
|
1.04
|
0.99 to 1.09
|
0.102
|
LP2.2 vs LP2.1
|
1.004
|
0.96 to 1.05
|
0.871
|
Venous
|
|
|
|
LP2.1 vs ABL
|
1.29
|
1.24 to 1.35
|
<0.001
|
LP2.2 vs ABL
|
1.29
|
1.23 to 1.35
|
<0.001
|
LP2.2 vs LP2.1
|
0.998
|
0.96 to 1.04
|
0.938
|
Healthy volunteers
|
Estimate (Ratio)
|
95 % CI
|
p-value
|
Arterial, LP2 vs ABL
|
0.96
|
0.93 to 1.002
|
0.063
|
Venous, LP2 vs ABL
|
1.07
|
1.03 to 1.11
|
0.001
|
Table 2 legend: The results presented in the table are based on mixed effects models with log transformed lactate as the dependent variable. The ratios presented are based on fixed effect coefficients, exponentially transformed to be interpretable as ratios. For example, arterial lactate measurements on the instrument LP2.1 are estimated as 3% higher than measurements on the ABL in ICU patients.
Sample site comparisons
In the ICU group, we found significant differences in both fingertip and earlobe compared to arterial blood using the LP2.1. Capillary blood lactate in the fingertip and earlobe was 47% (95% CI (29% to 68%), p<0.001) and 27% (95% CI (11% to 45%), p<0.001) higher than in arterial blood, respectively (Table 3). When comparing fingertip to earlobe we found that capillary blood lactate in the fingertip was 16% higher than in the earlobe (95% CI (2% to 32%), p=0.029). Bland Altman plots for comparison between capillary blood lactate in fingertip with arterial blood lactate on both handheld instruments (LP2.1 and LP2.2) in the ICU group are presented in Figure 3. We observed a rise in the lactate values in the fingertip from the first blood drop measured on LP2.1 to the second blood drop measured on LP2.2. We did not observe the same effect in the earlobe.
In the healthy volunteer group (n=11), we found that capillary blood lactate in the fingertip was 14% higher than arterial blood lactate (95% CI (4% to 24%), p=0.003) (Table 3). We found no significant difference between capillary blood lactate in the earlobe and arterial blood lactate (Table 3).
Table 3: Results of sample site comparisons in ICU patients and healthy volunteers.
ICU patients
|
Estimate (Ratio)
|
95 % CI
|
p-value
|
Finger LP2.1 vs Arterial LP2.1
|
1.47
|
1.29 to 1.68
|
<0.001
|
Earlobe LP2.1 vs Arterial LP2.1
|
1.27
|
1.11 to 1.45
|
<0.001
|
Finger LP2.2 vs Arterial LP2.2
|
1.85
|
1.55 to 2.21
|
<0.001
|
Earlobe LP2.2 vs Arterial LP2.2
|
1.22
|
1.03 to 1.46
|
0.024
|
Finger LP2.1 vs Venous LP2.1
|
1.14
|
1.01 to 1.30
|
0.040
|
Earlobe LP2.1 vs Venous LP2.1
|
0.99
|
0.87 to 1.12
|
0.871
|
Finger LP2.2 vs Venous LP2.2
|
1.43
|
1.20 to 1.69
|
<0.001
|
Earlobe LP2.2 vs Venous LP2.2
|
0.95
|
0.80 to 1.13
|
0.569
|
Healthy volunteers
|
Estimate (Ratio)
|
95 % CI
|
p-value
|
Finger vs arterial
|
1.14
|
1.04 to 1.24
|
0.003
|
Earlobe vs arterial
|
0.98
|
0.90 to 1.06
|
0.568
|
Table 3 legend: The results presented in the table are based on mixed effects models with log transformed lactate as the dependent variable. The ratios presented are based on fixed effect coefficients, exponentially transformed to be interpretable as ratios. For example, lactate measurements from the fingertip are estimated as 47% higher than arterial measurements in ICU patients.