The study enrolled 13 patients to initiate interstitial CGM monitoring simultaneously with an Accu-Chek Guide glucometer. One patient was withdrawn for having been transplanted before the end of follow-up. The baseline characteristics are described in Table 1.
Table 1 – Clinical and laboratory data
|
Study population n=12
|
Sex (M:F)
|
(8:4)
|
Age (years)
|
66.8 ± 8.0
|
Time in dialysis treatment (years)
|
4.7 ± 3.4
|
BMI with dry weight in V-3 (kg/m2)
|
25.5 ± 2.7
|
Loss volume per session (liters)
|
2.0 ± 0.5
|
Hematocrit (%)
|
36.7 ± 7.3
|
Hemoglobin (g/dl)
|
11.9 ± 2.2
|
Initial fasting blood glucose (mg/dl)
|
209.4 ± 66.6
|
Initial glycated hemoglobin (%)
|
7.3 ± 1.1
|
Intradialytic blood glucose variation (mg/dl)
|
52.9 ± 56.9
|
Hypoglycemia episodes in the first week
|
0
|
Data presented as mean with standard deviation.
Capillary blood glucose was similar to that of the interstitial sensor at HD session 0, 1, 2, 3 and 5. Only at session 4 there was a statistical difference between capillary and interstitial measurements (191.3 ± 36.3 mg/dl and 150.5 ± 50.6 mg/dl respectively, p=0.033). Comparison of capillary and interstitial glucose measurements after dialysis showed more statistically significant differences. Capillary differences were observed at session 1 (149.6 ± 28.1 mg/dl and 121.1 ± 31.0 mg/dl respectively, p=0.028), 2 (137.9 ± 26.8 mg/dl and 106.2 ± 28.0 mg/dl respectively, p=0.010) and 3 (144.6 ± 44.3 mg/dl and 104.8 ± 42.1 mg/dl respectively, p=0.010). There were also differences in the mean of all pre-dialysis capillary measurements with the sensor (188.3 ± 61.2 mg/dl and 163.9 ± 71.3 mg/dl respectively, p=0.029) and with the post-dialysis measurements (148, 8 ± 34.0 mg/dl and 123.0 ± 46.0 mg/dl respectively, p=0.000). (TABLE 2 and GRAPHIC 1 and 2)
Table 2 - Pre- and post-dialysis mean blood glucose measurements
|
Pre-dialysis
|
|
Post-dialysis
|
|
|
Capillary glucose (mg/dl)
|
Interstitial glucose
(mg/dl)
|
p
|
Capillary glucose (mg/dl)
|
Interstitial glucose
(mg/dl)
|
p
|
HD session 0
|
168.5 ± 55.5
|
155.2 ± 54.6
|
0.562
|
148.6 ± 36.1
|
128.3 ± 37.2
|
0.189
|
HD session 1
|
193.5 ± 80.31
|
191.0 ± 88.7
|
0.977
|
149.6 ± 28.1
|
121.1 ± 31.0
|
0.028
|
HD session 2
|
190.2 ± 73.9
|
173.1 ± 90.7
|
0.225
|
137.9 ± 26.8
|
106.2 ± 28.0
|
0.010
|
HD session 3
|
198.2 ± 65.8
|
159.0 ± 76.8
|
0.133
|
144.6 ± 44.3
|
104.8 ± 42.1
|
0.010
|
HD session 4
|
191.3 ± 36.3
|
150.5 ± 50.6
|
0.033
|
157.2 ± 37.9
|
122.0 ± 56.2
|
0.086
|
HD session 5
|
188.2 ± 54.2
|
155.8 ± 61.1
|
0.178
|
155.1 ± 31.5
|
111.5 ± 44.9
|
0.093
|
Overall
|
188.3 ± 61.2
|
163.9 ± 71.3
|
0.029
|
148.8 ± 34.0
|
123.0 ± 46.0
|
0.000
|
Data presented as mean with standard deviation. P significant when < 0.05
4.2.2 Specific glycemic assessment
Some clinical and laboratory parameters were evaluated to determine which clinical situations the sensor measure matches with the capillary blood glucose measure. In pre-dialysis measurements, capillary blood glucose showed a discrepancy from the interstitial during the second week of sensor use (192.6 ± 52.1 mg/dl and 154.7 ± 62.3 mg/dl respectively, p=0.007), in patient with a BMI greater than 24.9 kg/m2 (162.4 ± 31.7 mg/dl and 126.6 ± 38.7 mg/dl respectively, p=0.000), in patients who lose an average of more than 2 liters in the dialysis ultrafiltrate (187.7 ± 48.1 mg/dl and 154.3 ± 55.4 mg/dl respectively, p=0.002) and in patients with hemoglobin ≤ 12 (197.3 ± 52.0 mg/dl and 157 .5 ± 56.1 mg/dl respectively, p=0.006) (TABLE 3)
Table 3 – Glucose measurements in specific pre- and post-dialysis clinical situations
|
Pre-dialysis
|
Post-dialysis
|
|
Capillary glucose (mg/dl)
|
Interstitial glucose
(mg/dl)
|
p
|
Capillary glucose (mg/dl)
|
Interstitial glucose
(mg/dl)
|
p
|
1ª Week mean glucose (mg/dl) (n=36)
|
184.1 ± 69.5
|
173.1 ± 78.9
|
0.303
|
145.4 ± 30.25
|
118.58 ± 32.7
|
0.001
|
2ª Week mean glucose (mg/dl) (n=36)
|
192.6 ± 52.1
|
154.7 ± 62.3
|
0.007
|
152.3 ± 37.6
|
127.5 ± 56.4
|
0.032
|
Body mass index
|
|
|
|
|
|
|
> 24,9kg/m2 (n=36)
|
162.4 ± 31.7
|
126.6 ± 38.7
|
0.000
|
152.9 ± 30.8
|
113.8 ± 38.9
|
0.000
|
≤ 24,9kg/m2 (n=36)
|
214.2 ± 72.2
|
201.3 ± 77.0
|
0.466
|
144.8 ± 37.0
|
132.3 ± 51.1
|
0.237
|
Water loss (Ultrafiltrate)
|
|
|
|
|
|
|
> 2 liters (n=48)
|
187.7 ± 48.1
|
154.3 ± 55.4
|
0.002
|
144.9 ± 33.5
|
112.6 ± 43.5
|
0.000
|
≤ 2 liters (n=24)
|
189.5 ± 82.6
|
183.1 ± 94.0
|
0.805
|
156.8 ± 34.3
|
143.8 ± 44.7
|
0.265
|
Age
|
|
|
|
|
|
|
> 65 years (n=42)
|
205.9 ± 68.7
|
177.1 ± 82.6
|
0.086
|
144.3 ± 30.2
|
122.2 ± 48.1
|
0.014
|
≤ 65 years (n=30)
|
163.7 ± 37.7
|
145.5 ± 46.8
|
0.104
|
155.2 ± 38.4
|
124.2 ± 43.6
|
0.005
|
Time on dialysis treatment
|
|
|
|
|
|
|
> 4 years (n=36)
|
188.3 ± 56.6
|
164.9 ± 53.9
|
0.078
|
157.9 ± 39.6
|
131.7 ± 45.9
|
0.012
|
≤ 4 years (n=36)
|
188.4 ± 66.2
|
162.9 ± 86.0
|
0.164
|
139.8 ± 24.7
|
114.3 ± 45.1
|
0.004
|
Hemoglobin
|
|
|
|
|
|
|
>12 (g/dl) (n=42)
|
181.9 ± 66.9
|
168.6 ± 80.7
|
0.412
|
152.0 ± 35.5
|
129.5 ± 47.4
|
0.016
|
≤12 (g/dl) (n=30)
|
197.3 ± 52.0
|
157.5 ± 56.1
|
0.006
|
144.5 ± 31.9
|
113.9 ± 43.2
|
0.003
|
Glycated hemoglobin
|
|
|
|
|
|
|
>7% (n=42)
|
202.6 ± 72.0
|
167.6 ± 88.3
|
0.050
|
152.3 ± 35.9
|
119.2 ± 52.7
|
0.001
|
≤7% (n=30)
|
168.3 ± 33.4
|
158.8 ± 37.0
|
0.303
|
144.1 ± 31.2
|
128.4 ± 34.7
|
0.072
|
Data presented as mean with standard deviation. P significant when < 0.05. The n in parentheses correlates with number of measurements
Regarding post-dialysis measurements, capillary blood glucose measurements were different from those measured by the interstitial sensor in the first week (145.4 ± 30.25 mg/dl and 118.58 ± 32.7 mg/dl respectively, p=0.001) and in the second of treatment (152.3 ± 37.6 mg/dl and 127.5 ± 56.4 mg/dl respectively, p=0.032), in a patient with a BMI greater than 24.9 kg/m2 (152.9 ± 30.8 mg/dl and 113.8 ± 38.9 mg/dl respectively, p=0.000), in patients who lose more than 2 liters in the dialysis ultrafiltrate on average (144.9 ± 33.5 mg/dl and 112 .6 ± 43.5 mg/dl respectively, p=0.000), in patients older than 65 years (144.3 ± 30.2 mg/dl and 122.2 ± 48.1 mg/dl respectively, p=0.014) and under 65 (155.2 ± 38.4 mg/dl and 124.2 ± 43.6 mg/dl respectively, p=0.005), in patients with more than 4 years of dialysis treatment (157.9 ± 39, 6 mg/dl and 131.7 ± 45.9 mg/dl respectively, p=0.012) and there is less (139.8 ± 24.7 mg/dl and 114.3 ± 45.1 mg/dl respectively, p= 0.004), in a patient with hemoglobin greater than 12g/dl (152.0 ± 35.5 mg/dl and 129.5 ± 47.4 mg/dl respectively, p=0.016) and smaller (144.5 ± 31.9 mg/dl and 113.9 ± 43.2 mg/dl respectively, p=0.003) and glycated hemoglobin greater than 7% (152.3 ± 35.9 mg/dl and 119.2 ± 52.7 mg/dl respectively, p=0.001). (TABLE 4)
4.3 Clinical Verification
The mean relative absolute difference (MARD) was calculated using capillary blood glucose as a reference. The MARD of each hemodialysis session was compared with that of session 0 at the beginning and end of the dialysis session. Regarding pre-dialysis values, a progressive increase in MARD from the first session during each measurement was observed, reaching a statistically significant peak in session 3 (p=0.0013) and followed by a subsequent decrease until session 4 and 5 (9.4 ± 6.3%, 11.2 ± 12.8%, 15.3 ± 16.5%, 23.6 ± 17.0%, 21.5 ± 19.0% and 21, 1 ± 21.8% respectively).
The MARDs calculated in the post-dialysis showed a higher value in the first session with a drop in the subsequent session followed by a progressive increase, however without statistically significant differences throughout each session (23.5 ± 18.4%, 20.1 ± 11, 3%, 22.6 ± 14.8%, 26.9 ± 18.6%, 24.9 ± 19.2% and 36.4 ± 22.9% respectively). In the mean MARD per session without distinction between pre- and post-dialysis, the last session was the only one that showed significant differences from the first (p=0.037). The MARD with all values gathered was 21.4% (± 17.8) (TABLE 4 and GRAPH 3)
Table 4 – MARD of each dialysis session
|
Mean Absolute Relative Difference (MARD) per dialysis visit
|
|
|
Session 0
|
Session 1
|
Session 2
|
Session 3
|
Session 4
|
Session 5
|
Overall
|
Pre-dialysis MARD (%)
|
9.4 ± 6.3
|
11.2 ± 12.8
|
15.3 ± 16.5
|
23.6 ± 17.0
|
21.5 ± 19.0
|
21.1 ± 21.8
|
17.08 ± 16.6
|
p value*
|
|
0.686
|
0.644
|
0.013
|
0.056
|
0.099
|
|
Post-dialysis MARD (%)
|
23.5 ± 18.4
|
20.1 ± 11.3
|
22.6 ± 14.8
|
26.9 ± 18.6
|
24.9 ± 19.2
|
36.4 ± 22.9
|
25.7 ± 18.1
|
p value*
|
|
0.587
|
0.773
|
0.660
|
0.858
|
0.143
|
|
Overall MARD (%)
|
16.5 ± 15.2
|
15.6 ± 12.6
|
19.0 ± 15.8
|
25.3 ± 17.5
|
23.2 ± 18.8
|
28.8 ± 23.2
|
21.4 ± 17.8
|
p value*
|
|
0.838
|
0.583
|
0.071
|
0.179
|
0.037
|
|
*Comparative with first session
Data presented in mean absolute relative difference with standard deviation. P significant when < 0.05
Regarding the Clarke Error Grids, it was observed in the correlation with pre-dialysis measurements, the proportions of zone A, zone B, zone C, zone D and zone E were 72.2%, 22.2%, 0.0%, 5 .6% and 0.0%, respectively. In the correlation with post-dialysis measurements, the proportions of zone A, zone B, zone C, zone D and zone E were 52.8%, 31.9%, 0.0%, 15.3% and 0. 0%, respectively. Considering all measurements, the proportions of zone A, zone B, zone C, zone D and zone E were 62.5%, 27.1%, 0.0%, 10.4% and 0.0%, respectively. (TABLE 6 and GRAPH 4)
Table 6 – Percentage of points in Clarke and Parkes Grid zones
Percentage of points in Clarke and Parkes Grid zones, in pre, post-dialysis and overall
|
|
Zone A
|
Zone B
|
Zone C
|
Zone D
|
Zone E
|
Pre-dialysis Clarke Grid (%)
|
72.2
|
22.2
|
0
|
5.6
|
0
|
Post-dialysis Clarke Grid (%)
|
52.8
|
31.9
|
0
|
15.3
|
0
|
Overall Clarkes Grid (%)
|
62.5
|
27.1
|
0
|
10.4
|
0
|
Pre-dialysis Parkes Grid (%)
|
87.5
|
4.2
|
8.3
|
0
|
0
|
Post-dialysis Parkes Grid (%)
|
73.6
|
15.3
|
11.1
|
0
|
0
|
Overall Parkes Grid (%)
|
80.6
|
9.7
|
9.7
|
0
|
0
|
Percentage of points in each specific area of the graph with pre, post-dialysis and total measurements
In the Parkes grid measured in pre-dialysis, the proportions of zone A, zone B, zone C, zone D and zone E were 87.5%, 4.2%, 8.3% 0.0% and 0. 0%, respectively and in post-dialysis, the proportions of zone A, zone B, zone C, zone D and zone E were 73.6%, 15.3%, 11.1% 0.0% and 0. 0%, respectively. In the association of overall measures, the proportions of zone A, zone B, zone C, zone D and zone E were 80.6%, 9.7%, 9.7% 0.0% and 0.0%, respectively. (TABLE 6 and GRAPH 5)