The Journal of ExtraCorporeal Technology No 57-1 | Page 22

16 K . Gore et al .: J Extra Corpor Technol 2025 , 57 , 14 – 17
Table 2 . Baseline characteristics and reported comorbidities for renal replacement therapy in patients requiring mechanical circulatory support .
Terms
Estimates
Std Error
v 2
P values
Intercept
3.7
1.6
5.7
0.0171
Age
�0.07
0.02
12.5
0.0004 *
Sex , female
�0.04
0.26
2.5
0.1159
BMI
�0.01
0.04
0.2
0.6909
Insulin-dependent diabetes
1.4
1.4
1.0
0.3194
Chronic renal failure
0.3
0.3
0.5
0.4608
Chronic cardiovascular disease
0.4
0.3
1.5
0.2261
Immunomodulation
0.8
0.3
7.3
0.0067 *
Structural lung disease
�0.3
0.3
0.9
0.3343
Pacemaker / Internal cardiac defibrillator
0.6
0.3
4.2
0.0411 *
Atrial fibrillation
0.3
0.3
1.4
0.2430
Endocarditis
0.4
0.7
0.3
0.5936
Previous cardiac surgery
0.3
0.2
0.9
0.3357
Congestive heart failure
�0.6
0.3
3.0
0.0845
Peripheral vascular disease
�0.02
0.5
0
0.9611
(*) Denotes the baseline characteristics and comorbidities associated with the outcome of interest , the need for renal replacement therapy , are statistically significant .
Table 3 . Contingency table of the association of renal replacement therapy in patients with a history of immunomodulation during mechanical circulatory support .
Renal replacement therapy
Counts (%)
Yes
No
Total
Immunomodulation
Yes
12 ( 48 )
13 ( 52 )
25
No
45 ( 34 )
89 ( 66 )
134
Total
57
102
159
Table 4 . Contingency table of the association of renal replacement therapy in patients with pre-existing pacemaker or internal cardiac defibrillator ( ICD ) during mechanical circulatory support .
Renal replacement therapy
Counts (%)
Yes
No
Total
Pacemaker / ICD
Yes
25 ( 47 )
28 ( 53 )
53
No
32 ( 30 )
74 ( 70 )
106
Total
57
102
159
Table 5 . Contingency table of the association of renal replacement therapy by anticoagulant therapy during mechanical circulatory support .
study is potential bias due to confounding . However , the strength of this study was the statistical method used to adjust for all confounders through the application of machine learning against the outcome of interest , need for RRT . Machine learning performs better than traditional statistical analyses , especially when analyzing multifaceted data sets . The ability to utilize machine modeling provides a powerful tool to express information [ 14 ].
Renal replacement therapy
Counts (%)
Yes
No
Total
Interest groups
UFH
26 ( 43 )
35 ( 57 )
61
LMWH
0 ( 0 )
3 ( 100 )
3
None
4 ( 100 )
0 ( 0 )
4
Total
30
38
68
Conclusions
The incidence of RRT was high in this patient population . The mortality rate was high in patients requiring RRT . Moreover , these findings also suggest that other options for systemic anticoagulation during MCS should be considered . The novel associations of patients who have received prior immunotherapy or with pre-existing pacemaker / ICDs requiring MCS suggest an increased systemic inflammatory state exists that escalates the need for RRT . Further investigation into how these background inflammatory conditions contribute to the need for RRT during MCS is warranted .
Funding The authors received no funding to complete this research .
Conflicts of interest The authors declare no conflict of interest .
Data availability statement All available data are incorporated into the article .
Author contribution statement
Kelsey Gore , MA , BSRT , RRT : Design , Data harvest , Editorial review of the manuscript . Dean Linder , Jr . CCP , LP : Editorial review of the manuscript . Juan José Martinez Duque MD : Data harvest , Editorial review of the manuscript . Junxi Wang BS : Data harvest , Editorial review of the manuscript . Brett Wester , DO : Editorial review of the manuscript . Tiffany Otero , MD : Editorial review of the manuscript . Shaun Yockelson , MD : Editorial review of the manuscript . Adrian Alexis Ruiz , MD : Editorial review of the manuscript .