Download PDF
Methods Inf Med 1999; 38(02): 113-118
DOI: 10.1055/s-0038-1634177
Original Article
Schattauer GmbH

A Multistate Kalman Filter for Neonatal Clotting Time Prediction and Early Detection of Coagulation Disturbances

A. F. J. de Haan
1   Department of Medical Informatics, Epidemiology and Statistics, The Netherlands
,
J. van der Velde
2   Department of Neonatology, The Netherlands
,
W. B. Geven
2   Department of Neonatology, The Netherlands
,
C. Festen
3   Department of Pediatric Surgery, Faculty of Medical Sciences, University of Nijmegen and University Hospital Nijmegen, The Netherlands
,
A. L. M. Verbeek
1   Department of Medical Informatics, Epidemiology and Statistics, The Netherlands
› Author Affiliations
Further Information

Publication History

Publication Date:
08 February 2018 (online)

    Permissions and Reprints

Abstract

The multistate Kalman filter was applied to develop a heparin dose proposal system and to detect coagulation disturbances during neonatal extracorporeal membrane oxygenation (ECMO). A system containing this filter was based on the activated clotting time (ACT) values and the heparin doses administered every hour during ECMO. If the ACT value can be predicted accurately from the previous heparin dose, a heparin dose proposal can be given to achieve or maintain the required ACT level. The analysis was done on 6,356 ACT level measurements in 44 ECMO neonates. The multistate Kalman filter technique showed an unbiased prediction of ACT, with a standard deviation of 23 seconds. Two out of three cases of disseminated intravascular coagulation (DIC) were detected. ACT values were predicted sufficiently accurately by the multistate Kalman filter technique to justify a prospective study on the performance of the heparin dose proposal system and its ability to detect DIC.

Keywords

Multistate Kalman Filter - Extracorporeal Membrane Oxygenation - Computer-aided Monitoring
 

  • References

  • 1 Anderson HL, Edmunds Jr LH. Coagulation, Anti-coagulation and the interaction of blood and artificial surfaces. In: Zwischenlerger JB, Bartlett RH. eds. ECMO Extra corporeal Cardiopulmonary support in critical care . Michigan, USA: Ann Arbor; 1995
    Google Scholar
  • 2 Peverini RL, Sale M, Rhine WD, Fagan LM, Lenert LA. Anticoagulation therapy advisor: a decision-support system for heparin therapy during ECMO. Proc Annu Symp Comput Appl Med Care. 1992: 567-71.
    PubMedGoogle Scholar
  • 3 Smith AFM, West M. Monitoring renal transplants: an application of the multiprocess Kalman filter. Biometrics 1983; 39: 867-78.
    CrossrefPubMedGoogle Scholar
  • 4 Harrison PJ, Stevens CF. Bayesian forecasting (with Discussion). J Royal Statistical Soc, series B 1976; 38: 205-47.
    PubMedGoogle Scholar
  • 5 McDonald MM, Jacobson LJ, Hay Jr WW, Hathaway WE. Heparin clearance in the newborn. Pediatr Res 1981; 15: 1015-8.
    CrossrefPubMedGoogle Scholar
  • 6 Green TP, Isham-Schopf B, Steinhorn RH, Smith C, Irmiter RJ. Whole blood activated clotting time in infants during extracorporeal membrane oxygenation. Crit Care Med 1990; 18: 494-8.
    CrossrefPubMedGoogle Scholar
  • 7 Sheiner LB, Beal SL. Some suggestions for measuring predictive performance. J Pharmacokinet Biopharm 1981; 9: 503-12.
    CrossrefPubMedGoogle Scholar
  • 8 Sheiner LB, Beal S, Rosenberg B, Marathe VV. Forecasting individual pharmacokinetics. Clin Pharmacol Ther 1979; 26: 294-305.
    CrossrefPubMedGoogle Scholar
  • 9 Bartlett RH, Gazzaniga AB, Jefferies MR, Huxtable RF, Haiduc NJ, Fong SW. Extra-corporeal membrane oxygenation (ECMO) cardiopulmonary support in infancy. Trans Am Soc Artif Intern Organs 1976; 22: 80-93.
    PubMedGoogle Scholar