Abstract for presentation at 11th International Congress of Human Genetics

Validation of the Tulip Classification for Causes of Perinatal Mortality

  • Fleurisca Korteweg, University Medical Center Groningen, The Netherlands
  • Katelijne Bouman, University Medical Center Groningen, The Netherlands
  • JanJaap Erwich, University Medical Center Groningen, The Netherlands
  • Albertus Timmer, University Medical Center Groningen, The Netherlands
  • Sanne Gordijn, University Medical Center Groningen, The Netherlands
  • Klasina Bergman, University Medical Center Groningen, The Netherlands
  • Joke Ravise, University Medical Center Groningen, The Netherlands
  • Jozien Holm, University Medical Center Groningen, The Netherlands

    • A system for classification of causes of perinatal mortality is essential for evaluating quality of care, comparison and development of preventive strategies. Our objective was to validate the Dutch pathophysiological Tulip classification, a system aiming at identifying the underlying cause of perinatal mortality based on clinical and pathological findings.
    A random sample of late abortions, perinatal deaths, late neonatal deaths and perinatally related infant deaths occurring between 1999-2003 from our tertiary referral centre were classified according to the Tulip classification by 2 obstetricians, a resident, a pathologist and a neonatologist. Cause of death was defined as the pathophysiological entity initiating the chain of events that irreversibly led to death. After individual scoring based on case summaries, panel discussions for cases without consensus were held.
    The sample consisted of 411 cases ranging from 16 weeks of gestation up to 6 months after birth. Cases were classified into 6 causes of death: 1 Congenital anomaly 35% (chromosomal, syndrome, single or multiple organ system), 2 Placenta 27% (placenta-bed, placental pathology, umbilical cord complication, not otherwise specified(nos)), 3 Prematurity 23% (PPROM, preterm labour, cervical incompetence, iatrogenous, nos), 4 Infection 1% (fetal, neonatal), 5 Other 3% (hydrops of unknown origin, maternal disease, trauma, out of the ordinary), 6 Unknown 11%. Overall kappa coefficient for agreement was 0.81 (95% C.I. 0.80-0.83). Kappa after excluding guideline misinterpretation was: 0.86 (95% C.I. 0.84-0.87).
    Classifying perinatal mortality to compare performance over time and between centres is useful and necessary. Therefore, it is important to ensure that interpretation of classifications is consistent. The Tulip classification gives a good individual level of agreement with a low percentage of unknown causes and is easily applicable in a team of clinicians when Tulip guidelines are followed.

    Conference Organiser - ICMS Pty Ltd