Once a cesarean always a cesarean? A computer-assisted decision analysis

Oscar Sadan, Moshe Leshno, Ahuva Gottreich, Abraham Golan, Samuel Lurie

Research output: Contribution to journalArticlepeer-review


Objective: A previous decision analysis models for two strategic choices for trial of labor or repeated cesarean after prior cesarean concluded that the degree of wish for an additional future pregnancy appeared to be a major determinant for choice between the two strategic options. We had extended the analysis model to stillbirth and hypoxic-ischemic encephalopathy in addition to placental complications while updating most of the outcomes in the decision tree. Study design: A model was formulated using a decision tree based on reported probabilities for various outcomes and estimated utilities. The question asked was should trial of labor or repeated cesarean be performed after a prior cesarean, with a varying desire for an additional pregnancy. The highest expected outcome determines the preference of our model. Results: Our model favors repeated elective cesarean (0.9947) over trial of labor (0.9917) after a previous cesarean and is the preferred approach. This approach was preferable irrespective of the probability of additional pregnancy. Conclusion: In contrary to previous models, when taking into account the occurrence of a live infant birth, birth of an infant with hypoxic-ischemic encephalopathy stillbirth, neonatal death, abnormal placental implantation, hysterectomy and maternal death the preferred approach for women with previous cesarean is an elective repeated cesarean rather than trial of vaginal delivery.

Original languageEnglish
Pages (from-to)517-521
Number of pages5
JournalArchives of Gynecology and Obstetrics
Issue number5
StatePublished - Nov 2007


  • Cesarean section
  • Decision analysis
  • Vaginal birth after cesarean section


Dive into the research topics of 'Once a cesarean always a cesarean? A computer-assisted decision analysis'. Together they form a unique fingerprint.

Cite this