Article Text

Download PDFPDF

#442 Using data mining and data vision to improve the efficiency of a care pathway: an example for advanced adnexal cancers
  1. Delphine Hudry1,2,
  2. Nour Kheirbek3,1,
  3. Adrien Boscher4,
  4. Mathilde Duchatelet1,
  5. Houssein El Hajj5,
  6. Ali Hammoudi6,
  7. Franck Craynest7,
  8. Carlos Martinez Gomez1,
  9. Marie-Pierre Crépin8,
  10. Stéphanie Bécourt1 and
  11. Fabrice Narducci1,2
  1. 1Department of gynecologic oncology, Oscar Lambret Center, Lille, France
  2. 2Univ. Lille, Inserm, University Hospital of Lille, U1192 – Protéomique Réponse Inflammatoire Spectrométrie de Masse – PRISM, Lille, France
  3. 3Medicine faculty, Henri Warembourg, Lille University, Lille, France
  4. 4Department of Gynecology, Hospital Center, Dunkerque, Dunkerque, France
  5. 5Department of Surgery, Institut Curie, Hôpital René Huguenin, Saint-Cloud, France
  6. 6Office of Digital Uses, Oscar Lambret Center, Lille, France
  7. 7Information Systems Department, Oscar Lambret Center, Lille, France
  8. 8Care Management, Oscar Lambret Center, Lille, France


Introduction/Background The purpose of a care pathway is to standardize or streamline the sequence of different therapeutic steps. In oncology, such an organization promotes more efficient management for patients. Adnexal carcinoma is most often diagnosed at an advanced stage and its management will require several steps: from diagnosis to excision surgery, to chemotherapy and maintenance treatment. In our hospital, we have developed a clinical pathway for patients with advanced adnexal cancer. With the help of ’educated’ software, we were able to perform daily analyses of predefined indicators, and thus help us target the best quality of medical and surgical care.

Methodology A multidisciplinary team validated the key steps of the care pathway. Indicators were defined based on current European recommendations. The software was trained to automatically extract useful elements from the patient‘s electronic medical record. Medical and paramedical managers check the data on a regular basis. Indicators are updated daily and changes in practice are evaluated prospectively.

Results In total, until April 2023, 17 milestones have been progressively tested and defined, allowing for the analysis of 20 indicators. From January 2018 to March 2023, 497 patients were identified in the Turquoise pathway. The median times of the pathway were 6 days (5.5–8) from first call to first medical appointment, 12 days (1–69) from first appointment to diagnostic procedure, 14 days (1–46) from histopathological result to start of primary chemotherapy if indicated. The organization of appointments and the management of peritoneal biopsies performed by laparoscopy were modified.

Abstract #442 Figure 1

Turquoise Pathway: advanced ovarian care

Conclusion The use of machine learning has allowed to build a care pathway for patients requiring the same therapeutic steps with indicators available in real time that help to organize the care as efficiently as possible. The introduction of machine learning could save caregivers time and thus promote direct patient interactions, allowing patient-focused care.

Disclosures None disclosures

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.