Development of a Clinical Decision Support System With Artificial Intelligence for Cancer Care

Clinical Decision Support Systems (CDSSs) to augment clinical care and decision making. These are platforms which aim to improve healthcare delivery by enhancing medical decisions with targeted clinical knowledge, patient information, and other health information.

In view of the benefit of developing a CDSS, we sought to develop an alternative CDSS for oncologic therapy selection through a partnership with Ping An Technology (Shenzhen, China), beginning with gastric and oesophagal cancer. This would be done in a piecemeal fashion, with the prototype platform utilizing only international guidelines and high-quality published evidence from journals to arrive at case-specific treatment recommendations. This platform would then be evaluated by comparing its recommendations with that from the multidisciplinary tumour boards of several tertiary care institutions to determine the concordance rate.

Study Overview

• Status: Unknown
• Conditions: Gastric Cancer; Esophageal Cancer; Esophagogastric Junction Cancer
• Intervention / Treatment: Other: No intervention will be provided to the subject

Detailed Description

Management of cancer is a complex process which involves numerous stakeholders. In view of this, institutions worldwide have adopted the use of Multidisciplinary Tumor Boards (MTBs) for delivery of cancer care. By tapping on the collective specialized knowledge and experience of various specialties, MTBs have been shown in some studies to result in more appropriate recommendations and improved patient outcomes. At our institution, cancer cases are similarly discussed at regular MTBs which comprises surgeons, oncologists, pathologists and radiologists who review and recommend treatments.

However, in smaller centres or centres with limited resources and minimal multi-disciplinary expertise, delivery of timely and appropriate cancer care could be a challenge. Additionally, clinicians, with their busy schedule, may not be able to keep abreast of new developments in cancer research. With rapid advances in scientific research, this pool of knowledge is expected to continue to burgeon, making keeping up-to-date increasingly onerous.

To address this need, clinicians have adopted the use of Clinical Decision Support Systems (CDSSs) to augment clinical care and decision-making. These are platforms which aim to improve healthcare delivery by enhancing medical decisions with targeted clinical knowledge, patient information, and other health information. Various studies have shown CDSSs to be beneficial in selected settings such as patient safety and diagnosis [4], and to even increase adherence to clinical guidelines. In recent years, advancements in artificial intelligence have also seen its use expand to include oncologic therapy selection, with IBM's Watson for Oncology (WFO) being the most prominent and only platform in use to-date. In a 2018 study, WFO's ability to provide treatment advice for breast cancer was compared against recommendations from a multidisciplinary board, where it showed a high degree of concordance. Since then, several other studies have sought to examine WFO's ability to provide treatment recommendations for cancer such as ovarian, gastric, lung, cervical and colorectal cancers, with mixed results. In particular, both studies which examined the recommendations for gastric cancers showed a much lower concordance rate compared to other cancers.

In view of the above, we sought to develop an alternative CDSS for oncologic therapy selection through partnership with Ping An Technology (Shenzhen, China), beginning with gastric and esophageal cancer. This would be done in a piecemeal fashion, with the prototype platform utilizing only international guidelines and high-quality published evidence from journals to arrive at case-specific treatment recommendations. This platform would then be evaluated retrospectively and prospectively by comparing its recommendations with that from the multidisciplinary tumor boards of several tertiary care institutions to determine the concordance rate.

• Study Type: Observational
• Enrollment (Anticipated): 1000

Contacts and Locations

• Study Contact: Name: Bok Yan, Jimmy So; Phone Number: 24236 +65 6772 5555; Email: sursbyj@nus.edu.sg

Study Locations

• Singapore
  • Singapore, Singapore, 119228
    • Recruiting
    • National University Hospital
    • Contact: Jimmy So, MBChB; Phone Number: 24236 +65 6772 5555; Email: sursbyj@nus.edu.sg
    • Contact: Guowei Kim, MBBS; Phone Number: 28830 +65 6772 5555; Email: guo_wei_kim@nuhs.edu.sg

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 21 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

The population will be selected from hospitals that conduct tumor board for these cancers

Description

A. In discovery and internal retrospective validation part:

1. Patients with primary gastric adenocarcinoma including preinvasive carcinoma or
2. Patients with gastroesophageal junction cancers or
3. Patients with oesophageal cancer including adenocarcinoma, squamous cell carcinoma and preinvasive carcinoma subtypes.

B. In prospective validation part:

1. Patients with primary gastric adenocarcinoma including preinvasive carcinoma or
2. Patients with esophageal or gastroesophageal junction adenocarcinoma

Exclusion Criteria:

A. In discovery and internal retrospective validation part:

1. Patients with other primary cancers involving the stomach or oesophagus
2. Patients with other cancer subtypes
3. Patients with concomitant cancers of other organs

B. In prospective validation part:

1. Patients with esophageal squamous cell carcinoma
2. Patients who participate in clinical trials where the treatment modality is not standard of care

Study Plan

How is the study designed?

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Concordance Rate	Comparative agreement in recommendations between the two study groups, as measured by concordance rate	1 to 2 years

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Reason for Discordance	To identify the reason for the discordance	1 to 2 years

Collaborators and Investigators

• Sponsor: National University Hospital, Singapore

Publications and helpful links

General Publications

• Somashekhar SP, Sepulveda MJ, Puglielli S, Norden AD, Shortliffe EH, Rohit Kumar C, Rauthan A, Arun Kumar N, Patil P, Rhee K, Ramya Y. Watson for Oncology and breast cancer treatment recommendations: agreement with an expert multidisciplinary tumor board. Ann Oncol. 2018 Feb 1;29(2):418-423. doi: 10.1093/annonc/mdx781.

Study record dates

Study Major Dates

• Study Start (Actual): August 20, 2020
• Primary Completion (Anticipated): December 31, 2022
• Study Completion (Anticipated): December 31, 2022

Study Registration Dates

• First Submitted: December 15, 2020
• First Submitted That Met QC Criteria: December 15, 2020
• First Posted (Actual): December 19, 2020

Study Record Updates

• Last Update Posted (Actual): March 9, 2021
• Last Update Submitted That Met QC Criteria: March 7, 2021
• Last Verified: December 1, 2020

More Information

Terms related to this study

• Keywords: Gastric Cancer; Artificial Intelligence; Esophageal Cancer; Concordance; Clinical Decision Support System; Multidisciplinary Tumor Board
• Additional Relevant MeSH Terms: Digestive System Diseases; Neoplasms; Neoplasms by Site; Gastrointestinal Neoplasms; Digestive System Neoplasms; Gastrointestinal Diseases; Stomach Diseases; Head and Neck Neoplasms; Esophageal Diseases; Stomach Neoplasms; Esophageal Neoplasms
• Other Study ID Numbers: 2020/00493

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No