Initial Experience with Cancer Guidelines Navigator, a tool to standardize and improve the quality of cancer care in Sub-Saharan Africa, at Ocean Road Cancer Institute in Tanzania

e14106

Background: The IBM Cancer Guidelines Navigator (CGN) is a digital reference system to support treatment planning that allows clinicians to enter a cancer patient’s clinical characteristics and presents the corresponding treatment options in the NCCN Harmonized Guidelines (TM) for Sub-Saharan Africa. In October 2019, Ocean Road Cancer Institute (ORCI) in Tanzania became the first site in Africa to initiate a hospital-wide implementation of the tool to help clinicians reduce cancer treatment variability by increasing adherence to standard evidence-based care. We describe training and lessons learned from system introduction. Methods: Training for clinical staff at ORCI occurred over one week and included daily one-hour lectures, followed by personalized hands-on training. A survey was administered to assess usability and use cases of the tool. Results: Thirty-one ORCI clinical and IT staff members participated in training, and 12 completed the survey. Responses indicated that the most beneficial uses for CGN were at point of care and for self-learning. Participants indicated that the top benefits of the tool were quick access to guidelines and evidence (75%) and ease of use (58%). Expanding cancer coverage (42%), offline access and better integration into the workflow (25%) were identified as areas for improvement. Post-training, ORCI implemented easier access to CGN on each computer and tablet used for consultation and care management. Conclusions: CGN is a digital reference system that is designed to support easy and efficient access to regionalized cancer-treatment guidelines for point-of-care treatment planning and education. Expansion of this program has been planned for other hospitals in Tanzania. Future studies will examine whether CGN usage affects guideline adherence.

Use of machine learning to identify relevant research publications in clinical oncology

6558

Background: Finding high-quality science to support decisions for individual patients is challenging. Common approaches to assess clinical literature quality and relevance rely on bibliometrics or expert knowledge. We describe a method to automatically identify clinically relevant, high-quality scientific citations using abstract content. Methods: We used machine learning trained on text from PubMed papers cited in 3 expert resources: NCCN, NCI-PDQ, and Hemonc.org. Balanced training data included text cited in at least two sources to form an “on topic” set (i.e., relevant and high quality), and an “off-topic” set, not cited in any of the above 3 sources. The off-topic set was published in lower ranked journals, using a citation-based score. Articles were part of an Oncology Clinical Trial corpus generated using a standard PubMed query. We used a gradient boosted-tree approach with a binary logistic supervised learning classification. Briefly, 988 texts were processed to produce a term frequency-inverse document frequency (tf-idf) n-gram representation of both the training and the test set (70/30 split). Ideal parameters were determined using 1000-fold cross validation. Results: Our model classified papers in the test set with 0.93 accuracy (95% CI (0.09:0.96) p ≤ 0.0001), with sensitivity 0.95 and specificity 0.91. Some false positives contained language considered clinically relevant that may have been missed or not yet included in expert resources. False negatives revealed a potential bias towards chemotherapy-focused research over radiation therapy or surgical approaches. Conclusions: Machine learning can be used to automatically identify relevant clinical publications from biographic databases, without relying on expert curation or bibliometric methods. The use of machine learning to identify relevant publications may reduce the time clinicians spend finding pertinent evidence for a patient. This approach is generalizable to cases where a corpus of high-quality publications that can serve as a training set exists or cases where document metadata is unreliable, as is the case of “grey” literature within oncology and beyond to other diseases. Future work will extend this approach and may integrate it into oncology clinical decision-support tools.