ABSTRACT
In addition to randomized clinical trials, consideration of Real-World Evidence is necessary for mirroring clinical reality. However, processing such evidence for large numbers of patients often requires considerable time and effort. This is particularly true for rare tumor diseases such as multiple myeloma (MM) or for adverse effects that occur even more rarely. In such cases, artificial intelligence is able to efficiently detect patients with rare conditions. One of these rare adverse events, and the most discussed, following bone protective treatment in MM is medication-related osteonecrosis of the jaw (MRONJ). The association of bone protective treatment to MM outcome has been intensively studied. However, the impact of MRONJ resulting from such treatment on MM prognosis and outcome is poorly understood. In this retrospective study, we therefore investigated the long-term effects of MRONJ. We used natural language processing (NLP) to screen individual data of 2389 MM patients to find 50 out of 52 patients with MRONJ matching our inclusion criteria. To further improve data quality, we then performed propensity score matching. In comparison to MM patients without MRONJ, we found a significantly longer overall survival (median 126 vs. 86 months) despite slightly worse clinical features.
Subject(s)
Bisphosphonate-Associated Osteonecrosis of the Jaw , Bone Density Conservation Agents , Multiple Myeloma , Humans , Multiple Myeloma/drug therapy , Diphosphonates/adverse effects , Bisphosphonate-Associated Osteonecrosis of the Jaw/etiology , Bisphosphonate-Associated Osteonecrosis of the Jaw/diagnosis , Bone Density Conservation Agents/adverse effects , Retrospective Studies , Artificial IntelligenceABSTRACT
BACKGROUND: Natural language processing (NLP) is a powerful tool supporting the generation of Real-World Evidence (RWE). There is no NLP system that enables the extensive querying of parameters specific to multiple myeloma (MM) out of unstructured medical reports. We therefore created a MM-specific ontology to accelerate the information extraction (IE) out of unstructured text. METHODS: Our MM ontology consists of extensive MM-specific and hierarchically structured attributes and values. We implemented "A Rule-based Information Extraction System" (ARIES) that uses this ontology. We evaluated ARIES on 200 randomly selected medical reports of patients diagnosed with MM. RESULTS: Our system achieved a high F1-Score of 0.92 on the evaluation dataset with a precision of 0.87 and recall of 0.98. CONCLUSIONS: Our rule-based IE system enables the comprehensive querying of medical reports. The IE accelerates the extraction of data and enables clinicians to faster generate RWE on hematological issues. RWE helps clinicians to make decisions in an evidence-based manner. Our tool easily accelerates the integration of research evidence into everyday clinical practice.