LAUSR

Background: The diagnosis of rare diseases is often delayed by years [1]. The main factor for this delay is believed to be the lack of knowledge and awareness regarding rare diseases [2]. Probabilistic diagnostic decision support systems (DDSS) have the potential to accelerate rare disease diagnosis by highlighting differential diagnoses for physicians [3, 4]. DDSS’s are based on case input and incorporated medical knowledge. Objectives: We examine a probabilistic DDSS prototype and assess its potential to provide accurate rare disease suggestions early in the course of rare disease cases. Methods: Retrospectively, information from the medical records of 93 patients was transferred to the DDSS. Each of these patients had a confirmed rare inflammatory systemic disease. The accuracy of the DDSS disease suggestions was assessed for all documented visits over time. Time to correct top fit (TF) and top five fit (T5F) disease suggestion was assessed, as was the original time to clinical diagnosis (TD). TF/TD as well as T5F/TD were calculated to allow for comparison of TF respective T5F normalized to TD. Wilcoxon signed-rank test was conducted for TD-TF and TD-T5F. Results: The DDSS suggested the correct disease at a time earlier than the time of clinical diagnosis among the top five fit disease suggestions in 53.8% of cases (50 of 93), and as the top fit disease suggestion in 37.6% of cases (35 of 93). Median advantage of correct disease suggestions compared to the time point of clinical diagnosis was 3 months or 50% for top five fit respective 1 month or 21% for top fit. The correct diagnosis was suggested at the first documented patient visit among the top five fit disease suggestions in 33.3% (top five fit), respective 16.1% of cases (top fit). Wilcoxon signed-rank test shows a significant difference between the time to clinical diagnosis and the time to correct disease suggestion for both top five fit and top fit (z-score -6.68, respective -5.71, α=0.05, p-value <0.001). The DDSS suggested the correct rare disease at the time of diagnosis in 89% of cases (83 of 93) Conclusion: The DDSS was capable of providing accurate rare disease suggestions in most of the rare disease cases. In many cases it provided correct rare disease suggestions early in the course of the disease, sometimes in the very beginning of a patient’s journey. The interpretation of these results suggests that DDSS’s have the potential to highlight the possibility of a rare disease to physicians early in the course of a case. Limitations of this study derive from its retrospective and unblinded design, data input by a single user, and the optimization of the knowledge base during the course of the study. Whether the use of this DDSS leads to a reduced time to rare disease diagnosis in a clinical setting should be validated in prospective studies. References: [1] Blöß S, et al. Diagnostic needs for rare diseases and shared prediagnostic phenomena: Results of a German-wide expert Delphi survey. PLOS ONE. 2017feb;12(2):e0172532. Available from: http://dx.plos.org/10.1371/journal.pone.0172532. [2] Nationales Aktionsbündnis für Menschen mit Seltenen Erkrankungen (NAMSE). Nationaler Aktionsplan für Menschen mit Seltenen Erkrankungen. 2013. Available from: http://www.namse.de/images/stories/Dokumente/nationaler_aktionsplan.pdf. [3] Kostopoulou O, et al. Early diagnostic suggestions improve accuracy of family physicians: a randomized controlled trial in Greece. Family Practice. 2015jun;32(3):323–328. [4] Riches N, et al. The effectiveness of electronic differential diagnoses (DDX) generators: A systematic review and meta-analysis. PLoS ONE. 2016mar;11(3):e0148991. Available from: http://dx.plos.org/10.1371/journal.pone.0148991. Disclosure of Interests: Simon Ronicke Employee of: Ada Health GmbH, Berlin, Martin C. Hirsch Shareholder of: Ada health GmbH, Ewelina Türk Employee of: Ada Health GmbH, Berlin, Katharina Larionov: None declared, Daphne Tientcheu: None declared, Annette D. Wagner: None declared