P157 Serial daily recording of patient-reported outcomes (PRO2) might predict treatment change in the patients with Ulcerative Colitis.
Tsuchiya, K.(1,2)*;Nagahori, M.(3);Hanazawa, R.(2);Sasaki, M.(2);Sato, H.(2);Hirakawa, A.(2);Anzai, T.(4);Takahashi, K.(4);Kobayashi, Y.(5);Hibiya, S.(6);Takenaka, K.(6);Watanabe, M.(7);
(1)University of Tsukuba, Department of Gastroenterology, Tsukuba, Japan;(2)Tokyo Medical and Dental University, Division of Clinical Management, Tokyo, Japan;(3)Tokyo Medical and Dental University, Division of Clinical Management- Clinical Research Center, Tokyo, Japan;(4)Tokyo Medical and Dental University, M&D Data Science Center, Tokyo, Japan;(5)Hitachi ltd., Hitachi ltd., Tokyo, Japan;(6)Tokyo Medical and Dental University, Department of Gastroenterology, Tokyo, Japan;(7)Tokyo Medical and Dental University, Inflammatory Bowel Disease Laboratory, Tokyo, Japan;
Background
Patient-reported outcome (PRO2) score composed solely of the stool frequency and bleeding components from partial Mayo (pMayo) score are used for the evaluation of clinical symptom in the patients with Ulcerative Colitis (UC). Although pMayo and PRO2 score also perform as well as the full Mayo score to identify the disease activity and clinical response, it is unclear whether serial daily PRO2 score are relevant in clinical practice. Moreover, its clinical significance for long-term outcome has not been established. The aim of this study istherefore to clarify the relationship between daily PRO2 records using a smart device application and prognosis.
Methods
This study was approved by the ethics committee in Tokyo Medical and Dental University (TMDU). The application was originally developed by the cooperation with TMDU and Hitachi ltd. Patients with UC input their symptoms by themselves in accordance with PRO2 once a day using the application on their smart device. The alert for their symptom exacerbation is set according to our logic. Patient's condition including alert status is confirmed by the doctor once a week using doctor’s PC. Medical information such as therapeutics and blood examination, was extracted from the electronic medical chart in the hospital. The application data was combined with medical information on a daily basis. We defined a start or change of the treatment with prednisone or biologics as treatment change for the outcome of this study. The difference of PRO2 score before the date of treatment change is calculated for each patient and the odds ratio (OR) and AUC of the ROC curve are estimated by univariate logistic regression analysis.
Results
From July 2020 to March 2022, 183 UC patients participated in this study. Mean age was 46.6 years. Male to female ratio was 81:102. Mean observation period was 346 days. The number of the patients with alert of symptom exacerbation averaged 5.79/week. There were 27 patients with treatment change during the period. PRO2 score in the patients with treatment change was significantly increased compared with the patients with no change in treatment. The OR [95%CI] of the PRO2 change from 90 days to 5 days before treatment change was 7.708 [2.713-21.902], the AUC [95%CI] was 0.872 [0.772-0.973], the OR [95%CI] of the PRO2 change from 10 days to 5 days was 3.309 [1.601-6. 841] and AUC [95%CI] was 0.679 [0.569-0.789].
Conclusion
PRO2 variation over 90 days was highly reliable for predicting treatment change. Further analysis will clarify the clinical significance of serial daily PRO2 score recording.