Abstract: TH-PO1092
Incorporation of Chest Computed Tomography Quantification to Predict Outcomes in Hemodialysis Patients with COVID-19
Session Information
- COVID-19 - I
November 02, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Coronavirus (COVID-19)
- 000 Coronavirus (COVID-19)
Authors
- Xing, Haifan, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
- Li, Ze, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
- Fan, Ying, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China
Background
Patients undergoing maintenance hemodialysis are vulnerable to coronavirus disease 2019 (Covid-19) with a higher risk of hospitalizations and mortality. Early identification and intervention is important to prevent the disease progression in these dialysis patients with Covid-19.
Methods
A total of 186 Covid-19 patients from a hemodialysis center in Shanghai 6th People’s hospital, China were enrolled between March 2021 and June 2021. 70% of patients (n = 130) were randomly selected in the training set for the establishment of a prognostic nomogram. 30% (n = 56) were included for the validation of the predictive model. Artificial intelligence (AI) based parameters of chest computed tomography (CT) were quantitated; demographic features, comorbidities and laboratory examination items were screened using univariate and multivariate Cox regression analyses to construct a nomogram predicting the risk of death or transferring to ICU within 28 days in hemodialysis patients with COVID-19. The concordance statistics (C- statistics) and calibration curves were used to assess model performance.
Results
The recruited patients had a median age of 65 years (IQR 56-73 years) and 58.6% of participants were male. Median time on dialysis was 4 years (IQR 2-8 years). 81.2% dialyzed through arteriovenous fistula. Age, diabetes mellitus, serum phosphorus, lymphocyte count and thin-section CT scores were identified as independent prognostic factors (p < 0.05) which were further incorporated into the nomogram. The C-statistics were 0.865 and 0.748 in the training and validation sets, respectively. Calibration plots show good agreement between expected and actual outcomes.
Conclusion
This is the first study to develop a reliable nomogram using clinical indicators and AI based CT image parameters to predict outcome and survival probabilities in hemodialysis patients with COVID-19. This model could be helpful to clinicians in treating SARS-CoV-2 infection, managing serum phosphorus and adjusting the dialysis strategies in these patients to against severe-critical disease progression.
Funding
- Government Support – Non-U.S.