Session Type: ePosters
Session Title: ePosters
Authors(s): M. Lyu (1), H. Lai (2), Y. Zhou (1), Y. Wang (1), J. Tang (2), X. Chen (3), B. Ying (1)
Authors Affiliations(s): (1) Department of Laboratory Medicine, West China Hospital, Sichuan University, China, (2) West China Medical School/West China Hospital, Sichuan University, China, (3) Department of Respiratory and Critical Care Medicine, West China Hospital, Sichuan University, China
Background:
This study was conducted to explore the impact of rheumatoid factor (RF) on tuberculosis (TB)-related prognosis of patients with TB and rheumatoid arthritis (RA) and further establish predictive models for 2-month sputum smear conversion (2m-SSC).
Methods:From December 2013 to September 2019, initial and follow-up (every 6 months) medical records of TB-RA patients from West China Hospital were searched. Logistic and cox regression analysis were used to identify risk factors for 2m-SSC and TB relapse, respectively. Odds ratio (OR) or hazard ratio (HR) with corresponding 95% confidence interval (CI) were computed. Logistic regression and elastic net regression (ENR) were performed to build predictive models for 2m-SSC. Sensitivity, specificity and area under curve (AUC) were calculated. P<0.05 indicated a statistical significance.
Results:A total of 434 patients were finally included after reviewing 812 patients. Subjects were divided into high-RF and low-RF group according to the cut-off value (24). RF grading (OR=9.548, 95%CI: 5.390-16.913, p<0.001) was an independent risk factor for 2m-SSC. There was no significance in TB relapse risk between 2 groups (p=0.292). Multivariate models including RF grading were built to predict 2m-SSC. Sensitivity, specificity and AUC of logistic regression model were 57.0%, 78.3% and 66.6% in training set, while they were 53.8%, 92.9% and 74.9% in validation set. Using parameters selected by K fold cross-validation, ENR model showed 81.2% of sensitivity, 70.0% of specificity and 80.8% of AUC in training set, and 62.1% of sensitivity, 89.3% of specificity and 78.0% of AUC in validation set. When parameters were selected by minimum root mean squared error, sensitivity, specificity and AUC were 70%, 80.5% and 80.8% in training set, while they were 89.3%, 62.7% and 78.1% in validation set.
Conclusions:RF grading was an independent risk factor for 2m-SSC for TB-RA patients and contributed to the excellent performance of 2m-SSC predictive models. Grading RF level might assist clinicians to predict TB-related outcomes and thus take early measure to improve the prognosis of TB-RA patients.
Keyword(s): tuberculosis, rheumatoid factor, prognosisSession Type: ePosters
Session Title: ePosters
Authors(s): M. Lyu (1), H. Lai (2), Y. Zhou (1), Y. Wang (1), J. Tang (2), X. Chen (3), B. Ying (1)
Authors Affiliations(s): (1) Department of Laboratory Medicine, West China Hospital, Sichuan University, China, (2) West China Medical School/West China Hospital, Sichuan University, China, (3) Department of Respiratory and Critical Care Medicine, West China Hospital, Sichuan University, China
Background:
This study was conducted to explore the impact of rheumatoid factor (RF) on tuberculosis (TB)-related prognosis of patients with TB and rheumatoid arthritis (RA) and further establish predictive models for 2-month sputum smear conversion (2m-SSC).
Methods:From December 2013 to September 2019, initial and follow-up (every 6 months) medical records of TB-RA patients from West China Hospital were searched. Logistic and cox regression analysis were used to identify risk factors for 2m-SSC and TB relapse, respectively. Odds ratio (OR) or hazard ratio (HR) with corresponding 95% confidence interval (CI) were computed. Logistic regression and elastic net regression (ENR) were performed to build predictive models for 2m-SSC. Sensitivity, specificity and area under curve (AUC) were calculated. P<0.05 indicated a statistical significance.
Results:A total of 434 patients were finally included after reviewing 812 patients. Subjects were divided into high-RF and low-RF group according to the cut-off value (24). RF grading (OR=9.548, 95%CI: 5.390-16.913, p<0.001) was an independent risk factor for 2m-SSC. There was no significance in TB relapse risk between 2 groups (p=0.292). Multivariate models including RF grading were built to predict 2m-SSC. Sensitivity, specificity and AUC of logistic regression model were 57.0%, 78.3% and 66.6% in training set, while they were 53.8%, 92.9% and 74.9% in validation set. Using parameters selected by K fold cross-validation, ENR model showed 81.2% of sensitivity, 70.0% of specificity and 80.8% of AUC in training set, and 62.1% of sensitivity, 89.3% of specificity and 78.0% of AUC in validation set. When parameters were selected by minimum root mean squared error, sensitivity, specificity and AUC were 70%, 80.5% and 80.8% in training set, while they were 89.3%, 62.7% and 78.1% in validation set.
Conclusions:RF grading was an independent risk factor for 2m-SSC for TB-RA patients and contributed to the excellent performance of 2m-SSC predictive models. Grading RF level might assist clinicians to predict TB-related outcomes and thus take early measure to improve the prognosis of TB-RA patients.
Keyword(s): tuberculosis, rheumatoid factor, prognosis
Session Type: ePosters
Session Title: ePosters
Authors(s): M. Lyu (1), H. Lai (2), Y. Zhou (1), Y. Wang (1), J. Tang (2), X. Chen (3), B. Ying (1)
Authors Affiliations(s): (1) Department of Laboratory Medicine, West China Hospital, Sichuan University, China, (2) West China Medical School/West China Hospital, Sichuan University, China, (3) Department of Respiratory and Critical Care Medicine, West China Hospital, Sichuan University, China
Background:
This study was conducted to explore the impact of rheumatoid factor (RF) on tuberculosis (TB)-related prognosis of patients with TB and rheumatoid arthritis (RA) and further establish predictive models for 2-month sputum smear conversion (2m-SSC).
Methods:From December 2013 to September 2019, initial and follow-up (every 6 months) medical records of TB-RA patients from West China Hospital were searched. Logistic and cox regression analysis were used to identify risk factors for 2m-SSC and TB relapse, respectively. Odds ratio (OR) or hazard ratio (HR) with corresponding 95% confidence interval (CI) were computed. Logistic regression and elastic net regression (ENR) were performed to build predictive models for 2m-SSC. Sensitivity, specificity and area under curve (AUC) were calculated. P<0.05 indicated a statistical significance.
Results:A total of 434 patients were finally included after reviewing 812 patients. Subjects were divided into high-RF and low-RF group according to the cut-off value (24). RF grading (OR=9.548, 95%CI: 5.390-16.913, p<0.001) was an independent risk factor for 2m-SSC. There was no significance in TB relapse risk between 2 groups (p=0.292). Multivariate models including RF grading were built to predict 2m-SSC. Sensitivity, specificity and AUC of logistic regression model were 57.0%, 78.3% and 66.6% in training set, while they were 53.8%, 92.9% and 74.9% in validation set. Using parameters selected by K fold cross-validation, ENR model showed 81.2% of sensitivity, 70.0% of specificity and 80.8% of AUC in training set, and 62.1% of sensitivity, 89.3% of specificity and 78.0% of AUC in validation set. When parameters were selected by minimum root mean squared error, sensitivity, specificity and AUC were 70%, 80.5% and 80.8% in training set, while they were 89.3%, 62.7% and 78.1% in validation set.
Conclusions:RF grading was an independent risk factor for 2m-SSC for TB-RA patients and contributed to the excellent performance of 2m-SSC predictive models. Grading RF level might assist clinicians to predict TB-related outcomes and thus take early measure to improve the prognosis of TB-RA patients.
Keyword(s): tuberculosis, rheumatoid factor, prognosisSession Type: ePosters
Session Title: ePosters
Authors(s): M. Lyu (1), H. Lai (2), Y. Zhou (1), Y. Wang (1), J. Tang (2), X. Chen (3), B. Ying (1)
Authors Affiliations(s): (1) Department of Laboratory Medicine, West China Hospital, Sichuan University, China, (2) West China Medical School/West China Hospital, Sichuan University, China, (3) Department of Respiratory and Critical Care Medicine, West China Hospital, Sichuan University, China
Background:
This study was conducted to explore the impact of rheumatoid factor (RF) on tuberculosis (TB)-related prognosis of patients with TB and rheumatoid arthritis (RA) and further establish predictive models for 2-month sputum smear conversion (2m-SSC).
Methods:From December 2013 to September 2019, initial and follow-up (every 6 months) medical records of TB-RA patients from West China Hospital were searched. Logistic and cox regression analysis were used to identify risk factors for 2m-SSC and TB relapse, respectively. Odds ratio (OR) or hazard ratio (HR) with corresponding 95% confidence interval (CI) were computed. Logistic regression and elastic net regression (ENR) were performed to build predictive models for 2m-SSC. Sensitivity, specificity and area under curve (AUC) were calculated. P<0.05 indicated a statistical significance.
Results:A total of 434 patients were finally included after reviewing 812 patients. Subjects were divided into high-RF and low-RF group according to the cut-off value (24). RF grading (OR=9.548, 95%CI: 5.390-16.913, p<0.001) was an independent risk factor for 2m-SSC. There was no significance in TB relapse risk between 2 groups (p=0.292). Multivariate models including RF grading were built to predict 2m-SSC. Sensitivity, specificity and AUC of logistic regression model were 57.0%, 78.3% and 66.6% in training set, while they were 53.8%, 92.9% and 74.9% in validation set. Using parameters selected by K fold cross-validation, ENR model showed 81.2% of sensitivity, 70.0% of specificity and 80.8% of AUC in training set, and 62.1% of sensitivity, 89.3% of specificity and 78.0% of AUC in validation set. When parameters were selected by minimum root mean squared error, sensitivity, specificity and AUC were 70%, 80.5% and 80.8% in training set, while they were 89.3%, 62.7% and 78.1% in validation set.
Conclusions:RF grading was an independent risk factor for 2m-SSC for TB-RA patients and contributed to the excellent performance of 2m-SSC predictive models. Grading RF level might assist clinicians to predict TB-related outcomes and thus take early measure to improve the prognosis of TB-RA patients.
Keyword(s): tuberculosis, rheumatoid factor, prognosis