Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset
| dc.contributor.advisor | Gumedze, Freedom | |
| dc.contributor.advisor | Ntsekhe, Mpiko | |
| dc.contributor.author | Geffen, Hayli | |
| dc.date.accessioned | 2023-03-02T11:40:58Z | |
| dc.date.available | 2023-03-02T11:40:58Z | |
| dc.date.issued | 2022 | |
| dc.date.updated | 2023-02-20T12:47:22Z | |
| dc.description.abstract | Despite the recent global decline of tuberculosis infections, constrictive pericarditis, one of the most serious consequences of tuberculous pericarditis, continues to be a major cause of morbidity and mortality in sub-Saharan Africa. Currently, while the risk of constrictive pericarditis in individuals with tuberculous pericarditis does not appear to be uniform, there is no defined risk score available to predict an individual's baseline risk of constrictive pericarditis. Therefore the main aim of this research was to employ supervised learning classification using the data from 1400 participants enrolled in the first Investigation of the Management of Pericarditis randomised clinical trial to derive a risk score for constrictive pericarditis. While various supervised learning classification methods, including tree-based algorithms, support vector machines and artificial neural networks, were compared to stratify individuals according to low, medium and high risk for constrictive pericarditis, the final risk score was developed using logistic regression. Significant associations were found between constrictive pericarditis and the following predictors: HIV, New York Heart Association functional class, cardiac tamponade and effusive-constrictive pericarditis. Although prednisolone treatment was associated with a reduced relative risk of constrictive pericarditis in high (risk ratio = 0.59; 95% CI = 0.378 – 0.925) and medium (risk ratio = 0.12; 95% CI = 0.016 – 0.971) risk individuals, prednisolone treatment did not seem to benefit the individuals predicted to be at low risk (risk ratio = 0.92; 95% CI = 0.084 - 10.047) for constrictive pericarditis. These results confirm that the baseline risk of developing constrictive pericarditis in individuals with suspected or confirmed tuberculous pericarditis is not uniform. Importantly, interventions such as adjunctive prednisolone should only be recommended for individuals suspected to be at either medium or high risk for constrictive pericarditis as they are the most likely to benefit while prednisolone treatment should potentially be avoided in treating individuals with tuberculous pericarditis that are suspected to be at low risk for constrictive pericarditis as they are the least likely to derive any benefit. | |
| dc.identifier.apacitation | Geffen, H. (2022). <i>Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset</i>. (). ,Faculty of Science ,Department of Statistical Sciences. Retrieved from http://hdl.handle.net/11427/37154 | en_ZA |
| dc.identifier.chicagocitation | Geffen, Hayli. <i>"Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset."</i> ., ,Faculty of Science ,Department of Statistical Sciences, 2022. http://hdl.handle.net/11427/37154 | en_ZA |
| dc.identifier.citation | Geffen, H. 2022. Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset. . ,Faculty of Science ,Department of Statistical Sciences. http://hdl.handle.net/11427/37154 | en_ZA |
| dc.identifier.ris | TY - Master Thesis AU - Geffen, Hayli AB - Despite the recent global decline of tuberculosis infections, constrictive pericarditis, one of the most serious consequences of tuberculous pericarditis, continues to be a major cause of morbidity and mortality in sub-Saharan Africa. Currently, while the risk of constrictive pericarditis in individuals with tuberculous pericarditis does not appear to be uniform, there is no defined risk score available to predict an individual's baseline risk of constrictive pericarditis. Therefore the main aim of this research was to employ supervised learning classification using the data from 1400 participants enrolled in the first Investigation of the Management of Pericarditis randomised clinical trial to derive a risk score for constrictive pericarditis. While various supervised learning classification methods, including tree-based algorithms, support vector machines and artificial neural networks, were compared to stratify individuals according to low, medium and high risk for constrictive pericarditis, the final risk score was developed using logistic regression. Significant associations were found between constrictive pericarditis and the following predictors: HIV, New York Heart Association functional class, cardiac tamponade and effusive-constrictive pericarditis. Although prednisolone treatment was associated with a reduced relative risk of constrictive pericarditis in high (risk ratio = 0.59; 95% CI = 0.378 – 0.925) and medium (risk ratio = 0.12; 95% CI = 0.016 – 0.971) risk individuals, prednisolone treatment did not seem to benefit the individuals predicted to be at low risk (risk ratio = 0.92; 95% CI = 0.084 - 10.047) for constrictive pericarditis. These results confirm that the baseline risk of developing constrictive pericarditis in individuals with suspected or confirmed tuberculous pericarditis is not uniform. Importantly, interventions such as adjunctive prednisolone should only be recommended for individuals suspected to be at either medium or high risk for constrictive pericarditis as they are the most likely to benefit while prednisolone treatment should potentially be avoided in treating individuals with tuberculous pericarditis that are suspected to be at low risk for constrictive pericarditis as they are the least likely to derive any benefit. DA - 2022_ DB - OpenUCT DP - University of Cape Town KW - Statistical Sciences LK - https://open.uct.ac.za PY - 2022 T1 - Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset TI - Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset UR - http://hdl.handle.net/11427/37154 ER - | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11427/37154 | |
| dc.identifier.vancouvercitation | Geffen H. Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset. []. ,Faculty of Science ,Department of Statistical Sciences, 2022 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/37154 | en_ZA |
| dc.language.rfc3066 | eng | |
| dc.publisher.department | Department of Statistical Sciences | |
| dc.publisher.faculty | Faculty of Science | |
| dc.subject | Statistical Sciences | |
| dc.title | Development of a risk score for constrictive pericarditis using the Investigation of the Management of Pericarditis randomised clinical trial dataset | |
| dc.type | Master Thesis | |
| dc.type.qualificationlevel | Masters | |
| dc.type.qualificationlevel | MSc |