Abstract
Introduction/Background and aims
Over the past year numerous mortality prediction calculators have been created and validated for use in clinical settings to support healthcare workers identify patients most in need from SARS-CoV-2 viral infections. However, with the successful uptake of vaccines and other measures to control the spread of SARS-CoV-2 virus, we are faced with new challenges in other viral respiratory diseases such as Influenza A and B alongside SARS-CoV-2. Given that no validated risk-prediction tool exists for viral lung diseases other than SARS-CoV-2, and that the LUCAS tool predicts from very few generic inflammatory parameters, we evaluated the performance of LUCAS on patients with confirmed Influenza A and B on admission to hospital in a retrospective cohort study.
Methods/summary of work
LUCAS is based on lymphocyte count, urea, C-reactive protein (CRP), age and sex, with the option to include chest X-Ray results. This robust and highly accurate calculator [1] uses multivariable logistic regression that was developed on a database (NCCID) based on 19 hospitals within the UK. The Simple Logistic Regression based calculator is freely available at https://mdscore.net/.
De-identified and pseudo-anonymised patient data were obtained from data sets which were approved by the ethics committee as part of the existing Cardiac MRI Database NHS REC IRAS Ref: 222349 and University of Brighton REC (8011).
The LUCAS model was initially based on a subset of the NCCID dataset (23/01/20 to 7/12/20) temporally split between development and internal validation dataset with 1434 and 310 SARS-CoV-2 positive patients, respectively. The LUCAS mortality score included the five strongest predictors, which are available at any point of care with rapid turnaround of results.
Results/Discussion
Our simple multivariable logistic model showed high discrimination for SARS-CoV-2 fatal outcome with the AUC-ROC in development cohort 0.765 (95% confidence interval (CI): 0.738-0.790) and in internal validation cohort 0.744 (CI: 0.673-0.808). In an external validation set (NHS Greater Glasgow and Clyde; 2338 SARS-V-2 positive patients; 01/01/18 to 24/10/20), LUCAS predicted the survival of SARS-CoV-2 positive patients with 95.2% specificity (Table 1), from 635 patients positive for Influenza A, LUCAS demonstrated 95.8% specificity, and from 258 patients positive for Influenza B, LUCAS demonstrated 91.8% specificity.
Conclusion
LUCAS can be used to obtain valid predictions of survival of patients with Influenza A or B, in addition to the survival of patients within 60 days of SARS-CoV-2 RT-PCR results.
Over the past year numerous mortality prediction calculators have been created and validated for use in clinical settings to support healthcare workers identify patients most in need from SARS-CoV-2 viral infections. However, with the successful uptake of vaccines and other measures to control the spread of SARS-CoV-2 virus, we are faced with new challenges in other viral respiratory diseases such as Influenza A and B alongside SARS-CoV-2. Given that no validated risk-prediction tool exists for viral lung diseases other than SARS-CoV-2, and that the LUCAS tool predicts from very few generic inflammatory parameters, we evaluated the performance of LUCAS on patients with confirmed Influenza A and B on admission to hospital in a retrospective cohort study.
Methods/summary of work
LUCAS is based on lymphocyte count, urea, C-reactive protein (CRP), age and sex, with the option to include chest X-Ray results. This robust and highly accurate calculator [1] uses multivariable logistic regression that was developed on a database (NCCID) based on 19 hospitals within the UK. The Simple Logistic Regression based calculator is freely available at https://mdscore.net/.
De-identified and pseudo-anonymised patient data were obtained from data sets which were approved by the ethics committee as part of the existing Cardiac MRI Database NHS REC IRAS Ref: 222349 and University of Brighton REC (8011).
The LUCAS model was initially based on a subset of the NCCID dataset (23/01/20 to 7/12/20) temporally split between development and internal validation dataset with 1434 and 310 SARS-CoV-2 positive patients, respectively. The LUCAS mortality score included the five strongest predictors, which are available at any point of care with rapid turnaround of results.
Results/Discussion
Our simple multivariable logistic model showed high discrimination for SARS-CoV-2 fatal outcome with the AUC-ROC in development cohort 0.765 (95% confidence interval (CI): 0.738-0.790) and in internal validation cohort 0.744 (CI: 0.673-0.808). In an external validation set (NHS Greater Glasgow and Clyde; 2338 SARS-V-2 positive patients; 01/01/18 to 24/10/20), LUCAS predicted the survival of SARS-CoV-2 positive patients with 95.2% specificity (Table 1), from 635 patients positive for Influenza A, LUCAS demonstrated 95.8% specificity, and from 258 patients positive for Influenza B, LUCAS demonstrated 91.8% specificity.
Conclusion
LUCAS can be used to obtain valid predictions of survival of patients with Influenza A or B, in addition to the survival of patients within 60 days of SARS-CoV-2 RT-PCR results.
Original language | English |
---|---|
Publication status | Published - 7 Sept 2021 |
Event | Pharmacology 2021: today's science, tomorrow's medicines - Online, United Kingdom Duration: 7 Sept 2020 → 9 Sept 2021 https://meetings.bps.ac.uk/bpsevents/frontend/reg/thome.csp?pageID=21432&ef_sel_menu=250&eventID=40 |
Conference
Conference | Pharmacology 2021 |
---|---|
Country/Territory | United Kingdom |
Period | 7/09/20 → 9/09/21 |
Other | Conference celebrating the 90th anniversary of the British Pharmacological Society |
Internet address |