tetano
Editor, Senior Moderator
Ann Clin Lab Sci
. 2022 Nov;52(6):871-879.
Distinct Laboratory Test Result Profiles Between SARS-CoV-2 and Seasonable Influenza Infected Patients
Junyi Liu[SUP] 1 [/SUP], Lars F Westblade[SUP] 2 3 4 [/SUP], Fei Wang[SUP] 5 [/SUP], Amy Chadburn[SUP] 2 3 [/SUP], Richard Fideli[SUP] 2 3 [/SUP], Arryn Craney[SUP] 2 3 [/SUP], Sophie Rand[SUP] 2 [/SUP], Melissa M Cushing[SUP] 2 3 [/SUP], Jingjing Meng[SUP] 1 [/SUP], Zhen Zhao[SUP] #[/SUP][SUP] 6 3 [/SUP], He S Yang[SUP] #[/SUP][SUP] 7 3 [/SUP]
Affiliations
Abstract
Objective: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and influenza viruses are contagious respiratory pathogens with similar symptoms but require different treatment and management strategies. This study investigated the differences in laboratory test result profiles between SARS-CoV-2 and influenza infected patients upon presentation to emergency department (ED).
Methods: Laboratory test results and demographic information from 723 influenza positive (2018/1/1 to 2020/3/15) and 1,281 SARS-CoV-2 positive (2020/3/11 to 2020/6/30) ED patients were retrospectively analyzed. The dataset was randomly divided into a training/validation set (2/3) and a test set (1/3) with the same SARS-CoV-2/influenza ratio. Four machine learning models in differentiating the laboratory profiles of RT-PCR confirmed SARS-CoV-2 and influenza positive patients were evaluated. The Shapley Additive Explanations technique was employed to visualize the impact of laboratory tests on the overall differentiation. Furthermore, the model performance was also evaluated in a new test dataset including 519 SARS-CoV-2 ED patients (2020/12/1 to 2021/2/28) and the previous influenza positive patients (2018/1/1 to 2020/3/15).
Results: A laboratory test result profile consisting of 15 blood tests, together with patient age, gender, and race can discriminate the two types of viral infections using a random forest (RF) model. The RF model achieved an area under the receiver operating characteristic curve (AUC) of 0.90 in the test set. Among the profile of 15 laboratory tests, the serum total calcium level exhibited the greatest contribution to the overall differentiation. Furthermore, the model achieved an AUC of 0.81 in a new test set.
Conclusion: We developed a laboratory tests-based RF model differentiating SARS-CoV-2 from influenza, which may be useful for the preparedness of overlapping COVID-19 resurgence and future seasonal influenza.
Keywords: COVID-19; SARS-CoV-2; influenza; machine learning; routine laboratory tests.
. 2022 Nov;52(6):871-879.
Distinct Laboratory Test Result Profiles Between SARS-CoV-2 and Seasonable Influenza Infected Patients
Junyi Liu[SUP] 1 [/SUP], Lars F Westblade[SUP] 2 3 4 [/SUP], Fei Wang[SUP] 5 [/SUP], Amy Chadburn[SUP] 2 3 [/SUP], Richard Fideli[SUP] 2 3 [/SUP], Arryn Craney[SUP] 2 3 [/SUP], Sophie Rand[SUP] 2 [/SUP], Melissa M Cushing[SUP] 2 3 [/SUP], Jingjing Meng[SUP] 1 [/SUP], Zhen Zhao[SUP] #[/SUP][SUP] 6 3 [/SUP], He S Yang[SUP] #[/SUP][SUP] 7 3 [/SUP]
Affiliations
- PMID: 36564060
Abstract
Objective: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and influenza viruses are contagious respiratory pathogens with similar symptoms but require different treatment and management strategies. This study investigated the differences in laboratory test result profiles between SARS-CoV-2 and influenza infected patients upon presentation to emergency department (ED).
Methods: Laboratory test results and demographic information from 723 influenza positive (2018/1/1 to 2020/3/15) and 1,281 SARS-CoV-2 positive (2020/3/11 to 2020/6/30) ED patients were retrospectively analyzed. The dataset was randomly divided into a training/validation set (2/3) and a test set (1/3) with the same SARS-CoV-2/influenza ratio. Four machine learning models in differentiating the laboratory profiles of RT-PCR confirmed SARS-CoV-2 and influenza positive patients were evaluated. The Shapley Additive Explanations technique was employed to visualize the impact of laboratory tests on the overall differentiation. Furthermore, the model performance was also evaluated in a new test dataset including 519 SARS-CoV-2 ED patients (2020/12/1 to 2021/2/28) and the previous influenza positive patients (2018/1/1 to 2020/3/15).
Results: A laboratory test result profile consisting of 15 blood tests, together with patient age, gender, and race can discriminate the two types of viral infections using a random forest (RF) model. The RF model achieved an area under the receiver operating characteristic curve (AUC) of 0.90 in the test set. Among the profile of 15 laboratory tests, the serum total calcium level exhibited the greatest contribution to the overall differentiation. Furthermore, the model achieved an AUC of 0.81 in a new test set.
Conclusion: We developed a laboratory tests-based RF model differentiating SARS-CoV-2 from influenza, which may be useful for the preparedness of overlapping COVID-19 resurgence and future seasonal influenza.
Keywords: COVID-19; SARS-CoV-2; influenza; machine learning; routine laboratory tests.