A novel intubation prediction model for patients hospitalized with COVID-19: the OTO-COVID-19 scoring model

Abstract

Objective

The method for predicting the risk of intubation in patients with coronavirus disease 2019 (COVID-19) is yet to be standardized. This study aimed to introduce a new disease prognosis scoring model that may predict the intubation risk based on the symptoms, signs, and laboratory tests of patients hospitalized with the diagnosis of COVID-19.

Method

This cross-sectional retrospective study analyzed the intubation status of 733 patients hospitalized with COVID-19 diagnosis between March and December 2020 at Ondokuz Mayıs University Faculty of Medicine, Turkey, based on 33 variables. Binary logistic regression analysis was used to select the variables that significantly affect intubation, which constitute the risk factors. The Chi-square Automatic Interaction Detection algorithm, one of the data mining methods, was used to determine the threshold values of the important variables for intubation classification.

Results

The following variables found were mostly associated with intubation: C-reactive protein, lactate dehydrogenase, neutrophil-to-lymphocyte ratio, age, lymphocyte count, and malignancy. The logistic function based on these variables correctly predicted 81.13% of intubated (sensitivity), 99.52% of nonintubated (specificity), and 96.86% of both intubated and nonintubated (accurate classification rate) patients. The scoring model revealed the following risk statuses for the intubated patients: very high risk, 75.47%; moderate risk, 20.75%; and very low risk, 3.77%.

Conclusions

On the basis of certain variables measured at admission, the OTO-COVID-19 scoring model may help clinicians identify patients at the risk of intubation and subsequently provide a prompt and effective treatment at the earliest.

Keywords:

• COVID-19
• intubation
• risk scores

Transparency

Declaration of funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of financial/other relationships

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgements

None.

Data availability statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.