COVID19 Severity Scoring from CT — Primer for Radiologists

Suthirth Vaidya


The WHO Guide on ‘Use of Chest Imaging in COVID-19’ underscored the importance of imaging modalities such as chest CT in the management of COVID19 patients. Given the emphasis on CT’s role in visualizing lung damage, there is a pertinent need for an easy-to-use, reproducible method to standardize the communication of COVID-19 severity. In this article, we attempt to explain the 25-point severity score, CT severity score, or Total Severity Score (TSS) method.

This article has been divided into below sub-sections:

  • What is the Total Severity Score?
  • How do you calculate the Total Severity Score?
  • How does the Total Severity Score correlate with clinical features?
  • How does the Total Severity Score change over time?
What is the Total Severity Score?

The Total Severity Score (TSS) for COVID-19 allows for standardized communication of pulmonary involvement of COVID19-related abnormalities from thin-section CT imaging. The method can be adopted by radiologists without any additional tools and is fairly reproducible.

Papers using the Total Severity Score (TSS) can be seen in studies as early as the 2002 SARS epidemic. YC Chang et al. (Radiology 2005) had proposed the scoring method in their paper titled “Pulmonary Sequelae in Convalescent Patients after Severe Acute Respiratory Syndrome: Evaluation with Thin-Section CT”, showing that the score correlated with adult respiratory distress syndrome (ARDS). With the outbreak of the COVID-19 pandemic, however, there are numerous studies which have used the TSS method since February 2020.

Notably — the method was initially suggested by M Chung et al. (Radiology 2020) describing the CT findings of COVID-19 and most popularly used by F Pan et al. (Radiology 2020) on a study of the longitudinal changes on CT in recovered COVID19 patients. As of date, the latter has been cited over 600 times.

How do you calculate the Total Severity Score (TSS)?

The “total severity score” is a semi-quantitative scoring system used to estimate the pulmonary involvement of COVID19 related abnormalities on the basis of the area involved. Each of the five lung lobes is visually scored on a scale of 0 to 5 with:

  • 0: None (no involvement in lobe)
  • 1: < 5% of lobe
  • 2: 5%- 25% of lobe
  • 3: 26%- 49% of lobe
  • 4: 50%- 75% of lobe
  • 5: >75% of lobe

The total CT score is the sum of the individual lobar scores and ranges from 0 (no involvement) to 25 (maximum involvement).

Major CT findings in COVID19 as defined by the Fleischner Society Glossary include ground-glass opacity (GGO), crazy-paving pattern, and consolidation. The score is calculated using thin-section computed tomography images of the chest.

Figure borrowed from M Francone et al, (Eur Radiol 2020) that shows different CT scores of right lower lobe (RLL) involvement in COVID-19 pneumonia on axial, sagittal, and coronal images.

How does Total Severity Score correlate with clinical features?

At this point, it is important to note the clinical classification that was most commonly followed by these papers, as laid out by the Diagnosis and Treatment Plan of COVID19 issued by the National Health Commission of China. They define the types as:

  • Mild Type (minimal, light)
    mild clinical symptoms without pneumonia in imaging
  • Common Type (ordinary)
    fever, respiratory tract and other symptoms with pneumonia in imaging
  • Severe Type
    respiratory distress, respiratory rate >= 30 times/min; in resting state, oxygen saturation <= 93%; PaO2/FiO2 <= 300mm Hg
  • Critical Type
    respiratory failure requiring mechanical ventilation, shock and other organ failure requiring ICU monitoring and treatment

K Li et al. Eur Radiol (2020) in their study of 78 patients with 31% light type, 59% common and 9% of severe-critical type showed the below total severity score distribution on admission. The median TSS was 10 (range 8- 18) in the severe-critical-type group, significantly higher than that of common type with a median of 5 (range 1- 11).

Their evaluation also showed that the area under the curve (AUC) of TSS for diagnosing severe-critical type was 0.918 (95% CI: 0.843–0.994). They suggested that a TSS cutoff of 7.5 had 82.6% sensitivity and 100% specificity.

Figure from K Li et al. Eur Radiol (2020) shows the distribution of TSS for different clinical subtypes, with a median score of 10 for severe-critical type and a median score of 5 for common type.

Another study by K Li et al. (Invest Radiol. 2020) on 83 patients with 70% with ordinary type and 30% with severe-critical type showed similar cut-offs, with a median TSS score of 11 (interquartile range 8- 15.5) for severe-critical type and a median TSS score of 5 (interquartile range 2.5- 5) for the ordinary type.

Using a univariate logistic analysis of clinical and CT features in relationship to severe/critical pneumonia, they show that a CT score > 7 is statistically significant.

M Francone et al. (Eur Radiol 2020) studied the correlation of chest CT score with the short-term prognosis on their dataset of 130 patients with 61% mild-common, 32% severe and 7% critical. 15% of the patients died during the follow-up period. They carry out a Kaplan-Meier analysis and show that an estimated CT score cut-off of 18 significantly increases the risk of death over an observational period of 24 days.

Figure from M Francone et al. (Eur Radiol 2020) showing Kaplan-Meier curve showing an increased 24-day risk of death for patients with CT score greater than 18
How does Total Severity Score change over time?

Few groups have studied the progression of TSS using longitudinal CT scans in recovered patients. These studies excluded all patients who faced severe respiratory distress (respiratory rate >30 breaths per minute), a requirement of oxygen treatment or mechanical ventilation, or oxygen saturation as measured with pulse oximetry of less than 90% on room air.

According to F Pan et al. (Radiology 2020), on a sample size of 21 patients in Wuhan with 82 chest CT scans, the progression pattern is marked by an increase in score with the peak at around 10 days after onset of initial symptoms. While signs of improvement began at approximately 14 days after onset of initial symptoms. The below figure shows the change in severity for all patients (A) and a cubic interpolation of the observed data (B).

Figure from F Pan et al. (Radiology 2020) showing CT score progression in 21 recovered patients (A) and a cubic interpolation of the observed data (B)

The other group, X Ding et al (Eur Radiol 2020) studied the progression on a larger dataset of 112 patients again from Wuhan with 348 CT scans. They report a similar increase in CT score until Stage 3 (10–14days) after which the disease tended to be stable and lasted for a long time.

Figure from X Ding et al (Eur Radiol 2020) showing the distribution of total CT score based on stages, based on the time from onset of symptoms.

Given that the scores both increase due to the time progression of the disease as well as in the severe-critical type, it is important to have relevant clinical context, specifically the time from onset of symptoms, to arrive at the optimal therapeutic management for the patient. For instance, while a CT score of 8 could indicate severe-type in a stage-1 patient (days 0–4), it could fit well in the normative range for recovery in a stage-2 patient.

Studies showing CT score correlations in both these dimensions — outcomes as well as time could not be found, and can potentially be further useful in clinical routine practice for patient monitoring.

As a disclosure, our team at Predible Health has developed a COVID19 severity scoring on our LungIQ platform, enabling instant severity estimation from chest CT studies. Our analytics feature also lets you track patient progress over time as well as assign patient cohorts based on outcomes. If you would like to push the understanding of treatment protocols and management guidelines for COVID19 patients with CT imaging, we can help.

  1. Use of chest imaging in COVID-19 — A rapid advice guide. WHO Reference Number: WHO/2019-nCoV/Clinical/Radiology_imaging/2020.1
  2. Chang, Yeun-Chung, et al. “Pulmonary sequelae in convalescent patients after severe acute respiratory syndrome: evaluation with thin-section CT.” Radiology 236.3 (2005): 1067–1075.
  3. Chung, Michael, et al. “CT imaging features of 2019 novel coronavirus (2019-nCoV).” Radiology 295.1 (2020): 202–207.
  4. Pan, Feng, et al. “Time course of lung changes on chest CT during recovery from 2019 novel coronavirus (COVID-19) pneumonia.” Radiology (2020): 200370.
  5. Hansell, David M., et al. “Fleischner Society: glossary of terms for thoracic imaging.” Radiology 246.3 (2008): 697–722.
  6. Francone, Marco, et al. “Chest CT score in COVID-19 patients: correlation with disease severity and short-term prognosis.” European Radiology (2020): 1–10.
  7. Li, Kunwei, et al. “CT image visual quantitative evaluation and clinical classification of coronavirus disease (COVID-19).” European radiology (2020): 1–10.
  8. Li, Kunhua, et al. “The clinical and chest CT features associated with severe and critical COVID-19 pneumonia.” Investigative radiology (2020).
  9. Ding, Xun, et al. “Chest CT findings of COVID-19 pneumonia by duration of symptoms.” European Journal of Radiology (2020): 109009.

Read the original post
Schedule demo
Take a look at related news on this category...

No items found.