The main findings of our study can be summarized as follows: 1) CoSAS more accurately predicted 28-day mortality in an adult patient with severe covid-19 illness as compared to qSOFA (AUROC 0.78 vs 0.70); 2) Age of more than 55 years, male gender and having previous co-morbidities such as HTN, DM, IHD and CKD are all predictors of severe covid-19 illness. Our study has been conducted in Pakistan at a tertiary care center enrolling more than 200 patients to create an accurate prognostication tool for severe COVID-19 illness. The lack of a specific risk-scoring system for COVID-19 prompted the use of other previously validated screening tools such as qSOFA. Each carrying its innate advantages and disadvantages, CoSAS was hypothesized to be a risk-scoring tool in the evaluation of severely ill COVID-19 patient presenting to the emergency department. Although not validated, the information from a simple screening model may provide useful prognostic information to an Emergency Department and admitting clinicians, thereby potentially directing scarce personnel and medical resources towards those hospitalized individuals who are at the greatest risk of dying.
Multiple studies have been conducted globally on COVID-19, each providing a multitude of factors that can predict severity of covid-19 illness . Demographically many studies have suggested age > 55 years and the male sex both carry a higher predictive outcome for increased severity . Our study reflected similar results with regards to mean age for low risk (<6) being 58 years and high risk (≥6) associated with > 72 years of age. The male gender has also been a significant predictor of covid-19 illness. The American College of Cardiology along with the CDC have both stated that that male gender carries a higher risk of severe covid (20,21). Fatality rates were highest for cardiovascular disease (10.5%) compared with diabetes (7.3%), COPD (6.3%), hypertension (6.0%), and cancer (5.6%). In contrast, patients without pre-existing conditions had a fatality rate of <1% . A large analysis of 308,010 COVID-19 adults hospitalized at US academic centers showed that males have a higher rate of respiratory intubation and longer length of hospital stay compared to females and have a higher death rate even when compared across age groups, race/ethnicity, payers, and co-morbidity . Our study also supported these findings as a statistically significant (p<0.001) result for the male gender was noted to be a predictor for increased severity in covid-19 illness. CoSAS also has statistically significant results for patients that carry HTN (0.004), DM (<0.001), IHD (0.002) and CKD (0.011), respectively. A limitation in the demographic variables of our study is weight-based categorization. Obesity has been well-documented as a variable that causes increased risk of severe covid related illness . This is likely due to the emergency department unable to document weight during high-risk patient resuscitative procedures associated with large volumes and diminished resources.
qSOFA, consisting of three clinical variables (mental status, respiratory rate, and blood pressure), has been proposed as a rapid screening tool for infected patients . Some studies have concluded that qSOFA score and severity of covid illness have a positive correlation . Whereas others have negated this notion stating a score that is based on altered mentation and circulatory collapse is not created to accurately predict mortality in a virus that leads to ARDS [26,27]. Our screening tool utilized a Prognostic Multivariable Modelling Design based on data readily available in the first 24 hours of hospitalization to predict in-hospital mortality of COVID-19 patients. It was proven with stringent data analysis that CoSAS has a superior prognostic accuracy to qSOFA (shown in table 7) as proven by the ROC of CoSAS vs qSOFA as 0.78 vs 0.76, respectively. This stands true as CoSAS incorporated much more variables providing a statistically superior result as compared to qSOFA. qSOFA risk stratification scoring accurately predicted an association with age, CFS, NLR and D-Dimer levels whereas both CoSAS and qSOFA were unable to accurately predict length of stay (0.0601 vs 0.985 at CI of 95). Similarly, our study also confirmed that advanced age, male gender, elevated levels of CRP, and previous comorbidities were predictive of in-hospital mortality as was stated in other analyses [23,28–30]. D-dimer levels obtained on admission accurately predicted mortality which was seen in the CoSAS and qSOFA models. Although CoSAS takes into account 10 factors of any patient on arrival, the study was unable to find a statistically significant relation with CoSAS score and CFS (0.08), Oxygen Saturation at presentation (0.446), systolic blood pressure (0.31), NLR (0.79) or Ferritin (0.49). qSOFA although only requiring 4 initial values showed a statistical significance with CFS (0.004), NLR (<0.001), Ferritin (0.001) and D-Dimer levels (<0.001)
CoSAS (high score ≥6) for predicting 28-day mortality included: age, gender, clinical frailty score, oxygen saturation, co morbidities, systolic blood pressure, NLR, CRP, DID and ferritin showed an AUC of 0.78 with a sensitivity of 0.93 and specificity of 0.51. CALL score (high risk >10) for predicting clinical progression of Covid-19 illness included: co-morbidities, age, lymphocyte count, and lactate dehydrogenase was shown to have an AUC of 0.91 with a sensitivity of 0.45 and specificity of 0.97 . NOCOS calculator (high risk: >51.6%) for predicting 7-day survival included: serum blood urea nitrogen, age, absolute neutrophil count, red cell distribution width, oxygen saturation, and serum sodium and was shown to have an AUC of 0.82 with a sensitivity of 0.89 and specificity of 0.54 . qCSI score (high risk >4) for respiratory failure within 24 hours included: respiratory rate, minimum recorded pulse oximetry, and nasal cannula flow rate requirement was found to have an AUC of 0.81 with a sensitivity and specificity of 0.79 . 4C mortality score (high risk >9) for predicting in-hospital mortality included: age, sex, number of comorbidities, respiratory rate, pulse oximetry on room air, Glasgow coma scale, serum urea, and C-reactive protein showed an AUC of 0.78 with a sensitivity of 0.93 and specificity of 0.41 .
There are certain limitations to our study that have been identified. As it was a single tertiary care center study in Karachi, Pakistan we were not able to demonstrate whether race or ethnicity affected outcomes. The study was limited to the emergency department and was thus unable to follow up with these patients nor was there any inclusive variable of whether any these patients required Non-Invasive Mechanical Ventilation (NIMV) or Mechanical Ventilation. The sample size was also much smaller than those in other studies. Due to incomplete files, there was also some missing data. Attempts to counteract this limitation included increasing our sample size beyond the minimum 146. Another limitation is that our patient data is from the pre-vaccination era. It is not known how a vaccinated population would score on CoSAS.