ABC2-SPH score offers simple, objective assessment of in-hospital mortality in COVID-19

An easy-to-use rapid scoring system—named ABC2-SPH—that combines seven characteristics of COVID-19 patients commonly available at hospital presentation helps estimate in-hospital mortality and performs better than other existing tools, as reported in a recent study.

Described as simple, objective, and easily available, the score is based on age, number of comorbidities, blood urea nitrogen, C-reactive protein (CRP), platelet count, peripheral oxygen saturation (SpO2/FiO2 ratio), and heart rate.

The ABC₂-SPH score uses “well defined and routinely recordable variables … [and] may be used at bedside for earlier identification of in-hospital mortality risk and, thus, inform clinical decisions and the assignment to the appropriate level of care and treatment for COVID-19 patients,” according to a team of researchers based in Brazil.

In previous studies, the variables included in the ABC₂-SPH score have been shown to be associated with severe COVID-19, the researchers added. “[O]ur results are in line with the available evidence.”

To develop the score, the team looked at 3,978 consecutive adult patients (model cohort; median age 60 years, in-hospital mortality rate 20.3 percent) admitted to the participating hospitals in Brazil. They tested 20 out of 36 potential predictor variables they initially identified.

A generalized additive model (GAM) showed that the combination of seven variables mentioned earlier was the best predictor of in-hospital mortality. As the ABC₂-SPH risk score, the sum of the prediction scores ranged between 0 and 20, with a high score indicating higher risk.

In the model cohort, the score categorized the risk of in-hospital mortality as very high (≥9 score) in 69.4 percent of patients, high (5–8 score) in 32.0 percent, intermediate (2–4 score) in 11.4 percent, and low (0–1 score) in 2.0 percent. The score showed good overall performance (Brier score: 0.114) and high discriminative value (area under the receiver operating characteristic curve [AUROC], 0.844). [Int J Infect 2021;doi:10.1016/j.ijid.2021.07.049]

When the score was externally validated in 1,054 Brazilian patients and 474 Spanish patients, results were similar in terms of ABC₂-SPH’s discriminative ability. The corresponding AUROCS for predicting in-hospital mortality were 0.859 and 0.894.

Furthermore, the ABC₂-SPH score achieved better discriminative ability than other prediction scoring systems for COVID-19, pneumonia, and sepsis, as well as better clinical utility relative to the three existing scores for in-hospital mortality (4C Mortality Score, A-DROP and CURB-65), the investigators noted. [Critical Care 2020;24:1-13; BMJ 2020;370:m3339; Sci Rep 2016;6:22911]

“A major strength of the ABC₂-SPH score is its simplicity,” they pointed out, adding that the use of objective parameters helps to reduce inter-user variability and that such parameters can be easily measured upon patient presentation at the emergency department even in under-resourced settings.

The researchers were positive that the score could help clinicians identify high-risk patients from the triage phase, single out those who may benefit from therapy, and potentially guide recommendations for early palliative consultation.

While ABC₂-SPH may hold potential generalizability for other countries, the researchers acknowledged that prediction models are population specific and may produce varied results in different populations. They recommended further validation and recalibration of the score, considering the potential differences in COVID-19 population characteristics, thresholds for admission, outcome events, and patient management.

The ABC₂-SPH risk score is a web-based calculator (https://abc2sph.com/) and freely available to the public.