Current best evidence for clinical care (more info)
BACKGROUND: The 2019 novel coronavirus disease (COVID-19) has created unprecedented medical challenges. There remains a need for validated risk prediction models to assess short-term mortality risk among hospitalized patients with COVID-19. The objective of this study was to develop and validate a 7-day and 14-day mortality risk prediction model for patients hospitalized with COVID-19.
METHODS: We performed a multicenter retrospective cohort study with a separate multicenter cohort for external validation using two hospitals in New York, NY, and 9 hospitals in Massachusetts, respectively. A total of 664 patients in NY and 265 patients with COVID-19 in Massachusetts, hospitalized from March to April 2020.
RESULTS: We developed a risk model consisting of patient age, hypoxia severity, mean arterial pressure and presence of kidney dysfunction at hospital presentation. Multivariable regression model was based on risk factors selected from univariable and Chi-squared automatic interaction detection analyses. Validation was by receiver operating characteristic curve (discrimination) and Hosmer-Lemeshow goodness of fit (GOF) test (calibration). In internal cross-validation, prediction of 7-day mortality had an AUC of 0.86 (95%CI 0.74-0.98; GOF p = 0.744); while 14-day had an AUC of 0.83 (95%CI 0.69-0.97; GOF p = 0.588). External validation was achieved using 265 patients from an outside cohort and confirmed 7- and 14-day mortality prediction performance with an AUC of 0.85 (95%CI 0.78-0.92; GOF p = 0.340) and 0.83 (95%CI 0.76-0.89; GOF p = 0.471) respectively, along with excellent calibration. Retrospective data collection, short follow-up time, and development in COVID-19 epicenter may limit model generalizability.
CONCLUSIONS: The COVID-AID risk tool is a well-calibrated model that demonstrates accuracy in the prediction of both 7-day and 14-day mortality risk among patients hospitalized with COVID-19. This prediction score could assist with resource utilization, patient and caregiver education, and provide a risk stratification instrument for future research trials.
| Discipline / Specialty Area | Score |
|---|---|
| Infectious Disease | |
| Hospital Doctor/Hospitalists | |
| Internal Medicine | |
| Respirology/Pulmonology | |
| Intensivist/Critical Care | |
Promising initial data that still need to be validated in a larger sample.
Better papers and models available. For example, BMJ 2020;371:m3731, http://dx.doi.org/10.1136/bmj.m3731.
This article attempts to develop a prognostication tool for assessing who might be at increased risk for mortality due to COVD-19. This is useful for resource allocation and planning. It may also be helpful in states where a surge in infection stresses resources and requires resource allocation to be triaged to those who will benefit the most.
Great to have a risk assessment tool for COVID, but further work needed here before this is ready for `prime time.` There are other studies that are further along (e.g., recently published in BMJ from the UK).
Simple useful tool that will probably need further validation.
This is probably useful; although, mortality in recent cohorts may be better.
They did not do a good job in helping clinicians.
This study provides data to help predict outcomes due to Covid-19. The variables that predict mortality are not novel or surprising: age, comorbidities and markers of severe illness at time of presentation such as oxygen saturation. For me, the main value of this information is that it allows quantification of outcome prediction, as opposed to saying someone is at "high risk" or "low risk".
This study of prognosis finds that in patients hospitalized with COVID-19, a 4-variable model (age, mean arterial pressure, kidney dysfunction [creatinine at least 2x upper limit of normal], and severe hypoxia [needing more than 4 L/min]) predicted 7-day and 14-day mortality with moderate accuracy. There have now been several published attempts at such models. All studies struggle with limited sample size, which impairs validity of modeling approaches. This study addresses this problem with a less traditional approach (Chi-square automatic interaction detection) that may have been an advantage. The variables included in the model are all easily and rapidly obtainable. However, the model did not include other variables that other studies have selected (CRP, D-dimer, lymphocyte count, ferritin) that represent inflammation and hypercoagulability parameters different from the current model. No explanation was given. Forcing these in the model to test them would have added validity.
Should be widely distributed and used to allow additional validation.