HLA haplotypes were found to be associated with increased risk for viral infections or disease severity in various diseases, including SARS. Several genetic variants are associated with COVID-19 severity. Studies have proposed associations, based on a very small sample and a large number of tested HLA alleles, but no clear association between HLA and COVID-19 incidence or severity has been reported. We conducted a large-scale HLA analysis of Israeli individuals who tested positive for SARS-CoV-2 infection by PCR. Overall, 72,912 individuals with known HLA haplotypes were included in the study, of whom 6413 (8.8%) were found to have SARS-CoV-2 by PCR. A total of 20,937 subjects were of Ashkenazi origin (at least 2/4 grandparents). One hundred eighty-one patients (2.8% of the infected) were hospitalized due to the disease. None of the 66 most common HLA loci (within the five HLA subgroups: A, B, C, DQB1, DRB1) was found to be associated with SARS-CoV-2 infection or hospitalization in the general Israeli population. Similarly, no association was detected in the Ashkenazi Jewish subset. Moreover, no association was found between heterozygosity in any of the HLA loci and either infection or hospitalization. We conclude that HLA haplotypes are not a major risk/protecting factor among the Israeli population for SARS-CoV-2 infection or severity. Our results suggest that if any HLA association exists with the disease it is very weak, and of limited effect on the pandemic.

HLA haplotypes were found to be associated with increased risk for viral infections or disease severity in various diseases, including SARS. Several genetic variants are associated with COVID-19 severity. Studies have proposed associations, based on a very small sample and a large number of tested HLA alleles, but no clear association between HLA and COVID-19 incidence or severity has been reported. We conducted a large-scale HLA analysis of Israeli individuals who tested positive for SARS-CoV-2 infection by PCR. Overall, 72,912 individuals with known HLA haplotypes were included in the study, of whom 6413 (8.8%) were found to have SARS-CoV-2 by PCR. A total of 20,937 subjects were of Ashkenazi origin (at least 2/4 grandparents). One hundred eighty-one patients (2.8% of the infected) were hospitalized due to the disease. None of the 66 most common HLA loci (within the five HLA subgroups: A, B, C, DQB1, DRB1) was found to be associated with SARS-CoV-2 infection or hospitalization in the general Israeli population. Similarly, no association was detected in the Ashkenazi Jewish subset. Moreover, no association was found between heterozygosity in any of the HLA loci and either infection or hospitalization. We conclude that HLA haplotypes are not a major risk/protecting factor among the Israeli population for SARS-CoV-2 infection or severity. Our results suggest that if any HLA association exists with the disease it is very weak, and of limited effect on the pandemic.

OBJECTIVE: To illustrate the problem of subpopulation miscalibration, to adapt an algorithm for recalibration of the predictions, and to validate its performance.

MATERIALS AND METHODS: In this retrospective cohort study, we evaluated the calibration of predictions based on the Pooled Cohort Equations (PCE) and the fracture risk assessment tool (FRAX) in the overall population and in subpopulations defined by the intersection of age, sex, ethnicity, socioeconomic status, and immigration history. We next applied the recalibration algorithm and assessed the change in calibration metrics, including calibration-in-the-large.

RESULTS: 1 021 041 patients were included in the PCE population, and 1 116 324 patients were included in the FRAX population. Baseline overall model calibration of the 2 tested models was good, but calibration in a substantial portion of the subpopulations was poor. After applying the algorithm, subpopulation calibration statistics were greatly improved, with the variance of the calibration-in-the-large values across all subpopulations reduced by 98.8% and 94.3% in the PCE and FRAX models, respectively.

DISCUSSION: Prediction models in medicine are increasingly common. Calibration, the agreement between predicted and observed risks, is commonly poor for subpopulations that were underrepresented in the development set of the models, resulting in bias and reduced performance for these subpopulations. In this work, we empirically evaluated an adapted version of the fairness algorithm designed by Hebert-Johnson et al. (2017) and demonstrated its use in improving subpopulation miscalibration.

CONCLUSION: A postprocessing and model-independent fairness algorithm for recalibration of predictive models greatly decreases the bias of subpopulation miscalibration and thus increases fairness and equality.

OBJECTIVE: To illustrate the problem of subpopulation miscalibration, to adapt an algorithm for recalibration of the predictions, and to validate its performance.

MATERIALS AND METHODS: In this retrospective cohort study, we evaluated the calibration of predictions based on the Pooled Cohort Equations (PCE) and the fracture risk assessment tool (FRAX) in the overall population and in subpopulations defined by the intersection of age, sex, ethnicity, socioeconomic status, and immigration history. We next applied the recalibration algorithm and assessed the change in calibration metrics, including calibration-in-the-large.

RESULTS: 1 021 041 patients were included in the PCE population, and 1 116 324 patients were included in the FRAX population. Baseline overall model calibration of the 2 tested models was good, but calibration in a substantial portion of the subpopulations was poor. After applying the algorithm, subpopulation calibration statistics were greatly improved, with the variance of the calibration-in-the-large values across all subpopulations reduced by 98.8% and 94.3% in the PCE and FRAX models, respectively.

DISCUSSION: Prediction models in medicine are increasingly common. Calibration, the agreement between predicted and observed risks, is commonly poor for subpopulations that were underrepresented in the development set of the models, resulting in bias and reduced performance for these subpopulations. In this work, we empirically evaluated an adapted version of the fairness algorithm designed by Hebert-Johnson et al. (2017) and demonstrated its use in improving subpopulation miscalibration.

CONCLUSION: A postprocessing and model-independent fairness algorithm for recalibration of predictive models greatly decreases the bias of subpopulation miscalibration and thus increases fairness and equality.

At the COVID-19 pandemic onset, when individual-level data of COVID-19 patients were not yet available, there was already a need for risk predictors to support prevention and treatment decisions. Here, we report a hybrid strategy to create such a predictor, combining the development of a baseline severe respiratory infection risk predictor and a post-processing method to calibrate the predictions to reported COVID-19 case-fatality rates. With the accumulation of a COVID-19 patient cohort, this predictor is validated to have good discrimination (area under the receiver-operating characteristics curve of 0.943) and calibration (markedly improved compared to that of the baseline predictor). At a 5% risk threshold, 15% of patients are marked as high-risk, achieving a sensitivity of 88%. We thus demonstrate that even at the onset of a pandemic, shrouded in epidemiologic fog of war, it is possible to provide a useful risk predictor, now widely used in a large healthcare organization.

At the COVID-19 pandemic onset, when individual-level data of COVID-19 patients were not yet available, there was already a need for risk predictors to support prevention and treatment decisions. Here, we report a hybrid strategy to create such a predictor, combining the development of a baseline severe respiratory infection risk predictor and a post-processing method to calibrate the predictions to reported COVID-19 case-fatality rates. With the accumulation of a COVID-19 patient cohort, this predictor is validated to have good discrimination (area under the receiver-operating characteristics curve of 0.943) and calibration (markedly improved compared to that of the baseline predictor). At a 5% risk threshold, 15% of patients are marked as high-risk, achieving a sensitivity of 88%. We thus demonstrate that even at the onset of a pandemic, shrouded in epidemiologic fog of war, it is possible to provide a useful risk predictor, now widely used in a large healthcare organization.