described. 6 Using published cutoffs of 0.35 to rule out, and 0.67 to rule in, our main outcome was to validate these FAST score cutoffs in identifying NASH þ NAS 4 þ F2 in a cohort of US Veterans. Measures of diagnostic accuracy (ie, sensitivity, specificity, negative and positive predictive values [NPV and PPV], area under the receiver operating curve [C-statistic], and Youden index J) were assessed. The study was approved by the institutional review board. A 199 US Veterans (91% male), median (interquartile range) age of 61 (52-67) years, and body mass index of 32.8 (28.5-36.7) kg/m 2 were studied. We observed NASH in 57 (29%), NAS 4 in 68 (34%), and fibrosis stage 2 in 93 (47%). Combined NASH þ NAS 4 þ F2 was present in 28 (14.1%). The FAST score had a C-statistic of 0.75 (95% confidence interval, 0.69-0.81) and Youden index J of 0.44 (95% confidence interval, 0.31-0.54). The FAST score cutoff 0.35 to rule out 4 had a sensitivity and NPV of 1.00, and 0.67 to rule in 4 had a specificity of 0.69 with a PPV of 0.26 for NASH þ NAS 4 þ F2 (comparison with the published data in Table 1). The median FAST scores were similar for different quartiles of interval period from biopsy. For a fixed sensitivity of 0.90 in our cohort to rule out, the FAST score cutoff was 0.49, with a PPV of 0.24 and an NPV of 0.96. Similarly, for specificity of 0.90 to rule in, the FAST score cutoff was 0.85, with a PPV of 0.31 and an NPV of 0.87. To our knowledge, this is the first study to validate FAST score in stratifying high-risk NASH among the US Veterans. Our results are consistent with those published recently 4 despite an older, male-predominant population in a real-world clinical practice. Our sample size, when compared with the validation cohorts, 4 is the second largest study group. Although FAST score of 0.35 accurately excludes high-risk NASH, the rule in cutoff 0.67 had higher false negatives in our cohort for diagnosing high-risk NASH. Using newly generated cutoffs 0.49 to rule out and 0.85 to rule in, we identified 96% and 87% Veterans, respectively, who did not have high-risk NASH. It is possible that low PPV is likely caused by relatively low prevalence of NASH þ NAS 4 þ F2 in our cohort. In conclusion, the newly derived FAST score cutoffs are reliable noninvasive tests to classify subjects with low-and high-risk NASH. We propose to implement these cutoffs and anticipate to reduce the need for liver biopsies by 44% while maintaining a minimal falsepositive rate of 1.5% in evaluating US Veterans with NASH.