Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
We read the article by Harrison et al1 with great interest. The authors proposed an easily calculated composite score for predicting the risk of advanced fibrosis in patients with non-alcoholic fatty liver disease (NAFLD), called the BARD score: the weighted sum of the three variables (body mass index (BMI) ⩾28 = 1 point, aspartate aminotransferase/alanine aminotransferase ratio (AAR) ⩾0.8 = 2 points, diabetes = 1 point). When a BARD score of 2–4 was used, the area under the receiver operating characteristic curve (AUROC) was found to be 0.81 with an odds ratio (OR) of 17 (95% CI 9.2 to 31.9) for detecting advanced fibrosis. The positive predictive value (PPV) and negative predictive value (NPV) were 43% and 96%, respectively. We studied the reliability of the BARD score for identifying the risk of advanced fibrosis in Japanese patients with NAFLD.
A total of 122 patients (61% female; median age, 59 years) with NAFLD who underwent liver biopsy at our hospital were studied. Median BMI was 26 kg/m2 (range, 18–45); 33% of patients had a normal BMI of <25, whereas 46% were overweight (BMI 25–29), 21% were obese (BMI ⩾30) and 1% were extremely obese (BMI ⩾40). Common co-morbidities included hypertension (30%), diabetes (37%) and hyperlipidaemia (59%). Liver biopsy showed that 9 (7%) had simple steatosis, and 113 (93%) had non-alcoholic steatohepatitis, including 67 (55%) with mild fibrosis (stage 0–2) and 46 (38%) with advanced fibrosis (stage 3–4). When a BARD score of 2–4 was used, the AUROC was 0.73 (fig 1) with an OR of 4.9 (95% CI 2.2 to 10.8) for detecting advanced fibrosis. The PPV and NPV were 59% and 77%, respectively.
The BARD score was less predictive of advanced fibrosis in Japanese patients than in the study by Harrison et al, probably because of two major reasons. First, their subjects were predominantly Caucasians; only 2% were Asian Pacific Islanders. Although mean BMI is lower in Asian populations than in non-Asian populations, Asians have a higher percentage of body fat for a given BMI than non-Asians.2 In a study conducted in Japan,3 nearly half of the subjects with NAFLD were not overweight or obese, suggesting that different genetic and environmental factors are related to susceptibility to hepatic steatosis in the Japanese population. In our cohort, BMI was similar in patients with and without advanced fibrosis, as shown in table 1. Our results are consistent with those of Hashimoto et al,4 who found that older age, the presence of diabetes and elevated AAR were significantly associated with more advanced fibrosis in Japanese patients, whereas higher BMI was not. Secondly, Harrison et al assessed only AAR, glycated haemoglobin (HbA1c) and the quantitative assessment check index score among laboratory variables as potential risk factors for advanced fibrosis. Since decreased platelet count and decreased albumin concentration were significantly associated with more advanced fibrosis in our cohort (table 1), we assessed the value of the NAFLD fibrosis score,5 which includes these two variables, for the detection of advanced fibrosis. The AUROC for the NAFLD fibrosis score was 0.84, with a PPV and NPV of 59% and 89%, respectively (data not shown). Albeit slightly more complex, the NAFLD fibrosis score more accurately detected advanced fibrosis in our patients than the BARD score.
In summary, the BARD score can be easily derived from readily available clinical data, but may be less reliable for excluding the presence of advanced fibrosis in Japanese patients with NAFLD than in the study by Harrison et al.
Competing interests None.
Provenance and Peer review Not commissioned; not externally peer reviewed.