Busadee PratumvinitLeslie LamNarisorn KongruttanachokChinnawut HemthongGerald J. KostPanumas KamkangKanit ReesukumalChulalongkorn UniversityKing Chulalongkorn Memorial Hospital, Faculty of Medicine Chulalongkorn UniversityUniversity of California, DavisFaculty of Medicine, Siriraj Hospital, Mahidol UniversityNg Teng Fong General Hospital2020-01-272020-01-272020-01-01Clinica Chimica Acta. Vol.500, (2020), 172-17918733492000989812-s2.0-85076614689https://repository.li.mahidol.ac.th/handle/20.500.14594/49561© 2019 Elsevier B.V. Background: Anion gap (AG) aids the differential diagnosis of acid-base disorders. Its value has decreased, because of new analytical methods. Our goal was to compare AG reference intervals for different instruments and Southeast Asian populations. Methods: We studied AG at three hospitals. One used the cobas 8000; two others, the Architect c16000. We included consecutive adults ≥18 years whose samples were sent for electrolytes and creatinine. We assessed AG for all patients and patients with normal electrolytes. Results: AG means differed significantly (P < 0.001) between the three hospitals for all patients and the normal electrolyte subgroup. AG reference intervals from all patients were 9–19, 5–15, and 5–15 mmol/L, and for the normal electrolyte subgroup, 10–17, 6–14, and 5–12 mmol/L, respectively. Compared to the normal albumin group, hypoalbuminemia patients showed lower AG in two hospitals (P < 0.001, P = 0.03), whereas patients with hyperalbuminemia demonstrated higher AG in all three hospitals (P < 0.001). Conclusions: Different instruments produce different AGs. There is a weak correlation between albumin levels and AG. Laboratorians should verify reference intervals used when detecting laboratory errors and assisting clinicians in the differential diagnosis of acid base disorders and other medical conditions.Mahidol UniversityBiochemistry, Genetics and Molecular BiologyMedicineArticleSCOPUS10.1016/j.cca.2019.10.012