INTRODUCTION: Schistocytes are major signs of micro- and macroangiopathic haemolytic anaemia. The aim was to evaluate automated fragmented red cell (FRC) count compared to visual microscopy, and to assess FRCs in the presence of microcytosis and hypochromia using Sysmex automated counters. METHODS: Schistocytes were determined with visual microscopy after the observation of 1000 erythrocytes, and the automated counting with Sysmex XE-5000. Indices of microcytosis (%MicroR) and hypochromia (%Hypo-He) were also analysed in the XE-5000 analyser. RESULTS: Linear regression analysis showed a good correlation between automated and manual FRC% count (r = 0.824, P < 0.0001), but Bland-Altman's plot revealed an overestimation of FRC of 0.82%. There is a global correlation between %MicroR and FRCs. In subgroup analysis of %MicroR (reference value: 0.3-3%, mild microcytosis: 3.1-7.2% and severe microcytosis: 7.3-56.7%), no correlations with automated %FRC were noticed (P > 0.05). Based on %Hypo-He subgroups (mild hypochromia: 1.2-5.2%, and severe hypochromia: 5.3-35.4%), a significant correlation of automated %FRC with mild hypochromia was found (r = 0.621, P < 0.01). CONCLUSION: Despite the agreement between FRC count and the manual method, the overvaluation of FRC was high, leading to false-positive results. Microcytosis appeared to have no impact on FRC count, whereas mild hypochromia seemed to be related with FRC count. Particular attention is required to assess automated FRCs in samples with mild hypochromia.
INTRODUCTION: Schistocytes are major signs of micro- and macroangiopathic haemolytic anaemia. The aim was to evaluate automated fragmented red cell (FRC) count compared to visual microscopy, and to assess FRCs in the presence of microcytosis and hypochromia using Sysmex automated counters. METHODS: Schistocytes were determined with visual microscopy after the observation of 1000 erythrocytes, and the automated counting with Sysmex XE-5000. Indices of microcytosis (%MicroR) and hypochromia (%Hypo-He) were also analysed in the XE-5000 analyser. RESULTS: Linear regression analysis showed a good correlation between automated and manual FRC% count (r = 0.824, P < 0.0001), but Bland-Altman's plot revealed an overestimation of FRC of 0.82%. There is a global correlation between %MicroR and FRCs. In subgroup analysis of %MicroR (reference value: 0.3-3%, mild microcytosis: 3.1-7.2% and severe microcytosis: 7.3-56.7%), no correlations with automated %FRC were noticed (P > 0.05). Based on %Hypo-He subgroups (mild hypochromia: 1.2-5.2%, and severe hypochromia: 5.3-35.4%), a significant correlation of automated %FRC with mild hypochromia was found (r = 0.621, P < 0.01). CONCLUSION: Despite the agreement between FRC count and the manual method, the overvaluation of FRC was high, leading to false-positive results. Microcytosis appeared to have no impact on FRC count, whereas mild hypochromia seemed to be related with FRC count. Particular attention is required to assess automated FRCs in samples with mild hypochromia.