Summary: | In 2015, the newborn screening (NBS) programmes in England and Wales were expanded to include four additional disorders: Classical Homocystinuria, Isovaleric Acidemia, Glutaric Aciduria Type 1 and Maple Syrup Urine Disease, bringing the total number of analytes quantified to eight: phenylalanine, tyrosine, leucine, methionine, isovalerylcarnitine, glutarylcarnitine, octanoylcarnitine and decanoylcarnitine. Post-implementation, population data monitoring showed that inter-laboratory variation was greater than expected, with 90th centiles varying from 17 to 59%. We evaluated the effect of stable isotope internal standard (IS) used for quantitation on inter-laboratory variation. Four laboratories analysed routine screening samples (<i>n </i>><i> </i>101,820) using a common IS. Inter-laboratory variation was determined for the eight analytes and compared with results obtained using an in-house common IS (<i>n </i>><i> </i>102,194). A linear mixed-effects model was fitted to the data. Using a common IS mix reduced the inter-laboratory variation significantly (<i>p </i><<i> </i>0.05) for five analytes. For three analytes, the lack of significance was explained by use of individual laboratory “calibration factors”. For screening programmes where laboratories adhere to single analyte cut-off values (COVs), it is important that inter-laboratory variation is minimised, primarily to prevent false positive results. Whilst the use of a common IS helps achieve this, it is evident that instrument set-up also contributes to inter-laboratory variation.
|