Refining peripheral blood smear review rules for neonatal inpatients in a South African academic laboratory

Abstract

Peripheral blood smear review (PBSR) is a labour-intensive test, and skilled morphologists are in short supply. It is therefore helpful for laboratories to establish rules for PBSR to improve laboratory efficiency. Previously published guidelines in this regard are useful, but make few recommendations specific to neonates. Neonatal blood is characterized by several peculiarities which would be considered pathological if present in adults. Consequently, smear review rules (SRR) are often triggered in neonates without significant value being added on review. This study aimed to assess and fine-tune the SRR triggered in neonatal samples in order to improve laboratory efficiency. Full blood counts collected from 188 neonatal inpatients of the Chris Hani Baragwanath Academic Hospital in South Africa were retrospectively reviewed, the triggered rules documented, and the value added on PBSR determined. Smear review rules were triggered in 148 (78.7%) samples, with significant morphological abnormalities identified in 84 (54.4%), and a false-positive rate of 34.0%. In patients with unhelpful review, the commonest rules triggered were the flags querying the presence of abnormal lymphocytes, blasts or nucleated red blood cells. When one or more of these flags were triggered in the absence of any other SRR, PBSR was always noncontributory. Disregarding these flags in the current cohort would reduce both the review and the false-positive rates by >20% without increasing the false-negative rate. False-positive smear review is common in neonates, and minor modifications to SRR can substantially reduce the smear review rate without increasing the false-negative rate.