Skip to main content

Evaluation of Fourier-Transform Infrared Spectroscopy (IR BiotyperⓇ) for Rapid Serotyping of Streptococcus pneumoniae: A Prospective Validation in a Public Health Reference Laboratory

In brief

Infrared spectroscopy matched pneumococcal serotypes in 69 of 70 isolates

In blinded testing of 70 routine clinical isolates, an artificial neural network paired with infrared spectroscopy matched the Quellung reference method in 69 cases. Nine in 10 isolates received a high-reliability classification, and an untrained serotype was flagged for confirmation. The results support further use in surveillance, but the flagged isolate underscores the need for reference testing when results are uncertain.

Journal
International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases (Q1)
Published
19 November 2026
Study design
Unclassified
Evidence level
Level 5, Expert Opinion (CEBM 5)
Authors
Christian de Alencar Siebra, Michele Luiza Rodrigues Mehl Seibert, Sabrina da Conceição Barbosa, Luiza Souza Rodrigues, Thaís Muniz Vasconcelos, Samanta Cristine Grassi Almeida, et al.
PMID
43057611
DOI
10.1016/j.ijid.2026.109266

Why clinicians should know about it

  • Picked for Microbiology (medical) (top studies of the week, 22 November 2026): Prospective validation of rapid pneumococcal serotyping method

Abstract

OBJECTIVE: To validate an expanded Artificial Neural Network (ANN) classifier for S. pneumoniae serotyping using FT-IR spectroscopy and assess its prospective performance in a routine public health reference laboratory. METHODS: A diagnostic accuracy study was conducted at the Central Public Health Laboratory of Paraná - Brazil, based on the acquisition of 1,603 spectra, corresponding to 27 target groups/serotypes for training models and 70 independent routine clinical isolates (2025-2026) for prospective blind validation. RESULTS: The ANN classifier achieved 100% cross-validation accuracy across 27 target serogroups/serotypes in the derivation dataset. In prospective blind testing (N = 70 clinical isolates), the FT-IR/ANN workflow demonstrated 98.6% overall serotype concordance (69/70) with Quellung reaction (reference method), correctly differentiating high-priority vaccine-escape strains (19A, 3, and 6C). High-reliability classification (≥ 80.0) was obtained for 90.0% of isolates, with a mean Isolate Score of 94.6% (median 100.0%, IQR: 98.2-100.0%). An untested non-target serotype (35F) was flagged by a low Isolate Score (34.2), serving as a preliminary internal quality control indicator requiring confirmation by reference methods. CONCLUSION: FT-IR spectroscopy combined with ANN provides a rapid, accurate, and scalable alternative to the Quellung reaction for pneumococcal serotyping, supporting its use in routine public health laboratory surveillance.

Abstract as published, via PubMed.

View on PubMedFull text at the publisherOpen in the app

For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.