Laboratory characteristics associated with microbial identification quality: a national multicenter study based on...

Clinical microbiology laboratories are essential for infectious disease diagnosis and antimicrobial stewardship. However, data on laboratory capacity and quality gaps across tertiary hospitals in China remain limited. This multicenter cross-sectional study aimed to evaluate associations between laboratory characteristics and microbial identification performance, and to develop an exploratory prediction model for attainment of the program-defined mutual recognition threshold. Data were collected from 183 tertiary hospital microbiology laboratories across 31 provinces in China. Thirteen baseline laboratory characteristics, including institutional scale, staffing, equipment, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) availability, and ISO 15189 accreditation, were analyzed in relation to microbial identification performance assessed through a standardized national external quality assessment scheme. Correlation analysis, multivariable linear regression, K-means clustering, and logistic regression were used to evaluate associations, performance heterogeneity, and attainment of the program-defined mutual recognition threshold. MALDI-TOF MS availability showed the strongest and most consistent associations with identification performance and was positively associated in adjusted models with fastidious bacteria (β = 9.84, 95% CI 3.65-16.02; P = 0.002), rare bacteria (β = 7.23, 95% CI 4.03-10.43; P < 0.001), yeast (β = 7.16, 95% CI 1.66-12.66; P = 0.011), and total identification scores (β = 45.44, 95% CI 5.38-85.50; P = 0.026). Hospital bed capacity was associated only with yeast identification, while biosafety cabinet number was associated with Salmonella serotyping performance. Cluster analysis identified three performance tiers, supporting the presence of heterogeneity in microbial identification capability. The exploratory prediction model retained MALDI-TOF MS availability and conventional incubator number and showed moderate discrimination (AUC = 0.755), with a sensitivity of 92.9% and specificity of 50.0% at the optimal cutoff probability of 0.524. These findings indicate that technology availability and selected laboratory characteristics are associated with microbial identification performance and may help inform targeted laboratory quality improvement strategies.

Reliable microbial identification underpins infectious disease diagnosis, antimicrobial stewardship, and infection control. However, evidence directly linking routine laboratory resources to standardized identification performance remains limited. In this national multicenter study, standardized external quality assessment results were integrated with 13 baseline laboratory characteristics from 183 tertiary hospital microbiology laboratories across 31 provinces in China. matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) availability emerged as the laboratory characteristic most consistently associated with microbial identification performance, particularly for fastidious bacteria, rare bacteria, yeast, and overall identification scores. Exploratory clustering further demonstrated substantial heterogeneity in performance across laboratories, while a prediction model incorporating MALDI-TOF MS availability and conventional incubator number showed moderate discrimination for attainment of the mutual recognition threshold. These findings provide quantitative evidence that technology availability and broader laboratory capacity characteristics are associated with identification quality and may help guide targeted quality improvement and capacity-building efforts.