Multicentre evaluation of two multiplex PCR platforms for the rapid microbiological investigation of nosocomial pneumonia in UK ICUs: the INHALE WP1 study

Background ICU patients with hospital-acquired or ventilator-associated pneumonia (HAP or VAP) have high mortality, so broad-spectrum antibiotics are initiated at clinical diagnosis, then refined after 2-3 days, once microbiology results become available. Unfortunately, culture-based microbiological investigation is also insensitive, with aetiological agents remaining unidentified in many cases. This leads to extended over-treatment of patients with susceptible pathogens, whilst those with highly-resistant pathogens are treated inadequately for prolonged periods. Using PCR to seek pathogens and their resistance genes directly from clinical samples may improve therapy and stewardship. The INHALE study compared two PCR platforms for HAP/VAP diagnosis against routine microbiology (RM), identifying one to progress into a Randomised Controlled Trial (RCT).

Methods Surplus routine sputa, endotracheal tube exudates and bronchoalveolar lavages were collected from ICU patients about to receive new or changed antibiotics for hospital-onset lower respiratory tract infections at 15 UK hospitals. Samples were tested (or frozen for testing) within 72h of collection. Testing was performed using the BioFire FilmArray Pneumonia Panel (bioMérieux) and Unyvero Pneumonia Panel (Curetis). Agreement between machine- and RM-results was categorised as ‘full positive/negative concordance’, ‘partial concordance’ or ‘major/minor discordance’. Bayesian latent class (BLC) analysis was used to estimate the sensitivity and specificity of each test, incorporating information from both PCR panels, 16S rDNA analysis and RM.

Findings In 652 eligible samples; PCR identified pathogens in considerably more samples compared with RM: 60.4% and 74.2% for Unyvero and FilmArray respectively vs. 44.2%. Both tests also recorded more organisms per sample than routine culture, with the two PCR tests frequently in agreement with each other. For common HAP/VAP pathogens, FilmArray had sensitivity of 91.7-100.0% and specificity of 87.5-99.5%; Unyvero had sensitivity of 83.3-100.0%% except for Klebsiella aerogenes (50.0%) and Serratia marcescens (77.8%), and specificity of 89.4-99.0%. BLC analysis indicated that, compared with PCR, RM had low sensitivity, ranging from 27.0% to 69.4% for common respiratory pathogens. PCR detected more high-consequence antimicrobial resistance genes than would have been predicted by RM and susceptibility testing; around half the host strains could be detected when culture was repeated and they were sought assiduously.

Interpretation Conventional and BLC analysis demonstrated that both platforms performed similarly and were considerably more sensitive than RM, detecting potential pathogens in patient samples reported as culture negative. FilmArray had slightly higher sensitivity than Unyvero for common pathogens and was chosen for INHALE’s RCT, based on the balance of these results, a swifter turnaround time (75 min vs. 6h), and a smaller footprint. The increased sensitivity of detection realised by PCR offers potential for improved antimicrobial prescribing.