Specimen acceptability criteria and handling of unacceptable sputum specimens.

The dominant criterion for acceptable sputum was > 25 WBCs/LPF and <10 SECs/LPF (98.88%).49.83% of laboratories performed acceptability screening for BALF. Among them, 55.29% used <1% SECs as the criterion, while 44.71% used the same criteria as for sputum.

For unacceptable sputum specimens containing S. pneumoniae, identification and antimicrobial susceptibility testing were performed in only 24.60% of cases (811 out of 3,297), regardless of the bacterial load.

S. pneumoniae identification methods.

The identification methods for S.pneumoniae are listed in Table 1. Approximately 50% of the laboratories did not routinely conduct the Optochin disc and/or bile solubility assays.Among the 623 laboratories utilizing MALDI-TOF MS platforms not manufactured by bioMérieux, 113 (18.14%) did not adhere to the supplementary identification procedures stipulated in the operating documentation.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Proportion of kit/card/method identification of S. pneumoniae.

https://doi.org/10.1371/journal.pone.0353484.t001

S. pneumoniae AST methods.

The methodologies employed for the antimicrobial susceptibility testing of Streptococcus pneumoniae are depicted in Table 2. As per the specifications of the automated susceptibility testing instruments and their corresponding panels, the bioMérieux GP68 card necessitates confirmatory testing at a Minimum Inhibitory Concentration (MIC) of 2 µg/mL. This requirement is absent for panels from other manufacturers, as their MIC ranges adequately encompass all established breakpoints.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Proportion of AST kit/mehod of S. pneumoniae.

https://doi.org/10.1371/journal.pone.0353484.t002

Quality control for AST.

A significant proportion (44.10%, 1454/3297) did not perform routine QC for S. pneumoniae AST. Of these 1454 laboratories, 41.4% were from tertiary hospitals and 58.6% from secondary hospitals and others. For laboratories that performed routine QC, 63.65% were from tertiary hospitals and 36.35% from secondary hospitals and others. Among those not performing QC, 90.72% (1319/1454) lacked the specific QC strain.Only 49.65% (1637/3297) of all laboratories reported using ATCC 49619, the recommended QC strain for S. pneumoniae AST.

MALDI-TOF MS usage and QC.

30.15% (994/3297) of laboratories used MALDI-TOF MS for S. pneumoniae identification.Among these, 84.51%(840/994) were from tertiary hospitals and 15.49%(154/994) from secondary hospitals.Common brands: bioMérieux (37.32%), Bruker (24.14%), Autobio (21.23%).QC practices varied widely, with 13.38% relying solely on manufacturer-engineered calibration/QC.

This national survey provides a comprehensive overview of current laboratory practices for respiratory specimen processing and S. pneumoniae workup in CARSS hospitals across China.Our findings indicate that sputum remains the most frequently submitted specimen for bacterial culture (39.92%),a pattern consistent with previous reports from China where sputum is easily obtained and non-invasive.However, the over-reliance on sputum raises concerns about contamination and overdiagnosis, as lower respiratory tract specimens are often colonized by oropharyngeal flora.Encouragingly, a majority of laboratories perform routine Gram staining on sputum (71.55%) and BALF (66.88%), which is vital for initial assessment and guiding culture workup [8,16,17]. The reported criteria for sputum acceptability(>25 WBCs/LPF and <10 SECs/LPF) align well with established guidelines [18], suggesting good awareness of quality assessmen tamong most laboratories. Nevertheless, nearly half of the laboratories (49.83%) perform acceptability screening for BALF, and among those, 44.71% use the same criteria as for sputum—a practice that may not be optimal given the lower expected squamous epithelial cell content in BALF. This highlights an area for further guideline harmonization.

The identification of S. pneumoniae relies on a combination of traditional methods (Optochin/Bile Solubility testing), biochemistry-based automated systems, and modern techniques such as MALDI-TOF MS. The increasing adoption of MALDI-TOF MS (30.15%) represents a positive trend, as it enables rapid and accurate identification [19]. However, the adoption rate varies considerably by hospital level: tertiary hospitals reported a notably higher usage rate (84.51%) compared to secondary hospitals (15.49%), reflecting resource disparities.However, variations in calibration and quality control practices for MALDI-TOF MS highlight an area requiring standardization.Specifically, 13.38% of MALDI-TOF MS users relied solely on manufacturer-engineered calibration/QC without performing independent quality checks, and 18.14% did not follow supplementary identification procedures required by the manufacturer’s documentation. Given that some MALDI-TOF MS platforms cannot reliably distinguish S. pneumoniae from other Streptococcus mitis group species without additional tests (e.g., optochin susceptibility), this gap poses a risk of misidentification. Additionally, for the MALDI-TOF MS platforms used, it is essential to strictly adhere to the manufacturer's guidelines regarding necessary supplementary tests.

A concerning finding is the high percentage of laboratories (44.10%) not performing routine QC for S. pneumoniae AST, primarily due to the lack of the specific QC strain ATCC 49619. Reliable AST results are fundamental for guiding treatment and surveillance [20]. The absence of routine QC compromises the accuracy and reliability of susceptibility data, potentially impacting patient careand national resistance monitoring efforts through CARSS. Our data further reveal that among laboratories not performing QC, 90.72% explicitly cited the lack of the QC strain as the reason, suggesting that supply chain and cost barriers are major obstacles. This underscores an urgent need to promote and facilitate the consistent use of appropriate QC strains. Tertiary hospitals (65.28% performed routine QC) performed better than non-tertiary hospitals (44.17%), but both levels were suboptimal. This underscores an urgent need to promote and facilitate the consistent use of appropriate QC strains. National and regional public health authorities should consider subsidizing the distribution of ATCC 49619 or establishing a centralized QC strain supply network for CARSS member hospitals.

The diversity of antimicrobial susceptibility testing (AST) methods—including automated systems, disk diffusion, and gradient diffusion tests—reflects variations in resources and preferences across laboratories. Notably, the Vitek GP68 card requires confirmatory testing at an MIC of 2 µg/mL, but our survey did not assess adherence to this requirement; this represents a knowledge gap for future studies.While this methodological diversity can be managed with appropriate standardization and quality control (QC), the widespread lack of QC in a significant proportion of laboratories using these methods remains a serious concern.In fact, the QC deficit was observed across all AST method categories, not limited to any single platform. This suggests that the problem is systemic rather than technology-specific. In 2025, the Clinical and Laboratory Standards Institute (CLSI) introduced the concept of the Individualized Quality Control Plan (IQCP) for antimicrobial susceptibility testing (AST) in its M100 document (35th edition) [21]. This risk‑based framework allows laboratories to customize QC frequency and procedures based on their own risk assessments, offering a flexible alternative to traditional daily QC.Given that many laboratories face practical difficulties in obtaining routine QC strains—such as limited sources, long procurement cycles, and high costs—promoting the 2025 CLSI‑endorsed IQCP framework as a flexible and resource‑efficient alternative could be a pragmatic solution for those struggling with conventional QC strain acquisition.This study has limitations. The data are self-reported, which might introduce reporting bias. The survey captures practices at a specific point in time and may not reflect ongoing changes.