Staphylococcus aureus antimicrobial susceptibility trends and cluster detection in Vermont: 2012-2018

ABSTRACT

Objectives: This study presents demographic and temporal trends in the isolation of Staphylococcus aureus in Vermont clinical microbiology laboratories and explores the use of statistical algorithms and multi-resistance phenotypes to improve outbreak detection.

Methods: Routine microbiology test results downloaded from Vermont clinical laboratory information systems were used to monitor S. aureus antimicrobial resistance trends. The integrated WHONET-SaTScan software used multi-resistance phenotypes to identify possible acute outbreaks with the space-time permutation model and slowly emerging geographic clusters using the spatial-only multinomial model.

Results: Data were provided from seven hospital laboratories from 2012 to 2018 for 19,224 S. aureus isolates from 14,939 patients. Statistically significant differences (p ≤ 0.05) in methicillin-resistant S. aureus (MRSA) isolation were seen by age group, specimen type, and health-care setting. Among MRSA, multi-resistance profiles permitted the recognition and tracking of 6 common and 21 rare ‘phenotypic clones.’ We identified 43 acute MRSA clusters and 7 significant geographic clusters (p ≤ 0.05).

Conclusions: There was significant heterogeneity in MRSA strains between facilities and the use of multi-resistance phenotypes facilitated the recognition of possible outbreaks. Comprehensive electronic surveillance of antimicrobial resistance utilizing routine clinical microbiology data with free software tools offers early recognition and tracking of emerging resistance threats.

KEYWORDS:

• Staphylococcus aureus
• Vermont
• antimicrobial resistance surveillance
• laboratory-based surveillance
• WHONET

Article highlights

• Temporal and demographic trends in antimicrobial susceptibility of S. aureus over a 7-year period from patient samples from the clinical microbiology laboratory of seven Vermont health-care facilities collected in the inpatient, outpatient, emergency, and long-term care setting are presented.

• While most susceptibility rates were relatively consistent across gender, age group, specimen type, and healthcare facility setting, levofloxacin susceptibility was much lower in patients aged 65 and over, in acute-care (non-critical access) hospitals, in long-term care facilities, and in urine samples. Susceptibility rates for three antimicrobials were much lower in long-term care facilities. The proportion of MRSA isolates steadily decreased over the time period. This decrease was noted both at the species level and in the more detailed analysis of the most common multi-resistance profiles defined by susceptibility to a set of priority antimicrobial agents.

• SaTScan’s space-time permutation model detected a number of statistically significant patient clusters (p≤0.05), both in the facility-level analysis and in the multi-facility statewide analysis. The positive predictive value of such clusters would need to explore with prospective investigation and validation.

• SaTScan’s spatial multinomial model detected a number of statistically significant geographic clusters (p≤0.05).

• The prospective application of cluster detection algorithms to the monitoring of multi-resistance profile should improve the prompt recognition and containment of novel and emerging threats, though this will need to be validated through prospective work.

Author contributions

J Stelling, T O’Brien, and J Read were responsible for the conceptualization and design of the Vermont collaborative project, as well as for approval of the final manuscript submitted for publication. Data collection, processing, and data analysis were led by R Peters and M Bokhari under the supervision of J Stelling using software developed by J Stelling and A Clark. J Stelling, T O’Brien, J Read, M Bokhari, and R Peters were responsible for the interpretation of analysis results and development of study conclusions. J Stelling led the preparation of the manuscript with support from all authors, but especially from M Bokhari. All authors were involved in reviewing, improving, and approving the manuscripts drafts.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This work was partially supported by funds from the Vermont Department of Health, the Centers for Disease Control and Prevention (CDC) and research project grants [R01GM103525, RR025040, U01CA207167] from the National Institutes of Health (NIH). Ethical approval was received from the Partners Healthcare Institutional Review Board. The contents are solely the responsibility of the authors and do not necessarily represent the official views of VDH, CDC, NIH, or WHO. The CDC, NIH, and WHO had no role in study design, data collection and analysis, or decision to publish, or preparation of the manuscript.