Automated detection of infectious disease outbreaks in hospitals: a retrospective cohort study.

Division of Infectious Diseases and Health Policy Research Institute, University of California Irvine School of Medicine, Irvine, California, United States of America.
PLoS Medicine (Impact Factor: 14). 02/2010; 7(2):e1000238. DOI: 10.1371/journal.pmed.1000238
Source: PubMed

ABSTRACT Detection of outbreaks of hospital-acquired infections is often based on simple rules, such as the occurrence of three new cases of a single pathogen in two weeks on the same ward. These rules typically focus on only a few pathogens, and they do not account for the pathogens' underlying prevalence, the normal random variation in rates, and clusters that may occur beyond a single ward, such as those associated with specialty services. Ideally, outbreak detection programs should evaluate many pathogens, using a wide array of data sources.
We applied a space-time permutation scan statistic to microbiology data from patients admitted to a 750-bed academic medical center in 2002-2006, using WHONET-SaTScan laboratory information software from the World Health Organization (WHO) Collaborating Centre for Surveillance of Antimicrobial Resistance. We evaluated patients' first isolates for each potential pathogenic species. In order to evaluate hospital-associated infections, only pathogens first isolated >2 d after admission were included. Clusters were sought daily across the entire hospital, as well as in hospital wards, specialty services, and using similar antimicrobial susceptibility profiles. We assessed clusters that had a likelihood of occurring by chance less than once per year. For methicillin-resistant Staphylococcus aureus (MRSA) or vancomycin-resistant enterococci (VRE), WHONET-SaTScan-generated clusters were compared to those previously identified by the Infection Control program, which were based on a rule-based criterion of three occurrences in two weeks in the same ward. Two hospital epidemiologists independently classified each cluster's importance. From 2002 to 2006, WHONET-SaTScan found 59 clusters involving 2-27 patients (median 4). Clusters were identified by antimicrobial resistance profile (41%), wards (29%), service (13%), and hospital-wide assessments (17%). WHONET-SaTScan rapidly detected the two previously known gram-negative pathogen clusters. Compared to rule-based thresholds, WHONET-SaTScan considered only one of 73 previously designated MRSA clusters and 0 of 87 VRE clusters as episodes statistically unlikely to have occurred by chance. WHONET-SaTScan identified six MRSA and four VRE clusters that were previously unknown. Epidemiologists considered more than 95% of the 59 detected clusters to merit consideration, with 27% warranting active investigation or intervention.
Automated statistical software identified hospital clusters that had escaped routine detection. It also classified many previously identified clusters as events likely to occur because of normal random fluctuations. This automated method has the potential to provide valuable real-time guidance both by identifying otherwise unrecognized outbreaks and by preventing the unnecessary implementation of resource-intensive infection control measures that interfere with regular patient care. Please see later in the article for the Editors' Summary.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Most influenza surveillance is based on data from urban sentinel hospitals; little is known about influenza activity in rural communities. We conducted influenza surveillance in a rural region of China with the aim of detecting influenza activity in the 2009/2010 influenza season. The study was conducted from October 2009 to March 2010. Real-time polymerase chain reaction was used to confirm influenza cases. Over-the-counter (OTC) drug sales were daily collected in drugstores and hospitals/clinics. Space-time scan statistics were used to identify clusters of ILI in community. The incidence rate of ILI/influenza was estimated on the basis of the number of ILI/influenza cases detected by the hospitals/clinics. A total of 434 ILI cases (3.88% of all consultations) were reported; 64.71% of these cases were influenza A (H1N1) pdm09. The estimated incidence rate of ILI and influenza were 5.19/100 and 0.40/100, respectively. The numbers of ILI cases and OTC drug purchases in the previous 7 days were strongly correlated (Spearman rank correlation coefficient [r] = 0.620, P = 0.001). Four ILI outbreaks were detected by space-time permutation analysis. This rural community surveillance detected influenza A (H1N1) pdm09 activity and outbreaks in the 2009/2010 influenza season and enabled estimation of the incidence rate of influenza. It also provides a scientific data for public health measures.
    PLoS ONE 12/2014; 9(12):e115347. DOI:10.1371/journal.pone.0115347 · 3.53 Impact Factor
  • Biomédica: revista del Instituto Nacional de Salud 08/2013; 34:9. DOI:10.7705/biomedica.v34i0.2205 · 0.62 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Objective To evaluate the performance of China's infectious disease automated alert and response system in the detection of outbreaks of hand, foot and mouth (HEM) disease. Methods We estimated size, duration and delay in reporting HEM disease outbreaks from cases notified between 1 May 2008 and 30 April 2010 and between 1 May 2010 and 30 April 2012, before and after automatic alert and response included HFM disease. Sensitivity, specificity and timeliness of detection of aberrations in the incidence of HEM disease outbreaks were estimated by comparing automated detections to observations of public health staff. Findings The alert and response system recorded 106 005 aberrations in the incidence of HFM disease between 1 May 2010 and 30 April 2012 a mean of 5.6 aberrations per 100 days in each county that reported HEM disease. The response system had a sensitivity of 92.7% and a specificity of 95.0%. The mean delay between the reporting of the first case of an outbreak and detection of that outbreak by the response system was 2.1 days. Between the first and second study periods, the mean size of an HEM disease outbreak decreased from 19.4 to 15.8 cases and the mean interval between the onset and initial reporting of such an outbreak to the public health emergency reporting system decreased from 10.0 to 9.1 days. Conclusion The automated alert and response system shows good sensitivity in the detection of HEM disease outbreaks and appears to be relatively rapid. Continued use of this system should allow more effective prevention and limitation of such outbreaks in China.
    Bulletin of the World Health Organisation 09/2014; 92(9):656-63. DOI:10.2471/BLT.13.130666 · 5.11 Impact Factor

Full-text (2 Sources)

Available from
May 27, 2014