Comparison of previous external validation studies by retrospective medical chart review [23]
| Reference number (year) |
Surveillance system | Target HAI of ICU | Number of patients | Target period | Selection of hospitals or ICUs | Selection of patients | Investigators | Survey method |
|---|---|---|---|---|---|---|---|---|
| [6] (1998) | NNIS (National Nosocomial Infections Surveillance, precursor to NHSN) | UTI, BSI, PNEU | 1,136 | January 1991-June 1993 | Nine hospitals that reported a sufficient number of infections to permit an adequate sample | A stratified random sampling: 15 BSI, 15 PNEU, 25 SSI, 15 UTI; 65 high- and 15 low-risk non-HAI patients at each hospital | Four-to-six-member teams out of a total of 32 infection control professionals | Each chart was reviewed independently by two reviewers. If there was a disagreement, the two reviewers reread the chart together to reach a consensus. Two epidemiologists confirmed or rejected the discrepant cases. |
| [7] (2007) | KISS (Krankenhaus Infektions Surveillance System) | BSI, LRTI | 1,481 | Before August 2000 | Random selection of 20 ICUs from among 77 KISS ICUs | Fifteen BSI or LRTI and 60 non-HAI patients at each ICU | Trained physicians | In cases of discrepancy, a supervising epidemiologist was consulted. |
| [13] (2010) | SPIN-UTI (Italian Nosocomial Infections Surveillance in Intensive Care Units project) | PNEU, BSI, UTI, CRI | 832 | November 2006-May 2007 | Random selection of 8 ICUs from 49 SPIN-UTI ICUs | All medical records of 75 HAI patients and 757 non-HAI patients | Trained physicians | In cases of discrepancy, a supervising epidemiologist was consulted. |
| [8] (2009) | VICNISS (Victorian Hospital Acquired Infection Surveillance System) | CLABSI | 108 | January 2006-December 2006 | Random selection of 6 hospitals from among 18 hospitals in VICNISS | Up to 10 results reported as positive and 10 results reported as negative (of under 100 cases of bacteremia) Up to 20 results reported as positive and 20 as negative (among more than 100 cases of bacteremia) |
Trained ICP, surveillance system investigator, infectious diseases physician, senior ICP | Discussion with investigators and consensus was reached. |
| [9] (2010) | NHSN (National Healthcare Safety Network), Connecticut | CLABSI | 410 | October 2008-December 2008 | All 30 hospitals in Connecticut to report CLABSI cases | ICU patients with positive blood culture results | Two reviewers (a nurse microbiologist for primary chart review and a hospital epidemiologist) | All identified and reported CLABSI cases were also reviewed by a hospital epidemiologist. Discrepant cases were discussed with hospital IP. |
| [10] (2009) | NHSN, Oregon | CLABSI | 817 | January 2009-December 2009 | All 44 hospitals in Oregon to report CLABSI cases | All 76 patients with reported CLABSI and a systematic sample of 741 other patients with ICU-related bacteremia (all patients in hospitals with no more than 60 ICU-related bacteremia episodes, and a random sample up to 44 patients in hospitals with more than 60 ICU-related bacteremia episodes) | Five reviewers (a physician, an infection prevention nurse, a public health nurse, an epidemiologist, a research analyst) | Reviews were conducted by two to four reviewers. Multiple reviews of the same record were not routinely performed. From 2-4 weeks after the review, discordant findings were reviewed during a single telephone conference with each hospital. |
| [11] (2013) | NHSN, Colorado | CLABSI | 519 | January 2010-March 2010 | Selection of 43 hospitals from among the total of 60 hospitals in Colorado | A facility-stratified random sample of ICU patients with positive blood culture | Two reviewers (a nurse and an epidemiologist) | Reviews were conducted independently, discrepant cases were discussed, and consultation was sought. |
| [12] (2013) | NHSN, New Mexico | CLABSI | 123 | November 2009-March 2010 | Six pilot hospitals in New Mexico | ICU patients with positive blood culture results | Reviewer 1: an IP. Reviewer 2: Four members (an IP, an epidemiologist, a medical epidemiologist, and a nurse epidemiologist) |
Reviews were conducted independently by a pair of reviewers (reviewers 1 and 2). Final case determinations for discordant cases were made by two medical epidemiologists in conjunction with hospital IP. |
Abbreviations: HAI, healthcare-associated infection; ICU, intensive care unit; UTI, urinary tract infection; BSI, bloodstream infection; PNEU, pneumonia; LRTI, lower respiratory tract infection; CRI, catheter-related bloodstream infection; CLABSI, central line-associated BSI; IP, infection preventionist; ICP, infection control practitioner.
Korean J healthc assoc Infect Control Prev 2023;28:192~202 https://doi.org/10.14192/kjicp.2023.28.2.192
© Korean J healthc assoc Infect Control Prev