Establishing the most appropriate statistical analysis for patient safety data

Download files
Access & Terms of Use
open access
Copyright: Azim, Syed
Altmetric
Abstract
Background Healthcare associated infections (HAIs) are common adverse events that are often preventable and life threatening occurring in up to 10 out of 100 hospitalised patients in Australia. Therefore, hand hygiene (HH) in healthcare workers is aimed at reducing HAI, specifically methicillin resistant Staphylococcus aureus (MRSA) infections. Antibiotic resistance is a predictor of HAIs and improvement in hand hygiene compliance aims to reduce HAIs. Australia lacks a robust system for testing the reliability of the validity of the level of antibiotic resistance and the reliability of hand hygiene compliance rates. This thesis tests both and examines the hand hygiene compliance data for reliability. Methods This research is a collection of published and submitted, but yet un-published, peer-reviewed journal articles. Collectively, the work describes whether: (1) the current sampling method, usually over 1 month, used by the Australian Group on Antimicrobial Resistance (AGAR) to establish antimicrobial resistance patterns provides reliable estimates of resistance for five commonly used antibiotics (2) hand hygiene rates reported by Hand Hygiene Australia (HHA) for medical and nursing staff are reliable and the association between HH rates and Staphylcoccus aureus bloodstream infection (SABSI) are valid (3) the burden of HH for nurses and medical staffs impacts on HH compliance and (4) an automated HH monitoring system is a valid measure of compliance. Findings The current MRSA sampling methodology underestimated antibiotic resistance in outpatients with 45% of resistance due to under sampling of highly resistant phenotypes. Nurses’ HH is performed above the national threshold regardless of hospital size and artificially inflates the average compliance of a hospital and its medical staff. HH is still not performed at a sufficiently high level to impact the rate of SABSI. The burden of HH for nurses was three times higher than medical staff and their average weekly compliance was 1.5 times higher than medical staff. The automated system that measures HH compliance is an improved alternative to human auditing which has poor reliability and validity Conclusion The minimum sampling of MRSA should continue for at least 6 months to accommodate the impact that infrequent MRSA phenotypes have on resistance patterns. The number of HH opportunities required of medical staff is not burdensome. Automated auditing provides rapid daily feedback for medical staff that may positively impact on MRSA and HH compliance.
Persistent link to this record
Link to Publisher Version
Link to Open Access Version
Additional Link
Author(s)
Azim, Syed
Supervisor(s)
McLaws, Mary-Louise
Rahman, Bayzidur
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2016
Resource Type
Thesis
Degree Type
PhD Doctorate
UNSW Faculty
Files
download public version.pdf 1.52 MB Adobe Portable Document Format
Related dataset(s)