City Research Online

Abstract

BACKGROUND: Hand hygiene is monitored by direct observation to improve practice, but this approach can potentially cause information, selection, and confounding bias, threatening the validity of findings. The aim of this study was to identify and describe the potential biases in hand hygiene compliance monitoring by direct observation; develop a typology of biases and propose improvements to reduce bias; and increase the validity of compliance measurements.

METHODS: This systematic review of hospital-based intervention studies used direct observation to monitor health care workers' hand hygiene compliance.

RESULTS: Seventy-one publications were eligible for review. None was free of bias. Selection bias was present in all studies through lack of data collection on the weekends (n = 61, 86%) and at night (n = 46, 65%) and observations undertaken in single-specialty settings (n = 35, 49%). We observed inconsistency of terminology, definitions of hand hygiene opportunity, criteria, tools, and descriptions of the data collection. Frequency of observation, duration, or both were not described or were unclear in 58 (82%) publications. Observers were trained in 56 (79%) studies. Inter-rater reliability was measured in 26 (37%) studies.

CONCLUSIONS: Published research of hand hygiene compliance measured by direct observation lacks validity. Hand hygiene should be measured using methods that produce a valid indication of performance and quality. Standardization of methodology would expedite comparison of hand hygiene compliance between clinical settings and organizations.

Publication Type:	Article
Additional Information:	© 2018. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/
Publisher Keywords:	Bias, Hand hygiene, Hawthorne effect, Monitoring, Observation
Subjects:	R Medicine > RT Nursing
Departments:	School of Health & Psychological Sciences > Nursing
SWORD Depositor:	Symplectic Administrator

Export

Downloads