Comparison of human and electronic observation for the measurement of compliance with hand hygiene.

BACKGROUND: Monitoring of hand hygiene is an important part of the improvement of hospital quality indicators. METHODS: This study was prospectively performed over a 14-week (electronic observer) period from December 3, 2013-March 9, 2014, to evaluate hand hygiene compliance in an adult step-down unit. We compared electronic handwash counters with the application of radiofrequency identification (RFID - ZigBee; i-Healthsys, São Carlos, Brazil) (electronic observer), which counts each activation of the alcohol gel dispenser to direct observation (human observer) using the iScrub application. RESULTS: For the overall time period of simultaneous electronic and human observation, we found that the electronic observer identified 414 hand hygiene episodes, whereas the human observers identified 448 episodes. Therefore, we found 92% (95% confidence interval [CI], 90%-95%) overall concordance (414/448), with an intraclass correlation coefficient of .87 (95% CI, 0.77-0.92). CONCLUSION: Our RFID (ZigBee) system showed good accuracy (92%) and is a useful method to monitor hand hygiene compliance.

Positive deviance: a program for sustained improvement in hand hygiene compliance.

BACKGROUND: There is a paucity of data evaluating whether positive deviance (PD) can sustain improvement in hand hygiene compliance. METHODS: An observational study comparing the effect of PD on compliance with hand hygiene was conducted in two 20-bed step-down units (SDUs) at a private tertiary care hospital. In a 3-month baseline period (April-June 2008), hand hygiene counts were performed by electronic handwashing counters. Between July 1, 2008, and November 30, 2009, (East SDU) and between September 30, 2008, and December 2009 (West SDU), PD was applied in both units. RESULTS: There was more than a 2-fold difference in the number of alcohol gel aliquots dispensed per month from April 2008 (before PD) to November 2009 (last month in PD) in the East SDU. There was also a 2-fold difference in the number of alcohol gel aliquots dispensed per month from September 2008 (before PD) to December 2009 (last month in PD) in the West SDU. The difference in the rate of health care‒associated infections (HAIs) between the baseline period and 2009 was statistically significant in the East SDU (5.8 vs 2.8 per 1,000 device-days; P = .008) and in the West SDU (3.7 vs 1.7 per 1,000 device-days; P = .023). CONCLUSIONS: PD was responsible for a sustained improvement in hand hygiene in the inpatient setting and was associated with a decrease in the incidence of device-associated HAIs.