Article
No correlation between alcoholic hand rub consumption and compliance in 73 German ICU's
Keine Korrelation zwischen Verbrauch an Händedesinfektionsmittel und Compliance in 73 Intensivstationen
Search Medline for
Authors
Published: | June 2, 2010 |
---|
Outline
Text
Background: There are two systems available to determine hand hygiene (HH) compliance: direct measurement by observing of HH opportunities (HHO) in clinical practice and measurement of alcoholic hand rub consumption (AHC) as a surrogate parameter. Hospitals participating in the German national HH campaign are measuring AHC on a mandatory basis, while observation is voluntary.
Objectives: We correlated the AHC with the HH compliance of all ICU's providing both data in 2008.
Methods: The following data are provided annually per unit: consumption of AHC, the number of annual patient days (PD) per unit and type of unit. The AHC in ml per PD for individual units is fed back to the hospitals together with reference data stratified by type of unit. The AHC in pl per PD is fed back to the hospitals together with reference data stratified by type of unit. All participants used defined observation tools and were trained by campaign team members. A minimum of 200 observations per unit and 20 observations per indication was defined. Spearman correlation coefficient was used in order to define the relationship between the two parameters. Significance was tested using Wilcoxon rank sum test.
Results: Both datasets were available for 73 ICU's. There was no correlation neither for all units (ρ=0.019) nor for special types of units (surgery ρ=–0.068, internal medicine ρ=0.061, interdisciplinary ρ=0.234, neonatology ρ=-0.131).
Conclusion: Based on our data, there is no correlation between AHC and compliance rates. Observations to determine compliance rates are very challenging. Satisfactory inter rater reliability is hard to achieve. Evaluation of observation results revealed an inter rater reliability between 30% and 60%. Observation is an excellent tool to determine quality, e. g. whether HHO's are appropriately detected in clinical practice. However, without extensive training and evaluation, quantitative interpretations of compliance rates should be done very carefully. Measurement of AHC provides a feasible and perhaps more reliable system to detect quantitative changes in HH behaviour, provided that there is a stable and unit based order system of AHC.
(Table 1 [Tab. 1])