To the Editor—Hand hygiene compliance observation is an established quality indicator; however, current observation techniques only count correct indications according to the World Health Organization recommendation without assessing the quality of the hand disinfection performed. This study was designed to test the hypothesis that infection control staff are capable of correct classification of observed hand disinfections using subjective parameters suitable for clinical routine use rather than objective measurable parameters.

We studied 2 groups of observers; each group consisted of infection control practitioners and consultants in hospital epidemiology and infection control with >3 years of job experience. Group 1 observed 5 hand disinfections live (in person) and group 2 observed 5 hand disinfections via video link. Without technical aids (eg, a stop watch), all participants were asked to classify the hand disinfection as correct or incorrect considering time and skin coverage.

Test persons demonstrating hand hygiene were asked to perform hand disinfection either correctly or to make mistakes at their discretion. An independent observer measured the duration of the disinfection procedure, and 3 different observers estimated the skin coverage under black light by the fluorescent marker added to the disinfection solution. The test disinfection was classified as correct if >90% skin coverage of the hand was reached and at least 15 seconds passed after skin coverage (per the manufacturer’s instructions).

Table 1 shows the results of 81 observations. In group 1 (live observation), 97.5% of subjective observations were correct compared to 78.8% in group 2 (video observation). All incorrect disinfections were classified as such, resulting in a negative predictive value of subjective assessment of 100%. The positive predictive value for correct hand disinfections was only 92%. Thus, video observation is not a good substitute for live observation, likely because the fixed camera angle and artefacts imposed by light and shadow make the assessment of skin coverage difficult.

TABLE 1 Classifications of Observed Hand Disinfection Tests and Group-Specific ResultsFootnote ^a

^a Total observations, n=81.

The addition of an easy-to-use qualitative component to hand hygiene compliance observations and consecutive training efforts is important, given that <10% of all hand disinfections were performed correctly in an observational study by Tschudin-Sutter et al,Reference Tschudin-Sutter, Sepulcri, Dangel, Schuhmacher and Widmer ¹ who observed the 6-step technique. Appropriate hand-surface coverage was reached in only 7.9% of hand hygiene procedures observed by Park et al,Reference Park, Kim and Lim ² despite a high rate of compliance with the correct indications. Shah et alReference Shah, Patel, Shah, Phatak and Nimbalkar ³ performed a video observation of hand washing. Of 1,081 recordings, 403 (37.3%) were excellent, 521 (48.2%) were acceptable, and 157 (14.5%) were unacceptable.

A limitation of our study is the lack of bacterial counts, but the results of Riley et al,Reference Reily, Pice and Lang ⁴ who showed no correlation between hand coverage and bacterial counts with a 6-step technique compared to a 3-step approach, had not been published at the time of our experiment.Reference Reily, Pice and Lang ⁴ Another limitation is the small number of participants and the experimental setting of this proof-of-principle study. However, we believe that based on our results, the addition of dichotomous subjective quality assessment using the parameters time and skin coverage during live observation by experienced infection control staff is feasible and could be a valuable addition to conventional hand hygiene observation.