Have you ever read an IPC report that says something along the lines of “we did lots of observational hand hygiene compliance auditing, and compliance was 99%”? Well, we know that back in the real world hand hygiene compliance is typically 40%. Whilst we can do better with observational hand hygiene compliance audits, and make them a realistic starting point for improvement, there’s a real opportunity for automated hand hygiene monitoring systems to come in and shake up our understanding of hand hygiene compliance in a really positive way. But are the systems up to it? And are we ready for them? Today’s study in Infection Prevention in Practice from Denmark shows that an automated hand hygiene system can be implemented and generate some really useful hand hygiene surveillance information.

Why I chose this study

I’ve been following the concepts and evidence around automated hand hygiene closely over the years, because I think we desperately need a better approach to measuring hand hygiene! Observational audits are necessary (in my view – although this view is not universally held) but prone to several types of bias. Observation bias (aka the Hawthrone effect), observer bias, and selection bias, to name but a few! Automated systems offer the potential to slice through that rather complicated set of biases and give us some really objective information on hand hygiene compliance!

This study had a focus on using an automated system to not only measure hand hygiene, but also to improve it. There’s a lot of data around what works to improve hand hygiene (including a Cochrane review). My take on this is that pretty much any intervention will improve hand hygiene – but how to improve hand hygiene in a way that is sustained beyond the excitement of the intervention is the holy grail!

Design and methods

This 11-month prospective, interventional study was conducted between February 2021 and December 2021 at a large Danish hospital. Four inpatient wards in haematology / oncology were included in the study, with a total of 64 inpatient beds and 187 staff (doctors, nurses, and cleaners). The automated hand hygiene system used was Saninudge, which includes sensors on the alcohol gel dispensers, bed (as a proxy for the patient), and staff members badge. The system triangulated this data to give proxy compliance data for WHO moment 1 (before patient contact), 4 (after touching the patient), and 5 (after touching the patient surroundings). Following an pre-intervention period (period 1), group feedback was provided to staff (period 2), followed by both group and individual feedback (period 3), followed by a post intervention period (period 4).

Key findings

• A whopping 231,022 hand hygiene observations were recorded during the study. (Try getting that many observations of hand hygiene on four wards over 11 months using observational audits!)
• Overall compliance with hand hygiene was around the 30% mark during patient care, which has a ring of truth when compared with properly conducted observational audits of hand hygiene.
• Hand hygiene amongst nursing staff was generally higher than amongst doctors, which again has a ring of truth.
• Curiously, hand hygiene compliance was lowest in patient rooms, and highest in medication rooms.
• Whilst there were some interesting trends in various staff groups and settings, overall, the interventions (either group or individual feedback) made little impact in terms of hand hygiene compliance.

Points for discussion

• The authors attribute the lack of impact of the intervention to the limited scope of the intervention itself, which involved only feedback and no education or related activities.
• Data collection was undertaken during the COVID-19 pandemic – there was an increase in compliance identified in the followup period, which coincided with an increase in COVID-19 prevalence locally.

Strengths and limitations

• The study was undertaken in a large acute healthcare system, in four different wards.
• Impact in a multi-professional workforce was investigated.
• The system was evaluated in haematology / oncology, so findings may be less relevant in other specialities.
• The validity of the compliance measurements were not “validated” with observational audits of hand hygiene (although you could argue they are non-comparable in some ways).
• Some staff were unable to make the group feedback briefings due to differing shift patterns. Related to this, weekly feedback was not always achieved due to operational pressures. Also, some staff did not open emails that contained their individual feedback. Therefore, some staff had more exposure to group feedback than others.

What this means for IPC

• Based on the findings of this study, automated hand hygiene monitoring provides a viable alternative to observational audits of hand hygiene.
• Simple feedback of hand hygiene compliance data alone seems to make little impact in improving hand hygiene practice.

Finally, there will be an IPC Partners Journal Club on this article tomorrow (register here).