Do wearable alcohol-based handrub dispensers increase hand hygiene compliance? - a mixed-methods study

The study was conducted between April and June 2016 at the ED of the UHZ, Switzerland. The UHZ is a 950-bed tertiary hospital covering all medical specialties except pediatrics and orthopedics. The ED holds 17 beds, and a mean of 44′000 patients are admitted to the ED yearly. The ED staff consists of 100 nurses, 13 staff emergency physicians, 25 medical interns, and various other professions (e.g. maintenance and nursing assistants). Non-ED consulting physicians and surgeons regularly visiting the ED. The study population for HH observation and interviews consisted of ED staff only, ABHR consumption was measured overall, thus including also HCP visiting the ER. The ED has an open floor plan with a central nursing and working station, and single patient bays separated by privacy curtains. Mounted dispensers are available in or immediately outside every patient room/bay. In total, there are 34 wall-mounted or table-top dispensers available in close proximity, i.e. below 2 m, of each patient bed. Additional dispensers exist in the staff lounge, storage rooms, and restrooms.

This mixed-method quasi-experimental before-after study evaluated the effect of introducing wearable ABHR dispensers, i.e. the intervention, by multiple measures. The main outcome was the observed HH compliance, defined as the proportion of individual HH indications met by a HH action [12]. Secondary outcomes were the measured consumption of ABHR and self-reported HH compliance. In addition, we investigated the perception and attitude of ED HCP regarding wearable dispensers by self-applied questionnaires and semi-structured interviews.

The study consisted of a five-week baseline period and a seven-week intervention period that started with the introduction of wearable dispensers. The wearable dispenser (B Braun®, 100 ml, dimensions 7x8x2.2 cm) included clips to attach the dispensers to the hospital apparel. Promotion of the wearable dispensers occurred in several staff-meetings, by user information via email, and by distribution of dispensers to HCP in person. The number and position of the mounted dispensers remained unchanged.

HH compliance, the main outcome parameter, was assessed according to the WHO ‘My 5 moments for hand hygiene’ observation method through direct observation by two trained and validated members of the IPC team (MD and JK) [13, 14]. They recorded HH opportunities, HH actions, and HCP profession and sex in an anonymized form. During the last three weeks of the intervention period, the usage of either wearable dispensers or mounted dispensers was additionally noted during observations. Observations were done during weekdays at different time points between 8 am and 6 pm. Consumption of ABHR from mounted dispensers and wearable dispensers was assessed by weekly weighing the dispensers and counting the stock to calculate consumption.

Perceived ABHR availability, self-reported HH compliance, and knowledge of the WHO ‘My five moments for HH’ was assessed through a self-administered 3-item questionnaire using a 5-item Likert scale (Table 1). Additionally, participants indicated their profession and sex. Questionnaires were distributed to physicians’ physical mailboxes and were made available in the nurses’ common room after the baseline period and at the end of the intervention period.

The needed number of HH opportunities in baseline and intervention was calculated on the assumption that HH at baseline would be 50%, based on earlier observations using the same observation method. A clinically meaningful increase in HH compliance was judged to be at least 15 percentage points, i.e. from 50 to 65%. Requiring equal sample sizes for baseline and intervention, and assuming an average cluster size of three, a two-sided alpha of 0.05, and a power of 80% resulted in twice 339 opportunities.

Semi-structured interviews were conducted with a convenience sample of nurses and physicians working in the ED at the end of the intervention period. The interview guide is displayed in Table 2. All interviews were audio-taped after obtaining oral informed consent and transcribed verbatim. Data analysis of anonymized interviews was conducted inductively, following a grounded theory approach [15]. Two researchers (JK and AW) independently read transcripts and identified emerging themes. Emerging themes were then discussed and final categories were established through consensus. Quotes about barriers and facilitators were semi-quantified by a scoring system: three points for quotes mentioned by a majority of participants, two points for quotes mentioned by a minority but two or more, one point for quotes mentioned by a single participant.

A total of 811 HH opportunities were observed, 328 (40.4%) in the baseline period and 483 (59.6%) in the intervention period. The average HH compliance rose from 56% during baseline to 64% during intervention with a univariable odds ratio of 1.36 (1.02–1.81; p = .035). When adjusted for HH moment, sex and profession in the multivariable logistic regression model, odds ratio was 1.22 (0.89–1.66; p = .218) (Table 3). Weekly compliance across all indications are displayed in Fig. 1. In intervention, 7.3% (95% confidence interval (CI), 3.6–13.0) of observed HH actions were performed using wearable dispensers.

Figure 1 shows the weekly consumption of ABHR, which was 25.6 ml per patient admission during baseline, and 25.2 ml per patient admission during intervention. During the intervention period, 7.5% of the total ABHR consumption resulted from use of wearable dispensers, confirming the abovementioned 7.3% of observed use. Consumption of ABHR from wearable dispensers did increase over the seven-week intervention period (p < .001).

A total of 87 questionnaires were returned, 51 during baseline (20 (39%) by physicians, 27 (53%) by nurses, and 4 (8%) by others), and 36 during intervention (12 (33%) by physicians, 22 (61%) by nurses, and 2 (6%) by others).

The perceived availability of ABHR did not increase from baseline to intervention (4.47 vs 4.64, p = .31). Self-reported HH compliance did not differ between baseline and intervention (4.12 vs 4.03, p = .56), neither between sex (female vs male, 4.14 vs 3.97, p = .28), nor profession (physician vs nurse vs other, 4.09 vs 4.08 vs 4.0, p = .96). Mean self-reported knowledge of HH indications did not differ between baseline and intervention (4.27 vs. 4.28, p = .98). Overall, women’s perception of their own knowledge was better than men’s (4.40 vs 4.03, p = .024), whereas profession had no impact on the self-reported knowledge (physician vs nurse vs others, 4.31 vs 4.29 vs 4.0, p = .61).

Overall, 24 participants took part in the interviews, 14 (58%) nurses, nine physicians (38%), and one (4%) nursing assistant; six (25%) participants were male. The interviews had a mean duration of 10 min. Thirteen of the interviewees (54%) had a negative overall impression of the wearable dispensers, eight (33%) were neutral, and three (13%) were positive.

We identified four categories of barriers and facilitators for the use of the wearable dispensers: usability characteristics of the current wearable dispenser, availability of ABHR, cues to action/cognition (i.e. subconscious or conscious activation towards use), and perceptions about the safetyof wearable dispensers. Selected typical interview quotes are listed in Table 4. A majority of facilitators was categorized to “availability”. HCPs liked the principal idea of carrying the hand rub on them and also mentioned certain situations in which the wearable dispenser was of use, e.g. heading towards a patient from an area without mounted dispensers, and accompanying patient to wards where location of mounted dispensers was not familiar. In addition, interviewees perceived the wearable dispenser as a reminder to perform hand hygiene. Barriers were more often mentioned than facilitators, and arguments from all thematic groups emerged. Concerning usability, interviewees did not like the way the wearable dispensers were attached to the uniform because it made the dispenser dangle back and forth, and the fastening mechanism often failed, dropping the dispenser. HCP perceived the wearable dispensers as an additional item to carry around, while they already carried a multitude of other items such as stethoscopes and notes in their pockets, especially when wearing scrubs without coat. Wearable dispensers were mainly perceived to be unnecessary, given that a mounted dispenser was in easy reach during the vast majority of HH indications. HCPs were used to the mounted dispensers and habitually used them instead of the wearable dispensers. The outer surface of the wearable dispensers was perceived as risk factor for contaminating patients and hospital surroundings.

Interviewees made numerous suggestions about how to improve the wearable dispenser, including reducing its size to make it better fit in coat pockets, modifying the bottle outlet, e.g. by replacing it by a membrane, to avoid the need to open its cap, and improving the attaching mechanism.