Objective: To improve hand hygiene in two outpatient health care clinics through the introduction of a gel sanitizer and an informational poster.

Research questions. The primary research questions addressed by our study were as follows:

1. Could an interventional campaign in two outpatient health care clinics lead to improved hand hygiene?

2. Are there differences in the observed frequency of hand hygiene at baseline compared with that at one week and one month after introduction of the intervention?

3. Are there differences in the observed frequency of hand hygiene based on workers' sex, workers' profession, and timing (pre- and postpatient contact)?

4. Which hand hygiene tools do health care workers in these settings prefer?

5. Would the observed health care workers later report that either or both interventional tools were motivating and influenced their hand hygiene habits?

METHODS

Design and sample. This study used an interventional design, enrolling two outpatient clinics and performing direct observation of health care workers to measure hand hygiene opportunities and attempts at baseline, after the intervention, and during the follow-up period. To recruit the clinics, we first contacted the administration of a large midwestern health care system to determine whether there was mutual interest. Specific outpatient clinics were then chosen from within that system based on willingness to participate, whether the clinic layout would be conducive to direct observation, and whether the anticipated caseload during the study period would be sufficient. Because all observations and surveys were conducted anonymously, after review by the institutional review boards of the researchers' institution and the participating clinics, the study was given exempt status.

The first clinic was an outpatient oncology clinic having an open layout; sinks and foam sanitizers were available at the nurses' station but not in the immediate vicinity of individual patients. The second clinic was an outpatient gastrointestinal (GI) clinic with individual curtained rooms for patients. At each clinic, the observed health care workers included primarily physicians and nurses, as well as a few others. Sinks for handwashing were located at the nurses' station, and foam sanitizer was wall mounted at the entry to each curtained patient room. Gel sanitizer was not available at either clinic.

The intervention consisted of introducing alcohol-based gel sanitizer and a novel poster to each clinic. We included gel sanitizer in this campaign to provide health care workers with an alternative to soap and water and foam sanitizer.

An informational poster was created to increase health care workers' awareness of hand hygiene, provide information about when hand hygiene should be performed, and encourage them to take personal responsibility for reducing the spread of health care-associated infections. To create the poster, clinic administrators, nurse managers, and the research team first brainstormed ideas. Several media consultants from other universities that have conducted similar hand hygiene campaigns also provided input. A first draft was designed and shown to all of these people, and the final version of the poster incorporated their suggested changes for improvement.

Measures. For the purpose of this study, hand hygiene could be acceptably performed using either soap and water or hand sanitizer (foam or gel). We measured the number of hand hygiene attempts against the number of hand hygiene opportunities. Hand hygiene opportunities were defined as those occurring immediately before and after a health care worker makes direct contact with a patient; hand hygiene attempts were defined as an effort to perform hand hygiene during each opportunity. (The term "attempt" was used rather than "compliance," because we did not monitor hand hygiene technique for compliance with guidelines.) Direct observers monitored the opportunities and recorded the attempts. To establish intercoder reliability prior to the study, two observers were trained and then asked to observe and record the same hand hygiene opportunities; the overall percentage agreement between the observers was 96%. For the GI clinic, only those opportunities and attempts that gave observers full visibility through an open or partially open curtain were monitored and recorded.

Baseline data were collected on three nonconsecutive days, for four hours each day. Then approximately 10 pump bottles of gel sanitizer containing 65% ethyl alcohol were provided in each clinic on the nurses' station counter. The counters were long, extending the lengths of the clinics, and the bottles were spaced out so they lined up with patient areas or rooms as closely as possible. Twenty posters were strategically placed near sinks, at intervals along the nurses' station counter, and next to wall-mounted foam sanitizer dispensers. One influential physician in the GI clinic was displeased when the posters were hung, expressing concern that the tape used to mount them would damage the clinic's walls; the posters were rehung with painter's tape to ease his concern and prevent damage (the study design was not altered).

One week after the introduction of the intervention, direct observation of hand hygiene was performed on five nonconsecutive days, for four hours each day. This design allowed evaluation of the impact of a short intervention period. After these interventional observations were complete, the posters were removed at the request of the clinics' administration. The remaining gel sanitizer was left behind, but it too was removed by the administration, in keeping with the clinics' policy. One month after the last day of the interventional observations, follow-up direct observation of hand hygiene was done on three nonconsecutive days, for four hours each day. Neither posters nor gel sanitizers were provided during the follow-up period.

A survey was mailed to the nurse managers at both clinics and was distributed to 47 health care workers at the oncology clinic and to 41 health care workers at the GI clinic, three months after the last day of the follow-up observations. The survey was conducted to evaluate the workers' perceptions of the hand hygiene campaign, its motivational impact on their hygiene practice, perceived barriers to hand hygiene at their clinics, and their preferred hygiene tools (soap and water, foam sanitizer, or gel sanitizer). A five-item Likert scale (ranging from "strongly disagree" to "strongly agree") was used for questions regarding the motivational efficacy of the intervention tools.

Analytic strategy. Pearson's [chi]² analyses were used to compare the frequency of hand hygiene attempts during the three observation periods and to compare the precontact and postcontact frequency of such attempts. Each calculation had 1 degree of freedom. A P value less than or equal to 0.05 was considered significant. Only descriptive statistics were used to assess hand hygiene frequency by workers' sex and profession, because there were too few observations of male health care workers and physicians to permit comparative analyses. Descriptive statistics were also used for comparing the hand hygiene tools used, as product availability varied throughout the study. Because most observed health care workers were nurses and physicians, other health care workers were excluded from data analysis.

RESULTS

A total of 56 health care workers returned surveys (41 from the oncology clinic, 15 from the GI clinic). Fifty percent of all survey respondents agreed or strongly agreed that the campaign had increased their awareness about hand hygiene, with more workers at the GI clinic (80%) reporting this belief than at the oncology clinic (39%). Overall, 34% of all respondents agreed that the hand hygiene campaign had improved their hand hygiene practices.

At the oncology clinic, the usage rates for soap and water and foam sanitizer were almost equal at baseline (53% and 47% of attempts, respectively). But after the intervention's introduction of gel sanitizer, these usage rates dropped (to 21% and 40%, respectively). Gel sanitizer was used in 40% of hand hygiene attempts. During the follow-up observation period, the gel sanitizer provided by the researchers was removed. The usage rate for soap and water remained low (24%) while that for foam sanitizer rose (63%). Self-provided gel sanitizer was used also (13%). At the GI clinic, the baseline usage rate for foam sanitizer (59%) was higher than that for soap and water (40%). After the intervention's introduction of gel sanitizer, the usage rates for soap and water and foam sanitizer dropped (to 19% and 40%, respectively); gel sanitizer was used in 41% of hand hygiene attempts. During the follow-up period, the usage rate for soap and water remained low (34%), while that for foam sanitizer increased (62%). Self-provided gel sanitizer was used also (4%). (Some percentages don't sum to 100 because of rounding.)

Of the 56 survey respondents, 50% agreed or strongly agreed that the introduction of gel sanitizer was an effective motivator and led to improved hand hygiene frequency. Asked "Which do you prefer most as a means of hand hygiene?" the 41 respondents from the oncology clinic reported mixed preferences (soap and water, 58%; foam sanitizer, 24%; gel sanitizer, 22%), whereas 80% of the 15 respondents from the GI clinic reported a preference for soap and water, with the remainder preferring foam sanitizer (20%) and gel sanitizer (20%). (Percentages don't sum to 100 because some responders at both clinics reported preferences for more than one hand hygiene tool.) Asked "What is the most suitable place to keep sanitizer, to make its use more effective?" 66% of all respondents preferred placement in the immediate vicinity of the patients.

In the GI clinic, 93% of the 15 surveyed health care workers reported being aware of the new posters; of these, 36% agreed or strongly agreed that the posters were effective as a motivational tool for encouraging hand hygiene. In the oncology clinic, 49% of the 41 surveyed workers reported being aware of the posters, and of these, 45% agreed or strongly agreed that they were effective. Among all survey respondents, 42% reported finding neither the introduction of gel sanitizer nor the poster motivating, while 23% were motivated by both interventional tools.

Perceived barriers to hand hygiene among all survey respondents included "skin irritation" (34%), "forgetfulness" (32%), "insufficient time" (24%), "interference with patient care" (11%), "lack of facilities or inaccessible materials" (5%), and "insufficient training" (5%); some respondents selected multiple barriers. Twenty-three percent reported that none of the above was a barrier in their clinic. Although no objective data were collected to assess the incidence of skin irritation associated with use of the available hygiene tools, numerous health care workers in both clinics were overheard commenting during both observation periods that the gel sanitizer was less irritating than either the foam sanitizer or soap and water.

At baseline, the frequency of observed hand hygiene attempts was greater after patient contact than before patient contact (P = 0.001 in oncology clinic, P = 0.002 in GI clinic). This was also the case at one week after the intervention (P > 0.001 in each clinic) and during the follow-up observation period (P > 0.001 in each clinic). Hand hygiene improved from baseline to the intervention period for precontact and postcontact observations, and this improvement was sustained with no significant decreases in hygiene in both clinics during the follow-up period. A similar finding was seen in the survey responses (n = 53 for this question): more respondents reported "always" performing hand hygiene after contact with patients (83%) than reported "always" performing hand hygiene before contact with patients (57%).

DISCUSSION

Limitations and recommendations for further research. One limitation of this study is the possibility of the Hawthorne effect; some health care workers might have realized that they were being observed and cleansed their hands more frequently as a result. While using video cameras or training staff in covert observation might have minimized this possibility, the clinic layouts, time constraints, and the need to ensure patient privacy made such solutions less than desirable. To minimize possible bias, only clinic administrators, nurse managers, and medical directors were consulted regarding study design. The possibility of the Hawthorne effect notwithstanding, we recommend that future researchers, after collecting baseline data, seek the ongoing support and involvement of influential health care workers, who can also serve as role models for others.

Another limitation is that observations weren't tagged with health care workers' identities. It's possible that workers with excellent hand hygiene habits were observed with greater frequency than those with poor habits, thus skewing the data and the statistical analysis. Although health care workers probably prefer to remain anonymous during such observations, we recommend recording and associating worker identities with hygiene performance so that statistical analyses can be as accurate as possible.

Because resource limitations led to interventional and follow-up observation periods of different durations, and because of fluctuating patient caseloads during these periods, not all health care workers had the same number of opportunities for hand hygiene; future studies could schedule additional observation periods in order to achieve more equal sample sizes. Finally, leaving interventions in place for a longer time period and extending follow-up observations beyond one month (to three, six, and 12 months) would be useful in evaluating whether the interventions had sustained effects.

CONCLUSIONS AND PRACTICE IMPLICATIONS

This hand hygiene campaign showed that introducing a gel sanitizer and providing informational posters can yield modest short-term improvements in overall hand hygiene performance in outpatient clinics. To maximize the clinical impact of such a campaign, we suggest that administrators and influential health care workers work together to create an environment in which adherence to hand hygiene is expected; provide a variety of hand hygiene products-including gel sanitizer in either pocket-size or pump dispensers-in the immediate vicinity of patients; and encourage health care workers to create several motivational, setting-specific posters that can be rotated throughout the clinic on a regular basis. The findings of this study should also prompt increased attention to precontact hygiene in outpatient clinics, in order to minimize the potential spread of infection. Indeed, the intermittent evaluation of hand hygiene, using direct observation, is essential for identifying aspects that need improvement and maintaining adherence as a top priority.

REFERENCES