Hand-cleansing during Postanesthesia Care

Between June 19 and July 8, 2000, we conducted an observational study at the PACU of the University of Geneva Hospitals, a 2,300-bed healthcare center providing primary and tertiary care to Geneva and the surrounding area. The PACU is a 12-bed open unit with 25–35 daily admissions. Patients are admitted only after extubation. The nurse-to-patient ratio is 1:3. Three sinks with medicated soap and paper towels and individual bottles of handrub solution (alcohol-based preparation with 0.5% chlorhexidine gluconate and skin emollients) for hand-cleansing are available. 10 

Seven trained observers recorded potential opportunities for and actual performance of hand-cleansing during a care session. At least two observers worked simultaneously. Observations were not conducted during nights or weekends. Observers were visible but as unobtrusive as possible. 8–10PACU staff were aware of being monitored about infection control practices but did not have information on which aspect. Data were recorded on a specially designed report form that had been pretested and adjusted in previous studies. 8,10Two periods are reported: period I, immediately at PACU admission, and period II, starting at least 30 min after patient admission. Concordance among observers and sensitivity to detect opportunities for hand-cleansing were excellent when previously evaluated. 8,10 

The current study was approved by the hospital's ethical committee. Because data were recorded without any intervention and according to a protocol already used within the institution, 8including ICUs, 10,11authorization was given to waive informed consent. We informed department chairs about the upcoming study in May 2000. Personnel were not informed of which aspects of infection control practices would be evaluated, and during the study, they were not given feedback on their performance. In accordance with the requirements of the institutional review board, 8,10we did not identify staff members by unique identifier.

Patient care activities and indications for hand-cleansing were coded according to standard definitions 8,10,12and included direct patient contact, wound care, blood sampling, intravenous or arterial catheter care, respiratory tract care, handling of body fluids or secretions, and breaks in the sequence of patient care (i.e. , the HCW left the bedside to answer a telephone or to prepare materials or a drug for patient care). 13Compliance with hand-cleansing was defined as either washing hands with soap and water or rubbing hands with an alcohol-based solution. 10,12We focused the study on the type of care that generated indications for HCWs to perform hand-cleansing while directly caring for patients. Indications for hand-cleansing were categorized 8,10,11into those presenting a medium risk of cross-contamination (e.g. , after direct patient contact, intravenous/arterial care, urinary care, respiratory care, wound care, contact with biologic body fluid), and a high risk of cross-transmission (between care of a dirty and a clean body site, before intravenous/arterial care, urinary care, respiratory care, and wound care). It should be noted that both hand-cleansing after patient contact to prevent hand contamination between separate patients and hand-cleansing between a dirty and clean body site in the same patient to prevent cross-transmission within the same patient were considered as indications for hand-cleansing. 8,10,12Low-risk indications comprised indirect patient contact and hospital maintenance. Hand-cleansing was required regardless of whether gloves were used or changed. 12Failure to remove gloves after patient contact or contact between a dirty and a clean body site on the same patient was considered as noncompliance. 8,10,12 

Intensity of patient care was estimated by the number of patient care activities per hour for each HCW observed. The activity index 8,10,11was estimated by the number of observed indications at high or medium risk for hand-cleansing per hour for each HCW observation.

Potential confounders of hand-cleansing compliance at admission included HCW professional category, type of surgery, scheduled or unscheduled surgery, Altemeier class, 14American Society of Anesthesiologists physical status, anesthetic technique, level of risk for cross-contamination (high or medium risks), and intensity of patient care.

We compared categorical variables by use of the chi-square test or Fisher exact test when expected values in either cell were less than 5, and we compared continuous variables by use of the Student t  test. Nonparametric methods (rank-sum test) were used when departure from normality was observed (activity index). Continuous variables were expressed as the mean (± SD) or as the median (25th–75th percentiles). Association between continuous variables was graphically explored using nonparametric regression 15and summarized by linear regression if appropriate. This method was used to investigate the association between compliance and workload. 8,11The dependent variable was compliance with hand-cleansing. All variables from the patient's admission were first examined by univariate analysis using the Mantel-Haenszel method and logistic regression. Variables associated with the dependent variable with a probability equal to or less than 0.1 were further investigated in a multivariate logistic regression model. Measures of association are summarized by odds ratios, displayed with their 95% confidence intervals. We used robust estimates of variance (generalized estimating equation) by including HCW as a cluster. 8,10All tests were two-tailed, and P  values less than 0.05 were considered as statistically significant. We used Stata version 6 (Stata Corp., College Station, TX) for all analyses.

Data were collected from 240 periods, totaling 39 h of HCW observation (table 1). Eighty-seven HCW observations concerned 57 patients immediately at PACU admission (period I), and 153 concerned 130 patients already admitted in the PACU for at least 30 min (period II). Of 3,143 patient care activities recorded, 1,091 (35%) carried a medium or high risk for cross-transmission. Overall, hand-cleansing was performed 152 times during period I and 299 times during period II. Individual bottles of handrub solution were available at the bedside in 97% of observations.

Fifty-seven patients were admitted for a median length of stay in the PACU of 160 min (120–240 min) after orthopedic (40%), abdominal (30%), urogenital (21%), and thoracic (9%) surgical procedures. Observers recorded 1,212 patient care activities (median, 77 per hour; range, 60–104 per hour) among 61 nurses, 13 physicians, and 13 nurse assistants. Figure 1illustrates the intensity of patient care activities and indications for hand-cleansing at different times during postanesthesia care. Figure 1, A  illustrates the intensity of patient care activities and hand-cleansing indications on admission to the PACU. Among patient care activities, 485 (40%) carried a medium or high risk of cross-transmission and were thus considered as indications for hand-cleansing (table 2). A close relation between workload and the number of indications for hand-cleansing was observed (fig. 2).

Average compliance with hand-cleansing was 19.6% (table 2). Handwashing with soap and water was chosen in 21% of instances; handrubbing with the alcohol-based product was chosen in 79%. Compliance before initiating care of a new patient was 56%; it ranged from 0 to 22% during direct patient care activities in the same patient (table 2). There was no difference in hand-cleansing compliance according to the presence or absence of urinary bladder catheter, nasogastric tube, and abdominal or chest tubes. Similarly, no difference was observed according to American Society of Anesthesiologists physical status, type of surgery, or anesthetic technique (table 3). Among HCWs, physicians tended to be less compliant. In multivariate analysis (table 4), patient age less than 65 yr, recovery from clean/clean-contaminated surgery, and high workload were independently associated with noncompliance.

Observers recorded 1,931 patient care activities (median, 77 per hour; range, 55–96 per hour) among 120 nurses, 17 physicians, and 16 nurse assistants (table 1). A total of 606 activities were indications for hand-cleansing associated with medium or high risk of cross-transmission (table 2). Figure 1, B  illustrates the intensity of patient workload and hand-cleansing indications during PACU stay. On average, one third (31.1%) of patient care activities recorded were indications for hand-cleansing and average compliance, while caring for patients admitted in the PACU for at least 30 min was 12.5%. HCWs washed hands with soap and water in 17% of instances and used handrub with the alcohol-based product in 83%.

Our results showed a particularly low compliance with hand-cleansing in the PACU both at time of patient admission (19.6%) and during stay (12.5%). Overall, based on direct observation, indications for HCWs to cleanse their hands ranged between 31% of patient care activities during patient stays in the PACU and 40% of these activities at time of patient admission to the PACU.

Intensity of patient care in the PACU was extremely high, both on admission and during stay. Unsurprisingly, this has been repeatedly identified as a very, if not the most, important risk factor for low compliance both in the hospital ward and in the ICU. 8,10,11,16,17Within a short time period, PACU nurses must care for several acute medical events. For example, patients with excessive pain require substantial nursing care, especially when combined with agitation. 1Furthermore, unpredictable events such as critical respiratory or cardiac events add disproportionately to workload. In this study, there is a close relation between the intensity of patient care activities and the number of indications for hand-cleansing: The higher the workload, the higher the number of indications for hand-cleansing associated with a significant cross-transmission risk.

Other risk factors identified with noncompliance were older patient age and clean/clean–contaminated surgical procedures. Surprisingly, it has been reported that therapeutic effort decreases with increasing age in the ICU, 18,19and nursing workload also increases with patient age. 1Several investigators 6,8,10,20,21have previously observed that HCWs are more likely to comply better with infection control measures when they perceive a potential danger for themselves, e.g. , caring for patients recovering from contaminated or dirty surgical procedures.

A confluence of risk factors unique to the PACU renders improvement of compliance with hand-cleansing particularly challenging. First, HCWs may not be sufficiently aware of cross-transmission risks and may be inadequately trained in infection control. Second, admissions to the PACU are not distributed equally throughout the day. Workload increases rapidly when several patients are admitted at the same time. Third, although an open ward is optimal for patient observation, it is easier for HCWs to transmit organisms from patient to patient.

However, we acknowledge that we really do not know the impact of “contaminated” hands in the PACU. Indeed, we were unable to estimate cross-transmission and nosocomial infection rates and their possible relation with poor compliance because such a task would be too time and resource demanding. Although there are no reports to our knowledge of an outbreak in a PACU, the risk seems real. First, the risk of bacterial transfer on the hands of HCWs is proportional to the frequency with which a patient is touched, 13,22which is very high in the current study. Second, a large proportion of patient care activities were among those encountered in the ICU, a setting where outbreaks of infections related to pathogen cross-transmission from patient to patient or from contaminated device to patient are well documented. 4,5,23Third, a sustained improvement in compliance with hand-cleansing has been reported to coincide with a dramatic reduction of nosocomial infections and methicillin-resistant Staphylococcus aureus  transmission in a teaching hospital. 10Lastly, understaffing, overcrowding, high workload, and poor hygiene practices appear to play an important role in outbreak situations. 5,23–26 

A number of interventions aimed at increasing compliance with hand hygiene have been implemented, with varying degrees of success. 6,27–31Multifaceted approaches that combine education, continued feedback, reminders, and policy changes can have an important effect on hand-cleansing compliance. 6,10,11,30–32As shown by the relatively high number of hand-cleansing actions, HCWs in the PACU were not reluctant to clean their hands, and when correctly applied to high- and medium-risk indications, hand-cleansing compliance could have approached 50% in our PACU. Besides these fundamental aspects, education oriented on specific aspects of hand-cleansing during postanesthesia care and placement of equipment and furniture might contribute to help HCWs work more efficiently and comfortably. Recent experiences in hospitals, 6,8,10,12in particular in ICUs, 7,9,11,29suggest that appropriate location and easy access to alcohol-based handrub is associated with improved hand-cleansing practices. We propose to test this strategy in the PACU, supplemented by specific staff education to integrate such a system change in daily practice.

In conclusion, compliance with hand-cleansing is low in the PACU and is associated with factors identifiable at time of patient admission. The strong relation between the intensity of patient care and noncompliance argues that hand-cleansing should be viewed not only as problematic individual behavior, and it is essential that effective strategies be multimodal and consider system changes and new engineering solutions. Promotion should include educating PACU HCWs to recognize indications for hand-cleansing. Long-term change in behavior requires that all staff take responsibility for ensuring that hand-cleansing becomes an everyday part of clinical culture.

The authors thank the members of the staff of the Postanesthesia Care Unit, University of Geneva Hospitals, for their collaboration and Rosemary Sudan (Research Assistant, Infection Control Program, University of Geneva Hospitals, Geneva, Switzerland) for editorial assistance.