Hand hygiene knowledge, perception and practice of healthcare workers in a Turkish university hospital intensive care unit.

INTRODUCTION
While improvement of hand hygiene (HH) compliance is considered as the best approach to reduce healthcare-associated infections, the instructional interventions in HH among healthcare workers of intensive care unit (ICU) of our hospital was not successful enough. The following study was conducted to evaluate HH knowledge, perception, practice and effectiveness of the trainings among healthcare workers of ICU in our hospital.


METHODOLOGY
A cross-sectional study was conducted in the ICU containing 8 medical and 16 surgical beds with 284 filled questionnaires about HH knowledge and 1187 observed opportunities for HH compliance.


RESULTS
Overall observed HH compliance rate was 40.6%; lowest compliance was 21.7% for "before clean/aseptic procedure" indication and highest compliance was 68.6% for "after touching a patient" indication. Although > 90% healthcare workers correctly identified the World Health Organization's five indications for HH, 82 - 85% failed to recognize non-indications, i.e. when it was not necessary to perform HH. Our study showed that 15.1% of healthcare workers neither received nor felt the need for HH training.


CONCLUSIONS
Despite regular HH trainings, healthcare workers could not differentiate when HH was not required which suggested failure to understand HH rationale. This may explain poor HH compliance rates. A systematic study is needed in order to find out the reasons behind of this noncompliance and improve HH training methods for educating healthcare workers.


Introduction
Health care-associated infections (HCAIs) cause considerable morbidity and mortality in developing countries, putting an extra restraint on already limited resources [1]. It is widely accepted that prevention of HCAIs can be achieved through complete and continual compliance with hand hygiene (HH) by health care workers (HCWs) [2].
Despite our training efforts in HH, crosscolonization and subsequent infections with multi-drug resistant pathogens have frequently been observed in our intensive care unit (ICU) indicating a deficiency in HH compliance [3,4].
Assessing institution-or unit-specific barriers to HH compliance is recommended for formulating interventions that would be locally relevant [5,6]. Therefore, we aimed to determine the level of HH knowledge, perception and practices of HCWs working in the ICU of Marmara University, Pendik Training and Research Hospital, Istanbul, Turkey in order to improve future HH interventions.

Methodology
A cross-sectional study was conducted in the ICU containing 8 medical and 16 surgical beds. This study was approved by the institutional clinical research ethics committee of Marmara University, School of Medicine (file no: 1300252382). The study used selfadministered questionnaires and direct observation for data collection. A questionnaire based on the World Health Organization's (WHO) HH knowledge questionnaire was designed and pilot tested on five HCWs [7]. The questionnaire assessed sociodemographic characteristics, perception of HH compliance, previous HH trainings, perceived need for future training, and standard knowledge about HH and HH indications (Supplement). HH compliance of HCWs was evaluated through direct observation. The WHO's Five Moments (before touching a patient, before performing clean/aseptic procedure, after body fluid exposure risk, after touching a patient, and after touching patient's surroundings) Observation Form was used for data collection [8].
A group of final year medical students (interns) doing their public health clerkship were recruited as data collectors and observers. In the first phase of the study, they distributed the questionnaires to HCWs in the ICU and collected them. For the second (observational) phase, medical students were trained by the infection control physician on how to observe HH compliance using the WHO training video, and 40 HH opportunities were observed along with the infection control physician until a coherence was reached [9]. HH observations were carried out for a week in the ICU, from 8:30 to 10:30, Monday to Friday.
Instead of choosing a sample, all HCWs in the ICU were observed (12 nurses, 8 environmental services personnel (ESP), 8 physicians in medical and 23 nurses, 10 ESP, 10 physicians in surgical site of ICU). The anonymous questionnaire was completed by all HCWs in the ICU together with internists, surgeons and anesthesiologists who have access to the ICU (284 HCWs, response rate of 92%).
Descriptive data were presented as medians, interquartile range (IQR) and percentages. Categorical variables were compared with the chi-square and Fisher's tests. A p value of < 0.05 was set as the level of statistical significance.

Results
Of 284 questionnaire respondents, 45.8% were female and the median age was 29 (9) years. Within the study group, 72.3% were physicians, 18.8% were nurses and 8.9% were ESP.

Compliance perception
Of 281 participants, 62 (22.1%) indicated that their compliance to HH guidelines were insufficient. Factors such as age, sex, years in occupation (i.e. experience), prior HH training or recent (within 1 year) HH training were not associated with insufficient compliance perception. Only occupation was a significant factor: while 27.2% of the physicians expressed insufficiency with HH compliance, this rate was 9.1% for nonphysician HCWs (p = 0.001).

Perceived need for training
The proportion of HCWs who had received HH training was 76%. Of the participants, 68 (24.4%) currently felt the need for HH training. Among factors tested, prior HH training significantly lessened the perceived need for additional HH training (37.3% vs. 19.9%, p = 0.004) ( Table 1). Of 278 participants, 42 (15.1%) did not feel the need for HH training although they had not received training before.

HH knowledge
Over 95% of HCWs correctly answered six out of eight questions about HH knowledge including glove usage, HH requirement between patients and relation between HH and HCAI development. The application durations required for hand disinfection with alcoholbased hand rub and soap/water were the least correctly answered questions with 45.9% and 33.5% of HCWs giving the correct answers, respectively.

HH indication knowledge
Of the participants, > 93% identified the WHOrecommended five HH indications correctly. However, non-indications such as performing HH in the corridor before entering or after exiting patient room, or performing HH before touching patient's surroundings were also marked as correct by > 82% of HCWs. Prior HH training was not a significant factor in discerning non-indications (p > 0.05).

HH compliance
Overall observed HH compliance rate was 40.6% ( Figure 1A) with the lowest rate of 21.7% for "before clean/aseptic procedure" indication and the highest rate of 68.6% for "after touching a patient" indication ( Figure 1B).

Discussion
Our hospital is located on the Asian side of Istanbul with 650 beds serving a population of about 5 million. As a tertiary-level university hospital, we receive many critically ill patients referred from nearby hospitals. The ICU with a total of 24 medical and surgical beds are usually at > 90% occupancy. Despite our training efforts, cross-colonization among patients continues. In this study, we wanted to take a snapshot of the ICU in terms of HH knowledge, perception and practice to plan our future interventions.
In terms of knowledge, the overwhelming majority appeared to understand the importance of HH in the acquisition of HCAIs. Of the participants, > 93% identified the WHO-recommended five HH indications correctly, but failed to identify the non-indications. This finding might suggest that HCWs answered the questions intuitively rather than knowingly. It is critical that HCWs should understand the rationale behind HH and discern non-indications as well as indications, as failure to do so might lead to the belief that there are too many indications and it is impossible to comply with all. Whether this assertion is true should be studied in a future research.
Our study showed that HCWs' perception of HH compliance is much higher than what was observed. In the questionnaire, most (77.9%) HCWs believed that they had sufficient HH compliance. However, the direct observation showed an overall 40.6% compliance, which might even be an overestimation as observations were made during morning hours while the staff were rested and more in number; and open to the Hawthorne effect [10]. Moreover, the compliance to the indication before clean/aseptic procedures was 21.7%, lowest of all the indications. This indication is especially important in the prevention of colonization and infection of medical devices such as intravascular catheters or ventilators [11]. These findings may adequately explain the cross-colonization and deviceassociated HCAIs observed in the ICU. These points should be stressed during training sessions [12].
The HH compliance rates of ICU in our center are reported by the charge nurse quarterly. Their last 12 HH compliance rates for physicians, nurses and ESPs were at a median (IQR) of 56.5% (15%), 75.5% (20.8%), and 55% (12.2%), respectively, with an overall compliance of 57.4% (25.1%). Our study also showed that such self-reporting could be inaccurate as this rate was higher than the directly observed 40.6% compliance rate [10].
Physicians were more likely to perceive themselves as non-compliant. Moreover, there was a group of HCWs, who were neither trained nor felt the need to be trained. More studies are needed to find out the reasons behind this lack of interest and how to arouse motivation in HH [13].

Conclusion
Our findings showed that the observed HH compliance was much lower than HCWs' perception and there was a major deficiency in understanding the rationale behind HH indications despite our training efforts. To explore the reasons for poor HH compliance we need to understand the beliefs and attitudes that shape HH behavior in our cultural setting [14,15]. A qualitative research has been planned and finalized to address these questions [16].