Relationship between Systems-Level Factors and Hand Hygiene Adherence

Abstract

This study was a cross sectional descriptive survey of acute care hospitals in California to describe staff hand hygiene compliance and related predictors, and explore the relationship between hand hygiene adherence to health care-associated infections. Although there was a relatively small sample size, institutions with morning huddles reported a significantly higher proportion of ≥95% hand hygiene compliance. Huddles are an organizational tool to improve teamwork and communication and may offer promise to influence hand hygiene adherence.

Health care-associated infections (HAI) represent a serious threat to patients¹ and are a major source of morbidity and mortality, despite the fact that they are often preventable. Mortality in United States hospitals caused by or associated with HAI was estimated at 98,987 deaths in 2002.² The ability to prevent HAI is highly dependent on adherence to appropriate hand hygiene practices by health care workers.³ Current guidelines from the Centers for Disease Control and Prevention (CDC) recommend hand hygiene in the following circumstances: before and after patient contact, when hands are visibly contaminated, after contact with inanimate objects including medical equipment, before inserting invasive devices that do not require a surgical procedure, before donning sterile gloves for insertion of a central intravascular catheter, and after removal of gloves.⁴ However, published studies consistently demonstrate that health care workers practice hand hygiene incompletely and infrequently, with rates often reported as less than 50%.^5,6 The importance of hand hygiene is underscored by the incomplete protection gloves offer against cross-contamination.³ Hand hygiene is pivotal to the reduction of HAI and prevention of antimicrobial resistance.⁷

The specific mechanisms motivating health care professionals to practice hand hygiene are not fully understood.⁸ Hand hygiene practices are influenced by factors at both the individual and system levels.⁹ There is evidence to link hospital staffing variables and risk of HAI, but most studies have involved only nursing staff.¹⁰ Infection preventionists (IPs, previously called infection control nurses or professionals) are involved in many aspects of hospital infection control and prevention including disease surveillance, outbreak management, education, policy development and implementation. Infection preventionists frequently monitor and/or teach other clinicians how to monitor compliance with hand hygiene. Additionally, increasingly they take the lead in the implementation of processes aimed at reducing rates of infections.¹¹ Although IPs are an essential component of the health care team in the control of HAI, there has been minimal research describing the relationship between IP staffing, other systems-level factors, and reported hand hygiene practices of hospital staff.

Guided by Donabedian’s theory of quality¹² (ie, structure, process and outcomes), the aims of this study were to (1) describe hospital staff hand hygiene practices; (2) explore predictors of hand hygiene adherence, including hospital leadership involvement, size, teaching status, general organizational support, presence of a hospital epidemiologist, and IP staffing; and (3) examine the relationship between reported staff hand hygiene adherence and rates of central line-associated blood stream infection (CLABSI), ventilator-associated pneumonia (VAP), catheter-associated urinary tract infection (CAUTI).

METHODS

Design

This study is part of a larger parent project, The Changing Role of Infection Preventionists (http://www.cumc.columbia.edu/studies/pnice/chaipi/).¹³ The study is cross sectional and uses data from a web based survey that was conducted in 2008.

Setting and Sample

Eligibility criteria included all non-specialty acute care facilities in California. Psychiatric facilities, drug/alcohol rehabilitation centers, nursing homes, outpatient units, and children’s hospitals were excluded. In total, 350 hospitals were eligible to participate.

Recruitment Procedures

Following approval from Columbia University Medical Center’s Institutional Review Board, participants were recruited by the Association for Professionals in Infection Control and Epidemiology (APIC) and Columbia University School of Nursing research staff from October through December 2008 using a modified Dillman technique.¹⁴ One staff member from each hospital’s infection prevention and control department was asked to complete a web-based survey, preferably the department director or coordinator. Emails were sent directly to hospital infection prevention and control department contacts. Announcements were also included in APIC e-newsletters and contacts without email addresses were sent mailed letters describing this study. As an incentive to participate, eight weekly lotteries to win an APIC textbook or membership (each valued at $100) were offered to participants who completed the survey.

Instrumentation

The survey is a modified version of a previously developed and psychometrically tested instrument. Test-retest reliability of the survey instrument has been reported with a mean kappa of 0.88 for each item, SD+/−0.024.¹⁵ Criterion-referenced validity was assessed by comparing the institutional policies and data to survey responses; no discrepancies were found. For this study, minor modifications of the survey were made, and content validity was established by a panel of experts. In the survey, respondents were asked questions about the hospital structures, processes of care related to hand hygiene and HAI outcomes in a medical or medical/surgical (intensive care unit).

Hospital structural variables included the demographics such as type of setting, number of beds, number of eligible ICU beds, average length of stay, and teaching status/affiliation with a medical school. To measure infection control and prevention staffing, the number of full-time equivalent (FTE) IPs and physician epidemiologists were calculated as number per 100 beds, presuming a 40-hour work week.

The measure of organizational support was adapted from the Patient Safety Climate in Health care Organizations instrument.¹⁶ This instrument has 16 items and responses are indicated on a 5-point Likert scale. Scores range from 16–90, and a higher score represents a greater level of leadership involvement in infection prevention. Internal consistency was demonstrated with a Cronbach’s alpha of 0.92.

Processes of hospital wide infection prevention and control practices were assessed. The IP respondent in each hospital was asked to report whether or not routine monitoring of hand hygiene practices was conducted in their hospital, and participants who selected ‘yes’ were asked to estimate the proportion of time hand hygiene was practiced correctly on a 4-point ordinal scale ranging from all of the time (greater than 95%) to rarely/never (less than 25%). The remaining survey items included questions on hospital processes to monitor hand hygiene; type of hand hygiene agent(s) provided; location of antiseptic agents; and a variety of hospital-wide hand hygiene promotional and leadership activities (education for patients and visitors, use of administrative rewards or sanctions, the provision of real-time feedback to employees, modeling of hand hygiene during leadership rounds, inclusion of hand hygiene in strategic goals and staff huddles). Early morning huddles are brief (10–15 minute), informal meetings conducted on the clinical unit to improve patient outcomes by increasing communication and teamwork amongst frontline staff.

The outcomes measured were the incidence of a medical or medical/surgical ICU-specific CLABSI, VAP, and CAUTI events reported for a 3-month period, July to September 2008. Consistent with CDC National Health care Safety Network¹⁷ definitions (March 2009), rates of infection for CLABSI, VAP, and CAUTI were calculated by dividing the number of infections by the number of device days, and multiplying the result by 1,000.

STATISTICAL ANALYSIS

Data were imported into SPSS version 16 (Chicago, IL) and SAS version 9.2 (Cary, NC). SPSS was used to compute frequencies and descriptive statistics (mean, median, standard deviation) and to perform Chi-square analyses. SAS was used for multivariable regression analyses.

Descriptive statistics were used to summarize the structures, processes and outcomes of care. Bi-variate analyses were conducted to examine associations between structures/processes of care and hand hygiene compliance. In these analyses Pearson’s Chi-square, Fisher’s Exact two-sided test and the Wilcoxon signed rank test were computed based on cell size and level of measurement.

Negative binomial multi-variate regressions were used to examine the independent relationships of the structures and process of care and the incidence rates of VAP, CLABSI, and CAUTI. Statistical significance for inclusion in the regression models was defined as p value < 0.2. Relative risk ratios with 95^th% confidence intervals were computed.

FINDINGS

Of the 350 hospitals eligible for participation, 207 hospitals (a 59% response rate) participated. However, not all respondents provided all data elements. Table 1 provides the descriptive statistics and responses for hospital demographics, IP staffing, infection incidence rates in the ICU, organizational support and hospital wide infection prevention and control practices (hand hygiene agents, location of antiseptic agent, monitoring methods and hand hygiene compliance). The number of beds ranged from 10 to 952 (median = 202). The median length of stay was 4.30 days (range = 0.20 – 31), and median number of ICU beds was 12 (range = 3 – 47). The majority (79%, 163/207) of hospitals was affiliated with a medical school, and 43% (89/207) were located in an urban setting. Rural and suburban settings accounted for 25.1% (52/207) and 31.9 % (66/207) of the remaining facilities respectively. Infection prevention and control staff were predominantly non-physician professionals and the median number of IPs per 100 beds was 0.74 (range =0.01 to 5.71). The median number of reported ICU specific infections per 1,000 device days were 1.01 for CLABSIs (range = 0 –19.35), 1.7 for CAUTIs (range = 0 – 11), and 0 for VAPs (range = 0 – 19.23).

Table 1.

Demographics and infection control characteristics of health care facilities surveyed

	Median (range)	n (%)
Structures
Hospital demographics
Number of beds (n=206)	202 (10–952)
Eligible ICU beds (n=163)	12 (3–47)
Average length of stay in days (n=149)	4.3 (0.2–31)
Infection control staffing
Physician hospital epidemiologists per 100 beds (n=207)	0 (0–8)
Non-physician infection preventionists per 100 beds (n=194)	0.74 (0.01–5.71)
Organizational support score (n=148)	67 (28–79)
Processes
Hand hygiene compliance (n=149)
95–100% of the time		14 (9.4)
75–94% of the time		94 (63.1)
<75% of the time		43 (27.5)
Hand hygiene agents/placements (n=184)
Provides bland soap and water for general patient care		83 (45.1)
Provides antimicrobial soap for general patient care		136 (73.9)
Provides alcohol based sanitizer for general patient care		180 (97.8)
Provides antiseptic agent in patient rooms/high work load areas		177 (96.2)
Hand hygiene monitoring (n=184)
Direct observation of hand hygiene		174 (94.6)
Monitoring consumption of hand hygiene agent		54 (29.3)
Routine monitoring of staff hand hygiene compliance, any type		180 (97.9)
HAI Outcomes
Catheter-associated UTI (n=42)	1.7 (0–11)
Central line-associated blood stream infection (n=102)	1.01 (0–19.35)
Ventilator associated pneumonia (n=101)	0 (0–19.23)

^*Denominators vary due to missing data

ICU, intensive care unit; HAI, healthcare-associated infection; UTI, urinary tract infection

There was a 71.5% (148/207) response rate to the survey questions on organizational support. The median organizational support score was 67.0 (range = 28.0 – 79.0).

Only 72% (149/207) of facilities surveyed reported the proportion of time hand hygiene adherence was practiced correctly. Of these hospitals, 14 (9.4%) reported compliance ≥95% and 94 (63.1%) reported compliance of 75 – 94% of the time. The remaining 43 (27.5%) hospitals reported compliance of < 75% of the time. The types of hand hygiene agents used varied among facilities, but 97.8 % of 184 respondents reported the use of alcohol based sanitizer. Among 184 facilities with available data on infection prevention monitoring, 97.9% reported that hand hygiene was routinely monitored in some way, with 94.6% citing direct observation of hand hygiene practices and 29.3% reporting monitoring consumption of hand hygiene agent. The majority (96.2%) of sites provided an antiseptic agent for patient rooms/high work load areas (Table 1).

No associations were found between hand hygiene adherence and any of the demographic, infection prevention and control staffing, or hand hygiene agent variables (all p > 0.05, analyses not presented). There was no significant relationship between organizational support and hand hygiene adherence. Among the 149 institutions that reported the proportion of hand hygiene adherence, 8 (5.3%) had morning huddles. Of the 14 (9.4%) facilities reporting a higher proportion of ≥95% hand hygiene adherence, 3 (21.4%) conducted morning huddles (p=0.028) (Table 2). Other hand hygiene promotional activities were not significantly associated with hand hygiene adherence. In the regression models, there were no statistically significant associations between rates of CLABSI, VAP, or CAUTI and any of the hand hygiene practices examined (all p values > 0.10).

Table 2.

Bivariate Analyses Examining Relationships between Hand Hygiene Processes and Hand Hygiene Compliance

	Reported hand	Reported hand
	hygiene Compliance	hygiene Compliance
	<95%	≥95%	P-
	n=135	n=14	value*
Hand hygiene promotion activity
Leads morning huddles	5/135 (3.7)	3/14 (21.4)	0.028
Models hand hygiene during leadership     rounds	47/135 (34.8)	4/14 (28.6)	0.772
Conducts educational seminars or videos for     staff	74/135 (54.8)	11/14 (78.6)	0.098
Provides educational materials/instructions to     patients/visitors	91/135 (67.4)	10/14 (71.4)	0.999
Offers rewards or administrative sanctions	32/135 (23.7)	2/14 (14.3)	0.738
Provides real-time feedback to employees	91/135 (67.4)	8/14 (57.1)	0.439
Hospital characteristics
Includes hand hygiene in hospital strategic     goals	65/135 (48.2)	8/14 (57.1)	0.522
Hand hygiene agent
Provides bland soap and water	64/135 (46.7)	5/14(35.7)	0.404
Provides antimicrobial soap	101/135 (74.8)	10/14 (71.4)	0.754
Provides alcohol based hand sanitizer	133/135 (98.5)	13/14 (92.9)	0.258

^*Differences between reported hand hygiene levels assessed using Pearson’s χ test for independence or Fisher’s Exact two-sided test when cell sizes <5.

Data are reported as number (%).

DISCUSSION

The results of this study suggest that systems-level factors (in this case, staff huddles) may influence behaviors such as hand hygiene; health care facilities with huddles reported a significantly higher proportion of ≥95% hand hygiene compliance (p=0.028). Morning huddles are an organizational tool designed to improve teamwork and communication among frontline clinical staff, and arose from the need to expedite the work of improvement teams.¹⁸ Thus, huddles represent a systems-level initiative to improve patient safety and health care outcomes. Although the precise mechanism(s) by which huddles may have contributed to improved hand hygiene compliance in this study cannot be determined, there are a number of possible considerations, including improved hand-off communication, increased collaboration, and motivation for behavioral change among staff who participate in huddles.

Hand-off communication involves relaying up-to-date, patient-specific information from one caregiver or a team of caregivers to another and is an important component of patient safety.^19,20 Huddles may foster hand-off communication among clinical and administrative staff and contribute to patient safety by encouraging the health care team to communicate on the status of the patient, including the need for contact and or respiratory isolation and other infection control practices, as appropriate.

Huddles are likely to provide an efficient forum for collaboration among frontline staff by offering an opportunity to meet while remaining on the clinical unit and improving team coordination, described as a method to improve patient safety.²¹ Nursing and medical professionals caring for hospitalized patients often face time limitations, therefore finding it difficult to attend lengthy, off-unit meetings. Poor collaboration among staff is one of the factors that have been associated with HAI.¹ The presence of administrative staff in a huddle is an added benefit and offers the opportunity for staff to voice concerns regarding some of the reasons cited for poor hand hygiene, including high work load, inadequate time, and poor location of sinks or hand washing supplies.⁹

The dynamics of behavioral change are complex. Hand hygiene practices have been noted to be well established in childhood, and therefore it should not be surprising that these behaviors are resistant to change.²² However, huddles may offer an opportunity to reinforce the importance of hand hygiene for preventing infection, encouraging individuals to conform to hospital regulations and adding the element of peer influence and feedback. Motivation for correct hand hygiene practices through role modeling and peer pressure from senior clinical and administrative staff have been proposed as a potential target area for engineering persistent change ²³ and may warrant further examination.

In our study, health care facilities with huddles reported a significantly higher proportion of ≥95% hand hygiene compliance (p=0.028). However caution needs to be exercised in the interpretation of this finding due to the limited number of facilities conducting huddles. Among the 149 facilities that reported the proportion of hand hygiene compliance, only 8 conducted huddles, and a higher proportion of ≥95% hand hygiene compliance was demonstrated in 3 of these 8 institutions that had huddles. Although this result was statistically significant, it needs to be appraised in the context of the small sample size. Clinical utility of this finding would depend of further exploration with a larger sample size.

There were several possible limitations in the measurement of hand hygiene adherence that could have influenced the expected relationship between hand hygiene promotion activities, hand hygiene adherence, and infection rates. Although there was a good conceptual fit between the data and the objectives of this study, it is possible that the reported levels of hand hygiene compliance were not a reliable measure of hand hygiene practices due to the intrinsic difficulties in the measurement of hand hygiene adherence. These may include higher rates of adherence due to observer effects ²⁴ and intra-institutional variation in organizational support or culture,²⁵ which influence staff hand hygiene behaviors. The survey instrument did not compare hand hygiene adherence by category of health care professional, and variations in hand hygiene practices among health care disciplines has been reported.²⁶ Differences in observational strategies by infection prevention and control staff may also generate non-comparability between hospitals. Another limitation was the risk of bias due to self report by infection prevention and control staff. In this study, 72.5% of respondents reported the proportion of time that hand hygiene was practiced correctly to be 75% and greater, and these are substantially higher than previously reported hand hygiene rates of less than 50%.²⁷

There may be other characteristics of hospitals that conduct huddles which we did not measure, and huddles might in fact be a surrogate for other systems-level characteristics. Multimodal interventions have been recognized as a key strategy for improving hand hygiene adherence practices^7,28, and certainly huddles could be one such multimodal strategy. It will be important for future hand hygiene research studies to identify the most successful components of multifaceted systems-level interventions because to date there is minimal robust evidence to inform the choice of interventions aimed at improving hand hygiene practices.^29,30

The lack of a significant association between hand hygiene adherence and HAI rates may be attributable to loss of power due to missing data. Less than half of respondents reported their hospitals’ rates of HAI. As with any volunteer survey, it is not possible to determine the motivating factor(s) influencing whether or not any hospital elected to participate. Plausible explanations include hesitation to communicate unfavorable outcomes and or labor shortage precluding survey completion. The study’s relatively small sample size precluded the use of logistic regression to analyze the effects of various factors on reported hand hygiene compliance, thus it was not possible to analyze the joint effects of potential predictors of adherence or control for confounding. Also, because each potential predictor was tested using Chi-square, the overall risk of type II error was increased.

A second potential problem in evaluating the relationship between hand hygiene and HAI is the use of an institutional measure for adherence. There is a large body of data indicating that proper hand hygiene reduces the risks of transmission of pathogens. However, by virtue of measuring hand hygiene at an institutional level, individual level variables may confound the relationship between hand hygiene and infection, as well as variation in hand hygiene practices within an institution.³¹ In addition, the outcome of HAI was ICU-specific and all the other variables were hospital wide. Hospital-wide adherence practices, even if accurately measured, may not be a good proxy for hand hygiene practice in the ICU.

In this study, hand hygiene educational campaigns for hospital staff, patients and visitors were not a significant predictor of hand hygiene adherence (p>0.05). In general, hand hygiene education as part of a single or multifaceted intervention has not been associated with sustained improvements in hand hygiene compliance among health care workers.^32–34 Further study is necessary to evaluate more thoroughly what types of hand hygiene promotional activities are effective in promoting proper hand hygiene.

CONCLUSIONS

Adherence to correct hand hygiene practices is an essential component in the prevention of HAI.²⁹ Systems-level factors are likely to influence hand hygiene practices of staff. This study identified staff huddles as one possible influence, but this association requires further evaluation. These findings underscore the need for more methodologically robust studies to analyze this and other factors, including communication, collaboration, and behavioral influences such as motivation to change.