Central line-associated bloodstream infections (CLABSIs) unnecessarily cause mortality, increase morbidity and length of stay, and result in higher health costs (Rosenthal et al., 2006). The Joint Commission (n.d.) estimates that there are 250,000 annual CLABSIs that result in 30,000 deaths. This further comes at a cost of $45,000 per case (Zimlichman et al., 2013). Death tolls remain high even though it has been shown that virtually all CLABSIs are preventable (Pronovost et al., 2006).
The Comprehensive Unit-Based Safety Program (CUSP) is a widely used model that makes financial resources available and pools intellectual resources to strengthen safety within an organization (Pronovost & Vohr, 2010). The method to achieve optimal success through CUSP is described in great detail by Sawyer et al. (2010). CUSP was designed to improve teamwork and safety culture and to help clinical teams learn from mistakes through the integration of safety practices into daily work. It brings together administrators and clinicians to collectively identify and solve safety-related problems. CUSP is implemented at the unit level, with support at the hospital level, and represents a scalable intervention that can eventually be implemented throughout an organization. In a structured manner, CUSP aims to empower frontline staff to use teamwork to identify the greatest unit-level safety issues and removes obstacles that might otherwise contribute to safety problems (Pronovost & Vohr, 2010).
A CUSP team is interdisciplinary in nature. The team is formed of unit staff and a hospital executive leader, such as a CEO, CFO, or COO. The CUSP team typically includes at least one nurse, physician, and administrator from the hospital unit. Teams are often larger, but those are the minimum expectations. The team has the knowledge and skill to review empirical evidence for appropriate interventions that address the unit’s safety issues. The CUSP team meets regularly to discuss safety concerns and monitor progress toward their resolution.
CUSP has been shown as a successful model to reduce CLABSI rates. The first statewide CUSP project was launched in 2003 in Michigan. Within the first 18 months of the CUSP statewide launch in Michigan, an estimated 1,500 lives and $200 million were saved (Pronovost & Vohr, 2010), and gains have been largely sustained after 10 years (Pronovost, Watson, Goeschel, Hyzy, & Berenholtz, 2015). Notably, units in Michigan that participated in the CUSP program had significantly lower CLABSI rates than other units in the region (Lipitz-Snyderman et al., 2011). Currently, more than 1,000 hospitals in the United States use CUSP as a program to reduce infection rates.
Despite the success of the CUSP model, it has not been adopted in all hospital units. Explanations for why not all hospitals implement a CUSP program include the following: Some hospitals are not aware of the CUSP model, do not believe in its merits, or deem it too costly to implement. Because implementation of CUSP uses valuable workforce and resource scarcity dictates that hospitals must be efficient with resource allocation, cost may indeed hinder CUSP implementation in some hospitals.
The purpose of this study is to identify the factors of safety culture that are associated with a reduction or elimination of CLABSIs after CUSP implementation. Although previous studies showed that CUSP is an effective model to improve CLABSI rates, the relationship between specific factors of preimplementation safety culture and postimplementation CLABSI results is unknown. By identifying the baseline factors of safety culture that are the strongest predictors of CLABSIs following CUSP implementation, hospitals can target CUSP to those units expected to have the greatest odds of success. That will allow hospitals to maximize scarce resources and alleviate some of the CUSP program’s cost concerns if CUSP cannot be implemented in all units.
Conceptual Model and Hypotheses
High reliability organizations (HROs) are able to achieve nearly failure-free results despite operating in complex environments. HRO theory suggests that in order to achieve positive outcomes the organization needs to place a high priority on safety (Gaba & Cooper, 2000), procedures and resources should support safety (Sexton, Thomas, & Helmreich, 2000; Tucker & Edmondson, 2003), and there should be monitoring and learning from processes (Weingart, Ship, & Aronson, 2000). This theory guided our study, as we explored how factors of organizational safety culture affected CLABSIs. The CUSP model fits HRO theory in that introduction of CUSP signifies that the organization has placed a priority on safety. CUSP is designed so that financial and personnel resources are provided. Furthermore, CUSP is intended as a model to monitor safety issues and learn from actual or potential safety events.
Because the safety factors in our study have been linked to other healthcare-associated infections, there is reason to believe that they are also associated with CLABSI rates. The factors had shown some effect on HAI rates in previous studies (Vigorito, McNicoll, Adams, & Sexton 2011). We posit the following hypothesis:
Hypothesis 1. Units with stronger factors of safety culture prior to CUSP implementation have a lower number of infections after implementation.
Hypothesis 2. Units with stronger factors of safety culture prior to CUSP implementation have greater odds of zero infections after implementation.
The American Hospital Association’s Health Research and Educational Trust provided the data for this study. It is derived from the On the CUSP: Stop BSI program funded by the Agency for Healthcare Research and Quality. Hospital participation in the program was voluntary. Each state in the program had a median of 23% of its hospitals participate. This study was approved by the Ohio State University Institutional Review Board, with Protocol #B0297.
Contributing hospitals had a range of structural characteristics and unit involvement in CUSP. Hospitals were from the nonprofit, for-profit, and government sectors and ranged in size from small (minimum of 12 beds) to large (maximum of 1,452 beds). Hospitals selected the units that joined the CUSP CLABSI rate reduction program; not all hospital units participated. The range of participating units per hospital was 1–14, with a median of 1. Hospital intensive care units represent over 75% of participating units. Participating units in the study were in a range of two to six quarters. Actual survey respondents numbered 35 for each unit and included physicians, nurses, and administrative staff.
We included six cohorts of hospital units that had various start dates for initiating CUSP. The first cohort’s intervention period began May 1, 2009, and each subsequent cohort had a different start time; there was a typical gap of 4–5 months between cohort start dates. Hospital units were instructed to provide CLABSI rate information and survey responses on factors of safety culture from participating units. A total of 1,330 hospital units provided some CLABSI or safety culture data for the CUSP study during the period between May 2009 and June 2012; however, 681 of those units did not provide data on both CLABSI rates and factors of safety culture. A complete case analysis approach to missing data enabled us to use a final sample of 649 hospital units located in 435 hospitals and 38 states.
We used the Hospital Survey on Patient Safety Culture (HSOPS), first introduced by Agency for Healthcare Research and Quality in 2004, to derive the primary independent variables. Survey data were collected prior to CUSP implementation. The collection dates varied based on the start date of each cohort. Psychometric analysis conducted by multiple studies confirmed that the HSOPS dimensions, each composed of three to four related survey questions of perceptions of safety culture, are reliable measures valid at the individual, unit, and hospital levels and can be used by researchers to assess patient safety culture (Sorra & Dyer, 2010). The survey instrument and its dimensions can be found here: http://www.ahrq.gov/professionals/quality-patient-safety/patientsafetyculture/hospital/index.html.
The dimensions we studied were the baseline perceptions of hospital staff for each organizational safety culture factor. Supervisor support showed the priority a supervisor placed on safety. Organizational learning reflected continuous improvement regarding patient safety, in which mistakes led to positive changes and improvements were evaluated for their effectiveness. Teamwork within units exhibited the support and respect that people have for one another within a unit. Communication openness was the ability and willingness of staff to speak up against actions that adversely affect patient care. Error feedback was whether feedback was given after an error report and whether staff discussed ways to prevent recurrence of errors. Nonpunitive response to errors reflected whether the person who committed an error was punished. Management support was the prioritization and interest hospital management placed on safety. Staffing levels conveyed whether there was enough staff to appropriately handle patient care. Teamwork across units examined the coordination of patient care from one unit to another. Error reporting was derived from respondents’ perceptions about how often an error was reported. Questions about error reporting focused on near miss errors, errors that were made but did not result in harm to the patient. Successful handoffs were based on how well patient information was transferred. The focus was on transfers of patients to different units and the effect of shift changes on patient information transfer.
Information from 11 survey dimensions described immediately above formed our primary independent variables; each dimension was considered a unique variable. The independent variables of interest reflected perceptions of hospital staff about the factors of safety culture. We treated all primary independent variables as continuous, and they were composed of three to four related questions. The three to four questions that comprise each dimension are psychometrically reliable, with Cronbach alpha scores between .63 and .84 (Agency for Healthcare Research and Quality, 2014). The HSOPS survey used a 5-point Likert scale. We obtained the percent positive scores for each of the three to four related questions that comprise each organizational safety culture factor and averaged the scores for the questions within each safety culture factor. We used those scores to represent the variable value. We used the percent positive score as the variable value instead of the 5-point Likert scale mean primarily for interpretability reasons. Other researchers that studied safety culture used a similar technique (Pronovost et al., 2008; Singer, Lin, Falwell, Gaba, & Baker, 2009). Possible values for each variable ranged from 0 to 100.
We included the following hospital control variables: bed size, teaching, rural, for-profit, government, number of central line days, cohort number, and baseline CLABSI rate. Bed size was a continuous variable based on the number of hospital beds. The cohort number ranged from 1 to 6 and indicated when the hospital unit started the On the CUSP: Stop BSI program. Teaching, rural, for-profit, and government were binary variables. We also used the number of central line days to reflect the different duration of infection exposure for each hospital. By including the baseline CLABSI rate, we controlled for the number of CLABSIs per 1,000 central line days for the eight quarters preceding CUSP implementation.
We used different dependent variables depending on the two types of analysis conducted. In the logistic regression, the binary dependent variable was whether a hospital had zero infections in the postperiod of up to six quarters. Previous studies indicated that it is both optimal and feasible to achieve zero CLABSIs following CUSP implementation (Sexton et al., 2011). In our analysis of CLABSI rates, we used the number of CLABSIs in the postimplementation period as our dependent variable.
We ran two types of regression analyses: logistic regression and negative binomial regression. We used logistic regression to regress whether a hospital unit had zero infections in the postintervention period on the baseline factors. We ran separate regressions for each organizational factor to obtain odds ratios (ORs) for each factor, adjusting for the control variables. For our analysis that looked at the total number of CLABSIs, we used negative binomial regression. This type of analysis is commonly used in count data when there is an excess number of zeros in the dependent variable. We conducted power analysis and our sample size is sufficient for power of 90%.
All statistical analyses were performed using Stata: Release 12 software (StataCorp LP, College Station, TX).
Our sample distribution of hospitals was for-profit (11%), government (15%), teaching (46%), rural (25%), and with varying bed sizes. Eighty-two percent of hospitals sampled had a bed size of 100 or more. The attitude of hospital staff toward factors of safety culture was generally positive, as shown in Table 1. The average percentage of positive responses was greater than 50% for 9 of the 11 variables, although all variable averages were well below the 100% maximum. Table 2 also shows the correlations among the safety culture variables and indicates that there is a high level of correlation between some factors of safety culture.
Effect of CUSP Intervention on CLABSI Rates
Across hospital units in our study, the introduction of CUSP was followed by a major decline in CLABSI rates, as seen in Figure 1. On average, hospital units reported a significant reduction in CLABSI rates from 1.95 per 1,000 central line days at baseline to 1.04 per 1,000 central line days (p < .001) in the sixth quarter. Average CLABSI rates were fairly constant from Quarters 2 through 6, ranging from an average of 0.96 to 1.20 CLABSIs per 1,000 central line days for each quarter.
Median CLABSI rates per 1,000 central line days at baseline were 1.48, 1.16, 1.13, 1.19, 1.16, and 0.83 for Cohorts 1–6, respectively. In each of Cohorts 2–6, the participating units achieved a median score of zero CLABSIs beginning in the first quarter and sustained that median rate of zero for the remainder of the six quarters. Cohort 1 achieved a median of zero in Quarter 2 and sustained that rate for the remainder of the study.
Results of the negative binomial regression of the factors of safety culture on CLABSI rates include three variables of statistical significance and demonstrate support for Hypothesis 1, that units with stronger factors of safety culture prior to CUSP implementation have a lower number of infections after implementation. An increase in perceptions of the baseline factors of safety culture was associated with a decrease in the number of CLABSIs in the postperiod. Table 3 reflects the impact of a one percent change in the safety culture factors to the incident rate ratio (IRR) of the expected number of CLABSIs. Three of these associations were statistically significant: the factors organizational learning (IRR = 0.992, p < .05), communication openness (IRR = 0.993, p < .05), and teamwork across units (IRR = 0.991, p < .01).
Predictors of Zero Infections Following CUSP Implementation
We found support for Hypothesis 2, that units with stronger factors of safety culture prior to CUSP implementation have greater odds of zero infections after implementation. Of the 649 hospital units under study, 207 had zero infections in the six quarters following CUSP implementation. Controlling for baseline CLABSI rate and multiple hospital structure variables, an increase in perceptions of each baseline organizational factor of safety culture was associated with an increase in the odds of zero infections in the postperiod. Table 4 reflects the impact of a one percent change to the percent positive score for each dimension and demonstrates support for Hypothesis 2. Two of these associations were statistically significant: the factors communication openness (OR = 1.014, p < .05) and staffing levels (OR = 1.015, p < .05).
Considering that there are more than 30,000 preventable deaths from CLABSIs each year, hospitals need to eliminate these infections. A significant reduction in CLABSIs is a feasible goal. CUSP is a program that has been associated with reduced CLABSI rates, yet is a program that requires leadership commitment of financial and human resources. Other evidence from the literature indicates that more frequent executive visits are associated with better safety culture (Frankel et al., 2008). However, the time commitment incurred by the executive and other staff members on a CUSP team is not without cost. Partly as a result, the historic rollout of CUSP at hospitals has been slow and staggered. Our findings suggest that hospital units with higher levels of safety culture can be targeted for early implementation of the CUSP program in order to achieve zero CLABSIs. Specifically, units that perceive excellence in communication openness, organizational learning, staffing, and teamwork across units should be the focus of early implementation of CUSP if CUSP cannot be immediately implemented in the entire hospital. These factors have the greatest association with reduced CLABSI rates after implementation of the CUSP program.
We identified that hospital units with a stronger perceived safety culture prior to CUSP implementation are better prepared to fully employ the CUSP model. Safety culture takes time to change, because it is a combination of individual and group attitudes, values, perceptions, and patterns of behavior (Elder, Brungs, Nagy, Kudel, & Render, 2008). Although CUSP is designed to improve safety culture, it appears that the model’s impact on CLABSI rates is greater in units that already have a strong perceived safety culture.
The CUSP model requires frequent interdisciplinary meetings (Pronovost & Vohr, 2010), so it is not surprising that sufficient staffing and open communications are a prerequisite of successful CUSP implementation. Having enough people on staff a unit increases the ability for those staff to attend meetings. Frequent meetings that involve interdisciplinary staff are a necessary component to the success of the program. Open communications that include the willingness of staff to speak up in those meetings even when others have more authority is critical. Leader inclusiveness predicts engagement in quality improvement work (Nembhard & Edmondson, 2006). Without that engagement, a quality improvement program such as CUSP is not as successful when it is introduced.
It is also important to have a strong culture around organizational learning or continuous improvement. Organizational learning encompasses learning from mistakes and an active desire to improve safety. The CUSP model asks where the next patient will be harmed. To get at that information, the unit must have a frank discussion of where past mistakes have occurred and have ability and desire to learn from those mistakes.
Similarly, teamwork is a critical component of the CUSP model. Although CUSP is a model designed to improve teamwork, our study indicates that units that perceive there is a strong foundation of teamwork in place prior to CUSP implementation have increased odds of reducing CLABSIs. The importance of teamwork across units may result from the flexible and joint cooperation and creative problem-solving that occur. Although the CUSP team is formed at the unit level, that unit still has interactions with other units that play a critical role in patient safety and the elimination of CLABSIs.
If hospitals choose to improve teamwork, communication openness, or learning activities prior to initiation of CUSP, there are evidence-based methods to do so. Examples for teamwork include teamwork training, use of team huddles, interdisciplinary rounds, and the introduction of focus groups designed to identify teamwork issues (Farley, Sorbero, Lovejoy, & Salisbury, 2010; Kalisch, Curley, & Stefanov, 2007; O’Leary et al., 2010). To facilitate open communications, organizations have eliminated hierarchies and increased employee identification with the workgroup (Tangirala & Ramanujam, 2008). To improve organizational learning, safety rounds have proven effective (Campbell & Thompson, 2007).
Limitations and Suggestions for Future Research
One limitation of this study may be the generalizability of the findings. We obtained factors of safety culture from survey data from hospital units that volunteered to participate in the On the CUSP: Stop BSI program. In comparison to all American Hospital Association’s hospitals, our study units were from hospitals generally larger than the national average and with a teaching focus. Although it is possible that our results do not apply to all hospitals, the large number of participating hospitals and the geographic and size distribution of these facilities lead us to believe that our results may be generalized to all U.S. hospitals affected by CLABSIs.
A second limitation is the missing data. We compared the attributes of the hospitals with missing data to those used in our analysis and found that the hospitals with missing data were less likely to be teaching hospitals (33% vs. 46%, p ≤ .001) and were smaller (average bed size of 234 vs. 281, p = .003). Neither variable was a statistically significant predictor of the dependent variables of interest.
A third limitation is that we were unable to assess the degree in which the CUSP model was followed and whether other safety interventions were simultaneously being conducted. It is possible that both were sources of outcome variation, but we were unable to test for either, given our data.
Future research should examine the CUSP implementation process. As more hospital units implement CUSP, it is important to gain insight how the best performers managed staffing and teamwork. It may be helpful to understand the optimal staffing level for success under the CUSP model. To overcome one of our study’s limitations, future research can also ascertain the degree to which the CUSP model is followed and how that impacts measurable patient safety outcomes. A small, controlled mixed methods study may be beneficial.
Findings from our study can be expanded to examine other healthcare-associated infections to see if the results are similar. In fact, an initiative that modifies the CUSP model to focus on prevention of catheter-associated urinary tract infections (CAUTIs) is underway. It is worth studying whether higher baseline levels of factors of safety culture similarly increases the odds of zero CAUTIs, following CUSP implementation. Future research should also examine how CUSP affects safety culture and whether the change to safety culture impacts CLABSI or CAUTI rates.
CLABSIs are an unfortunate part of our health system but can be effectively eliminated through the proper focus and commitment. It is important for managers to understand that zero CLABSIs is an attainable goal for hospital units.
Given that CUSP is a relatively recent model, there is an opportunity to expand the program to hospital units worldwide that will achieve optimal success. In a large sample, our study confirmed that it is possible to achieve zero CLABSIs in a short period of time, and we found that the largest drop occurred within the first 6 months. This suggests that the CUSP program can be an effective tool to rapidly lower CLABSI rates.
Hospitals that want to eliminate CLABSIs through the introduction of the CUSP program should target implementation toward units with the highest factors of safety culture. Specifically, hospitals should prioritize introduction of CUSP to units that have better perceived staffing, communication openness, organizational learning, and teamwork across units. These factors of safety culture are associated with better CLABSI rates after CUSP implementation. In an ideal state with unlimited resources, every hospital would implement CUSP, reduce CLABSIs, and gain other patient safety benefits. However, hospitals have competing priorities and need to prioritize resource expenditure. A targeted implementation of CUSP affords hospitals the greatest opportunity to efficiently reduce or eliminate CLABSIs, thus conserving resources in a time of competing demands on them. Hospital units with a poorer safety culture should take steps to improve that culture and then implement CUSP.
We would like to thank the Agency for Healthcare Research and Quality for access to the data used in this study.
Campbell D. A. Jr., Thompson M. (2007). Patient safety rounds: Description of an inexpensive but important strategy to improve the safety culture. American Journal of Medical Quality
, 22(1), 26–33. doi:10.1177/1062860606295619
Elder N. C., Brungs S. M., Nagy M., Kudel I., Render M. L. (2008). Intensive care unit nurses’ perceptions of safety after a highly specific safety intervention. Quality & Safety in Health Care
, 17(1), 25–30.
Farley D. O., Sorbero M. E., Lovejoy S. L., Salisbury M. (2010). Achieving strong teamwork practices in hospital labor and delivery units
. Santa Monica, CA: Center For Military Health Policy Research, RAND Corp.
Frankel A., Grillo S. P., Pittman M., Thomas E. J., Horowitz L., Page M., Sexton B. (2008). Revealing and resolving patient safety defects: The impact of leadership WalkRounds on frontline caregiver assessments of patient safety. Health Services Research
, 43(6), 2050–2066. doi:10.1111/j.1475-6773.2008.00878.x
Gaba D., Cooper J. (2000). Landmark report published on patient safety. Journal of Clinical Monitoring and Computing
, 16(3), 231–232.
Kalisch B. J., Curley M., Stefanov S. (2007). An intervention to enhance nursing staff teamwork and engagement. The Journal of Nursing Administration
, 37(2), 77–84.
Lipitz-Snyderman A., Needham D. M., Colantuoni E., Goeschel C. A., Marsteller J. A., Thompson D. A., Pronovost P. J. (2011). The ability of intensive care units to maintain zero central line-associated bloodstream infections. Archives of Internal Medicine
Nembhard I. M., Edmondson A. C. (2006). Making it safe: The effects of leader inclusiveness and professional status on psychological safety and improvement efforts in health care teams. Journal of Organizational Behavior
, 27(7), 941–966.
O’Leary K. J., Ritter C. D., Wheeler H., Szekendi M. K., Brinton T. S., Williams M. V. (2010). Teamwork on inpatient medical units: Assessing attitudes and barriers. Quality and Safety in Health Care
, 19(2), 117–121.
Pronovost P., Needham D., Berenholtz S., Sinopoli D., Chu H., Cosgrove S., Goeschel C. (2006). An intervention to decrease catheter-related bloodstream infections in the ICU. The New England Journal of Medicine
, 355(26), 2725–2732. doi:10.1056/NEJMoa061115
Pronovost P., Vohr E. (2010). Safe patients, smart hospitals: how one doctor's checklist can help us change health care from the inside out
. New York, NY: Hudson Street Press.
Pronovost P. J., Berenholtz S. M., Goeschel C., Thom I., Watson S. R., Holzmueller C. G., Sexton J. B. (2008). Improving patient safety in intensive care units in Michigan. Journal of Critical Care
, 23(2), 207–221. doi:10.1016/j.jcrc.2007.09.002
Pronovost P. J., Watson S. R., Goeschel C. A., Hyzy R. C., Berenholtz S. M. (2015). Sustaining reductions in central line-associated bloodstream infections in Michigan intensive care units: A 10-year analysis. American Journal of Medical Quality
, 31, 197–202. doi:10.1177/1062860614568647
Rosenthal V. D., Maki D. G., Salomao R., Moreno C. A., Mehta Y., Higuera F., … International Nosocomial Infection Control Consortium. (2006). Device-associated nosocomial infections in 55 intensive care units of 8 developing countries. Annals of Internal Medicine
Sawyer M., Weeks K., Goeschel C. A., Thompson D. A., Berenholtz S. M., Marsteller J. A., Pronovost P. J. (2010). Using evidence, rigorous measurement, and collaboration to eliminate central catheter-associated bloodstream infections. Critical Care Medicine
, 38, S292–S298.
Sexton J. B., Berenholtz S. M., Goeschel C. A., Watson S. R., Holzmueller C. G., Thompson D. A., Pronovost P. J. (2011). Assessing and improving safety climate in a large cohort of intensive care units. Critical Care Medicine
, 39(5), 934–939.
Sexton J. B., Thomas E. J., Helmreich R. L. (2000). Error, stress, and teamwork in medicine and aviation: Cross sectional surveys. BMJ
, 320(7237), 745–749.
Singer S., Lin S., Falwell A., Gaba D., Baker L. (2009). Relationship of safety climate and safety performance in hospitals. Health Services Research
, 44(2 Pt. 1), 399–421. doi:10.1111/j.1475-6773.2008.00918.x
Sorra J. S., Dyer N. (2010). Multilevel psychometric properties of the AHRQ hospital survey on patient safety culture. BMC Health Services Research
, 10, 199. doi:10.1186/1472-6963-10-199
Tangirala S., Ramanujam R. (2008). Employee silence on critical work issues: The cross level effects of procedural justice climate. Personnel Psychology
, 61(1), 37–68. doi:10.1111/j.1744-6570.2008.00105.x
Tucker A. L., Edmondson A. C. (2003). Why hospitals don't learn from failures. California Management Review
, 45(2), 55–72.
Vigorito M. C., McNicoll L., Adams L., Sexton B. (2011). Improving safety culture results in Rhode Island ICUs: Lessons learned from the development of action-oriented plans. Joint Commission Journal on Quality and Patient Safety
, 37(11), 509–514.
Weingart S. N., Ship A. N., Aronson M. D. (2000). Confidential clinician-reported surveillance of adverse events among medical inpatients. Journal of General Internal Medicine
15(7), 470–477. http://doi.org/10.1046/j.1525-1497.2000.06269.x
Zimlichman E., Henderson D., Tamir O., Franz C., Song P., Yamin C. K., Bates D. W. (2013). Health care-associated infections: A meta-analysis of costs and financial impact on the US health care system. JAMA Internal Medicine
Keywords:Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved
culture; infections; patient safety; quality improvement