It is estimated that nearly 400,000 deaths per year are attributable to medical errors, making this the third leading cause of death in the United States.1 Although our healthcare delivery system has made some progress in reducing patient harm in our healthcare facilities, and much work has been carried out to create a safety culture that aims to reduce errors, there is still much to be done.2 It has been suggested that the safety culture of an organization establishes the behavioral norms that promote or fail to embrace the notion that a reduction in patient harm is, indeed, possible.3 Previous research has identified important aspects of a hospital's organizational culture that are strongly correlated with a robust safety climate. In general, organizational strategies that promote a “group orientation” and reduced hierarchy, including the use of multidisciplinary team decision-making, the use of continuous quality improvement tools, and fair and consistently applied human resource disciplinary polices help promote a strong patient safety culture.4 Leaders in healthcare recognize the need for moving from a culture of blame to a culture of safety, but the real challenge that they face is that cultures are very slow to change. Sadly, the prototypical response in healthcare organizations to performance problems, “near misses,” or other failures to achieve desired outcomes has been silence.5 The alternative to a “Blame Culture” is a “Just Culture”, which is characterized by an environment supportive of free and open dialog to facilitate and reward safer practices.6 In the prototypical “Just Culture” model, members feel empowered to question existing practices, to freely express concerns or dissent, and to admit mistakes without fear of ridicule or unfair discipline.7
There is mounting evidence to suggest that organizational culture is strongly correlated with the evolution of a patient safety climate and that frontline staff continue to report concerns about a culture of blame that has emerged in our nation's hospitals.8 Several authors have asserted that a blame culture has evolved from the hierarchical “top-down” management systems currently prevalent in health care and that for health care organizations to succeed in reducing patient harm, they must move to a management system that encourages the blame-free reporting of errors.6,9 There are additional concerns about how to strike a proper balance between a blame-free approach to reporting and the need for staff accountability.10 Although the “Just Culture” concept offers the potential for creating an environment that supports blame-free reporting and provides a format to treat individuals involved in an incident fairly, it is unclear whether all organizations are ready for this methodology.
This study aims to answer the following questions: (1) Does the introduction of formal training in the “Just Culture” methodology on a medical-surgical unit change the perception of the safety culture? (2) Is there an association between the organizational culture and the readiness of an organization to benefit from “Just Culture” training? (3) Is there an association between job classification and the perception of the safety culture?
The author will present two case studies of the effects of “Just Culture” training in promoting a “Just Culture” in two hospitals of similar size and mission, but having marked differences regarding performance on Centers for Medicare & Medicaid Services quality metrics and work environment.
This study aimed to determine whether there was an association between the organizational culture of a hospital and the hospital's ability to benefit from a training program in “Just Culture”. The study further aimed to determine whether there was an association between job classification and the perception of the safety culture in the hospital.
The “Just Culture” Assessment Tool (JCAT) and the Competing Values Framework (CVF) surveys were distributed to the nurse managers, registered nurses, certified nurse assistants, pharmacists, pharmacy technicians, and unit secretaries on the patient care units that were studied. Surveys were also distributed to the chief executive officer, the chief operating officer, and the chief nursing officer of the two hospitals where the study was conducted. Both surveys were also administered to physician hospitalists on the unit. The total number of participants sampled was 172.
The two hospitals in which the study was performed were located in suburban areas of Riverside County in Southern California. The study was performed on medical-surgical units at two hospitals, which were comparable on many metrics. Both hospitals had a total bed size of approximately 200, were for-profit entities, and drew from a catchment area that had a similar payor mix and socioeconomic makeup. Hospital 1 was acquired from another hospital system and has had challenges with employee retention, quality outcomes, and patient satisfaction. Hospital 2 was newly opened in 2015 and has performed at or above the top decile in patient care outcomes as reported to Centers for Medicare & Medicaid Services and has had patient satisfaction scores in the top quartile. This hospital was named as one of the 10 best places to be used in the county.11 Although both hospitals have conducted patient safety campaigns and quality campaigns in the past, neither hospital had offered a formal “Just Culture” training before.
All study participants received an e-mail from the principal investigator indicating the background of the project and were asked to return the JCAT and CVF surveys within 10 days. A list of participants who were willing to complete the surveys (along with their e-mail addresses) was obtained from the nurse manager, the pharmacy director, and the lead hospitalist. Completion of the surveys by participants was completely voluntary. In addition, participants were reassured that the data would be collected anonymously (the only identifying information would include the job classification, work site, years at the organization, and years employed in the profession). Immediately after sending out the study invitation e-mail, the nurse manager, pharmacy director, and lead hospitalist received training in “Just Culture” during a 60-minute presentation adapted from the “Just Culture” Training Manual for Healthcare Managers.12 A 45-minute PowerPoint presentation (available on request from the author) was delivered by the principal investigator. All presentations were identical, with a 15-minute period devoted to questions and answers. The nurse manager, pharmacy director, and lead hospitalists were given copies of the “Just Culture” Algorithm v3.012 (Fig. 1) and were encouraged to use the algorithm to guide meetings regarding adverse events, root cause analyses, opportunities to counsel employees regarding errors, and practitioner peer review. Six weeks after the initial training, another e-mail was sent to the same group by the principal investigator requesting completion of both study instruments. Survey requests and responses were completed electronically using Research Electronic Data Capture (REDCap), which is a secure web application used for building and managing online surveys and data collection instruments.13
Study participants were surveyed using two assessment tools. The first was the JCAT, which was designed to measure whether or not individuals believed that they would receive fair and just treatment when involved in an adverse event (i.e., if a “Just Culture” existed in an organization).2 The JCAT is a 27-item survey that uses a seven-point Likert scale. It aims to provide an assessment across the following six integral dimensions: feedback and communication (items 1, 2, 3); openness of communication (items 4, 5, 6, 7, 8); balance (items 9, 10, 11, 12, 13); the quality of the event reporting process (items 14, 15, 16, 17, 18); support of continuous quality improvement (items 19, 20, 21, 22); and the establishment of trust (items 23, 24, 25, 26, 27). In this study, the JCAT was used in its original form, without alteration.
The organizational culture of the organization was assessed using the CVF. This tool characterized the relationship between an organization and its environment along two perpendicular dimensions: structure (flexibility, control) and focus (internal, external). By classifying organizations on these two dimensions, an organization's culture was characterized as group, entrepreneurial, hierarchical, or rational.14 The version of the CVF that was used in this study had the respondent score each item on a five-point Likert scale, measuring agreement or disagreement on how well the statement described their facility. The reported Cronbach's α values for this version of the CVF were reported to have ranged from 0.68 to 0.85.15
The responses to each group of four statements on the organizational culture survey were reported as mean and standard deviation for each hospital. Responses were also subgrouped based on the job classification of the respondent. Likewise, responses to the questions on the JCAT were analyzed by hospital. “Problematic responses” included answering “strongly agree,” “agree,” “strongly disagree,” or “disagree” to items, depending on whether or not the question was asked in the affirmative or the negative. Data were reported as the mean and standard deviation of the percentage of problematic responses (PPRs) by hospital. Data were further subgrouped based on the job classification of the respondent. In addition, the PPRs was individually assessed along each of the six integral dimensions that the survey aimed to assess. The data set from the surveys administered before the training intervention and after the training intervention were assessed to determine whether the assessment of organizational culture changed as a consequence of the “Just Culture” training, whether there was an association between the percentage problematic response on the JCAT and the archetypal culture that best described the hospital, and, finally, whether the degree of change (if any) in the PPR on the JCAT was associated with the archetypal culture of the hospital. At baseline, the results for each survey tool (by hospital and job classification) were assessed using an independent t test. After the training intervention, the paired t test was used to determine whether there was a significant change in the responses (by dimension) to the JCAT based on the archetypal organizational culture of the organization. The data management and analysis were performed using International Business Machines Statistical Package for the Social Sciences (IBM SPSS) 25. The level of significance (α) was set at 0.05. All statistical tests were two-sided. The changes were considered statistically significant if P values from the statistical tests were less than 0.05. The Z score was used to determine whether there was a significant difference in the percentage change from the pretraining to the posttraining value. The null hypothesis was rejected if Z was greater than 1.96 or less than −1.96.
The responses to both survey tools were anonymous. There were no patient-specific questions on the tools. The study was reviewed by the University of Mississippi Medical Center Institutional Review Board and was deemed to be “exempt.”
Response to the Survey Tools
In the preintervention phase, 172 survey requests were sent out by a link in REDCap. Eighty-five requests were sent to hospital 1, whereas 87 requests were sent to hospital 2. Although the number of years in the profession was similar for respondents from hospital 1 and hospital 2, the years at the facility were much higher for hospital 1 than hospital 2. This was expected as hospital 2 was opened only 4 years ago, whereas hospital 1 has been in operation for more than 40 years. Because both survey tools were linked, 100% of respondents completed both surveys, and all respondents answered all of the questions on the surveys. The distribution of job classifications comprising the study cohorts at both hospitals was similar, and there was no significant change in the response by job classification when comparing the pretraining and posttraining survey waves.
Preintervention Survey Data
Before the “Just Culture” training intervention, the overall PPR rate was significantly higher (P < 0.001) for respondents from hospital 1 (58.58%) as compared with hospital 2 (24.51%). The difference in the PPR rate was largest for the domain of “Establishment of Trust” (72.58% for hospital 1 as compared with 26.62% for hospital 2). The PPR rates were significantly higher (P < 0.001) across all six survey domains for hospital 1, as compared with hospital 2. In addition, at baseline, there was a significant difference in the PPR rates by job classification across all six domains. Overall, registered nurses had the highest rate of problematic responses, whereas hospital administrators had the lowest rate of problematic responses. In addition, at baseline, survey responses on the CVF tool indicated that hospital 1 had significantly higher scores on the questions that indicated a hierarchical and rational culture and a significantly lower score on the questions that indicated a group culture (Fig. 2). Before the training intervention, there was no significant difference in the characterization of the hospital culture by respondents from different job classifications.
Postintervention Survey Data
The same survey tools were administered 6 weeks after completion of “Just Culture” training. For hospital 1, there was essentially no change in the composite PPR rates on the JCAT, and, likewise, there was essentially no change in the PPR rates across each of the six domains. The PPR rates went up slightly after the “Just Culture” training in the domains of “Feedback and Communication,” “Openness of Communication,” and “Balance.” For hospital 2, there was a significant reduction in the PPR rates across all six domains (Z > 5.00 for all domains). The percentage of change in the PPR rates by hospital, after the completion of “Just Culture” training is shown in Table 1. The PPR rates significantly improved after “Just Culture” training across all job classifications.
After “Just Culture” training, survey responses on the CVF survey tool indicated that there was an overall increase in the Group Culture Score for both hospitals; however, the magnitude of change was significantly higher for hospital 2 as compared with hospital 1 (Z = 2.16). In addition, there was an overall increase in the Entrepreneurial Culture Score for both hospitals; however, the increase was not significantly different for hospital 2 as compared with hospital 1 (Z = 1.21). There was a decrease in the Rational Culture Score for both hospitals. The magnitude of change was significantly higher for hospital 2 as compared with hospital 1 (Z = 4.26). For hospital 2, there was a reduction in the Hierarchical Culture Score, whereas there was a small increase in the Hierarchical Culture Score for hospital 1. The difference in the percentage change was statistically significant (Z = 3.71). The percentage change in the CVF assessment, by hospital, after the completion of “Just Culture” training is shown in Table 2.
Despite having far more potential for improvement, the lower functioning hospital (hospital 1) did not show a greater improvement in the assessment of the existence of a “Just Culture”, but instead had a lower percentage of improvement than the higher functioning hospital (hospital 2). These results support the contention that to institute a Culture of Safety in a lower functioning hospital, a bundled approach of interventions is needed.
The CVF survey confirmed that hospital 1 had a much more Hierarchical Culture, whereas hospital 2 had a Group-oriented Culture. There was essentially no change in the assessment of the existence of a “Just Culture” after training for hospital 1, whereas there was a significant improvement in the assessment of a “Just Culture” for hospital 2.
Implications of the Findings
The direct and indirect costs of implementing a “Just Culture” training program are significant, and the published literature is inconsistent in demonstrating a tangible benefit to the organization after the training is completed. This study supports the contention that the underlying organizational culture plays a critical role in determining whether or not a “Just Culture” training program will be effective. It would be imperative to address the organizational culture before attempting to implement a “Just Culture” program in a healthcare organization. An organization with an archetypal Group Culture would seem to benefit more from a “Just Culture” program than an organization with a Hierarchical Culture.
Perhaps most importantly, future research needs to address the clinical effects that the introduction of a “Just Culture” has on patient outcomes and the patient experience. Very little has been published in this area, because the most relevant outcomes are hard to define, are sometimes difficult to quantify, and always take a considerable length of time to measure. There remains a lingering concern that a culture that encourages blame-free reporting may also lead to unintended patient harm if caregivers feel that this makes them less accountable for personal decisions. What is certain is that patient harm still occurs at an unacceptably high rate in our hospitals and that the efforts, to date, to reduce avoidable errors have not met the mandates set by the Institute of Medicine Report16 nearly two decades ago.
Limitations of the Study
A response rate of approximately 73% for both survey waves is high and may not be achievable in future studies. The fact that the author was in a position of organizational leadership could be a source of bias in the study. The leadership position of the author, despite the guaranty of anonymity, could act to decrease the likelihood of providing a “problematic response”; however, the fact that the problematic responses were clearly skewed toward one hospital over the other suggests that this was not a significant factor affecting the candor of responses. One could argue that merely focusing on the Culture of Safety through any type of intervention conducted by a hospital leader (the placebo effect) could account for some of the benefit that was demonstrated after the training intervention. Arguing against this is the fact that the same presentation was given to both organizations, and only one appeared to benefit from it. The fact that organizations can demonstrate an improvement in performance simply based on being observed (the Hawthorn effect) needs to be considered. However, such bias should exist in both hospitals and, in theory, would be expected to be even greater in the poorly performing hospital. This was not borne out by the results of the study. The study was not powered to determine whether the magnitude of change was significantly affected by job classification.
Although the improvement in responses on the JCAT after “Just Culture” training would seem to add support to the use of this tool to measure the existence of a “Just Culture” in an organization, this is only the second study that has reported on its use, and it remains to be determined how well this tool measures “Just Culture”. In addition, although the hospitals that participated were quite similar in case mix index, geographical location, size, and governance, it is uncertain if another baseline variable that was not accounted for could have biased the results. Finally, a six-week period is a relatively short period in which to determine an effect of a training program on something as complex as the existence of a “Just Culture” and the assessment of organizational culture (which is typically very slow to change). However, in a study which produced a negative result, the sustainability of results is not relevant, and it would be unlikely that the lower performing hospital would improve its performance if the period for the study was extended.
Recommendations for Future Research
It would be important to determine what the optimal length of time to determine the ultimate effect of “Just Culture” training on the perception of the existence of a “Just Culture”, and studies with a longer interval between training and reassessment of the culture should be performed. There also has been little if anything published about the sustainability of “Just Culture” after training and if there is a need for ongoing training and how often that training needs to be performed. Particularly, as it pertains to this investigation, it would be important to determine whether the improvements in safety climate after a very brief intervention and measurement period are sustained. Because there are many tools that have been developed to measure the patient safety culture, further research needs to be carried out to determine which combination of tools provides the most accurate representation of the existence of a “Just Culture”. Specifically, future research efforts should focus on the correlation between the perception of a “Just Culture” and employee engagement (as commonly measured by staff engagement surveys), as efforts aimed at improving the perception of the Culture of Safety may also be effective in improving staff engagement. Because the population sampled was skewed toward nurses, it would be valuable to sample a population with a higher number of non-nurses to determine whether “Just Culture” training is truly effective across disciplines and whether training in “Just Culture” effectively improves the perception of the culture of safety and the willingness to report errors or “near misses” across all job classifications.
It was determined that there was significantly greater benefit from a “Just Culture” training intervention in a hospital with a highly developed Group Culture as opposed to a comparable hospital with a more Hierarchical Culture. Given the cost and effort required to conduct “Just Culture” training (or any significant institutional program that address the culture of safety) in an organization, it seems to be important to formally address the organizational culture before the implementation of this type of training. In organizations with a poorly developed Group Culture or a strongly Hierarchical Culture, it would seem that efforts to address the organizational culture should be undertaken before implementation of a large-scale patient safety initiative.