The challenges of achieving safer health care have become increasingly recognized during the last decades, after several reports on the frequency of preventable adverse events.1–3 The importance of developing a culture of safety has been emphasized because investigations into failures in health care have identified weak patient safety culture as a common contributing factor to adverse events.4 Safety culture is an aspect of organizational culture that refers to how safety is viewed and treated by the members in organizations.5
It has been argued that safety culture cannot be assessed quantitatively and that patient safety climate should be evaluated instead as a quantifiable surrogate measure of safety culture.6 Patient safety climate can be considered as a snapshot of the patient safety culture, representing the surface features of the underlying culture.7 However, the concepts “climate” and “culture” are often applied interchangeably in patient safety research. Instruments developed to assess patient safety culture most likely capture the climate rather than the culture. In this study, we consistently use the term “culture.”
Although instruments to measure patient safety culture vary with regard to precise content, the concept of patient safety culture is usually broken down into dimensions such as management commitment to safety, safety systems, work pressure, communication, teamwork, nonpunitive response to errors, and leadership.8 The two most widely applied patient safety culture instruments in hospital settings are the Safety Attitude Questionnaire and the Hospital Survey on Patient Safety Culture.8 According to Sorra and Battles (2014), Hospital Survey on Patient Safety Culture has been adopted for use in 45 countries,8 although a smaller number of countries have published studies on the psychometric properties,8 and it has been translated into more than 20 languages, including Swedish.9
Patient safety culture surveys such as HSPSC are widely used not only in research but also in practice, where they are seen as an important management tool.10 Surveys can be used to raise staff awareness about patient safety, assess the current status of and trends in the patient safety culture, and identify strengths and areas for improvement. The patient safety culture has been surveyed regularly in Sweden since 2011 using the Swedish version of HSPSC. The county councils, which are responsible for providing health care in Sweden, have received payment for undertaking these surveys as part of a government-supported financial incentive plan to improve patient safety.11
Using the HSPSC, this study investigates the patient safety culture in all hospitals in Sweden. Earlier studies based on HSPSC have focused predominantly on assessing the patient safety culture in specific hospitals or care units12–15 and/or among various professional groups.16,17 Study populations have generally been small. Managers' perceptions are rarely investigated despite that management commitment to safety and leadership are considered important for achieving a favorable patient safety culture. We have only been able to identify 2 previous national-level studies18,19 that have investigated the patient safety culture across many different hospitals or primary care units.
The aim of this study is to investigate the patient safety culture in all Swedish hospitals using the HSPSC. More specifically, the objectives are to compare the patient safety culture among managers, physicians, registered nurses, and enrolled nurses and to identify factors associated with high overall patient safety.
Design and Setting
This national study has a correlational design. It is based on cross-sectional data using the Swedish adaptation of the HSPSC and was conducted in Sweden during 2012 and 2014. All 21 county councils in Sweden, which are all tax funded, participated in the survey. Ethics approval was obtained from the Ethical Review Board in Stockholm (no. 2010/820-31/5).
Participants and Data Collection
A 4-year government-supported financial incentive plan concerning patient safety in health care was carried out between 2011 and 2014 under the strategic leadership of the Swedish Association of Local Authorities and Regions. Patient safety culture was an important aspect of the plan. The research team behind this study was granted access to a national database administered by the Swedish Association of Local Authorities and Regions. The data in this study were obtained from the surveys conducted in all hospitals in Sweden from November 2012 to September 2014.
The survey encompassed all health care practitioners employed in Swedish health care. Questionnaires were answered individually using a digital system whereby no personal identification was possible. The participants could choose to answer during work time or from home.
A total of 182,413 questionnaires were administered, of which 111,478 were answered, yielding a response rate of 64.7%. For the present study, the following 3 work areas were selected: general wards, emergency care, and psychiatry care (n = 75,282). The following 4 staff positions in these work areas were selected: managers, registered nurses, enrolled nurses, and physicians (n = 58,671). Questionnaires with any data missing were excluded, and the final sample consisted of 23,781 participants.
The HSPSC questionnaire20 was translated into Swedish in 2007 and has been validated in a study.9 The questionnaire was translated by a professional independent translator and then controlled by 4 health care and patient safety experts. It was then back-translated by another translator. Discrepancies between versions were solved by collaboration between the experts and translator.9 The Swedish version differs slightly from the original HSPSC, featuring 7 additional items. The additional items are a further “outcome” question, which concerns the number of reported risks, 4 questions about information and support to patients and family who have had an adverse event, and 2 questions regarding information and support to staff who have been involved in an adverse event.9 However, the present study included only variables from the original HSPSC.
The HSPSC includes a number of background demographic variables. The following 6 background variables were used in the present study: sex, age, professional experience (years), work time (hours per week), work area (general wards, emergency care, and psychiatry care), and staff position (manager, registered nurses, enrolled nurses, and physicians).
The HSPSC also includes 12 patient safety culture dimensions (Table 1), encompassing a total of 42 items, with 3 or 4 items per dimension.20 All items are based on a 5-point Likert-type response scale of agreement (“strongly disagree” to “strongly agree”) or frequency (“never” to “always”). When needed, items were reversed so that a higher score always indicated a higher value. The items in each dimension were then summed and linear transformed onto a 0 to 100 scale ([raw scale score − lowest possible score]/[possible score range] × 100) (100 = highest value for rated patient safety culture).
All dimensions demonstrated satisfactory internal consistency reliability in the present study, estimated with Cronbach α (0.73–0.86). One exception was “communication openness,” which had a lower Cronbach α (0.68) but still close to the desirable α level of greater than 0.7 (Table 1).
One single item in the HSPSC provides an overall grade of patient safety, with a 5-point Likert-type scale response (“failing” to “excellent”). This item has been advocated by Sorra and Dyer20 as the most reliable outcome variable. The overall patient safety item was used as an outcome variable in the present study, similar to numerous other studies.15,16,18,21
Descriptive statistics were used to present background and study variables. Pearson χ2 test and 1-way analysis of variance were used to compare background characteristics and the reported patient safety culture among the different staff groups. Bonferroni-corrected P values were used as post hoc test in the analysis of variance.
A series of logistic regression analyses were conducted to determine the association between the explanatory variables (6 background variables and 12 dimensions of patient safety culture) and the outcome variable (overall patient safety).
The 12 dimensions were entered in the model as continuous variables. Sex was entered as a dichotomous variable with male sex as the reference category. Age, professional experience, work time, work area, and staff position were entered as dummy variables with the following as reference categories, respectively: 18 to 34 years, 1 to 5 years, less than 20 hours per week, general wards, and managers. The single item about overall patient safety was entered as an outcome variable after it was dichotomized into high (response options “excellent” and “very good”) and low (response options “acceptable,” “fair,” and “failing”) overall patient safety.
In a first step, simple logistic regression analyses were conducted to determine the bivariate association between each explanatory variable and the outcome variable. In a second step, all explanatory variables were entered in the model to determine the multivariate associations between the variables.
The Hosmer-Lemeshow test was used to evaluate the overall goodness of fit of the logistic regression models. All regression models demonstrated a nonsignificance test (P > 0.05), which indicates a good model fit. According to the variance inflation factor, no multicollinearity problems were detected across the explanatory variables in the multivariate logistic model (mean variance inflation factor = 1.87).
All data were analyzed with Stata 14.1 for Windows (StataCorp, College Station, Tex). Overall, P values of less than 0.05 were considered significant.
The sample consisted of 23,781 participants, of which 77% were women and 23% were men. Most participants were registered nurses (51%), followed by enrolled nurses (23%), physicians (17%), and managers (8%). Sixty percent worked in general wards (medicine, surgical, obstetrics, pediatrics, and rehabilitation care). The rest worked in emergency care units, which included emergency care, intensive care units, and operating theater and anesthesiology (25%), or psychiatric care (15%). Furthermore, most participants were between the ages of 35 and 54 years (55%), had professional experience of more than 15 years (54%), and worked 20 to 59 hours per week (96%, Table 2).
Patient Safety Culture Dimensions
The highest rated patient safety culture dimensions were “teamwork within units” (mean, 73.5), “nonpunitive response to error” (mean, 67.2), “supervisor/manager expectations and actions promoting safety” (mean, 66.7), and “communication openness” (mean, 66.3). Lowest rated patient safety culture dimensions were “management support for patient safety” (mean, 47.9) and “staffing” (mean, 51.9).
Patient safety culture was rated significantly differently across the 4 staff positions. Managers scored the highest patient safety culture in all dimensions, whereas physicians scored lowest in all dimensions except for “management support for patient safety,” “communication openness,” “overall perception of patient safety,” and “teamwork across units.” Although there were significant differences between the staff positions, the post hoc tests showed that ratings from registered nurses, enrolled nurses, and physicians were less divergent than those from managers (Table 2, Fig. 1).
Overall Patient Safety
The mean rating for overall patient safety was 3.3 for all participants, but the rating differed significantly between the staff positions (P < 0.001), except enrolled nurses and physicians (which both scored 3.3). Managers scored highest overall patient safety (mean, 3.7), and registered nurses scored the lowest overall patient safety (mean, 3.2; Table 2).
Factors Associated With Overall Patient Safety
The bivariate analysis showed that all 12 dimensions of patient safety culture were significantly associated with overall patient safety (Table 3). A higher level of patient safety culture implied increased probability for high overall patient safety. Higher age, longer professional experience, and being manager were also associated with high overall patient safety. There were also some associations between work area and overall patient safety. The probability of reporting a high level of overall patient safety was greater for respondents working at emergency care units compared with general wards. However, no such association was shown for psychiatric care. Sex and work time were not significantly associated with overall patient safety.
The findings for the multivariate analysis including all explanatory variables were similar to those for the bivariate analysis regarding the 12 patient safety culture dimensions (Table 3). Hence, a higher level of patient safety culture implied increased probability for high overall patient safety. The opposite was shown for “nonpunitive response to errors,” that is, a higher level implied decreased probability for high overall patient safety. In addition, “teamwork across units” was no longer significantly associated with overall patient safety. In contrast to the bivariate analysis, higher age decreased the probability for high overall patient safety.
There were still some associations between professional experience and overall patient safety. Long professional experience (>15 years) implied increased probability for high overall patient safety. In contrast to the bivariate model, no such association was shown for 6 to 15 years of professional experience. There was also an association between work areas and overall patient safety. Compared with general wards, the probability for high overall patient safety increased for emergency care but decreased for psychiatric care. In contrast to the bivariate model, the probability for high overall patient safety was higher for both enrolled nurses and physicians compared with managers. No such association was shown for registered nurses. As in the bivariate analyses, sex and work time were not associated with overall patient safety (Table 3).
This study sets out to investigate various aspects of patient safety culture. The results are based on 23,781 completed HSPSC questionnaires by managers, physicians, registered nurses, and enrolled nurse in Swedish hospitals. We used descriptive statistics and conducted logistic regression analyses to identify whether background variables and patient safety culture dimensions were associated with higher overall patient safety, a recommended outcome variable.20
The descriptive results show that the patient safety culture dimensions with highest scores were “teamwork within units,” “nonpunitive response to error,” “manager expectation and actions promoting safety,” “communication openness,” and “feedback and communication about errors.” Our results are consistent with other findings concerning specific areas of strengths.19,21 These dimensions are well recognized as important aspects of patient safety culture.8 Indeed, open and blame-free communication, teamwork and management engagement, and support are all included in the most widely used instruments to capture patient safety culture.20,22–24
“Staffing” was one of the lowest rated dimensions for patient safety culture. The process of hiring, positioning, and overseeing employees in an organization, that is, staffing, is a well-known and important challenge for attaining a favorable patient safety culture. Staffing has been identified as one of the dimensions of patient safety culture that is most strongly associated with overall patient safety,15 and there is research that demonstrates strong connections between staffing and various patient safety outcomes.25 The relevance of staffing for patient safety culture and patient safety outcomes underscores the importance of work environment conditions for patient safety.26,27 Other studies on patient safety culture have pointed to staffing as an important area for improvement.19
In the descriptive findings, “manager expectation and actions promoting safety” was rated highly, whereas “management support for patient safety” was the lowest rated dimensions. This finding suggests that the respondents are fairly content with what is currently being carried out at the department level but believe top-level managers could do more to support patient safety. Some of the patient safety challenges facing frontline health care providers cannot be solved at the floor level; they need to be addressed at higher levels of the organizations.
With regard to managers in relation to the 3 other professional categories, our findings demonstrate that managers scored all dimensions and overall patient safety higher than registered nurses, enrolled nurses, and physicians. This pattern has been shown previously.17,28 Hammer et al29 have stressed that managers' views on patient safety culture are important; they believe that there is a need for a modified questionnaire from their point of view.
The managers' favorable scores on patient safety culture and overall patient safety were reflected in the bivariate regression model, which also showed managers to be far more likely than the 3 professional categories to report high overall patient safety. However, the multivariate regression model revealed a different picture; the odds for high overall patient safety were 1.32 times higher for enrolled nurses and 1.24 times higher for physicians than managers. This marked difference between the bivariate and multivariate models indicates that there are underlying factors other than having a manager position, which are associated with high overall patient safety.
Professional experience of 15 years or more seems to be important for attaining high overall patient safety, as measured with the HSPSC. A study by Okuyama et al30 found that confidence based on experience made it easier for health care providers to express their opinions on issues of patient safety, suggesting that experience breeds confidence, which is important for patient safety culture.
Considering the importance of professional experience, our findings with regard to age in relation to overall patient safety seem to be somewhat paradoxical, because higher age was not associated with high overall patient safety. To the contrary, the odds for high overall patient safety were 1.23 times lower for staff aged 35 to 54 years and 1.40 times lower for staff older than 54 years compared with staff aged 18 to 34 years. Thus, it seems that age in itself is not as important as professional experience. In addition, the effects of the background variables (reflected by the odds ratio) were small in relation to the patient safety culture dimensions. Together with the large sample size, even small associations will be significant. This may also explain the divergent results about staff position between the bivariate and multivariate models.
The difference between the results of the bivariate and multivariate regression models highlights the risk of presenting findings from surveys of this type without proper statistical analysis. Although HSPSC has been advocated for use as a management tool in local patient safety improvement efforts,10 including Sweden, our findings clearly demonstrate that the instrument requires considerable knowledge and proper statistical analyses if relevant conclusions are to be drawn.
This study has shortcomings that must be considered when interpreting the results. One potential limitation is the high number of missing data. Only complete questionnaires with no missing data were included. Despite this, the final sample was large enough to conduct the binary logistic regression analyses. The sample size was also sufficiently large to detect small and not always clinically significant differences. We could not control for different nationalities in the sample although the knowledge of Swedish was likely acceptable because we included only fully answered questionnaires.
The study investigated the patient safety culture in hospitals with a focus on nurses, physicians, and enrolled nurses, which make up approximately 60% of the staff in Swedish hospitals. Further studies must be conducted to obtain a more complete picture of the patient safety culture by also accounting for other staff categories in Swedish health care.
The safety culture dimensions of HSPSC contributed more to overall patient safety than the background characteristics, suggesting that these dimensions are important in efforts to improve overall patient safety culture.
1. Kohn LT, Corrigan JM, Donaldson MS. To Err Is Human: Building a Safer Health System
. Washington, DC: National Academy Press; 2000.
2. Leape LL. Institute of Medicine medical error figures are not exaggerated. JAMA
3. Soop M, Fryksmark U, Köster M, et al. The incidence of adverse events in Swedish hospitals: a retrospective medical record review study. Int J Qual Health Care
4. Francis R. Final Report of the Mid Staffordshire NHS Foundation Trust: Public Inquiry
. London: The Stationery Office; 2013.
5. Hofinger G. Sicherheitskultur. In: Schrappe M, Hölscher U, eds. Proceedings Dachkongress Medizintechnik and Patientensicherheit
. Münster: Universität Münster; 2008:6–7.3.
6. Guldenmund FW. The nature of safety culture: a review of theory and research. Saf Sci
7. Cox S, Flin R. Safety culture: philosopher's stone or man of straw? Work Stress
8. Waterson P. Patient Safety Culture: Theory, Methods and Application
. Farnham, Surrey: Ashgate Publishing; 2014.
9. Hedsköld M, Pukk-Härenstam K, Berg E, et al. Psychometric properties of the hospital survey on patient safety culture
, HSOPSC, applied on a large Swedish health care sample. BMC Health Serv Res
10. Mannion R, Konteh FH, Davies HTO. Assessing organisational culture for quality and safety improvement: a national survey of tools and tool use. Qual Saf Health Care
12. Agnew C, Flin R, Mearns K. Patient safety climate
and worker safety behaviours in acute hospitals in Scotland. J Safety Res
13. Campbell EG, Singer S, Kitch BT, et al. Patient safety climate
in hospitals: act locally on variation across units. Jt Comm J Qual Patient Saf
14. Hamdan M, Saleem AA. Assessment of patient safety culture
in Palestinian public hospitals. Int J Qual Health Care
15. Olsen E. Workers' perceptions of safety culture at a hospital [in Norwegian]. Tidsskr Nor Laegeforen
16. Karimi S, Yaghoubi M, Rahi F, et al. Patient's safety culture from the viewpoint of nurses working at selected charity, private, and public hospitals of Isfahan. Int J Health Syst Disast Manage
17. Nordin A, Theander K, Wilde-Larsson B, et al. Health care staffs' perception of patient safety culture
in hospital settings and factors of importance for this. Open J Nurs
18. El-Jardali F, Dimassi H, Jamal D, et al. Predictors and outcomes of patient safety culture
in hospitals. BMC Health Serv Res
19. Vlayen A, Hellings J, Claes N, et al. A nationwide hospital survey on patient safety culture
in Belgian hospitals: setting priorities at the launch of a 5-year patient safety plan. BMJ Qual Saf
20. Sorra JS, Dyer N. Multilevel psychometric properties of the AHRQ hospital survey on patient safety culture
. BMC Health Serv Res
21. Wagner C, Smits M, Sorra J, et al. Assessing patient safety culture
in hospitals across countries. Int J Qual Health Care
22. Sexton JB, Helmreich RL, Neilands TB, et al. The Safety Attitudes Questionnaire: psychometric properties, benchmarking data, and emerging research. BMC Health Serv Res
23. Singer S, Meterko M, Baker L, et al. Workforce perceptions of hospital safety culture: development and validation of the Patient Safety Climate
in Healthcare Organizations Survey. Health Serv Res
24. Ginsburg L, Gilin D, Tregunno D, et al. Advancing measurement of patient safety culture
. Health Serv Res
25. Aiken L, Douglas M, Sloane D, et al. Nurse staffing and education and hospital mortality in nine European countries: a retrospective observational study. Lancet
26. Welp A, Meier LL, Manser T. The interplay between teamwork, clinicians' emotional exhaustion, and clinician-rated patient safety: a longitudinal study. Crit Care
27. Kuntz L, Mennicken R, Scholtes S. Stress on the ward: evidence of safety tipping points in hospitals. Manage Sci
28. Singer SJ, Gaba DM, Geppert JJ, et al. The culture of safety: results of an organization-wide survey in 15 California hospitals. Qual Saf Health Care
29. Hammer A, Ernstmann N, Ommen O, et al. Psychometric properties of the Hospital Survey on Patient Safety Culture
for hospital management (HSOPS_M). BMC Health Serv Res
30. Okuyama A, Wagner C, Bijnen B. Speaking up for patient safety by hospital-based health care professionals: a literature review. BMC Health Serv Res