Patient Safety Culture in Hospitals Across Countries
Patient Safety Culture in Hospitals Across Countries
For this study, we were looking for countries with well-developed health-care systems, similar educated professional groups and a focus on improving patient safety. The Netherlands, Taiwan and the USA fulfil these criteria, come from different parts of the world, were already using the same safety culture questionnaire on a broader scale and had empirical data.
The Hospital Survey on Patient Safety Culture (Hospital SOPS) was originally developed, pilot-tested and revised by Westat in the USA and then released by the Agency of Healthcare Research and Quality (AHRQ). The psychometric properties of the US version have been published. The questionnaire has been used in various countries besides the USA. The survey was designed to assess opinions of hospital staff about patient safety issues, medical error and event reporting and includes 42 items measuring 12 dimensions of patient safety culture (Table A1). Respondents are asked to rate each item of a dimension on a five-point Likert scale of agreement (strongly disagree, disagree, neutral, agree and strongly agree) or frequency (never, rarely, sometimes, most of the time, always). The survey includes two questions asking respondents to provide an overall grade on patient safety for their work area/unit and to indicate the number of events they have reported over the past 12 months. Respondents are asked to provide limited background information about themselves.
For the Netherlands and Taiwan, the Hospital SOPS was translated using forward and backward translation to check the quality of the translation. In both countries, the questionnaire was pilot-tested and an expert panel was used to discuss the intelligibility and applicability of the items. In case an item was found to be confusing or unclear, the translation was improved. There were no items that were inapplicable. The results of the validation process have been described in more detail elsewhere.
The sample selection for this study was not planned for an international comparative study upfront. Each country administered independently a national survey of safety culture using the Hospital SOPS questionnaire. Nevertheless, the data provide a good opportunity to compare the safety cultures of three countries by investigating commonalities and differences.
The Netherlands The questionnaire was administered on paper in 45 out of 96 Dutch hospitals from June 2005 till December 2007. Hospitals differed by teaching status and were spread equally over the whole country.
A total number of 171 hospital units participated in the study. Units and hospitals were not randomly selected. In each unit, a random sample of about 30 health-care providers was drawn. Respondents had to be involved in the care process, work for at least 6 months at the unit and work for more than 12 h a week. These criteria were thought to be important to get reliable answers. A total of 3779 respondents filled out the questionnaire. The response rates of 125 of the 171 units were registered. In the other units, the distribution of the questionnaire was not carried out by the researchers and calculating response rates for these units was not possible. The average response rate of the 125 units was 69%.
Taiwan The questionnaire was administered on paper in 74 out of 566 Taiwanese hospitals from June 2007 till August 2008. Most hospitals were selected using stratified sampling for 11 types of hospital ownership. Another 10 hospitals participated voluntarily in the study. Participating hospitals represented a range of bed sizes and types of hospitals. Hospitals were spread across the country. Most hospitals were teaching hospitals (57%) and general private hospitals (61%). Units and respondents were not randomly selected. The mean number of units per hospital was 10.6 (range 1–15). A total of 10 146 respondents in 782 hospital units filled in the questionnaire. The average response rate was 87%.
The USA For this article, we have used 2008 data from the Hospital SOPS Comparative Database. The hospitals were not randomly selected; however, the characteristics of the database hospitals are fairly consistent with the distribution of US hospitals registered with the American Hospital Association (AHA). The data report the results from a total of 622 hospitals with 4894 units and 196 462 hospital staff respondents. The participating hospitals represented a range of bed sizes and geographic regions. Most hospitals were non-teaching (69%) and non-government-owned (voluntary/non-profit or proprietary/investor-owned) (78%). The average response rate was 52%. Most hospitals (74%) administered the survey to all staff or a sample of all staff from all hospital departments.
Percentages of positive responses were calculated for each item and each dimension. Negatively worded items were recoded. The percentages represent the average percentage of positive responses across hospitals. To obtain the dimension scores, item percent positive scores were computed first and then the scores were averaged, which gives equal weight to each item in a composite. Questionnaires with missing responses on all of the non-demographic items were excluded from the analyses.
To describe the safety culture across countries, descriptive statistics with 95% confidence intervals were used to determine statistical significance for differences in the dimension scores at country level. For each item, ranges of positive responses were calculated at hospital level to determine variation between hospitals within countries. The units of analysis were countries and hospitals within countries. Data were analysed using SPSS 15.0.
Methods
Selection of the Three Countries
For this study, we were looking for countries with well-developed health-care systems, similar educated professional groups and a focus on improving patient safety. The Netherlands, Taiwan and the USA fulfil these criteria, come from different parts of the world, were already using the same safety culture questionnaire on a broader scale and had empirical data.
Patient Safety Culture Survey
The Hospital Survey on Patient Safety Culture (Hospital SOPS) was originally developed, pilot-tested and revised by Westat in the USA and then released by the Agency of Healthcare Research and Quality (AHRQ). The psychometric properties of the US version have been published. The questionnaire has been used in various countries besides the USA. The survey was designed to assess opinions of hospital staff about patient safety issues, medical error and event reporting and includes 42 items measuring 12 dimensions of patient safety culture (Table A1). Respondents are asked to rate each item of a dimension on a five-point Likert scale of agreement (strongly disagree, disagree, neutral, agree and strongly agree) or frequency (never, rarely, sometimes, most of the time, always). The survey includes two questions asking respondents to provide an overall grade on patient safety for their work area/unit and to indicate the number of events they have reported over the past 12 months. Respondents are asked to provide limited background information about themselves.
For the Netherlands and Taiwan, the Hospital SOPS was translated using forward and backward translation to check the quality of the translation. In both countries, the questionnaire was pilot-tested and an expert panel was used to discuss the intelligibility and applicability of the items. In case an item was found to be confusing or unclear, the translation was improved. There were no items that were inapplicable. The results of the validation process have been described in more detail elsewhere.
Data Collection
The sample selection for this study was not planned for an international comparative study upfront. Each country administered independently a national survey of safety culture using the Hospital SOPS questionnaire. Nevertheless, the data provide a good opportunity to compare the safety cultures of three countries by investigating commonalities and differences.
The Netherlands The questionnaire was administered on paper in 45 out of 96 Dutch hospitals from June 2005 till December 2007. Hospitals differed by teaching status and were spread equally over the whole country.
A total number of 171 hospital units participated in the study. Units and hospitals were not randomly selected. In each unit, a random sample of about 30 health-care providers was drawn. Respondents had to be involved in the care process, work for at least 6 months at the unit and work for more than 12 h a week. These criteria were thought to be important to get reliable answers. A total of 3779 respondents filled out the questionnaire. The response rates of 125 of the 171 units were registered. In the other units, the distribution of the questionnaire was not carried out by the researchers and calculating response rates for these units was not possible. The average response rate of the 125 units was 69%.
Taiwan The questionnaire was administered on paper in 74 out of 566 Taiwanese hospitals from June 2007 till August 2008. Most hospitals were selected using stratified sampling for 11 types of hospital ownership. Another 10 hospitals participated voluntarily in the study. Participating hospitals represented a range of bed sizes and types of hospitals. Hospitals were spread across the country. Most hospitals were teaching hospitals (57%) and general private hospitals (61%). Units and respondents were not randomly selected. The mean number of units per hospital was 10.6 (range 1–15). A total of 10 146 respondents in 782 hospital units filled in the questionnaire. The average response rate was 87%.
The USA For this article, we have used 2008 data from the Hospital SOPS Comparative Database. The hospitals were not randomly selected; however, the characteristics of the database hospitals are fairly consistent with the distribution of US hospitals registered with the American Hospital Association (AHA). The data report the results from a total of 622 hospitals with 4894 units and 196 462 hospital staff respondents. The participating hospitals represented a range of bed sizes and geographic regions. Most hospitals were non-teaching (69%) and non-government-owned (voluntary/non-profit or proprietary/investor-owned) (78%). The average response rate was 52%. Most hospitals (74%) administered the survey to all staff or a sample of all staff from all hospital departments.
Statistical Analysis
Percentages of positive responses were calculated for each item and each dimension. Negatively worded items were recoded. The percentages represent the average percentage of positive responses across hospitals. To obtain the dimension scores, item percent positive scores were computed first and then the scores were averaged, which gives equal weight to each item in a composite. Questionnaires with missing responses on all of the non-demographic items were excluded from the analyses.
To describe the safety culture across countries, descriptive statistics with 95% confidence intervals were used to determine statistical significance for differences in the dimension scores at country level. For each item, ranges of positive responses were calculated at hospital level to determine variation between hospitals within countries. The units of analysis were countries and hospitals within countries. Data were analysed using SPSS 15.0.
Source...