Psychometrics evaluation of the Persian version of Attitudes toward Patient Safety Questionnaire (APSQ-III) in nursing students Authors: Omolhoda Kaveh, Hamid Sharif-Nia, Zeinab Hosseini, Harpaljit Kaur, Vida Shafipour Published: 2024 ABSTRACT BackgroundThe topic of patient safety and attitudes toward patient safety among health professionals is a main priority of healthcare systems globally. This study aims to investigate the psychometric properties of APSQ-III among Iranian nursing students.MethodsThere were 307 undergraduate nursing students who participated in the cross-sectional methodological study, from December 2022 to January 2023. Content validity, Construct validity and reliability of the APSQ-III was evaluated.ResultsIn the exploratory factor analysis, three factors were identified from 11 items, accounting for 58% of the total variance. The confirmatory factor analysis indices confirmed the model fit (χ2(38) = 91.851, p < 0.001, χ2/df = 2.417, CFI = 0.966, PNFI = 0.652, IFI = 0.966, TLI = 0.951 and RMSEA = 0.068). The results of the reliability evaluation indicated that the factors demonstrated a satisfactory level of internal consistency, as evidenced by Cronbach's alpha (α) and McDonald's omega (ω) values. Additionally, the Composite Reliability (CR) and Maximal Reliability H were both above 0.8, further supporting the reliability of the factors.ConclusionThe results of this study support the valid and reliable APSQ-III with 11 items and three factors. This questionnaire can be used to assess patient safety attitude among nursing students. Background Patient safety refers to preventing harm to patients by identifying and addressing potential dangers, evaluating nursing performance, and enhancing healthcare standards globally [1, 2]. Providing a safe and comfortable care environment for patients is one of the most important tasks and priorities of healthcare systems worldwide [3, 4]. Nursing plays a crucial role in keeping patients safe. Studies show that focusing on patient safety is important to prevent medical errors and ultimately harm to patients and strain on the healthcare system [5–8]. Significant rise in fatalities attributed to medical errors over recent decades has evolved into a critical worldwide challenge [8, 9]. Creating a patient safety culture is important to prevent harm to patients. Targeted strategies are needed to maintain and improve patient safety [1, 10, 11]. Implementing a patient safety program in hospitals involves using a critical incident reporting system to reduce medical errors and improve patient care. This system allows clear guidelines for reporting errors and provides feedback through meetings, paper reports, or email alerts to raise awareness [12]. The study from Tunisia suggests that improving patient safety culture in healthcare settings should prioritize enhancing leadership skills. This involves promoting open communication, creating a blame-free atmosphere, encouraging teamwork, and fostering a culture of continuous learning within the organization [13]. Improving patient safety requires a change in healthcare culture with a focus on better communication among all healthcare team members. Patient safety education plays a key role in shifting attitudes and culture towards better outcomes [14–16]. Furthermore, enhancing safety culture education to decrease patient harm is a worldwide priority [17]. The key to patient safety culture is how doctors handle medical errors, including admitting mistakes and taking responsibility. Researchers recommend more training in medical school to promote this attitude. Some studies suggest that education should be a priority at all levels of the healthcare system, while others focus on incorporating patient safety education in students' programs [18, 19]. Nurses are a crucial part of the healthcare team, responsible for coordinating care and monitoring patients. They have a significant impact on patient safety and their training is essential for ensuring safety in a complex medical setting [3, 20, 21]. However, a prevailing issue that exists is that the concept of patient safety in nursing education has not been clearly institutionalized [22]. It is important to teach patient safety principles to health science professionals during their training to encourage the use of safety practices. Understanding nursing students' perspectives on patient safety and how it is addressed in clinical classes is crucial. Promoting patient safety in healthcare is a key goal of nursing education, with some universities in Europe already implementing patient safety policies in their programs to raise awareness among students [23]. It is important to periodically check in with nursing students to understand their perspectives on patient safety, as they work in different environments and undergo diverse training. Since nursing students have a temporary role on the treatment team, they may offer valuable insights into patient safety. Their opinions can help identify areas for improvement in education and healthcare settings, and promote a culture of patient safety among future nurses [24, 25]. Therefore, there is a need for a valid questionnaire to examine students' attitudes towards patient safety. The APSQ-III was designed by Carruthers (2009) and was employed amongst medical students and tutors dent. The questionnaire contains 26 items, measuring 9 different aspects of patient safety [26]. It has been employed in different countries, including Australia, America, Germany, Spain, Saudi Arabia [27–31], to investigate students' perspectives about patient safety. Although many studies have explored the attitudes of medical students regarding patient safety, there is a scarcity of research on the attitudes of nursing students towards patient safety, particularly among Iranian students who operate within a healthcare system influenced by cultural and educational disparities. Moreover, given the potential threat posed by clinical errors to patient safety, the current study recognized the need to evaluate the psychometric properties of the Persian version of Attitudes toward patient safety questionnaire among nursing students. Methods Study design and participants This methodological cross-sectional design was used in the current study to evaluate the psychometric properties of a Persian version of attitudes towards patient safety among Iranian nursing students. The survey was conducted among 307 Iranian undergraduate nursing students at from December 2022 to January 2023. Sample selection was based on convenience sampling. The participants were selected from the nursing students of Mazandaran University of Medical Sciences and the questionnaires were provided to them in person. Ethical considerations The current study has been sanctioned by the Ethics Committee of Mazandaran University of Medical Sciences (IR.MAZUMS.REC.1401.278). This study strictly followed all ethical guidelines, including informing participants about the study's objectives and procedures, maintaining patient confidentiality, ensuring participants understood the voluntary nature of their involvement, and obtaining informed consent from all study participants. Measurement In this study, a Persian version of attitudes toward patient safety questionnaire (APSQ-III), a self-report, short, and easily administered questionnaire to assess nursing students' attitudes about patient safety, developed by Carruthers et al. [26] was utilized. The first part of the questionnaire comprised of demographic characteristics of the respondents, such as age, gender, marital status, educational level, and employment status. The students were also asked to state the number of students and patients in the internship. In the second section, the 26 items APSQ-III in nine domains (patient safety training received, error reporting confidence, working hours as an error cause, error inevitability, professional incompetence as an error cause, disclosure responsibility, team functioning, patient involvement in reducing error, importance of patient safety in the curriculum) were used to measure the students’ attitudes towards patient safety. Students were asked to indicate whether they agree with each statement using a 5-point Likert-type scale ranging from 1 (totally disagree) to 5 (totally agree). Items 11,14,15,16,17 from the original questionnaire have been reverse scored [26]. A higher score indicates a stronger positive attitude among students towards patient safety. The translation process Translation of the APSQ-III was carried out following the guidelines set by the World Health Organization (WHO). Initially, authorization was sought from the scale's author or developer. Subsequently, two translators proficient in both English and Persian independently translated the APSQ-III from English to Persian. The two independently translated questionnaires were then reviewed and evaluated by a panel of experts, which included two professional translators to construct one Persian translation of APSQ-III. Subsequently, this version of the questionnaire was translated back into English independently by two translators proficient in Persian and English. The accuracy of the translation was verified by a panel of experts [32]. Content validity In order to validate the content both qualitative and quantitative methods, the APSQ-III questionnaire was distributed to ten nursing experts specializing in psychiatric nursing to gather their input and insights on the appropriateness of item allocation, item wording, and representativeness of the items. In addition, the experts rated the necessity of the items and the Content Validity Ratio (CVR) and Content Validity Index (CVI) were obtained. The minimum acceptable value for CVR was 0.62 whereas for CVI was 0.7 [33]. Construct validity The construct validity on the psychometric evaluation of the APSQ-III was assessed by conducting both exploratory factors analysis (EFA) and confirmatory factor analysis (CFA). The dataset of the study was randomly separated into two parts. The first dataset (n = 150) was used for EFA through the maximum likelihood and Promax rotation using SPSS 22. In order to confirm the appropriateness of the data for factor analysis, we evaluated the Kaiser–Meyer–Olkin (KMO) measure, which exceeded 0.8, and performed Bartlett's test of Sphericity, resulting in a significant outcome (p < 0.05) [34]. The determination of the number of factors was guided by Horn’s parallel analysis, and factor extraction relied on eigenvalues exceeding 1 and communalities surpassing 0.2 [35]. Eigenvalues (λ) represent the sum of squared factor loadings (SSL) across all items (k) for each factor, indicating the variance in each item that can be explained by the analysis. To determine the percentage of total variance explained by the factor, divide the eigenvalue by the total number of items [36]. To assess the structural factors, a CFA was conducted utilizing the maximum likelihood method and standard goodness-of-fit indices. Analysis of Moment Structure (AMOS) 24 was used for CFA evaluation. The model fitness was assessed by Root Mean Square of Error of Approximation (RMSEA < 0.08), Comparative Fit Index (CFI > 0.90), Parsimonious Comparative Fit Index (PCFI > 0.50), and Parsimonious Normed Fit Index (PNFI > 0.50), Incremental Fit Index (IFI > 0.90) and CMIN/DF < 3 [35]. Convergent and discriminant validity In order to evaluate the convergent and discriminant validity of APSQ-III according to Fornell and Larcker's perspective, various metrics were analyzed. These metrics included the average extracted variance (AVE), maximum shared squared variance (MSV), and composite reliability (CR). AVE values exceeding 0.5 or CR values surpassing 0.7 are typically considered indicative of appropriate convergent validity. Furthermore, if the AVE value is higher than the MSV value, it confirms the instrument's discriminant validity. AVE is widely recognized as a reliable measure of convergent validity. Moreover, a CR value above 0.7 is commonly utilized to evaluate convergent validity in psychological research [37]. Reliability Internal consistency of APSQ-III was evaluated using Cronbach’s alpha (α), McDonald’s omega (Ω) and CR of more than 0.7 and Average Inter-Item Correlation (AIC) coefficient. The value of AIC between 0.2 to 0.4 was accepted for internal consistency [38] and, Maximal Reliability H (the value of greater than 0.7 was good) was also utilized Fig. 1.Fig. 1The results of the CFA and factor loadings The results of the CFA and factor loadings Multivariate normality and outliers Skewness (± 3) and kurtosis (± 7) were used to analyze potential deviations from a normal distribution in the data [39]. Additionally, multivariate normality was assessed by calculating the Mardia coefficient of multivariate kurtosis (< 8). The Mahalanobis d-squared test (p < 0.001) was utilized to identify any multivariate outliers [34]. EFA was performed using the pairwise deletion method to address missing data. Results The study sample included 147 females (47.9%) and 160 males (52.1%). The average age of the participants was 23 years, with a standard deviation of 2.29. The study revealed that the majority of participants, 266 (86.6%), were single. Additionally, 47 (15.3%) were in the 6th semester, while the majority, 260 (84.7%), were in the 7th and 8th semesters. The results show that, on average, each department had four students and 17 patients. Furthermore, the findings indicate that only 35.8% of students had previous work experience. Content validity was assessed by calculating the CVR and CVI. Based on the results, all items demonstrated adequate content validity and thus all of them entered to EFA. The results of KMO (0.846) and Bartlett's Test of Sphericity (χ2 = 1605.989) shows an adequate sampling for factor analysis. During the EFA, three factors were identified, explaining 58% of the total variance. Fifteen items were excluded from the final structure of the questionnaire based on factor loadings below 0.3, communalities below 0.2, or the presence of cross-loading. This indicates that these three factors capture a significant portion of the variability present in the observed variables. In Table 1, the eigenvalues and the percentage of variance explained by each of the three factors should be presented. This information is crucial for understanding the contribution of each factor to the overall variability in the data set. In the CFA with a sample size of 150 participants, the results confirmed and validated the factor structure derived from the EFA. The CFA results indicated a good model fit for the three-factor measurement model, with the following fit indices: χ2[38] = 91.851, p < 0.001, χ2/df = 2.417, CFI = 0.966, PNFI = 0.652, IFI = 0.966, TLI = 0.951 and RMSEA = 0.068. These indices suggest that the three-factor model is a good fit for the data, supporting the validity of the underlying factor structure. Table 1Exploratory factors extracted of APSQ-III (n = 150)FactorsQn. ItemFactor loadingh2*λ*Variance%The importance of patient safety education in the curriculumQ24. Patient safety issues cannot be taught and can only be learned by clinical experience when qualified ®0.8910.702.5022.7%Q26. Learning about patient safety issues is not as important as learning other more skill-based aspects of being a doctor0.8300.60Q21. Patients have an important role in preventing medical errors0.6630.58Q20. Teaching teamwork skills will reduce medical errors0.6360.51Q10. The number of hours doctors work increases the likelihood of making medical errors0.3680.35Working hours as error causeQ9. By not taking regular breaks during shifts doctors are at an increased risk of making errors0.8850.702.1019.0%Q7. I feel confident I could report an error I had made without feeling I would be blamed0.8350.67Q8. Shorter shifts for doctors will reduce medical errors0.7220.63Patient safety training receivedQ1. My training is preparing me to understand the causes of medical errors0.8130.601.8016.3%Q3. My training is preparing me to prevent medical errors0.7440.59Q2. I have a good understanding of patient safety issues as a result of my undergraduate medical training0.7380.57*h2: CommunalitiesRindicates reverse scoring Table 1Exploratory factors extracted of APSQ-III (n = 150)FactorsQn. ItemFactor loadingh2*λ*Variance%The importance of patient safety education in the curriculumQ24. Patient safety issues cannot be taught and can only be learned by clinical experience when qualified ®0.8910.702.5022.7%Q26. Learning about patient safety issues is not as important as learning other more skill-based aspects of being a doctor0.8300.60Q21. Patients have an important role in preventing medical errors0.6630.58Q20. Teaching teamwork skills will reduce medical errors0.6360.51Q10. The number of hours doctors work increases the likelihood of making medical errors0.3680.35Working hours as error causeQ9. By not taking regular breaks during shifts doctors are at an increased risk of making errors0.8850.702.1019.0%Q7. I feel confident I could report an error I had made without feeling I would be blamed0.8350.67Q8. Shorter shifts for doctors will reduce medical errors0.7220.63Patient safety training receivedQ1. My training is preparing me to understand the causes of medical errors0.8130.601.8016.3%Q3. My training is preparing me to prevent medical errors0.7440.59Q2. I have a good understanding of patient safety issues as a result of my undergraduate medical training0.7380.57*h2: CommunalitiesRindicates reverse scoring Exploratory factors extracted of APSQ-III (n = 150) *h2: Communalities Rindicates reverse scoring The results show that all three factors had CR higher than 0.7, however the first factor had AVE less than 0.5 (0.498), whereas the other two factors had AVE more than 0.5. Based on Fornell and Larcker's recommendation, if the CR exceeds 0.6 for psychological constructs, convergent validity can still be established even if the AVE for individual factors is slightly below the recommended threshold of 0.5. The results in Table 2 likely demonstrate that all three constructs achieved convergent validity based on the CR criterion, despite the AVE for the first factor being slightly less than 0.5. This suggests that the constructs displayed sufficient internal consistency reliability to support their convergent validity. For discriminant validity, the Fornell and Larcker criterion showed that the MSV was less than AVE, which proved that these factors had good discriminant validity. Cronbach's alpha, MacDonald's omega, CR, AIC, and maximum reliability of the three factors extracted from the APSQ-III to evaluate factor’s internal consistency and construct reliability were above the given threshold, hence indicating good internal consistency and good construct reliability (Table 2). Table 2The indices of the convergent, discriminant validity, and internal consistency of APSQ-III (n = 150)CRAVEMSVMaxR(H)αΩAICThe importance of patient safety education in the curriculum0.8280.4980.3830.8410.8430.8490.520Working hours as error cause0.8550.6620.3830.8550.8520.8630.661Patient safety training received0.8100.5870.3230.8120.8060.8130.587 Table 2The indices of the convergent, discriminant validity, and internal consistency of APSQ-III (n = 150)CRAVEMSVMaxR(H)αΩAICThe importance of patient safety education in the curriculum0.8280.4980.3830.8410.8430.8490.520Working hours as error cause0.8550.6620.3830.8550.8520.8630.661Patient safety training received0.8100.5870.3230.8120.8060.8130.587 The indices of the convergent, discriminant validity, and internal consistency of APSQ-III (n = 150) Discussion In recent years, the issue of patient safety and any activity in the field of promoting the culture of patient safety among health professionals has gained significant attention and has become a central concern for healthcare systems globally. Hence, considering the critical significance of this matter within the context of the present study, the researchers investigated the psychometric properties of the APSQ-III among nursing students. The results revealed a three-factor structure of scale ("The importance of patient safety education in the curriculum", "working hours as error cause", "patient safety training received") comprising of 11 items (Table 3) as compared to the original version that was done among medical students which included 26 items and 9 factors. These three factors explained 58% of the total variance of the APSQ-III among the Iranian nursing students. These factors include "the importance of patient safety education in the curriculum" (5 items), working hours as error cause (3 items), patient safety training received (3 items) and the factor loadings ranged from 0.368 to 0.891 as compared to the study of Carruthers et al., whereby the 26 items and 9 factors showed an Overall alpha of 0.73 from 0.82 to 0.39 [26]. Table 3Persian version of Attitudes toward patient safety Questionnaire (APSQ-III)FactorItemStrongly AgreeAgreeNeutralDisagreeStrongly DisagreeThe importance of patient safety education in the curriculum1.Patient safety issues cannot be taught and can only be learned by clinical experience when qualified (R)2.Learning about patient safety issues is not as important as learning other more skill-based aspects of being a doctor3.Patients have an important role in preventing medical errors4.Teaching teamwork skills will reduce medical errors5.The number of hours doctors work increases the likelihood of making medical errorsWorking hours as error cause6.By not taking regular breaks during shifts doctors are at an increased risk of making errors7.I feel confident I could report an error I had made without feeling I would be blamed8.Shorter shifts for doctors will reduce medical errorsPatient safety training received9.My training is preparing me to understand the causes of medical errors10.My training is preparing me to prevent medical errors11.I have a good understanding of patient safety issues as a result of my undergraduate medical trainingRIndicates reverse scoring Table 3Persian version of Attitudes toward patient safety Questionnaire (APSQ-III)FactorItemStrongly AgreeAgreeNeutralDisagreeStrongly DisagreeThe importance of patient safety education in the curriculum1.Patient safety issues cannot be taught and can only be learned by clinical experience when qualified (R)2.Learning about patient safety issues is not as important as learning other more skill-based aspects of being a doctor3.Patients have an important role in preventing medical errors4.Teaching teamwork skills will reduce medical errors5.The number of hours doctors work increases the likelihood of making medical errorsWorking hours as error cause6.By not taking regular breaks during shifts doctors are at an increased risk of making errors7.I feel confident I could report an error I had made without feeling I would be blamed8.Shorter shifts for doctors will reduce medical errorsPatient safety training received9.My training is preparing me to understand the causes of medical errors10.My training is preparing me to prevent medical errors11.I have a good understanding of patient safety issues as a result of my undergraduate medical trainingRIndicates reverse scoring Persian version of Attitudes toward patient safety Questionnaire (APSQ-III) RIndicates reverse scoring The results of the CFA affirmed the goodness of fit of the model. Furthermore, the high values of CR, Cronbach's alpha, and McDonald's omega indicated that the three factors comprising the scale exhibited acceptable levels of internal consistency and reliability. These findings contribute to the overall validity and robustness of the scale in measuring the constructs of interest. The Persian version of APSQ-III revealed acceptable psychometric properties; thus APSQ-III is valid and reliable to examine the field of patient safety among nursing students. Based on the EFA results, the Persian version of APSQ-III consists of three factors which were different to the original scale. The results of the present study are consistent with the study done by Cervera-Gasch et al. among Spanish nursing students that has adequate psychometric properties [31]. The number of items in the Persian version of the questionnaire has decreased compared to the Spanish versions. It should be noted that, the items of the Spanish questionnaire had significant modifications compared to the original version too. The Spanish version comprises 22 items categorized into 6 dimensions: responsibility, organization and communication, teamwork, training, notification, and consciousness. Upon evaluating the items and dimensions of both versions, it was observed that they cover analogous concepts [31]. However, the limitation, similar found in the Spanish study, was that there is a need to address the cross-cultural adaptation, a more rigorous content validity analysis and a larger sample could improve the results [31]. Some studies demonstrated that cultural differences affects patient safety prevailing in the country of Iran [40, 41]. In addition, it appears that the studied population (nursing students) on one hand and the idiosyncrasies of patient safety [42] on the other hand, contribute to the variations in the quantity of factors and items. The first factor of the APSQ-III is "The importance of patient safety education in the curriculum", encompasses of error inevitability, team functioning and importance of patient safety in the curriculum. This factor revolves around the recognition of potential mistakes made by any individual, irrespective of their experience and competence; coordination of different treatment teams; the need to teach teamwork function to students; and the prioritization of educating students about patient safety as well as instructing them on pertinent safety issues. Numerous studies have emphasized the necessity and importance of incorporating patient safety culture within the educational curriculum [43, 44]. The second factor of the APSQ-III is "working hours as error cause" which includes shorter shifts, taking regular breaks during shifts and increase working hours which is consistent with the original version [26]. The current study indicated that escalated work pressure contributes to higher medical errors, posing a threat to patient safety. Various studies have indicated an association between hours worked per week with the patient safety culture and medical errors and death of the patient [9, 45]. Finally the third factor of APSQ-III is "patient safety training received", indicating readiness to understand the causes of medical errors, good understanding of patient safety issues and preventing medical errors by following training which is in line with the original version and Spanish version [26, 31]. Hanifia et al. in their study showed that the patient safety educational program could improve some safety culture and the findings of the present study supports it [46]. In addition, nurses’ patient safety culture needs to be supported by teamwork and continuous organizational learning is in line with the researches done by Tlili et al. in 30 primary healthcare centers in Tunisia and Ammouri et al.[1, 13]. Another study highlighted the importance of nurses’ education level and training as factors that influence patient safety outcomes [47]. Finnish researchers highlighted the necessity of preparing and training nursing students in areas encompassing open communication, patient safety and error reporting, and interprofessional teamwork [48]. Strengths and limitations The strength of this study is that the appropriate sample size used in this study was able to properly reflect the level of patient safety among nursing students in the Persian context of Iran. In addition, clinical educators and nursing managers can utilize the findings of this study in preparing the nursing students. The study has some limitations. The self-report method was used in data collection, which can lead to errors in reporting. Furthermore, the study was conducted in a specific cultural and geographical area; hence the generalization of the results of this study to wider geographical areas should be looked into. In addition, future studies should consider different cultures and geographical regions. Conclusion Based on the results of the present study, the Persian version of APSQ-III revealed acceptable psychometric properties; therefore, APSQ-III is valid and reliable for studying in the field of patient safety among nursing students. The researchers suggested that this questionnaire could be used as a pre- and post-measure to assess the success of a patient safety culture that has been incorporated into the curriculum.