Adaptation and validation of safety attitudes questionnaire (SAQ-SF ) for use in Poland

Introduction

Patient safety in the healthcare system is defined as the prevention and minimization of adverse events affecting recipients of healthcare services. Safe and appropriate care is one of the fundamental patient rights [1]. The issue of patient safety is particularly important in the context of improving the quality of healthcare systems, especially in hospital care, where the consequences of adverse events may pose greater risks for patients’ health and lives compared with outpatient care. The primary task of healthcare institutions is to provide services to patients while ensuring minimal or no risks that may compromise their safety [2]. An adverse event is defined as a harm that occurs during treatment, or as a result of treatment not related to the natural course of the disease or the patient condition, which may result in serious health consequences, need for hospitalization or its prolongation [3]. An adverse event may also occur when an intended nursing intervention leads to unintended negative outcomes [4]. Improving patient safety should encompass two essential dimensions. The first is technical, related to the professionalism of actions, including practical skills and staff experience, adequate staffing in relation to the patient needs and health status, formal competencies, sanitary and hygienic conditions, pharmaceutical storage, and technical condition of medical equipment. The second dimension, referred to as relational professionalism, concerns communication with patients in the broad sense. This includes empathy and understanding, accurate transmission of information about health status and treatment methods, dedicating sufficient time to the patient as well as considering their individual needs [5].
The term “safety climate” generally refers to measurable components of safety culture, such as managerial behavior, safety systems, and employees’ perceptions of safety [6, 7]. Several self-administered questionnaires have been developed as means to quantitatively measure key aspects of the safety climate. These tools assess healthcare personnel’s perception of patient safety climate in hospital settings [8, 9].

Origins and conceptual background of the safety attitudes questionnaire (SAQ)

The safety attitudes questionnaire (SAQ) was originally developed for hospitals on the base of the flight management attitudes questionnaire (FMAQ), an exploratory human factors questionnaire used in commercial aviation [10, 11]. The FMAQ was designed in response to observations indicating that the majority of aviation accidents resulted from disruptions in interpersonal cooperation, such as inadequate communication, poor teamwork, leadership deficiencies, or problems in decision-making. The SAQ was subsequently adapted to healthcare needs and has been applied in a variety of clinical contexts [6]. It was developed by Sexton et al. in 2006 in a multi-step process, and validated by exploratory and confirmatory factor analysis (CFA) employed in data of 10,843 respondents from 203 clinical areas in the United States, the United Kingdom, and New Zealand. Exploratory factor analysis (EFA) was used to investigate the existence of a hidden structure of elements, which indicated six factors. Psychometric properties of this tool have been analyzed and described, demonstrating evidence of its validity and reliability [7].

Assumptions of the adaptation process in Poland

The adaptation of the SAQ-SF to Polish conditions involved translation and validation (psychometric evaluation). The objective was to ensure maximum equivalence with the original version across five dimensions [12, 13]:
1. Facial (appearance of the test, instructions, and scoring method),
2. Linguistic (question content and difficulty level of wording),
3. Functional (applicability to the same research purpose),
4. Reconstructive (methods of testing reliability and validity),
5. Psychometric (reliability, cross-version consistency, and correlations with external scales).
The adaptation retained the original graphical layout, number, and format of questions as well as the response scale. Simultaneously, the relevance of items was assessed in the context of specific characteristics of the Polish healthcare system to ensure cultural and semantic validity [7, 13].
The adaptation process of the SAQ-SF questionnaire consisted of the following stage [7, 13, 14]:
1. Obtaining permission from the original authors: The first step involved obtaining formal consent from the original authors to use the questionnaire (via direct contact with the authors).
2. Preparation of the Polish language version: The original English version of the SAQ-SF was independently translated by two professional English translators whose native language was Polish. Both translators held degrees in English philology and worked professionally in translation. Based on the two independent translations, a preliminary Polish version of the questionnaire was developed. The project leader then reviewed all items, identifying potentially problematic phrases or terms that could cause ambiguity or interpretation issues. The next step involved back translation, i.e., re-translating the Polish version into English by a native English speaker who was fluent in Polish.
3. Expert panel review for translation equivalence: A five-member panel of researchers evaluated the comparability of the translated version with the original SAQ-SF. This was the first level of expert review, aiming at optimizing translation accuracy. Necessary revisions were made, and the final Polish version was subsequently reviewed by English-speaking translators to ensure that the original meaning of the questionnaire items was preserved.
4. Content and cultural equivalence assessment by expert judges: The next stage employed the method of “expert judges”. A group of 17 experts participated: six practicing nurses, two academic nurses, three physicians, three physiotherapists, and three paramedics. All experts were proficient in both Polish and English, had professional experience in the Polish healthcare system, and had been actively employed within five years preceding the study. It was assumed that experts with clinical experience had the best understanding of the meaning and practical relevance of the items. The experts completed an online survey, each receiving an individual link. Every item was evaluated in terms of:
• Content (accuracy of translation meaning),
• Conceptual relevance (applicability to the Polish healthcare context), using a three-point scale: 1 – not relevant, 2 – somewhat relevant, 3 – highly relevant. Additionally, for each question, the experts indicated whether the translation was semantically and technically equivalent to the original item (Yes/No). Finally, a five-member research team analyzed the feedback and refined selected items to ensure cross-cultural consistency and practical applicability of the SAQ-SF in the Polish healthcare setting [7, 13, 14].
The validation procedure of the SAQ in Poland was conducted in 2018, and was approved by the Center for Healthcare Quality and Safety Team of the University of Texas at Houston-Memorial Hermann. Ethical approval was also obtained from the Bioethics Committee of the Jagiellonian University (No. 122.6120.286.2016). The study population included 322 employees of one of Krakow’s hospitals, including nurses, physicians, physiotherapists, paramedics, and healthcare assistants. Before starting the research, the translation was assessed, with accuracy of 0.98 [14].

Statistical analysis

Theoretical validity of the scale was evaluated using the principal components analysis (PCA) with Varimax rotation. The number of components was determined based on the Kaiser criterion (eigenvalue > 1) and interpretability. Factor loadings above 0.6 within sub-scales were considered acceptable. Sampling adequacy was verified with the Kaiser-Meyer-Olkin (KMO) coefficient, while internal consistency was evaluated with Cronbach’s a (≥ 0.7 = satisfactory, ≥ 0.8 = good) [14].

Psychometric properties of the Polish version of the SAQ reliability

Cronbach’s a coefficients across sub-scales ranged from 0.66 to 0.95. For the teamwork climate sub-scale, the reliability was 0.66, and in terms of the safety climate sub-scale, the reliability was 0.74. For job satisfaction, Cronbach’s a was 0.82. The reliability coefficient for the stress recognition sub-scale was 0.86, while for working conditions it was 0.75. For the perception of management sub-scale, Cronbach’s a was 0.95 for ward managers and 0.93 for hospital directors (Table 1) [14].

Validity

Before conducting the validity analysis of the Polish adaptation of the SAQ, the Kaiser test was applied to verify whether the data met the requirements for factor analysis. The KMO value, serving as a measure of sampling adequacy, was estimated at 0.87 (df = 8630, p < 0.001). PCA analysis identified nine components with eigenvalues greater than 1. However, testing solutions with varying numbers of factors indicated that the eight-factor model was the closest to the original factor structure of the instrument, allowing the clearest interpretation of the factors. This model explained 68% of the total variance of the analyzed dataset [14].

Description of the scale, scoring, and interpretation of results

The short version of the questionnaire (SAQ) contains 41 items in total, divided into two parts: the first consists of 36 questions, and the second collects demographic data (age, sex, professional experience, nationality). Completion of the entire questionnaire takes approximately 10-15 minutes. Responses are given on a five-point Likert’s scale:
• A – 1: Strongly disagree,
• B – 2: Slightly disagree,
• C – 3: Neutral,
• D – 4: Slightly agree,
• E – 5: Strongly agree.
Additionally, an “X: Not applicable” option is available. Some questions are negatively worded and require reverse scoring during analysis (Table 2) [7].
The SAQ measures healthcare workers’ attitudes across six dimensions:
1. Teamwork climate: Perceived quality of collaboration within the team (items 1-6),
2. Safety climate: Perceived organizational commitment to safety issues (items 7-13),
3. Job satisfaction: Positive attitude towards one’s work (items 15-19),
4. Stress recognition: Awareness of the impact of stress on work quality (items 20-23),
5. Perceptions of management: Assessment of administrative actions and support at the ward and hospital levels (items 24-28),
6. Working conditions: Evaluation of work environment and logistical support (items 29-32) [7, 14, 15].
Additional questions (items 14, 33-36) refer to the evaluation of supervisors and interprofessional collaboration [7].

Scoring and interpretation

The mean score for each sub-scale is calculated according to a pre-defined formula (Table 2). The resulting score ranges from 0 to 100. A score ≥ 75 is considered positive, corresponding to the response “Slightly agree” on the original Likert’s scale [7, 15].

Overview and practical remarks

Measuring patient safety culture is a crucial component of quality management systems in healthcare. Available tools, including standardized questionnaires, allow not only for diagnosing the level of safety culture within a given institution, but also for identifying areas that require corrective actions. Systematic monitoring of these aspects enables the implementation of targeted interventions aimed at reducing the risk of adverse events. The effectiveness of such actions is strengthened by the consistent application of evidence-based medicine (EBM) guidelines and adherence to internal organizational standards. Implementing comprehensive patient-safety-oriented strategies contributes not only to reducing the number of adverse events, but also to improving the quality, continuity, and efficiency of healthcare [7, 14].
Numerous studies indicate that the SAQ demonstrates strong psychometric properties [4, 11, 14-28]. Its translation into 16 languages supports the stability of these properties, including Polish [14], Albanian [16], Amharic [4], Arabic [17], Chinese [18], Croatian [19], Danish [11], Dutch [20], German [21], Italian [22], French [23], Malaysian [24], Norwegian [25], Slovenian [26], Swedish [27], and Turkish [28]. The Polish adaptation of the questionnaire provides access to a linguistically accurate and reliable research tool, enabling internationally comparable results (Table 3) [4, 11, 14, 16-28].
The SAQ has proven to be a valuable instrument for assessing attitudes toward patient safety within the Polish healthcare system. The results of the adaptation process indicate satisfactory psychometric properties, confirming its applicability both in clinical practice and scientific research. Use of the SAQ helps to better understand how healthcare professionals perceive safety issues, teamwork, and relationships with management. This enables not only the diagnosis of organizational strengths and weaknesses, but also the planning of actions targeted at improving quality and reducing the risk of adverse events [14].
In clinical practice, it is particularly important to treat the questionnaire results as a tool for fostering a culture of openness and continuous improvement, rather than as means of evaluating individual employees. Regular use of the SAQ may facilitate monitoring of safety-related changes, assessing the effectiveness of training programs or management interventions, and initiating corrective actions in areas with the highest risks. In the Polish context, where staffing and organizational resources are often limited, this tool may constitute an important element in building a systemic approach to patient safety and quality improvement.

Conclusions

The SAQ-SF-PL is a tool that enables valid and reliable assessment of the safety climate in healthcare institutions, including hospitals, thereby allowing the identification of areas, which require improvement or reinforcement.

Information for questionnaire users

The author of the Polish version grants permission for the use of the questionnaire in research (Appendix 1).

Disclosures

This review paper received no external funding.
Approval of the Bioethics Committee was not required.
The authors declare no conflict of interest.
Appendix 1 is available on the journal’s website.

References