1. Introduction
As an integral part of health policies, patient safety is a primary goal for all organisations, invariably associated with the quality of healthcare systems [
1,
2]. The right to health protection is a fundamental social right, protected under the terms of Article 64 of the Constitution of the Portuguese Republic [
3], as well as by the Basic Health Law, approved by Decree-Law No. 095/2019 of 4 September. In order to guarantee this right and taking into account the recommendations set by the World Health Organization (WHO) [
4], the Portuguese National Patient Safety Plan (PNSD) 2021–2026 [
5] reinforces the need to endorse the importance of patient safety to consolidate and promote safety in healthcare provision. To this end, it emphasises the principles that sustains it, such as patient safety culture, communication, and the continued implementation of safe practices in environments that tend to be more complex [
5].
Given the complexity and criticality of the healthcare provided, hospitals are institutions that are highly susceptible to incidents and adverse events [
6,
7]. In fact, recently, Chatzi and Malliarou [
8] dove deeper into the definition of patient safety, claiming that it is the state in which harm to patients from nursing practice is eliminated or abridged so far as reasonably practicable through a continuing process of adverse events’ identification. The WHO refers to adverse events as errors or failures that occur during the provision of healthcare, which can result in harm to the patient, unrelated to their underlying condition [
4,
9]. These are unintentional and unexpected occurrences that can lead to temporary or permanent harm to patients and are one of the critical indicators of patient safety [
10]. Studies estimate that 0.9% to 5.2% of deaths in hospital settings were potentially preventable [
11]. In fact, the worldwide reporting of adverse events or incidents shows a very low prevalence of around 5% to 17%, of which 60% were preventable [
12,
13]. Evidence presented by the WHO [
4] points to 134 million adverse events, contributing to 2.6 million deaths per year, so according to recent estimates, the social cost of patient harm is estimated at around USD 1 to 2 trillion per year. The Portuguese reality is in line with this trend, since 42% to 66% of the 11% of occurrences were preventable [
12,
13].
Patient safety culture is therefore a complex and multidimensional concept seen as the set of values, beliefs, norms, competencies, and individual and organisational behaviour patterns that determine commitment, style, and action regarding patient safety issues [
2,
4]. Referring to organisational behaviour, Fernandes and Queirós [
14] highlight the enormous influence of nurses, since they are the largest professional group in Portuguese hospitals, and as a result of their job, have the most direct contact with patients. Given their proximity to patients, their recognised differentiated skills and their fundamental role in maintaining and consolidating a culture of patient safety [
7,
15], the professional nursing group is a key player in the process of continuously improving the quality of healthcare [
15,
16]. Thus, demonstrating nurses’ perceptions regarding patient safety culture can provide an overview of organisational conditions in this area and support for a more realistic analysis and understanding of individual and collective behaviour patterns that can have a positive or negative impact on healthcare. Therefore, this knowledge can enable the improvement of processes, flows, bundles, strengthening the focus on preventive measures, the notification of adverse events, and permanent and continuing education [
17].
Several international studies point to various factors influencing nurses’ decisions to report adverse events, namely, the ability of the manager to respond, the fear of measures resulting from the report, the support of the team, and the existence or not of a safety culture [
13,
18,
19,
20]. Transposing this to the assessment of patient safety culture in national hospital institutions, the results are in line with the international overview [
21]. In fact, the PNSD 2021–2026 is a reinforcement of the importance of promoting patient safety as part of a structured collective effort involving administrations, managers, and health professionals to raise public awareness of this issue [
22]. In this regard, the Hospital Survey on Patient Safety Culture (HSOPSC), developed by the Agency for Healthcare Research and Quality (AHRQ) and translated and validated in Portugal by Eiras et al. [
23], is a tool that allows a multidimensional assessment of safety culture [
23,
24]. In addition, the HSOPSC allows internal and external benchmarking exercises to be carried out: being widely applied in more than 93 countries, these provide a reliable international database on patient safety culture [
24,
25,
26]. In Portugal, as a result of the evaluations carried out over the last 10 years, areas with potential for improvement have been identified, specifically the non-punitive response to error, the frequency of events reported, staffing, management support for patient safety, teamwork across units, communication openness, feedback and communication about errors, and overall perceptions of patient safety and transitions [
26].
Although a transparent policy is currently advocated, with the recognition and analysis of reported adverse events, there is still a culture of blame in many organisations, discouraging professionals from reporting incidents, essentially due to concerns about liability and/or fear of their image in the eyes of their peers [
27,
28,
29]. The consequences are under-reported and the expected learning from adverse events not occurring on a large scale [
29]. Several studies found that supervisor/leadership responsiveness encompasses respect on the part of leaders and a sense of justice in the face of an employee’s exposure and reporting of an incident [
30]. Similarly, a lack of feedback on the reported adverse event is seen as a factor in the reporting process [
19] and the lack of time and nurses as an impediment to the adverse event reporting process [
31].
This scarcity of adverse event records is a pattern in various health systems, resulting in biased knowledge of the facts and subsequently making it difficult to devise effective strategies [
22]. The culture of reporting adverse events is a crucial element for critical patient safety culture (CPSC); therefore, involving healthcare professionals and their organisations in defining and implementing corrective and preventive interventions has a positive generative effect on incident reporting [
31]. These and other initiatives contribute to strengthening the safety culture in healthcare organisations, constituting an important pillar for the sustainability of a learning environment and for building a sphere of trust in these organisations [
32]. These authors also add that assessing CPSC allows continuous improvement processes to focus on the areas identified as priorities. In line with this, another study revealed that hospital units whose professionals rate patient safety more positively have an equally favourable assessment of healthcare satisfaction from the patients’ perspective [
33]. These authors therefore state that investing in a patient safety culture can positively enhance the patient’s experience of healthcare.
This study arose from the need to promote a CPSC by analysing nurses’ perceptions on this matter, in order to draw up an overview of the organisational situation and define strategies for continuous quality improvement and risk management. That said, this study aimed to assess patient safety culture according to nurses’ perceptions in the context of providing healthcare to the people in critical condition (PCC) using the validated HSPSC, identifying associated factors and contributions to guiding its improvement. Based on the literature review, the following investigation questions were formulated. What is the perception of patient safety culture among nurses within the critical care environment? Are sociodemographic and professional characteristics associated with the CPSC?
2. Materials and Methods
2.1. Type of Study
This was an observational, cross-sectional study, written according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations [
34] and using descriptive and inferential statistics.
2.2. Study Site
The hospital studied is a national health service (NHS) institution located in the centre of Portugal. The study focussed on the nurses at this institution whose departments included PCC, namely, the Intensive Care Unit (ICU), Operating Theatre (OR), and Emergency Department (ED).
2.3. Period
Data collection took place between September and December 2023.
2.4. Population
The population included the nursing teams of the ICU, OR, and ED, totalling 268 nurses, 11 of whom were absent due to illness or parenthood, giving a target population of 257 nurses.
2.5. Selection Criteria
The inclusion criteria were as follows: belonging to the nursing professional group, working in the units/departments being studied, and agreeing to take part in the study voluntarily.
2.6. Sample
The sample was non-probabilistic by convenience, as it included only nurses from the selected units/departments who volunteered to take part in the study.
2.7. Variables
The dependent variable corresponds to the CPSC dimensions and the independent variables to the sociodemographic and professional context variables (gender, educational qualifications, unit/department, experience in the unit/department, and experience in the organisation), as well as the variables of number of adverse event reports and perceived level of patient safety.
2.8. Data Collection Instrument
The questionnaire used corresponds to version 1.0 of the HSPSC, originally developed by the AHRQ [
35], translated, validated, and provided by the authors [
23] and adopted by the Direção-Geral da Saúde (DGS) under Regulation No. 005/2018 of 20 February 2018. The survey applied consists of two parts, the first of which covers the characterization of the sample, whose variables correspond to age, gender, educational qualifications, unit/department, years of experience in the unit/department, years of experience in the organisation, knowledge of whether the hospital/unit is accredited and/or is undergoing an external quality assessment process, and previous response to the survey.
The second part corresponds to the HSPSC, with 42 items, assessed using a Likert scale and evaluating 12 dimensions of patient safety culture: 1. teamwork; 2. supervisor/manager expectations and actions promoting patient safety; 3. organisational learning—continuous improvement; 4. management support for patient safety; 5. overall perceptions of patient safety; 6. feedback and communication about errors; 7. communication openness; 8. frequency of events reported; 9. teamwork across units/departments; 10. staffing; 11.transitions; and 12. non-punitive response to error [
26,
35,
36]. In addition to these dimensions, the HSPSC also includes an assessment of the level of critical patient safety perceived by nurses and the number of adverse events reported in the last 12 months. It should be noted that all the questions were compulsory, and as such, no completed questionnaire was excluded for missing data.
2.9. Data Collection
The survey was made available electronically via institutional email, using Microsoft Forms®, sent through the Chief Nursing Officer of each unit/department within the scope of the study, along with a reminder to complete the questionnaire every month. Throughout the process, anonymity was guaranteed and only one response was given for each nurse, automatically linking the response submitted (which could not be filled in again) to the respondent’s institutional email address.
2.10. Data Processing and Analysis
To analyse and interpret the data, the indications in the AHRQ guide were used as a reference [
35]. Given that the questionnaire includes negatively worded items (A5, A7, A8, A10, A12, A14, A16, A17, B3, B4, C6, F2, F3, F5, F6, F7, F9, and F11), the guidelines issued by the AHRQ suggest inverting the scale to facilitate analysis, and these items have been identified with an “r” (
Table 1) [
16,
23,
35,
37].
Then, they were recoded and grouped into 3 categories in order to facilitate analysis: positive (encompasses the answers “agree”, “strongly agree”, “most of the time”, and “always”), neutral (includes the answers “neither agree nor disagree” and “sometimes”) and negative (“strongly disagree”, “disagree”, “never”, and “rarely”) [
16,
35].
Once the items had been grouped into the aforementioned safety culture dimensions [
35], the “not applicable” answers (6 on the Likert scale) were omitted, the rating for each dimension was calculated to allow a proper comparative analysis between them and subsequently to analyse the possible statistical relationships between variables and trends through the respective median (Me). The AHRQ benchmark was also applied, which suggests that positive ratings of 75% or more indicate “strong” dimensions regarding safety culture, while positive responses of 50% or less correspond to weak dimensions that require improvements [
17,
38].
The data were analysed using SPSS
® version 29.0. In the field of inferential statistics, non-parametric tests are used, since the necessary postulates to apply parametric tests are not guaranteed and it is not possible to invoke the central limit theorem [
39]. Therefore, the following tests were used: Kruskal–Wallis, Spearman, and Mann–Whitney U, considering a 95% confidence interval, as well as the use of frequencies (relative and absolute), standard deviation, mean, and median.
Concerning the analysis of the sociodemographic and professional characterization items, since the classes inherent to the variables “educational qualifications” and “unit/department” had frequencies of fewer than 5, they were grouped into other classes within their variable [
39]. Thus, new variables emerge: the “professional category”, which includes the class of specialist nurse (combining the other speciality classes) and (generalist) nurse, and the variable “departments”, which includes 3 grouped classes (ED, OR, and UCI).
2.11. Ethical Aspects
Authorization was first sought from the authors who translated and validated version 1.0 of the HSOPSC for the Portuguese population. To carry out the study, it was essential to obtain a favourable opinion from the ethics committee, the legal support unit, and the board of directors of the institution under study.
Participation in this research required reading and accepting an informed consent form attached to the survey. In this way, participation was voluntary without prejudice to withdrawing at any time, guaranteeing the confidentiality of the data.
5. Conclusions
This study has made it possible to assess nurses’ perceptions of safety culture in units where patients are in critical condition. In general terms, the teamwork dimension stands out, having been identified as the only strong factor in the CPSC of the healthcare institution. Dimensions 2 (expectations of the supervisor/manager and actions promoting patient safety) and 3 (organisational learning—continuous improvement), although better positioned with positivity between 50% and 75%, should be the target of interventions so that they can be considered factors that enhance CPSC. On the other hand, the remaining dimensions assessed were below 50%, with the most emerging and requiring priority intervention being dimensions 12 (non-punitive response to error), 10 (staffing), 8 (frequency of events reported), and 4 (management support for patient safety).
The study carried out highlighted gaps related to the CPSC of the institution in question, helping to provide an overview in these matters. Weaknesses and positive aspects were identified, representing an opportunity for hospital management to adjust the risk management matrix and quality processes and flows so that strategies can be drawn up to promote and consolidate the CPSC.
The promotion and dissemination of the importance of a systematic evaluation of the CPSC in order to diagnose areas of improvement is fundamental in all healthcare organisations, guaranteeing the quality of healthcare to PCC. It is therefore recommended that a study be carried out regarding CPSC including a more representative sample, promoting adherence in advance through awareness-raising actions and strategies aimed at health professionals, and focusing on the dimensions identified as weak and therefore in need of improvement. In addition to what has been suggested, the promotion of a CPSC, in line with a process of continuous improvement through communication, training, and awareness-raising actions, must be evident and include the involvement and responsibility of top hospital management and leaders of healthcare institutions.