A qualitative exploration of patient safety in a hospital setting in Spain: Policy and practice recommendations on patients' and companions' participation

Abstract Introduction Patients' and companions' participation in healthcare could help prevent adverse events, which are a significant cause of disease and disability. Before designing interventions to increase participation, it is first necessary to identify attitudes to patient safety. This study aimed to explore patients' and companions' perceptions, attitudes and experiences of patient safety, taking into account contextual factors, such as cultural background, which are not usually captured in the literature. Methods We conducted a qualitative study with a theoretical sampling of 13 inpatients and 3 companions in a university hospital in Barcelona, Spain. Information was obtained from individual and triangular interviews. A descriptive thematic content analysis was conducted by four analysts and a consensus was reached within the research team on the key categories that were identified. We also conducted a card‐sorting exercise. Results All informants emphasized the role of good communication with health professionals, a calm environment and the need for patient education. Discursive positions differed by cultural background. Informants from a Pakistani–Bangladeshi background emphasized language barriers, while those from European and Latin‐American backgrounds stressed health professionals' lack of time and the need for more interdisciplinary teamwork. The card‐sorting exercise identified several opportunities to enhance participation: checking patient identification and medication dispensation, and maintaining personal and environmental hygiene. Conclusion This exploration of informants' discourse on patient safety identified a wide variety of categories not usually considered from institutional perspectives. The findings of this study could enrich interventions in areas with diverse cultural backgrounds, as well as current frameworks based exclusively on institutional perspectives. Patient or Public Contribution The results of the study were communicated to patients and accompanying persons via telephone or email. Similarly, a focus group was held with a patient forum to comment on the results. In the design of subsequent interventions to improve patient safety at the hospital, the proposals of patients and companions for their participation will be included together with healthcare professionals' opinions.


| INTRODUCTION
According to the World Health Organization, adverse events due to unsafe care are one of the 10 leading causes of death and disability worldwide. 1 In high-income countries, it is estimated that 1 in 10 patients experience an adverse event while receiving hospital care, 2 of which almost 50% are preventable. 3 This is also true of Spain. 4 In this regard, inequities in the safety of care and a higher risk of adverse events have been found in ethnic minorities. 5 Recommendations for hospital quality improvement, and especially for the prevention of adverse events, usually originate from and are directed towards healthcare teams. 6,7 However, it is becoming increasingly important to involve patients directly in interventions to enhance patient safety, which could reduce the burden of harm. 6,8 Although data on the effectiveness of patient participation in patient safety can vary depending on the specific aims pursued, 9 there is evidence that adverse events can be reduced by involving patients and their companions in medication monitoring, [10][11][12] and by preventing pressure ulcers, patient falls and surgical infections. 6,13,14 In contrast, studies assessing patient involvement in hand hygiene promotion among physicians show they are particularly reluctant to speak out when they observe poor hygiene practices. 7 Patient participation in safety during admission is a key element of patient empowerment and a hospital culture that promotes a new, more proactive patient role. 15 Both could be achieved by simply enquiring about patients' experiences during hospitalization. 9 However, before increasing patient participation, there is a need to identify patients' characteristics, such as cultural background, socioeconomic position, gender and age, 15 as they may influence attitudes towards participation. Indeed, specific mechanisms are required to ensure participation among certain groups, such as ethnic minorities. 5

| METHODS
This qualitative study is part of a larger project conducted in our hospital aiming to design a set of interventions to improve patient safety in hospital care by involving patients and their companions.
The project is called 'Improving patient safety through the active involvement of patients and companions'.
This study was conducted from design to analysis from the perspective of critical theory and equity. This was essential because the research team was aware that the area served by the hospital has inequalities influencing participation and is more deprived than other areas of the city. 16

| Study setting and design
The Hospital del Mar is one of four public university hospitals in the city of Barcelona and attends medium-and high-complexity diseases in a catchment area of more than 300,000 inhabitants. The hospital has more than 400 conventional beds, and 12 operating rooms, and annually assesses more than 95,000 patients in the Emergency Department.
This exploratory and interpretative qualitative study used a naturalist-comprehensive paradigm. The perspective adopted was socioconstructionist: forms of participation are proposed from particular historical, social and individual contexts, and the aim of researchers is to critically interpret the social positions underlying ABIÉTAR ET AL. | 1537 these proposals. 17 Data collection techniques consisted of individual and triangular semistructured interviews with patients and their companions. Triangular interviews (also known as triangular groups) are composed of a maximum of three people. They are intended to delve deeper into some topics, because, in a larger group, due to group pressure, the discourse would be less rich. Our triangular groups consisted of interviewing patients and companions together. 17 In accordance with the project's objective of proposing actions to improve patient safety in the hospital, the degree of data interpretation was high.
An external company gave support in the transcription and coding stages. The external company is a strategic consultancy specializing both in innovation and qualitative studies. They have led trend studies, ethnographic studies, digital ethnography and interviews with experts. A coding procedure was agreed upon before the start of the study. Coding quality and control mechanisms were also established, and the coding was finally reviewed by the research team. A process of sharing the research objectives and the methodological approach was necessary since the company was not involved in the process of designing the study protocol. Patients and companions were told they could ask to stop the interview if they became tired or uncomfortable. Likewise, if interviewers perceived any difficulty, they spontaneously asked if the informants wanted to stop the interview.

| Study population
To guarantee appropriate health status for participation, eligible participants were hospitalized patients who were close to discharge and who belonged to European, Pakistani-Indian-Bangladeshi or Latin-American communities and were older than 18 years, irrespective of the reason for admission. Those selected were also able to understand and provide consent to participate. Patients' companions who were present at the time of the interview were approached after the patient met the criteria and were invited to participate and to provide consent.

| Recruitment and sampling
First, participants were identified a priori through hospital records in November 2021, on the basis of general characteristics such as age, sex and place of birth. Among those patients who were expected to meet the heterogeneity criteria, potential participants were approached by two investigators (D. G. A. and N. S.) and by a cultural interpreter (when necessary), face-to-face and a short survey was conducted to characterize their profile with the heterogeneity criteria described in the paragraph below. A convenience sample of participants was recruited in-house (D. G. A.) after the short survey was conducted. Sample diversity and the equity perspective were ensured by using the following heterogeneity criteria: age (18-34, 35-64 and >65 years); a sense of belonging to a cultural community (European, Latin American, Indian-Pakistani-Bangladeshi and Maghrebi); educational attainment (none, primary education, secondary education, vocational training and higher education); sex assigned at birth (male, female or intersex) and gender identity (feminine, masculine or nonbinary*: nonbinary people are those who do not identify with the male-female binomial). People from non-Western cultural communities were specifically included to avoid underrepresentation due to their lack of access to effective participation.
The cultural communities selected are the four most frequent in the health areas served by the hospital. Only one potential participant refused to participate in the study because she felt dizzy due to her health status. We used a semistructured topic guide ( Figure 1) to conduct the interviews, based on findings from a nonsystematic review of qualitative and quantitative studies examining patients' perspectives and experiences of patient safety in hospitals, created by four members of our team. Because most previous studies have been presented within frameworks consistently based on professional perspectives, in this study, we deliberately chose not to use a specific theoretical framework, as a means to explore how patients and their companions conceptualized patient safety in their own words. During the interviews, and after question 5, a card-sorting exercise was used to facilitate visualization of safety-related scenarios ( Figure 2) and to try to overcome language and health literacy barriers. The informants were asked to rank the images from those representing the least safe situation to that representing the safest situation and to explain their reasons as a way to help them to think about these situations. All the information collected with this exercise was included in the textual corpus for the descriptive thematic content analysis.

| Data collection
All interviews took place in December 2021 and January 2022 and were audio recorded with the participants' permission. Data were collected until saturation criteria were reached. We considered the discourse 'saturated' when discursive positions were clearly defined and no new elements were identified for any of them. Because the discourse of the selected cultural communities had not been previously explored in our setting, we did not decide on a specific conceptual framework beforehand for the data analysis.  Nine informants (66%) were men and 7 (44%) were women. None of the participants identified as nonbinary or trans (gender identity) or reported being intersex (sex at birth). Patients were from different hospital services: general surgery, gastroenterology, neurology, neurosurgery, urology, oncology, traumatology and internal medicine. Patient characteristics are shown in Table 1.

| Conceptualization of patient safety and its key moments
For all informants-independently of cultural background, age, sex assigned at birth and educational level-their conceptualization of patient safety during hospitalization was broader and more ambiguous than the concept commonly used by preventive medicine services ('absence of errors in healthcare') and was specifically defined in subjective, relational terms.

| Subjective factors
When asked for a definition of patient safety, informants reported, on the one hand, positive experiences of trust and well-being between patients and health professionals, mostly with nurses and, on the other hand, also negative experiences, such as shame during wound care or a lack of communication by physicians. Therefore, they emphasized the value of the subjective component of safety. In their discourse, they assigned little importance to the usual definition of patient safety, understood as an awareness of errors and experiences of adverse events, because in one patient's words, 'we trust the professionals, otherwise we wouldn't be here'. In other words, if informants trusted professionals, they trusted them not to make mistakes.

| Safety understood as quality
To a certain extent, the informants' conceptualization of safety was similar to the concept of 'quality', and they had both positive and negative perceptions during hospitalization. For participants, patient safety was mostly related to two dimensions: professionals' behaviour (e.g., 'humane manners, professionalism, sincerity and honesty' and coordination as a team), as well as a comfortable hospital environment (e.g., 'clean and quiet space', and 'being supported'). They reported that both items protected them from errors and made them feel safe. Perceptions were closely linked to the discursive positions shown in Figure 3.

| Patients' and companions' proposals for their participation in patient safety
From the descriptive thematic content analysis, various activities and moments were considered appropriate for participation.

| Checking correct identification
The informants believed that the identities of both patients and professionals should be checked, whenever the latter establish contact with them or with accompanying persons, as mentioned by T A B L E 2 Most relevant proposals for participation by study informants and equivalent proposals are discussed in the literature reviewed.

Proposals for patient and companions' participation in this study Equivalent proposals and their rationale in other studies
Checking correct identification of patients and professionals, whenever the latter establishes contact with them or with accompanying persons.
− Also mentioned in Vaismoriadi 20 and Park. 6 Ensuring the medication dispensed is correct before taking it, writing down notes about its indication and dosage.
− To engage both patients and companions in treatment surveillance, de Jong 10 . − Examples of medication management have been described by Gabe et al. 21 and Jordan et al. 22 − Successful interventions to encourage patient participation in the monitoring and self-management of medication in hospitals have been described by Hall et al. 23 Proactively helping with nursing care whenever possible. − This proposal could be considered together with patients' and companions' roles in preventing pressure ulcers. 4 Always expressing doubts and being communicative with professionals, especially when patients or their companions perceive errors.
− Patient participation relies on patients being encouraged to raise doubts without fear of offending healthcare staff. Agreement on how patients should ask these questions would encourage patientprovider trust. 24 − To encourage patient participation, the management system must be supportive and continuously identify and correct all the weaknesses and failures that arise in the system. 25 Management should also be committed to supporting and empowering patient involvement and challenging power inequities. 24 Maintaining personal and environmental hygiene.
− Most concerns are about environmental factors such as noise at night and poor bathroom facilities. 5 − This was also considered important for surgical infections. 4 The informed consent process should be communicated entirely orally and face-to-face to allow patients to express doubts.
− Communication and cooperation between patients and healthcare professionals are important resources for patient participation in quality improvement projects. 7 Professionals ask for patients' safety perceptions at discharge to improve the processes.
− Successful quality improvement interventions can be simple. Direct patient feedback can pinpoint areas of harm not previously noticed and can embolden healthcare professionals' to report harm, leading to changes at the microsystem level. 7 − For this type of intervention, professionals might need training. 24 − Active intervention is required, whether at the individual or collective level, to create an environment where patients are listened to, and their views are taken seriously and acted on. Productive communication does not occur by itself. 5 − Many problems and interventions in healthcare are complicated or complex, but effective safety interventions can also be straightforward. For example, to encourage patients and their families to report harm, the introduction of a simple, real-time bedside questionnaire enhanced the ward's overall safety culture. 7 Participation requires resources (especially time) for professionals. Knowing the approximate time of the nurse's appointment time during the admission process could help and their companions.
− Healthcare managers are responsible for providing an appropriate and positive environment for nurses to engage patients in patient safety. 24 − Research has identified multiple barriers that need to be identified and considered by organizations when managing quality improvement efforts. Such barriers include healthcare system financing, competing organizational changes and the work environment, for example, time constraints, staffing, routines, educational skills and the existing attitudes and culture. 7 Finally, informants were asked to propose specific measures to be considered for our context, to avoid ambiguity over how patients should participate in patient safety and to prioritize areas of action, which is in agreement with other studies. 6,20 The proposals are summarized and compared with those in similar studies in Table 2. In these proposals, the effect of the card-sorting exercise should be considered, as the informants' collective beliefs have been fuelled by the images proposed as moments when patient safety is important.

| Comparison with existing literature
To a lesser extent, the perception of patient safety for patients and companions was defined more on the basis of care structures or outcomes of care than on the healthcare process, which is consistent with the existing systematic reviews on quality. 24 As in other studies, informants were aware of the existence of errors but did not consider them as the main issue in patient safety and, when they did occur, they mostly blamed poor communication before and after the error (as did health professionals). 6 safe. This process could be interpreted as an act of handing over collective responsibility to protect patient safety to the institution and its professionals. Therefore, these elements could be key in promoting a shift to a more proactive attitude of informants to participate.
One of the most notable results of this study is that there was agreement that transfers were key unsafe moments during hospitalization. However, informants expressed different needs according to

Proposals for patient and companions' participation in this study Equivalent proposals and their rationale in other studies
The role of companions' is conceived as acting as mediators with professionals when misunderstandings occur or when patients must responsibility.
− This role has been previously conceived, especially when a patient's ability to participate is reduced by illness. 24 − Companions are expected to speak up at a time when patients themselves are too vulnerable or unwell to act as their own advocates. 5 Interaction with other inpatients in physical spaces designed for patients and companions, so that they can share experiences during admission and generate self-help links. Patient associations raise doubts about their usefulness and possible hidden interests.
− It is important to generate a participation culture. 13 − It is also important to consider patient participation from a collective perspective. 5 Well-designed collective forums for the patient and public participation can be a motor for change. The power of the collective contrasts with the inequalities between patientprofessional communication at the point of care, when patients are particularly vulnerable. In collective forums, patients could potentially work together to achieve stronger influence than is possible in individual-level interventions, which are inherently asymmetrical. 5 − An easy way to improve quality would be to facilitate face-to-face meetings, encourage participants to listen to each other and to reflect, and encourage the development of these relationships and cooperation methods. 7 their profiles.

| Strengths and limitations
This study has some strengths and limitations. First, we did not include any inpatients from Maghreb, which is the third largest non-Spanish cultural community in the hospital healthcare area.
However, we were able to include informants from Asia Minor, Latin America and Europe, and we ensured a diversity of profiles with different educational levels, cultures and gender. There was also a lack of younger participants, given that inpatients are usually restrictions, visits were strictly limited, which hampered our ability to interview more companions.
Second, we conducted the interviews during hospital admission, which may have limited participants' responses to more positive opinions about professionals and the healthcare process in general.
Moreover, as previously mentioned, limits on visits made it difficult to interview more companions. The inpatient setting could also have led to some processes and structures emerging more frequently than outcomes as indicators of patient safety.
Finally, it was difficult to find quiet spaces for the interviews, although, as previously mentioned, informants could stop the interview at any time. However, performing the interview during the admission facilitated the participation of all patient profiles and avoided the loss of information due to memory bias.

| CONCLUSION
The results of this study indicate that identifying patients' and companions' perspectives on patient' safety and their participation is needed before beginning to design an intervention. Examples shared in our paper show the need for a broader conceptualization of patient safety in hospitals, an explicit description of the cultural characteristics of the communities served and a particular focus on communication between patients, companions and professionals.
Both patients and companions are able to identify needs particular to their diverse profiles and key aspects for their involvement in hospital safety.

| IMPLICATIONS FOR POLICY AND PRACTICE
Because informants defined patient safety as the subjective percep-