Effect of Accreditation on Patient Safety Culture: Insights from a Brazilian Multicenter Cross-Sectional Study

Abstract

Introduction

As healthcare organizations strive to improve the quality and safety of their services, there is growing recognition of the importance of fostering a patient safety culture to enhance patient safety and improve patient care outcomes. This study aims to evaluate healthcare professionals’ perceptions of patient safety culture in accredited vs nonaccredited hospitals within a network of 68 hospitals in Brazil.

Methods

This cross-sectional, multicenter study included 68 hospitals from a private network. The Hospital Survey on Patient Safety Culture (HSOPSC) was administered across all participating hospitals in September 2022. Hospitals that had been formally recognized for their quality and safety standards were compared with nonaccredited hospitals. Scores for various dimensions of patient safety culture were compared between groups. A logistic regression model was applied to assess the association between the frequency of event reporting in the past 12 months and participant characteristics.

Results

A total of 31,919 healthcare professionals responded to the survey. Compared with nonaccredited hospitals, accredited hospitals reported higher scores in communication openness (3% higher, p = 0.04), frequency of events reported (4% higher, p = 0.02), and overall perception of patient safety (4% higher, p = 0.02). Accreditation was associated with a reduced likelihood of event underreporting (odds ratio = 0.80; 95% CI, 0.74–0.87), and physicians were more likely to underreport compared with nursing staff.

Conclusion

Although accreditation enhances patient safety culture, its effect may be more limited in healthcare networks with robust quality management systems already in place. To drive meaningful improvements, policymakers should go beyond accreditation and prioritize the reinforcement of ongoing institutional safety initiatives. Particular attention should be given to persistent challenges, such as fostering a nonpunitive approach to errors and addressing underreporting of adverse events. A graphical abstract is provided in the supplemental material.

INTRODUCTION

The quality of healthcare services has received significant global attention over the years. A prevalent strategy used to enhance quality is the implementation of the accreditation process, which applies quality management systems. Healthcare organizations often pursue accreditation as a means of self-regulation, enabling them to uphold high standards in healthcare delivery while simultaneously gaining recognition for excellence in patient care. Accreditation programs serve to strengthen the structural and procedural aspects of healthcare services, ultimately leading to improved clinical outcomes. Furthermore, these programs often foster positive and sustained transformations in both organizational and clinical practices, ensuring compliance with adopted standards and promoting ongoing improvement.[1]

Different studies have been carried out to assess the effect of accreditation as a mechanism for improvement within healthcare organizations.[2–4] However, the findings have been inconclusive, suggesting that the association between accreditation and outcomes is weak.[3,5,6] Research on the effects of accreditation programs on quality and safety, particularly in relation to patient safety culture, is scarce.[7] This raises the question of whether the introduction of accreditation programs has influenced staff perceptions of patient safety culture across varying cultural and socioeconomic contexts.[8]

Hospital accreditation is still a relatively recent development in Brazil, with estimates indicating that less than 5% of hospitals hold an accreditation certificate.[9] Despite its growing importance, few studies have assessed the effect of accreditation programs on healthcare professionals’ perceptions of patient safety culture within the country.[9] Standardized processes, which are integral to accreditation, have been shown to reduce variability in care delivery, which may facilitate a progressive maturation among healthcare professionals regarding the reporting of near misses and adverse events,[10] thereby promoting continuous learning and enhancing perceptions of safety. However, it is crucial to note that hospital accreditation alone does not guarantee the establishment of a robust patient safety culture,[11,12] highlighting the need for thorough evaluation of this tool. The methodologic challenges associated with measuring the effects of accreditation are compounded by the complexity and heterogeneity of the hospital organizations.

The aim of this study is to evaluate perceptions of patient safety culture among healthcare professionals in accredited hospitals (AH) and nonaccredited hospitals (NAH) within a healthcare network of 68 private hospitals in Brazil.

METHODS

This cross-sectional study investigated the relationship between hospital accreditation and patient safety culture using the Hospital Survey on Patient Safety Culture (HSOPSC) dimensions. The HSOPSC version 1.0, adopted to Portuguese for the Brazilian context,[13,14] was used to assess organizational and patient safety culture in September 2022 (Supplemental Tables S1 and S2, available online). Ethical approval was obtained from the Hospital São Luiz & Rede D´or and Affiliated Teaching Hospitals Research Ethics Board (protocol #70474123.0.0000.0087, assent CEP 6.143.110). Additionally, this study complied with Resolution 466/2012 of the Brazilian National Health Council. All responses were collected anonymously, and no identifiable personal data were recorded. The survey was conducted through a secure digital platform, ensuring that individual responses could not be traced back to participants. Only aggregated data were analyzed and reported to maintain confidentiality and comply with ethical research guidelines.

Study Population and Survey

The study population comprised healthcare professionals from 68 private general hospitals across five macro-geographic regions of Brazil. The survey included 42 items, with 33 assessed on a five-point Likert scale (ranging from totally agree to totally disagree) and nine items evaluated on a five-point frequency scale (from never to always). Patient safety ratings were categorized as excellent, very good, good, fair, or poor, reflecting healthcare providers’ perspective on patient safety culture.

A web-based platform was developed for administering the HSOPSC survey, allowing participants to complete the questionnaire online at their convenience. Study participation was voluntary; hospitals used simple random sampling to distribute surveys among employees. Apart from the hospital’s name, survey respondents were anonymous.

To evaluate patient safety culture using the HSOPSC, we calculated the percentage of positive responses for each positively framed dimension using the following formula: percentage of positive responses = (number of positive responses to the items within the assessed dimension ÷ total number of valid responses to the items within the assessed dimension [positive, neutral, and negative data]) × 100%. The percentage of negative responses was similarly calculated. Strong areas of patient safety were defined as items with 75% or more positive responses, or 75% or more negative responses for negatively framed items. Fragile areas requiring improvement were those with 50% or less positive responses, and 50% or less negative responses for negatively framed items. The average of all items within a dimension of patient safety culture was calculated to determine the dimensional score.

Hospitals were categorized into two groups based on their accreditation status during the study period (AH or NAH). AH were certified by one of the following organizations: Joint Commission International (JCI),[15] Qmentum International (Accreditation Canada),[16] or Organização Nacional de Acreditação (ONA, the Brazilian accreditation organization).[17] JCI accreditation follows rigorous standards emphasizing patient safety, clinical effectiveness, and continuous quality improvement. Qmentum International assesses healthcare organizations based on leadership, risk management, and person-centered care, ensuring compliance with global best practices. ONA accreditation, structured in three progressive levels, evaluates hospitals on their adherence to safety protocols, process standardization, and institutional maturity in quality management. The choice of accreditation model was determined by each institution based on its strategic priorities, resource availability, and alignment with specific quality and safety goals.

For the final analysis, only fully completed questionnaires were included. This report adheres to the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) guidelines[18] (Supplemental Table S3).

Statistical Analysis

The response rate for the survey was calculated by dividing the number of completed surveys by the total number of surveys distributed. Descriptive statistics were used to present the frequencies of sample characteristics and factors related to patient safety climate. To ensure consistency in interpretation, negatively worded items were inverted so that positive responses reflected a higher score. Consequently, a higher score indicated a more favorable attitude among healthcare workers towards patient safety culture.

Differences in dimension scores between AH and NAH were reported as mean with SD. The magnitude of the effects was assessed using the Cohen d (defined as the difference between two means divided by a standard deviation for the data), which measures the standardized mean difference between two groups, along with its 95% CI. This effect size helped determine the practical significance of differences observed in patient safety culture dimensions. Following conventional interpretation, values of the Cohen d were categorized as small (≤ 0.2), moderate (≈ 0.5), or large (≥ 0.8), where larger values indicated a greater distinction between groups.[19] A logistic regression model was used to assess the association between the frequency of event reporting in the past 12 months and participant characteristics. A multivariate logistic regression analysis was conducted using the direct method to compute adjusted odds ratios (ORs) with corresponding 95% CIs. In this approach, all independent variables were entered into the model simultaneously, without assuming a predefined hierarchy or relative importance. Statistical significance was determined by p-values less than 0.05. All statistical analyses were conducted using R software, version 4.1.2 (R Foundation for Statistical Computing).

RESULTS

Among the 68 participating hospitals, 55 (80%) were accredited (AH) and 13 (20%) were not (NAH). A total of 31,919 responses were analyzed, with geographic distribution across Brazil as follows: 62.2% from the southeast, 27.3% from the northeast, 8.6% from the central-west, 1.2% from the south, and 0.7% from the north. Among respondents, 16,213 (50.8%) were full-time employees, and 11,381 (35.6%) have been working at their organizations for 5 years or more.

Regarding professional roles, physicians comprised 6218 respondents (19.5%) and nursing staff account for 18,851 (59.0%). The most represented work units included intensive care units (25.1%; n = 8030), medical nonsurgical units (22.5%; n = 7205), surgical centers (15.3%; n = 4892), and emergency units (14.4%; n = 4601). Notably, 89.9% (27,711) of the responses came from AH, which account for 13 (19%) of the hospitals included in the survey. The overall response rate was 91.4%.

The overall patient safety culture score was 65.1% (SD = 29.8). Positive response rates exceeded 75%, indicating a strong area of safety, were observed for management support for patient safety (77.5% [SD = 6.6]) and organizational learning (81.8% [SD = 4.7]), highlighting these as strengths in patient safety culture. Conversely, nonpunitive response to errors (43.2% [SD = 5.2]) had the lowest positive response rate, identifying it as the weakest area assessed. Patient safety evaluations showed that 69.3% (n = 22,123) of respondents rated patient safety as excellent or very good, 26.8% (n = 8577) deemed it acceptable, and only 3.8% (n = 1219) rated it as poor.

AH and NAH exhibit similar areas of strengths and weaknesses. Figure 1 presents a heat map comparing patient safety culture dimension scores between AH and NAH, visually highlighting these similarities and differences. AH exhibited significantly higher scores in communication openness (58.5% vs 55.7%; p = 0.04; Cohen d = 0.64; 95% CI, 0.03–1.25), frequency of events reported (61.0% vs 56.7%; p = 0.02; Cohen d = 0.74; 95% CI, 0.12–1.35), and overall perception of patient safety (64.5% vs 60.6%; p = 0.02; Cohen d = 0.72; 95% CI, 0.10–1.34), indicating a stronger patient safety culture in these domains. No significant differences were observed in other dimensions, such as feedback and communication about errors, teamwork across and within units, handoffs and transitions, and staffing (p > 0.05 for all) (Table 1).

Figure 1.

Comparison of patient safety culture dimension scores between accredited and nonaccredited hospitals in a private network of 68 hospitals in Brazil. Total number of responses = 31,919. *p < 0.05.

Table 1.

Participants’ perceptions of patient safety culture—Hospital Survey on Patient Safety Culture (HSOPSC) in Brazil

Safety Culture Dimension	Overall (n = 31,919)	Hospital Accreditation Status		p-value	Cohen d (95% CI)a
		NAH (n = 3208)	AH (n = 28,711)
Communication openness	58.0 (4.5)	55.7 (3.8)	58.5 (4.5)	0.04	0.64 (0.03 to 1.25)
Feedback and communication about errors	61.6 (7.0)	58.7 (6.6)	62.3 (7.0)	0.10	0.22 (−0.09 to 1.13)
Frequency of events reported	60.1 (5.6)	56.7 (5.6)	61 (5.8)	0.02	0.74 (0.12 to 1.35)
Management support for patient safety	77.6 (6.6)	76.6 (6.8)	77.8 (6.6)	0.55	0.19 (−0.42 to 0.79)
Teamwork across units	64.3 (6.6)	64.2 (5.9)	64.3 (6.9)	0.99	0.003 (−0.60 to 0.61)
Teamwork within units	73.3 (4.6)	73.0 (4.5)	73.4 (4.7)	0.81	0.07 (−0.53 to 0.68)
Handoffs and transitions	56.0 (8.3)	55.1 (7.5)	56.3 (8.5)	0.65	0.14 (−0.46 to 0.75)
Organizational learning	81.8 (4.7)	80.7 (4.5)	82.0 (4.8)	0.37	0.28 (−0.33 to 0.88)
Supervisor or manager expectations and actions promoting safety	74.4 (5.7)	74.0 (5.2)	74.4 (6.0)	0.99	−0.002 (−0.61 to 0.60)
Overall perception of patient safety	63.7 (5.5)	60.6 (6.6)	64.5 (5.0)	0.02	0.72 (0.10 to 1.34)
Nonpunitive response to errors	43.2 (5.2)	42.5 (5.7)	43.4 (5.1)	0.55	0.18 (−0.42 to 0.79)
Staffing	52.0 (7.3)	51.0 (7.6)	52.3 (7.2)	0.55	0.19 (−0.42 to 0.79)

Values expressed as mean (SD).

^aCohen d was calculated as the difference between the means of two groups divided by the pooled standard deviation.

AH: accredited hospital; NAH: nonaccredited hospital.

Figure 2 shows the distribution of reported patient safety events over 12 months, comparing responses between AH and NAH. A significant proportion of respondents reported no patient safety events, with a higher percentage observed in NAH (46.2%) compared with AH (41.5%; p < 0.001). When asked how often mistakes that are identified and corrected before reaching the patient are reported, 60.3% (n = 19,252) indicated that this occurs always or most of the time (AH 60.6 vs NAH 57.6%; p < 0.001). When asked how often mistakes that have no potential to harm the patient are reported, 63.0% (n = 20,120) indicated that this occurs always or most of the time (AH 63.0 vs NAH 61.0%; p = 0.01).

Figure 2.

Distribution of reported patient safety events over the past 12 months by hospital accreditation status in a private network of 68 hospitals in Brazil. Total number of responses = 31,919. *partitioned χ² p < 0.05.

A multivariable logistic regression analysis was performed to assess the effect of study variables on event reporting frequency (Table 2). AH have a significantly lower likelihood of no event reporting compared with NAH, with an adjusted OR = 0.80 (95% CI, 0.74–0.86). Among professional categories, support staff had the highest likelihood of not reporting patient safety events (adjusted OR = 4.79; 95% CI, 4.19–5.48), followed by emergency staff physicians (adjusted OR = 4.05; 95% CI, 3.45–4.76) and administrative or reception staff (adjusted OR = 4.02; 95% CI, 3.50–4.62). In contrast, nursing staff, who represent the largest proportion of respondents, had a lower likelihood of nonreporting (adjusted OR = 0.82, 95% CI, 0.60–0.92). A higher weekly workload was associated with an increased likelihood of event reporting. Respondents working more than 40 hours per week had significantly higher odds of reporting events compared with those working less than 20 hours (adjusted OR = 0.30, 95% CI, 0.27–0.34).

Table 2.

Logistic regression analysis of frequency of event reporting in Brazil

Variable	Frequency of Event Reporting in the Past 12 Months		Unadjusted OR (95% CI)	p-value	Adjusted OR (95% CI)	p-value
	1 or more events n (%)	0 events n (%)
Hospital accreditation status
Nonaccredited	1729 (9.3)	1479 (11.0)	1.00		1.00
Accredited	16,819 (90.7)	11,892 (89.0)	0.83 (0.77–0.89)	<0.001	0.80 (0.74–0.86)	<0.001
Staff position
Supervision and management	1753 (9.4)	611 (4.6)	1.00		1.00
Administrative and reception	1652 (8.9)	1748 (13.0)	3.04 (2.71–3.40)	<0.001	4.02 (3.50–4.62)	<0.001
Surgical staff physician	136 (0.7)	201 (1.5)	4.24 (3.35–5.38)	<0.001	2.90 (2.26–3.74)	<0.001
Clinical staff physician	187 (1.0)	178 (1.3)	2.73 (2.18–3.42)	<0.001	2.33 (1.84–2.96)	<0.001
Pharmacist	424 (2.2)	119 (0.8)	0.81 (0.64–1.00)	0.057	1.02 (0.80–1.30)	0.86
Emergency staff physician	484 (2.6)	943 (7.0)	5.59 (4.85–6.45)	<0.001	4.05 (3.45–4.76)	<0.001
Radiology staff physician	79 (0.4)	100 (0.7)	3.63 (2.67–4.96)	<0.001	3.00 (2.16–4.18)	<0.001
Intensive care medical staff	523 (2.8)	724 (5.4)	3.97 (3.44–4.60)	<0.001	3.19 (2.72–3.74)	<0.001
Nursing staff	11,542 (62.2)	6689 (50.0)	0.86 (0.65–0.95)	<0.001	0.82 (0.60–0.92)	<0.001
Nutritionist	356 (1.9)	110 (0.8)	0.89 (0.70–1.12)	0.31	1.16 (0.89–1.49)	0.11
Support	1412 (7.6)	1948 (14.5)	3.96 (3.53–4.44)	<0.001	4.79 (4.19–5.48)	<0.001
Work area or unit
Emergency department	2478 (13.3)	2123 (15.9)	0.86 (0.80–0.92)	<0.001	1.03 (0.95–1.13)	<0.001
Intensive care unit	4632 (24.9)	3398 (25.4)	1.09 (0.85–1.40)	0.40	1.38 (1.06–1.79)	0.46
Laboratory	138 (0.7)	129 (0.9)	0.67 (0.62–0.72)	<0.001	0.87 (0.80–1.1)	0.01
Medicine (nonsurgical)	4572 (24.6)	2633 (19.7)	0.87 (0.80–0.95)	<0.001	1.02 (0.90–1.15)	0.001
Radiology	973 (5.2)	876 (6.5)	1.05 (0.94–1.17)	0.66	0.82 (0.72–1.4)	0.46
Rehabilitation	170 (0.9)	273 (2.0)	1.87 (1.54–2.29)	<0.001	1.13 (0.91–1.42)	0.27
Surgery	2936 (15.8)	1956 (14.6)	0.78 (0.72–0.84)	<0.001	1.07 (0.97–1.17)	0.17
Othera	2649 (14.2)	1983 (14.8)	1.00		1.00
Work experience (y)
<1	2438 (13.1)	3373 (25.2)	1.00		1.00
1–5	9096 (49.0)	5631 (42.1)	0.45 (0.42–0.48)	<0.001	0.47 (0.44–0.50)	<0.001
6–10	3952 (21.3)	2345 (17.5)	0.43 (0.40–0.46)	<0.001	0.46 (0.42–0.50)	<0.001
11–15	1739 (9.4)	1096 (8.2)	0.46 (0.42–0.50)	<0.001	0.47 (0.42–0.53)	<0.001
16–20	743 (4.0)	468 (3.5)	0.46 (0.40–0.52)	<0.001	0.44 (0.38–0.51)	<0.001
>20	580 (3.1)	458 (3.4)	0.57 (0.50–0.65)	<0.001	0.50 (0.42–0.59)	<0.001
Experience in current position (y)
<1	795 (4.2)	1037 (7.7)	1.00		1.00
1–5	5860 (31.5)	4365 (32.6)	0.57 (0.52–0.63)	<0.001	0.94 (0.83–1.05)	0.27
6–10	4783 (25.8)	3129 (23.4)	0.50 (0.45–0.56)	<0.001	0.93 (0.82–1.05)	0.23
11–15	3572 (19.2)	2281 (17.0)	0.49 (0.44–0.54)	<0.001	0.97 (0.85–1.10)	0.60
16–20	1921 (10.3)	1,84 (9.6)	0.51 (0.46–0.58)	<0.001	1.05 (0.92–1.21)	0.45
>20	1617 (8.7)	1275 (9.6)	0.60 (0.54–0.68)	<0.001	1.23 (1.06–1.43)	0.005
Weekly workload (h)
Less than 20	895 (4.9)	1897 (14.2)	1.00		1.00
20–40	7275 (39.2)	5639 (42.2)	0.37 (0.34–0.40)	<0.001	0.40 (0.36–0.44)	<0.001
More than 40	10,378 (56.0)	5835 (43.6)	0.27 (0.24–0.29)	<0.001	0.30 (0.27–0.34)	<0.001
Geographic region
Central-west	1516 (8.2)	1243 (9.3)	1.00		1.00
North	133 (0.7)	94 (0.7)	0.86 (0.65–1.13)	0.28	0.79 (0.59–1.06)	0.25
Northeast	4595 (24.8)	4109 (30.7)	1.09 (1.00–1.19)	0.04	1.22 (1.11–1.34)	<0.001
Southeast	12,068 (65.0)	7800 (58.3)	0.79 (0.73–0.85)	<0.001	0.78 (0.72–0.85)	<0.001
South	236 (1.3)	125 (0.9)	0.65 (0.51–0.81)	<0.001	0.67 (0.52–0.84)	<0.001

Odds ratios (ORs) and 95% CIs were estimated using multivariate logistic regression with the direct method.

^aPharmacy, hospital supplies, mental health service, and nutrition.

DISCUSSION

Enhancing the perception of patient safety culture within healthcare is critical for improving quality, efficiency, and overall patient outcomes. Our study highlights that key dimensions that influence patient safety culture, such as nonpunitive responses to errors and open communication, still deserve attention. Although AH and NAH exhibited similar areas of strength and weakness, three dimensions showed notable differences: communication openness, frequency of events reported, and overall perception of patient safety. In a private healthcare network with standardized protocols and quality management systems, NAH often develop patient safety culture comparable to that of their accredited counterparts, suggesting that formal accreditation may not be the sole determinant of a strong patient safety culture. Additionally, institutional benchmarking, participation in quality improvement networks, and adherence to national safety regulations may contribute to improvements in patient safety even in the absence of accreditation.

Although studies in Brazil’s public healthcare system have reported a frail patient safety culture across most analyzed dimensions,[20,21] our study in the private sector identified two key areas in need of improvement. Differences between these systems, particularly regarding resource allocation for quality programs, may explain these findings, as the private sector typically implements more robust patient safety initiatives.[20,21] However, research linking accreditation to improvements in quality and patient safety culture in Brazil remains limited, likely due to the developing maturity of quality and safety science, highlighting the need for continued investigation.

Accreditation is a multifaceted intervention that should be tailored to different healthcare environments, ensuring both immediate performance improvements and long-term process tracking. In our findings, AH scored higher in communication openness than NAH, a critical dimension that reflects the ability of teams to question decisions when patient safety is at risk. However, evidence suggests that many hospital staff members lack the confidence or competence to effectively voice concerns.[22,23] Interestingly, a comparative analysis among nurses revealed that communication openness was rated lower in AH, suggesting a paradox where accreditation does not necessarily enhance communication practices.[24]

Moreover, the dimension of nonpunitive response to errors received the lowest score across both AH and NAH in our study. This finding aligns with previous research that identifies this dimension as a persistent challenge in promoting patient safety management.[25–27] A cross-sectional study conducted in South American countries observed that nonpunitive response to errors was the dimension with the lowest score (37.5%).[28] The lack of differences between AH and NAH in this dimension underscores the need for further initiatives and research to address this important area of safety.

A punitive culture, often reinforced by hierarchical structures and fear of professional repercussions, can discourage healthcare professionals from reporting errors, limiting opportunities for systemwide learning and improvement. Establishing a just culture, in which errors are analyzed constructively rather than punitively, is essential to strengthening patient safety practices. Strategies such as leadership training in psychological safety, clear institutional policies promoting learning from mistakes, and anonymous reporting mechanisms may help overcome these barriers and foster a more open patient safety culture.

The effect of these challenges is further reflected in our finding that 41.9% of professionals did not report any patient safety events over the past 12 months. This low reporting rate may be partially explained by the fear of repercussions, particularly in environments where nonpunitive responses to errors are not well established. Additionally, a lack of awareness or training on what constitutes a reportable event may lead professionals to overlook certain incidents. Another key factor is the perceived lack of effect, where professionals may believe that reporting does not lead to meaningful improvements.

Our results also indicate an opportunity for improvement by shifting the focus of events reporting from blame to problem-solving and generating actionable solutions. The tendency for underreporting necessitates further investigation to identify barriers that inhibit staff from reporting patient safety events.[29] Notably, NAH exhibited lower scores for frequency of reported events compared with AH, implying that accreditation process may foster a culture of event reporting and analysis.[30] Additionally, our study found that nurses reported more incidents than physicians, with greater consistency in reports between nurses and patients than those between nurses and physicians or between physicians and patients.[31] This tendency for underreporting among physicians has been corroborated by numerous studies that indicate that physician reporting is often neither sensitive nor specific in detecting adverse events.[31–34]

Beyond individual-level factors, systemic barriers such as infrastructure limitations, resource availability, and institutional culture play a critical role in event reporting. Insufficient access to user-friendly reporting systems, lack of feedback on the management of previously reported events, and the absence of organizational incentives for error reporting can discourage healthcare professionals from documenting incidents. Additionally, variations in hospital infrastructure, particularly in institutions with limited digitalization of reporting processes, may further contribute to underreporting. Addressing these systemic issues requires targeted interventions, including improved reporting mechanisms, staff training, and leadership engagement in fostering a patient safety culture.

One hypothesis examined in this investigation was whether professionals in accredited institutions, due to their involvement in a more standardized healthcare system, would demonstrate enhanced critical thinking and a deeper understanding of patient safety culture. This could potentially lead to a more discerning perception of patient safety culture, resulting in lower ratings compared with NAH. However, our findings did not support this concern, as the questionnaire results indicated a significantly higher overall perception of patient safety than NAH.

An observational study conducted in South Korea identified a statistically significant association between patient safety culture and the accreditation process.[35] Conversely, several reviews have highlighted insufficient evidence regarding the effect of accreditation on measurable changes in care quality, health outcomes, and patient satisfaction.[36–38] In this context, an important distinction exists between AH and NAH: the former undergo a rigorous process of improvement and standardization prior to obtaining external validation. However, within this network of private hospitals—where 80% are already accredited—there is a widespread standardization process that may equalize perceptions of patient culture, making them independent of accreditation status.

Accreditation not only enhances quality performance but also provides its greatest benefit when organizations integrate standards into their routine workflows. This integration helps avoid last-minute preparations for surveys and ensures consistent application of evidence-based practices for every patient encounter. Ultimately, although accreditation serves as one of several tools for promoting patient safety, our findings suggest it is not significantly necessary to hasten healthcare professionals’ perception of patient safety culture. Despite the statistically significant differences in scores across three dimensions between AH and NAH, these differences were less than 5 percentage points, indicating limited practical relevance.[26] The relatively small differences (< 5%) observed in certain dimensions, such as communication openness and event reporting, suggest that accreditation alone may not be a transformative factor but rather one component within a broader framework of patient safety culture improvement.

A key constraint for hospitals is the cost of accreditation, a resource-intensive process. Given this, an open question remains regarding the extent to which any positive effects of accreditation, if evident, are sustainable over time. Devkaran and O’Farrell[39] have argued that rigorous empirical studies assessing whether hospitals maintain compliance with quality and patient safety standards throughout the accreditation cycle are lacking. However, an interrupted time series analysis has shown that the tangible effect of accreditation can sustain improvements over time and that once a high level of quality compliance is achieved—following the initial accreditation visit—it is highly likely to be maintained.[40] Additionally, repeated accreditation surveys help reduce variations in quality compliance, improving standardization of best practices, performance, and quality of care over time, thereby reinforcing an organization’s progression commitment to highly reliable practices.[40] Future research should further explore the long-term effects of repeated accreditation cycles on patient safety culture.

This study has several limitations. First, all participating hospitals belong to a private healthcare network, which may limit the generalizability of the findings to broader healthcare systems and introduce potential bias. Secondly, 80% of the hospitals in this study were already accredited, which could have skewed the comparative analysis by reducing observable differences between accredited and nonaccredited institutions. Additionally, there is a potential selection bias, as hospitals that voluntarily seek accreditation may already be more invested in quality improvement initiatives. Another important limitation is the reliance solely on self-reported perceptions, without objective verification of actual safety practices, which could affect the accuracy of the findings. Furthermore, response bias may be present, as participants might have provided socially desirable answers, particularly given the voluntary nature of the survey. The professional representation within the sample was also uneven, with nursing staff contributing 59% of responses, whereas physicians accounted for only 19.5%, potentially influencing the reported perceptions of patient safety culture. Lastly, although some differences were statistically significant, the small effect sizes suggest that the clinical significance of the findings may be limited and do not necessarily indicate areas of substantial practical improvement.

CONCLUSION

This large, cross-sectional, multicenter study demonstrates that accreditation status positively influences perceptions of patient safety culture, particularly in areas such as communication openness and event reporting. The relatively small differences observed between AH and NAH suggest that accreditation alone may not be the primary driver of a strong patient safety culture. Our findings indicate that whereas accreditation serves as a structured framework for improving patient safety, its effect may be more pronounced in settings lacking well-established quality management systems.

To maximize the benefits of accreditation, policymakers and hospital administrators should not only promote accreditation processes but also implement complementary institutional strategies to strengthen patient safety culture. Leadership engagement, continuous staff training, and nonpunitive reporting mechanisms are essential to sustaining improvements beyond accreditation and ensuring long-term safety enhancements across healthcare institutions.

Data Availability

The datasets generated and/or analyzed in the current study are available in Mendeley Data (Muniz da Silva, Leopoldo. Safety Culture Data.2024. Mendeley Data, V1, DOI: 10.17632/p5zfv4ty3m.1). The files associated with this dataset are licensed under an Attribution Noncommercial 3.0 Unported license (CC BY NC 3.0).

Supplemental Material

Supplemental materials are available online with the article.