Abstract

Introduction

Clinical sustainability in healthcare ensures long-term delivery of high-quality care. A ‘care bundle’ is a set of interventions specifically designed for managing a particular condition, differing from standard care checklists by comprising evidence-based best practices. Care bundles like sepsis management are essential for improving patient outcomes and hospital accreditation processes. We assessed the clinical sustainability of sepsis care bundle use in practice.

Methods

The study conducted in a single department at a tertiary public hospital in Oman utilized the clinical sustainability assessment tool (C-SAT). This cross-sectional study involved healthcare professionals at Royal Hospital, Oman, using the C-SAT to evaluate sepsis care bundle sustainability. Data were collected via a digital survey between June and October 2023. Descriptive statistics summarized demographic characteristics and domain scores. Internal consistency was assessed with Cronbach’s alpha, and multivariable regression analysed the impact of profession on sustainability scores.

Results

A total of 121 (n = 245) healthcare professionals participated in the survey (49% response rate), including 97 nurses (80%) and 24 physicians (19%). Gender distribution showed 86% female and 14% male respondents. Nurses had more experience (12 years) compared to physicians (8 years). Cronbach’s alpha indicated strong internal consistency across various domains, with values ranging from 0.9117 to 0.9541. Clinical sustainability scores for sepsis bundles showed no significant variation between nurses and physicians. Multivariable regression analysis revealed that nurses had significantly lower scores in Engaged Staff & Leadership compared to physicians (mean difference: −4.16, P = 0.014).

Conclusions

The study findings emphasize the impact of various factors, such as workload and organizational structure, on sustainability perceptions, highlighting the need for targeted interventions to enhance collaboration and sustainability in sepsis care bundle delivery.

Introduction

Sepsis is a significant cause of morbidity and mortality in healthcare settings, necessitating effective clinical management strategies such as sepsis care bundle, a set of evidence-based practices designed to improve patient outcomes, has been widely adopted [1]. However, the sustainability of these practices within clinical context and settings remains a critical concern. A published clinical sustainability assessment tool (C-SAT) developed by Washington University in St. Louis [2] sets out with the following: the first objective of implementing the C-SAT assessment is to evaluate clinical sustainability practices in sepsis management and address influencing factors. The second objective is to make recommendations and propose actions to empower sustainability of evidence-based practices related to sepsis control. These recommendations might guide towards improved long-term implementation strategies together with scaling up the use of this tool. By achieving these two objectives, the study not only contributes to the literature on clinical sustainability but also serves as a guide for the hospital’s ongoing commitment to delivering high-quality care for patients with sepsis. Through this research, we anticipate shedding light on the dynamics that underpin the care practices and setting a direction for subsequent sustainability assessments in various clinical contexts.

Healthcare sustainability is a specialized application of this concept within the healthcare sector. This includes efficient use of resources, reducing waste, and integrating sustainable practices in healthcare operations [3, 4]. While clinical sustainability focuses even more narrowly on maintaining and improving clinical practices that ensure the continuous delivery of high-quality patient care, it involves implementing and adhering to clinical guidelines that promote effective, efficient, and safe healthcare services. Sustaining these practices requires ongoing training, evaluation, leadership, and adaptation to emerging evidence and technologies [5, 6]. Healthcare organizations continually adopt innovative ideas, system modifications, and new initiatives that provide favourable outcomes. Despite the success of these changes, organizations frequently encounter challenges in maintaining their new initiatives, regardless of the efficacy of these reforms. Failure to sustain the initiative changes in the healthcare setting may lead to deteriorated outcomes of patients [7].

Research studies usually focus on the initiation and the implementation phases, and sustainability is frequently disregarded [8]. Indeed, the scope of research on sustainability is constrained by the need for long-term investigation [9]. Nevertheless, a clinical sustainability assessment tool instrument has been created by Washington University in St. Louis [2] for assessing clinical sustainability—the capacity of an organization to maintain newly implemented clinical practices over extended periods of time [2].

This pursuit of excellence is encapsulated in the concept of clinical sustainability, particularly within the scope of care bundles [10]. These are integral to achieving and maintaining healthcare accreditations, serving as benchmarks for quality and safety in patient care.

Care bundles, such as those developed for the management of sepsis, are instrumental in this context. They represent a synthesis of best practices that, when applied systematically, have proven to significantly enhance patient outcomes [11]. The integration of these care bundles into clinical routines is not just a medical requirement but is increasingly recognized as a vital component of hospital accreditation processes. Accrediting bodies assess the implementation of such care bundles to ensure that healthcare providers meet established standards of care, reflecting an institution’s commitment to continuous improvement [12].

Methods

Study design, participants, and the tool

The cross-sectional study implemented involves healthcare professionals at the department of medicine (mainly, nurses, physicians, and medical residency programme physicians compromising a total of about 245 practitioners) in the Royal Hospital, Oman, where the sepsis care bundle was used for several years. The sepsis care bundle was used as an example for the clinical sustainability assessment because the sepsis care bundle is an important hospital performance indicator. The survey was conducted between June and October 2023. One department (Department of Medicine) in the hospital was assessed to act as a pilot for the practicality of the use of this assessment tool. Data were obtained from physicians and nurses working at the Department of Medicine and were invited to participate and fill in the questionnaire C-SAT. The hospital’s Department of Medicine was selected as a pilot site to evaluate the practicality of using the C-SAT. Data were collected from physicians and nurses in the Department of Medicine, who were invited to complete the validated C-SAT questionnaire developed by Washington University in St. Louis. The CSAT was administered by first converting it into a Google Form. This digital format facilitated ease of access and response collection. The Google Form link was then distributed via a WhatsApp group managed by focal points (one physician and one nurse from the Department of Medicine), who are members of this research authorship team and are working in the Department of Medicine. The C-SAT responses were scored on a 7-point Likert scale, with items ranging from 1 (to a little or no extent) to 7 (to a very great extent). The scores were averaged for each domain to identify sustainability strengths and challenges, guiding subsequent action planning.

Statistical analysis

Descriptive statistics were used to summarize the demographic characteristics of respondents and the scores of individual and overall domains reported as means with standard deviation (SD) or median with interquartile range, for parametric and non-parametric distributed scores. Categorical data such as gender and profession were described as frequency and percentages. Internal consistency was measured using Cronbach’s alpha for individual construct. A value of more than 0.7 indicates high reliability of the items to measure the key concept/ domain. Stratification by profession was conducted and supported by t-test, Mann–Whitney, or chi-squared to identify univariable association. Multivariable linear regression analyses were performed to measure the effect of profession on clinical sustainability scores after adjusting for predefined confounders, years of experience, and gender. A P-value of less than 0.05 is considered significant statistically.

Ethical approval

Ethical approval was granted by the Research and Ethics Committee of Royal Hospital, Oman (SRC#26 908). The study adhered to institutional and national ethical standards, including the Helsinki Declaration. Consent was included in the survey introduction, informing participants that their participation was optional.

Results

Demographic characteristics

A total of 121 healthcare professionals out of 245 participated in the survey, with an overall response rate of 49%, including 97 (80%) nurses and 24 (19%) physicians. The gender distribution revealed a significant difference, with 104 (86%) of respondents were female compared to 17 (14%) who were male, and the average years of experience was higher among nurses, 12 (6%) years, compared to physicians, 8 (7) years (Table 1).

Table 1.

Demographic characteristics of respondents by clinical profession

 NursesPhysiciansTotalP-value
Total: n (%)97 (80)24 (20)121 (100)
Gender: n (%)
Female90 (87)14 (13)104 (100)0.000
Male7 (41)10 (59)17 (100)
Years of experience. Mean (SD)12 (6)8 (7)0.01
 NursesPhysiciansTotalP-value
Total: n (%)97 (80)24 (20)121 (100)
Gender: n (%)
Female90 (87)14 (13)104 (100)0.000
Male7 (41)10 (59)17 (100)
Years of experience. Mean (SD)12 (6)8 (7)0.01
Table 1.

Demographic characteristics of respondents by clinical profession

 NursesPhysiciansTotalP-value
Total: n (%)97 (80)24 (20)121 (100)
Gender: n (%)
Female90 (87)14 (13)104 (100)0.000
Male7 (41)10 (59)17 (100)
Years of experience. Mean (SD)12 (6)8 (7)0.01
 NursesPhysiciansTotalP-value
Total: n (%)97 (80)24 (20)121 (100)
Gender: n (%)
Female90 (87)14 (13)104 (100)0.000
Male7 (41)10 (59)17 (100)
Years of experience. Mean (SD)12 (6)8 (7)0.01

Internal consistency and reliability

The Cronbach’s alpha calculation yielded values of 0.9117 for staff and leadership engagement, 0.9403 for stakeholder engagement, 0.9297 for organizational readiness, 0.9467 for workflow integration, 0.9369 for implementation and training, 0.9429 for monitoring and evaluation, and 0.9541 for outcomes and effectiveness.

Clinical sustainability scores for sepsis bundle

The average scores yielded by nurses and physicians ranged from 4.4 to 4.7 for nurses and 4.5 to 4.9 for physicians (Figures 1 and 2). The mean difference in the overall score was 2.08 (1.89), which was lower among nurses compared to physicians, with a P-value of 0.273. Supplementary Tables 1–7 display the mean scores for each individual domain stratified by clinical profession, which shows statistically insignificant variation between the two groups.

Overall total and mean scores yielded from the clinical sustainability assessment tool among nursing professionals.
Figure 1

Overall total and mean scores yielded from the clinical sustainability assessment tool among nursing professionals.

Overall total and mean scores yielded from the clinical sustainability assessment tool among physicians.
Figure 2

Overall total and mean scores yielded from the clinical sustainability assessment tool among physicians.

Factors affecting C-SAT scores

Multivariable regression analysis showed strong evidence of an association between CSAT mean score for Engaged Staff & Leadership and respondents’ profession. Mean score is 4.16 units lower in nurses compared to physicians, with 95% confidence interval: −7.48 to −0.85, P-value of 0.014 (Table 2).

Table 2.

ANOVA table with multivariable linear regression analysis

SourceSum of squares (SS)Degrees of freedom (df)Mean squares (MS)Number of observations=119
F (3, 115)= 0.59
Model97.3795975332.4598658Prob > F= 0.6200
Residual6281.61211554.622713R2= 0.0153
Adjusted R2= −0.0104
Total6378.991611854.0592508Root mean squared error= 7.3907
Outcome and effectCoefficientStandard Errort-valueP>|t|(95% CI)
Profession−2.0833451.891911−1.100.273−5.830856–1.664167
Years of experience0.02493440.10683080.230.816−0.1866769–0.2365457
Gender2.0786652.1364180.970.333−2.153168–6.310497
Cons22.861023.1786587.190.00016.56471–29.15733
SourceSum of squares (SS)Degrees of freedom (df)Mean squares (MS)Number of observations=119
F (3, 115)= 0.59
Model97.3795975332.4598658Prob > F= 0.6200
Residual6281.61211554.622713R2= 0.0153
Adjusted R2= −0.0104
Total6378.991611854.0592508Root mean squared error= 7.3907
Outcome and effectCoefficientStandard Errort-valueP>|t|(95% CI)
Profession−2.0833451.891911−1.100.273−5.830856–1.664167
Years of experience0.02493440.10683080.230.816−0.1866769–0.2365457
Gender2.0786652.1364180.970.333−2.153168–6.310497
Cons22.861023.1786587.190.00016.56471–29.15733
Table 2.

ANOVA table with multivariable linear regression analysis

SourceSum of squares (SS)Degrees of freedom (df)Mean squares (MS)Number of observations=119
F (3, 115)= 0.59
Model97.3795975332.4598658Prob > F= 0.6200
Residual6281.61211554.622713R2= 0.0153
Adjusted R2= −0.0104
Total6378.991611854.0592508Root mean squared error= 7.3907
Outcome and effectCoefficientStandard Errort-valueP>|t|(95% CI)
Profession−2.0833451.891911−1.100.273−5.830856–1.664167
Years of experience0.02493440.10683080.230.816−0.1866769–0.2365457
Gender2.0786652.1364180.970.333−2.153168–6.310497
Cons22.861023.1786587.190.00016.56471–29.15733
SourceSum of squares (SS)Degrees of freedom (df)Mean squares (MS)Number of observations=119
F (3, 115)= 0.59
Model97.3795975332.4598658Prob > F= 0.6200
Residual6281.61211554.622713R2= 0.0153
Adjusted R2= −0.0104
Total6378.991611854.0592508Root mean squared error= 7.3907
Outcome and effectCoefficientStandard Errort-valueP>|t|(95% CI)
Profession−2.0833451.891911−1.100.273−5.830856–1.664167
Years of experience0.02493440.10683080.230.816−0.1866769–0.2365457
Gender2.0786652.1364180.970.333−2.153168–6.310497
Cons22.861023.1786587.190.00016.56471–29.15733

Discussion

The study indicated differences in the perception of sustainability between nurses and physicians, highlighting areas of strength and those requiring improvement. The internal consistency of the C-SAT in evaluating clinical sustainability of evidence-informed practices in sepsis management was high; Cronbach’s alpha above 0.9 for individual domains. This finding indicates that C-SAT is highly reliable in measuring what it measures among an inpatient hospital setting. This is a secondary finding for the C-SAT as no similar findings using the C-SAT have been published during the study period.

The significant findings include the difference in C-SAT scores between nurses and physicians, particularly in the domain of Engaged Staff and Leadership. Nurses reported lower engagement compared to physicians, indicating challenges in maintaining sustained engagement among nursing staff. Other studies [13] have highlighted that nurse managers’ supportive leadership and organizational justice are crucial in enhancing nurses’ work engagement. A lack of leadership support and recognition may contribute to disengagement, especially among nurses, who face higher workloads and fewer leadership opportunities. Addressing these issues can improve nurse engagement, job satisfaction, and overall healthcare quality [13].

Furthermore, it has been shown that communication and collaboration between nurses and physicians significantly impact work engagement and overall job satisfaction [14]. Nurses often face higher levels of stress and burnout due to demanding workloads and insufficient staffing, which negatively affect engagement levels [14]. Leadership plays a crucial role in prioritizing work engagement, with ‘supportive leadership styles’ being strongly associated with higher levels of engagement among nurses [15]. In addition, it has been emphasized on the importance of communication between healthcare professionals in encouraging teamwork and reducing burnout [14, 15], highlighting that transformational leadership has the greatest positive impact on job satisfaction, further reinforcing the need for strong, supportive leadership [15]. The lack of leadership opportunities for nurses can contribute to lower engagement scores, as opportunities for professional growth and involvement in decision-making are essential for maintaining high engagement [15]. Empowering nurses through participation in clinical and organizational decisions increases their sense of value and commitment. Together, these strategies lead to a more motivated, satisfied nursing workforce [14, 15].

The inconclusive evidence regarding the effect of profession, years of experience, and gender on C-SAT domain scores suggests that sustainability perceptions are shaped by a complex mix of factors. Some studies [16] highlight that sustaining professionals’ adherence to guidelines (e.g. Sepsis Care Bundle) often requires ongoing efforts, as adherence tends to decline over time without continuous reinforcement and the need to understanding that sustainability in healthcare practice is dynamic [16]. Moreover, another study [17] emphasized that organizational structure and role expectations, particularly within gendered professions like nursing, can create distinct experiences of stress, workload, and motivation [17]. These studies [16, 17] support the idea that sustainability perceptions are heavily influenced by contextual factors related to both professional and organizational norms.

In addition, clinical sustainability involves not only delivering quality care but also considering the broader impact of healthcare practices, such as improving resource efficiency [18]. Hence, by integrating sustainability into clinical care, healthcare systems can improve patient outcomes while aligning with the overarching goals of continuous improvement in areas like sepsis management.

Furthermore, it has been shown that the implementation of evidence-based sepsis care bundles significantly improves outcomes, reducing ICU stays and mortality rates [19]. This approach ensures that healthcare organizations not only improve immediate outcomes but also create a culture of sustainable and evidence-based clinical practices.

This study’s use of the C-SAT provides a structured approach to evaluating clinical sustainability, offering valuable insights into organizational readiness, stakeholder engagement, and the effectiveness of training and implementation strategies. These insights can inform targeted interventions to improve the long-term sustainability of the sepsis care bundle.

The sepsis care bundle [19], as an example, shows the critical role of structured care practices in achieving sustainable healthcare outcomes. A study [20] that emphasizes the importance of early detection and timely intervention through the Surviving Sepsis Campaign (SSC) guidelines, which have proven to reduce sepsis-related morbidity and mortality (22), confirms the criticality of a structured approach to management and the importance of sustaining the approach. In contrast, it is important to address the sustainability paradox which highlights the challenge of sustaining health interventions in the middle of constant change [21]. The Dynamic Sustainability Framework (DSF) emphasizes continuous adaptation and improvement of health interventions to ensure they remain effective in changing real-world contexts [21]. Unlike static models, which assume interventions are optimized before implementation, DSF promotes an ongoing process of refinement, ensuring interventions fit evolving organizational, cultural, and ecological settings [21].

The sepsis care bundle represents a small but critical component of healthcare accreditation processes across various accrediting bodies. Utilizing the C-SAT in conjunction with other care bundles or standards can provide a comprehensive view of sustainability within healthcare practices. This approach can inform healthcare professionals about the current state of these practices and identify areas for improvement. Moreover, the results obtained from the C-SAT tool might demonstrate similarities across different care bundles due to the shared principles and methodologies involved in their implementation. By assessing sustainability across multiple bundles, healthcare organizations can adapt a more holistic understanding of their operational effectiveness and ensure alignment with best practices. This, in turn, facilitates the adaptation of strategies that enhance patient care and promote continuous quality improvement across various clinical settings.

It is also important to share here some of the challenges that were faced while trying to implement the C-SAT tool, such as a good number of users found it challenging to comprehend. Hence, there might be a need to contextualize the use of the C-SAT tool. Maybe even opt for group-based response which has also been suggested by the owners of the actual C-SAT [2]. The objective is not to generate individualized responses over an electronic platform but the intention is to make system-based resolution regarding the ability to sustain certain practices.

The study has several limitations, including a small sample size from a single hospital department, which may limit generalizability. Specifically, our study highlights that nurse engagement in sepsis care is influenced by factors such as workload, burnout, and leadership opportunities—challenges that are not unique to our setting. While these factors may manifest differently across hospitals and departments, they remain critical to the sustainability of sepsis care initiatives. By considering how similar workforce dynamics impact clinical sustainability in other institutions, our findings offer broader applicability. Future research could explore these variables in multiple departments or hospitals to further validate and expand upon our insights.

Conclusion

The study showed that the nurses reported lower engagement possibly due to factors like workload, burnout, and limited leadership opportunities. The findings suggest that sustainability perceptions are influenced by a complex interplay of factors, including workload, organizational structure, and supportive leadership, which varies across professional groups. A recommendation of the study is the importance of aligning clinical practices with sustainability goals, which can enhance healthcare practices and improve patient outcomes. Furthermore, practical solutions to address the identified gaps include implementing leadership training programmes to empower nursing staff and developing targeted strategies to reduce nurse burnout, such as flexible scheduling and wellness initiatives. Additionally, hospitals could benefit from conducting regular assessments of workload distribution and staffing levels to ensure adequate support for clinical teams. In addition, integrating feedback mechanisms that allow nurses to voice concerns and participate in decision-making can also enhance engagement and improve job satisfaction.

Acknowledgements

We acknowledge the survey participants for taking the time to fill the questionnaire.

Author contributions

Jehan Al Fannah (Conceptualization, Methodology, Writing—Original Draft), Hiba Al Naabi (Writing—Original Draft, Visualization), Thuraiya Al Harthi (Formal analysis, Data Curation), Samiha Al Habsi (Investigation), Fatma Al Fahdi (Investigation), Salah Al Awaidy (Review & Editing).

Supplementary data

Supplementary data is available at IJQHC Communications online.

Conflict of interest

None declared.

Funding

No funding was received for this study.

Data availability

Supplementary data attached (Supplementary C-SAT).

References

1.

Ramsdell
 
TH
,
Smith
 
AN
,
Kerkhove
 
E
.
Compliance with updated sepsis bundles to meet new sepsis core measure in a tertiary care hospital
.
Hosp Pharm
 
2017
;
52
:
177
86
.

2.

Malone
 
S
,
Prewitt
 
K
,
Hackett
 
R
 et al.  
The clinical sustainability assessment tool: measuring organizational capacity to promote sustainability in healthcare
.
Implement Sci Commun
 
2021
;
2
:77.

3.

Mehra
 
R
,
Sharma
 
MK
.
Sustainability analytics and modeling
.
Sustain Anal Model
 
2021
;
1
:100001.

4.

Ament
 
SMC
,
de Groot
 
JJA
,
Maessen
 
JMC
 et al.  
Sustainability of professionals’ adherence to clinical practice guidelines in medical care: a systematic review
.
BMJ Open
 
2015
;
5
:e008073.

5.

Proctor
 
E
,
Luke
 
D
,
Calhoun
 
A
 et al.  
Sustainability of evidence-based healthcare: research agenda, methodological advances, and infrastructure support
.
Implement Sci
 
2015
;
10
:88.

6.

Pali-Schöll
 
I
,
Hermuth-Kleinschmidt
 
K
,
Dramburg
 
S
 et al.  
An EAACI review: go green in health care and research. Practical suggestions for sustainability in clinical practice, laboratories, and scientific meetings
.
Allergy
 
2021
;
78
:
2606
22
. doi:

7.

Lymperakaki
 
V
,
Sarafis
 
P
,
Malliarou
 
M
.
Communication, work engagement and caring provision differences between nurses and physicians
.
Int J Caring Sci
 
2021
;
14
:
100
3
.

8.

Dunne
 
H
,
Rizan
 
C
,
Jones
 
A
 et al.  
Effectiveness of an online module: climate-change and sustainability in clinical practice
.
BMC Med Educ
 
2022
;
22
:682.

9.

Chaudoir
 
SR
,
Dugan
 
AG
,
Barr
 
CH
.
Measuring factors affecting implementation of health innovations: a systematic review of structural, organizational, provider, patient, and innovation level measures
.
Implement Sci
 
2013
;
8
:
1
20
.

10.

Lavallée
 
JF
,
Gray
 
TA
,
Dumville
 
J
 et al.  
The effects of care bundles on patient outcomes: a systematic review and meta-analysis
.
Implement Sci
 
2017
;
12
:Article142.

11.

Paul
 
N
,
Buse
 
ER
,
Knauthe
 
A-C
 et al.  
Effect of ICU care bundles on long-term patient-relevant outcomes: a scoping review
.
BMJ Open
 
2023
;
13
:e070962.

12.

Green
 
SA
,
Bell
 
D
,
Mays
 
N
.
Identification of factors that support successful implementation of care bundles in the acute medical setting: a qualitative study
.
BMC Health Serv Res
 
2017
;
17
:Article120.

13.

Kim
 
E
,
Lee
 
JY
,
Lee
 
SE
.
Associations among leadership, resources, and nurses’ work engagement: findings from the fifth Korean Working Conditions Survey
.
BMC Nurs
 
2023
;
22
:191.

14.

Van Servellen
 
G
.
Communication Skills for the Health Care Professional: Concepts, Practice, and Evidence
.
Sudbury, MA, USA
:
Jones & Bartlett Publishers
,
2009
.

15.

Specchia
 
ML
,
Cozzolino
 
MR
,
Carini
 
E
 et al.  
Leadership styles and nurses’ job satisfaction
.
BMC Health Serv Res
 
2021
;
21
:603.

16.

Fleiszer
 
AR
,
Semenic
 
SE
,
Ritchie
 
JA
 et al.  
An organizational perspective on the long-term sustainability of a nursing best practice guidelines program: a case study
.
BMC Health Serv Res
 
2015
;
15
:
1
16
.

17.

Baduge
 
MSDSP
,
Garth
 
B
,
Boyd
 
L
 et al.  
Barriers to advancing women nurses in healthcare leadership: a systematic review and meta-synthesis
.
Eclinicalmedicine
 
2023
;
67
:102354.

18.

Palmer
 
J
.
Sustainability in clinical care: integrating environmental health and patient safety
.
Nat Sustain
 
2022
;
5
:
638
9
.

19.

Sayed
 
ZA
.
Effect of evidence based sepsis care bundle on patient outcome in medical intensive care unit
.
Egypt J Health Care
 
2020
;
11
:
826
34
.

20.

Shakeel
 
S
,
Iffat
 
W
,
Nesar
 
S
 et al.  
Prompt identification of sepsis on hospital floors: are healthcare professionals ready for the implementation of the hour-1 bundle?
 
Trop Med Infect Dis
 
2022
;
7
:291.

21.

Chambers
 
DA
,
Glasgow
 
RE
,
Stange
 
KC
.
The dynamic sustainability framework: addressing the paradox of sustainment amid ongoing change
.
Implement Sci
 
2013
;
8
:117.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

Supplementary data