Strengths and limitations of this study

The study’s cross-sectional design and large sample of 497 ICU physicians provide robust data for assessing knowledge, attitudes, and practices (KAP) regarding acute respiratory distress syndrome, enhancing the overall validity of the findings.

The application of mediation analysis allows for a detailed understanding of how various factors influence KAPs, informing targeted educational interventions.

The cross-sectional design limits the ability to establish causal relationships between variables, reducing the study’s inferential strength.

Self-reported data may introduce bias and be less reliable than objective measures, impacting the validity of the findings.

Introduction

Acute respiratory distress syndrome (ARDS) is a life-threatening inflammatory lung injury characterised by poor oxygenation, pulmonary infiltration and an acute onset in seriously ill patients.1 Despite regional differences, ARDS accounts for approximately 10.4% of intensive care unit (ICU) admissions worldwide, with a hospital mortality rate of 40%.2 According to the Berlin definition, there are three types of ARDS based on the ratio of the patient’s arterial blood oxygen level (PaO2) to the fraction of inspired oxygen (FiO2): mild (200 mm Hg < PaO2/FiO2≤300 mm Hg), moderate (100 mm Hg < PaO2/FiO2≤200 mm Hg) and severe (PaO2/FiO2≤100 mm Hg).3 In addition to pulmonary infections or aspiration, several extrapulmonary factors can trigger ARDS by initiating an inflammatory cascade that leads to pulmonary damage. These risk factors include sepsis, trauma, extensive blood transfusions, drowning, drug overdoses, fat embolisms, exposure to toxic fumes and pancreatitis.4

The treatment of ARDS involves a combination of respiratory support, fluid management and general supportive measures.5 6 Mechanical ventilation is a fundamental component of management, intending to minimise ventilator-induced lung injury.5 6 Guidelines for mechanical ventilation in ARDS recommend six key interventions, including low tidal volume ventilation (LTV), prone positioning ventilation (PPV), oscillatory ventilation at high frequencies, deciding on higher versus lower positive end-expiratory pressure (PEEP), lung recruitment manoeuvres and extracorporeal membrane oxygenation (ECMO).7 It is worth noting that each of these interventions comes with its own uncertainties and controversy, which can complicate their clinical application.5 6 In a recent prospective longitudinal study conducted across 18 ICUs in mainland China, it was found that the incidence of ARDS appeared to be low. However, both mortality and the withdrawal of life-sustaining care were high. Despite a high level of compliance with the recommended lung-protective strategy, adjunct treatments such as neuromuscular blockade and recruitment manoeuvres were not well implemented. These findings suggest that ARDS management in China needs to be improved.8

Knowledge, attitude and practice (KAP) surveys are commonly used in the health sciences to assess individuals’ understanding, beliefs and actions related to a specific topic of interest.9 Previous KAP studies focused on specific points, such as the use of neuromuscular blockers10 11 or practice, but not on knowledge or attitudes.12–14 KAP studies are particularly useful in determining the gaps, misunderstandings and misconceptions that constitute barriers to the optimal implementation of a set of actions.9 Identifying those barriers could allow the design of tailored educational interventions to improve the KAP of ICU physicians towards ARDS, which could translate into better patient outcomes. According to the KAP theory, knowledge is the basis for practice, and attitudes are the force driving practice.9 The entire KAP continuum should be examined since improving knowledge and attitudes should result in improved practice.

Therefore, this study aimed to assess the KAP of ICU physicians in China towards ARDS. The results could help improve the management of ARDS.

MethodsStudy design and participants

This cross-sectional study involved ICU physicians between September and November 2022. The inclusion criteria were (1) active employment in an ICU, (2) possession of a valid physician’s license and (3) a minimum of 6 months of work experience in an ICU. The exclusion criteria were (1) being on maternity or sick leave during the study period and (2) temporary employment status. This study was approved by the Ethical Review Committee of the First Affiliated Hospital of Guangxi Medical University. All participants provided informed consent. The study data were handled anonymously. All data were kept on the hospital’s server, and access was restricted to the study team.

Procedures

A self-administered questionnaire was developed based on established guidelines, including the ‘Management of Acute Respiratory Distress Syndrome’,7 ‘Acute Respiratory Distress Syndrome Advances in Diagnosis and Treatment’,15 ‘ARDS Clinical Practice Guideline 2021’,16 ‘Guidelines on the Management of Acute Respiratory Distress Syndrome’17 and ‘Intensive Care Medicine Course (edited by Haibo Qiu and Xiangdong Guan)’.18 The questionnaire was revised for content value according to feedback from four experts, consisting of three intensive care medicine specialists and one epidemiologist. A small-scale pretest involving 41 participants was conducted before the official launch of the questionnaire to evaluate face value. The Cronbach’s alpha was 0.811. The pretest participants were asked to identify unclear questions, and the questionnaire was modified accordingly.

The final version of the questionnaire comprised four dimensions (online supplemental file 2): (1) demographic information (age, gender, education level, work experience, hospital classification and type, ICU type and experience in treating ARDS patients), (2) knowledge dimension, (3) attitude dimension and (4) practice dimension. The knowledge dimension included 17 questions, with one point assigned for correct answers and zero points assigned for incorrect or unclear answers, resulting in a score range of 0–17 points. The attitude dimension contained 19 questions related to adopting a patient-centred approach to evidence-based management. Questions 1, 2, 4–11, 13 and 15 were rated on a scale from ‘strongly agree/very likely’ (five points) to ‘strongly disagree/very unlikely’ (one point). The score for question 5 was determined by averaging the scores of the five questions. The total score for this dimension ranged from 12 to 60 points. The practice dimension consisted of 21 questions. Questions 1 (mild/moderate), 2 (moderate/severe), 4 (severe), 7 (moderate/severe), 8 (severe), 9 (severe), 11–20, 22 and 23 were assigned two points each for a correct response and zero points for an incorrect response. Question 21 awarded one point for each option chosen. Thus, the total score for this dimension ranged from 1 to 48 points. Scores exceeding 80% of the total for KAP indicated sufficient knowledge, positive attitudes and active practices. Scores falling between 60% and 80% were considered moderate. Scores below 60% indicated insufficient knowledge, negative attitudes and passive practices.

The study used a convenience sampling method to recruit participants without a specific hospital selection criterion. To ensure that all participants were indeed ICU physicians, recruitment efforts were carried out through various channels. These included reaching out to the researcher’s physician classmates, friends and WeChat groups specialising in critical care medicine, such as those focused on critical care ultrasound and qualification training groups. To increase geographic diversity, physician classmates and friends from several hospitals across the country were invited to assist in distributing the questionnaires. These hospitals included the People’s Hospital of Hunan Province, the First Affiliated Hospital of Sun Yat-sen University and the Second Affiliated Hospital of Harbin Medical University. Additionally, participants from other regions, such as Chongqing, Shaanxi, Sichuan, Zhejiang, Beijing, Inner Mongolia, Shanxi and Shandong, also completed the questionnaires. Electronic questionnaires were distributed through the social media platform WeChat (Tencent, China) using a QR code generated by Wenjuanxing (Changsha Ranxing Information Technology Co., Ltd.).19 Participants were informed about the study’s objectives and requirements before filling out the survey. They were requested to voluntarily complete the questionnaire and submit it to the WeChat group of ICU physicians within their respective hospitals. To ensure data quality, each IP address was limited to one submission, and it was mandatory for participants to complete all questionnaire items. The research team carefully reviewed each questionnaire for completeness, internal consistency and logical coherence.

Sample size calculation

The sample size was calculated using the formula for cross-sectional studies: α=0.05, , where when α=0.05, the assumed degree of variability of p=0.5 maximises the required sample size and δ is an admissible error (which was 5% here). The theoretical sample size was 480, which included an extra 20% to allow subjects to be lost during the study.

Statistical analysis

Statistical analysis was performed using Stata 17.0 (Stata Corporation, College Station, TX, USA). Before analysis, the data underwent normality testing. Continuous data were expressed as mean±SD and compared using analysis of variance. Categorical data were presented as frequency (percentage). The Pearson correlation coefficient was applied to assess the correlations between knowledge scores, attitude scores and practice scores. Univariable regression analyses were performed to identify the factors associated with the KAP dimension scores. The variables with p<0.05 were included in the multivariable regression analyses to identify the factors independently associated with the KAP dimension scores (used as the dependent variables). Mediation analysis was used to elucidate the association between demographic characteristics and KAP scores. In the first round of mediation analysis, all variables that showed significant differences in table 1 were included. Then, non-significant variables were removed to streamline the analysis and ensure the robustness of the results.20 Model fit was evaluated using the root mean square error of approximation (RMSEA), standardised root mean square residual (SRMR), Tucker-Lewis Index (TLI) and Comparative Fit Index (CFI). Two-sided tests with p<0.05 were considered statistically significant.

Table 1

Baseline characteristics and KAP scores

Patient and public involvement

No patients were involved.

ResultsDemographic characteristics

A total of 497 ICU physicians participated in the study, with 258 (51.91%) being males and 282 (56.74%) falling within the age range of 30–40 years. The KAP scores averaged 11.89±2.64 (ranging from 0 to 17), 44.73±4.85 (ranging from 12 to 60) and 18.26±3.43 (ranging from 1 to 48), respectively. These scores indicate that participants have a moderate level of knowledge and a positive attitude but showed relatively inactive practices towards ARDS. Significant differences in KAP scores were observed concerning age, education, hospital classification, hospital type, ICU type and prior experience with ARDS patients (p<0.05; table 1).

Knowledge, attitude and practice dimensions

Participants demonstrated insufficient knowledge of whether severe non-thoracic trauma could result in ARDS and the key components of lung-protective ventilation (accuracy rate=14.89% and 12.27%, respectively; online supplemental table S1). In the attitude dimension, most participants agreed with the key components of lung recruitment and PPV in ARDS treatment. Towards the preference for lung recruitment manoeuvre, 146 (29.38%) physicians chose sustained inflation, 95 (19.11%) chose high-pressure controlled ventilation, 16 (3.22%) chose intermittent sighs, 6 (1.21%) chose incremental sighs and 234 (47.08%) chose incremental PEEP. Furthermore, most (81.49%) of the physicians supported the idea that PPV should be combined with lung recruitment therapy as early as possible. Only 5.83% (5.23%+0.60%) disagreed with the statement, “A small dose of methylprednisolone intravenous injection is used as a treatment in the early phase of severe ARDS” (online supplemental table S1).

In the case of a patient with severe ARDS, the majority of physicians (60.56%) would opt for IPPV to improve patient oxygenation. In the practice assessment, over 90% of participants responded positively towards LVT, monitoring important ventilation parameters during LVT, assessing recruitability before lung recruitment therapy and administering PPV. However, the practice score was notably lower, primarily due to physicians’ limited engagement in practice related to determining the optimal timing for tracheostomy and managing ventilation parameters such as PEEP, PaCO2 and tidal volume (online supplemental table S1). Insufficient working conditions and a lack of work experience accounted for more than 50% of the reasons for not implementing LTV, lung recruitment and PPV (online supplemental table S1).

Participants working in tertiary hospitals and teaching hospitals, as well as those who had previous encounters with ARDS patients, outperformed their counterparts in all three categories (p<0.001). This was particularly evident in their attitudes towards lung recruitment, PPV, the combination of PPV with ECMO therapy, engagement in training and their practices in treating both mild and severe ARDS (online supplemental table S2).

Pearson correlation analysis

Pearson correlation analysis revealed positive correlations between knowledge and attitude, knowledge and practice, as well as attitude and practice. The correlation coefficients were 0.367, 0.582 and 0.314, respectively (p<0.001, table 2).

Table 2

Pearson correlation analysis

Univariable and multivariable logistic regression analyses

Table 3 presents the univariable regression analyses of the factors associated with the KAP dimension scores. The multivariable regression analysis showed that a master’s degree (OR=4.70, 95% CI 1.87 to 11.8, p=0.001), a doctorate degree (OR=7.25, 95% CI 2.09 to 25.1, p=0.002), working in public hospitals below the tertiary level (OR=0.44, 95% CI 0.25 to 0.79, p=0.006), working in a specialised ICU (OR=0.47, 95% CI 0.28 to 0.79, p=0.004) and never treated patients with ARDS (OR=0.36, 95% CI 0.13 to 0.98, p=0.047) were independently associated with the knowledge scores. Working in a specialised ICU was independently associated with the attitude scores (OR=0.36, 95% CI 0.20 to 0.65, p=0.001). The knowledge scores (OR=1.32, 95% CI 1.17 to 1.50, p<0.001) and ages 30–40 (OR=2.12, 95% CI 1.22 to 3.67, p=0.007) were independently associated with the practice scores (table 4).

Table 3

Univariable logistic regression analysis based on the knowledge, attitude and practice dimensions

Table 4

Multivariable logistic regression analysis based on the knowledge, attitude and practice dimensions

Mediation analysis

Mediation analysis found that education (β=0.66, p<0.001), work experience (β=0.35, p<0.001), hospital classification (β=−1.27, p<0.001), ICU type (β=−0.72, p=0.003) and ARDS experience (β=−1.90, p<0.001) had direct effects on knowledge. Knowledge (β=0.46, p<0.001), gender (β=−0.83, p=0.033) and hospital type (β=−1.14, p=0.008) had direct effects on attitude. Education (β=0.30, p=0.001), work experience (β=0.16, p=0.002) and hospital classification (β=−0.58, p<0.001) had indirect effects on attitude. ICU type had direct (β=−1.41, p=0.003) and indirect (β=−0.33, p=0.008) effects on attitude. ARDS experience also had direct (β=−2.89, p<0.001) and indirect (β=−0.88, p<0.001) effects on attitude. Attitude (β=0.07, p<0.001) and hospital type (β=−0.84, p=0.005) had direct effects on practice. Knowledge had direct (β=0.68, p<0.001) and indirect (β=0.03, p=0.019) effects on practice. Education (β=0.47, p<0.001), work experience (β=0.25, p<0.001), hospital classification (β=−0.91, p<0.001), ICU type (β=−0.61, p=0.001) and ARDS experience (β=−1.57, p<0.001) had indirect effects on practice (table 5 and figure 1). The RMSEA was 0.014, the SRMR was 0.015 and both the TLI and the CFI were above 0.995, suggesting a good fit for the model (online supplemental table S3).

Table 5

Mediation analysis

Figure 1

Mediation analysis. ARDS, acute respiratory distress syndrome; ICU, intensive care unit.

Discussion

In the present study, we provided a comprehensive overview of ARDS management through a cross-sectional study. The results showed that ICU physicians in China possessed inadequate knowledge, moderate attitudes and passive practices towards ARDS. These findings suggest that the practices of treating ARDS patients may be improved by enhancing the knowledge and fostering positive attitudes among ICU physicians in China.

ARDS is typically caused by direct lung injuries or indirect lung injuries, such as sepsis, pancreatitis and non-thoracic trauma.5 6 However, only 14.89% of the respondents in our study were aware of this fact. In addition, 87.73% of the participants were unaware of the essential components of protective mechanical ventilation in ARDS. It is important to note that protective mechanical ventilation with low tidal volumes is the established standard of care for patients with ARDS.21 According to the ARDS Network formula, the tidal volume should be reduced to 6 mL/kg of predicted body weight.22 Unfortunately, only approximately 20% of the respondents correctly answered the questions related to LVT, suggesting a pressing need for reinforcement of knowledge in this area. Those results are supported by previous studies that showed variable practices regarding ARDS management.12–14

A study involving 312 emergency medicine residents in the USA revealed that their educational experience and proficiency in managing ventilated patients were positively correlated with their knowledge of mechanical ventilation and comfort level.23 Consistently, we found in the mediation analysis that ICU physicians with prior experience in consulting on ARDS patients outperformed their peers in all three categories, with their experience predicting high scores in practical abilities, although the independent association was observed only for knowledge in the regression analysis. Weiss et al24 reported that physicians believed 92.5% of their patients with ARDS should be treated with LVT, but only a small fraction (7.4%) received this treatment. In another study, by the same research group, the percentage of ARDS patients receiving LVT increased to 19.3%, possibly because clinicians began emphasising the importance of LVT for most ARDS patients.25 These findings suggest a significant gap between perceptions and actual practices towards LVT administration, possibly due to the under-recognition of ARDS and various process-related barriers.2 Additionally, adjunct interventions such as recruitment manoeuvres and PPV have been shown to reduce the mortality of ARDS.4 26 However, Liu et al13 reported that these adjunct interventions were not widely used in mainland China, where only 35.5% of ARDS patients received recruitment manoeuvres and 8.7% of severe patients received PPV. This low utilisation might be attributed to a lack of appropriate equipment or uncertainty about the effectiveness of these interventions. Interestingly, despite the limited knowledge about LVT, a significant majority (93.16%) of respondents initiated LVT in patients with mild ARDS. Furthermore, 93.96% and 97.99% of respondents initiated lung recruitment and PPV, respectively. Lack of experience and inadequate working conditions were identified as the most common reasons for not initiating these interventions. To increase the use of adjunct interventions in China, it is advisable to improve the availability of necessary instruments and supplies. Additionally, expanding opportunities for healthcare professionals to gain experience in managing ARDS patients, possibly through training in higher-level hospitals, could enhance their proficiency in implementing these interventions effectively.

In ARDS patients receiving invasive mechanical ventilation, myorelaxants are used to inhibit patient-initiated active respiration and reduce high tidal volumes.27 According to the Intensive Care Medicine Rapid Practice Guideline,28 it is recommended to administer a continuous infusion of myorelaxant for 48 hours in patients who require deep sedation. This facilitates lung-protective ventilation or PPV. However, a recent epidemiological study conducted in 50 countries has shown that neuromuscular blockade is not widely used in patients with ARDS.2 A study showed that intensivists kept neuromuscular blockers when the patient was deeply sedated.10 A cross-sectional survey conducted in the USA revealed that 82% of ICU physicians support myorelaxant administration after a trial of deep sedation, and 59.34% support it after initiating PPV.10 A study in the UK also showed wide variations among physicians regarding the use of neuromuscular blockers.14 In this study, 59.56% of respondents either strongly or somewhat agreed with the administration of myorelaxants during PPV. These findings suggest that ICU physicians should have a deep understanding of the guidelines and be willing to adjust their clinical practices accordingly.

During the recovery phase of ARDS, many patients require both a tracheostomy and a percutaneous feeding tube. A tracheostomy is beneficial as it facilitates the process of weaning patients off mechanical ventilators and simplifies the clearance of respiratory secretions. Typically, a tracheostomy is performed around 2–3 weeks after the onset of ARDS.1 However, early tracheostomy within the first 7 days after ARDS is associated with a shorter median duration of mechanical ventilation and a shorter median ICU stay.29–31 This early intervention, particularly during the COVID-19 pandemic, has been shown to improve patient outcomes and optimise ICU capacity without increasing mortality.31 A study conducted across ICUs in 50 countries found that, on average, tracheostomies were performed approximately 14 days after the onset of ARDS, with only 27.8% of patients undergoing the procedure on or before the seventh day. It is important to note that while tracheostomy may prolong patient survival, it does not significantly impact 60-day or 90-day mortality.32 Consequently, less than half of the participants in the study supported early tracheostomy, possibly due to its uncertain benefits.

Our mediation and multivariable regression analyses highlight the complex interplay of factors influencing KAP among ICU physicians towards ARDS. The mediation model fit indices confirmed the robustness of the mediation analysis. Treating ARDS patients negatively influenced both knowledge and attitude, possibly due to the complexity and high-stress nature of ARDS management. When physicians treat ARDS patients, they are often confronted with the limitations of current medical knowledge and the difficulties in applying theoretical principles to real-world scenarios. This can lead to a more critical self-assessment, where physicians recognise gaps in their understanding and skills, resulting in lower self-reported knowledge scores. Additionally, the challenging and often frustrating nature of ARDS treatment can impact patients' attitudes negatively, as they may feel overwhelmed or disillusioned with the effectiveness of available treatments and resources. This combination of recognising knowledge gaps and experiencing the intense demands of ARDS management can contribute to the negative direct influence on both knowledge and attitude. Additionally, the negative impact of ICU type and hospital classification on knowledge points to systemic issues in certain hospital settings that may hinder optimal ARDS management. Future efforts should focus on developing comprehensive training modules that not only increase theoretical knowledge but also provide hands-on experience in ARDS management. This could involve simulations, workshops and clinical rotations in specialised ICUs. Furthermore, policy changes at the hospital administration level to support continuous education and practice improvements in ARDS management are essential.

This study has certain limitations. First, being a cross-sectional study, it is not possible to establish causal relationships between variables. Consequently, associations between certain factors and KAP related to ARDS management may be bidirectional or influenced by unmeasured confounders. Second, the data were self-reported, which may introduce less reliability compared with medical records and laboratory measurements due to self-report bias. Furthermore, the collected data were somewhat limited, and some factors, such as educational programmes at occupational sites or physicians’ self-imposed requirements, were not collected. Thirdly, the study’s participant pool was limited to a single geographic region, which restricts the generalisability of our findings to a broader population. Furthermore, participants may differ from non-respondents in their ARDS management experiences and KAP. Although we designed a rigorous questionnaire based on established guidelines, its external validity should be verified in a more extensive and diverse population. Additionally, the use of convenience sampling lacks a comprehensive representation of ICU and participant data. The fact that only 42.05% of respondents had received intensive care training could introduce bias into our results. Therefore, while our study offers valuable insights, it may not fully reflect the broader population due to these limitations in our sampling methodology. Future studies should address these issues.

Conclusions

The results of this study indicate that ICU physicians in China have inadequate knowledge, moderate attitudes and passive practices towards managing ARDS patients. To improve the situation, we recommend the implementation of educational interventions and training programmes. Such initiatives have the potential to significantly enhance physicians’ KAP in this area. Furthermore, we suggest conducting future multicentre studies in different geographic regions. This approach can help broaden the applicability and generalisability of our findings.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statementsPatient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by the Medical Ethics Committee of the First Affiliated Hospital of Guangxi Medical University (2022-E322-01). The requirement for informed consent was waived by the Institutional Review Board of the First Affiliated Hospital of Guangxi Medical University because of the retrospective nature of the study. This work has been carried out in accordance with the Declaration of Helsinki (2000) of the World Medical Association. I confirm that all methods were performed in accordance with the relevant guidelines. Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We would like to express our appreciation to Professor Pinhu Liao, Chief Physician Chaoyan Chen from the Department of Critical Care Medicine of the First Affiliated Hospital of Guangxi Medical University and Associate Professor Tang Xianyan from the School of Public Health of Guangxi Medical University for their valuable input on the questionnaire design. We would also like to extend our appreciation to all survey respondents.