Prevalence of suicidal ideation, attempts and associated risk factors in oral cancer patients: a systematic review and meta-analysis

Abstract

Background

Oral cancer, including malignancies of the lip, and oral cavity, is associated with significant psychological distress, increasing the risk of suicidal ideation (SI) and suicide attempts (SA). Despite its severe morbidity and mortality, the prevalence of SI and SA and their associated risk factors remain unclear. This systematic review and meta-analysis aimed to quantify the prevalence of SI and SA among oral cancer patients and identify key psychosocial risk factors.

Methods

A comprehensive literature search was conducted across PubMed, Embase, and Web of Science databases, with an initial search on August 13, 2024, and an update on December 19, 2024. Observational studies reporting SI and SA prevalence in oral cancer patients were included. Study quality was assessed using a modified Newcastle-Ottawa Scale (NOS), and statistical analyses were performed using R version 4.4. Pooled prevalence rates were calculated using a random-effects model, heterogeneity was assessed with I² statistics, sensitivity analysis was performed using a leave-one-out approach, and publication bias was evaluated using Doi plots and LFK index values.

Results

From 5,145 articles, 13 studies from Taiwan, South Korea, Canada, Germany, and the United States, involving 899,765 participants, were included. The pooled prevalence of SI was 13.08% (95% CI, 7.41–22.03) with moderate heterogeneity (I² = 65%), while SA prevalence was 0.32% (95% CI, 0.1–0.7) with high heterogeneity (I² = 99%). Key risk factors included depression (OR = 42.83), dysphoria (OR = 8.06), prior cancer history (OR = 4.53), rural residence, and male gender.

Conclusion

This review revealed a substantial psychological impact faced by oral cancer patients, with notable prevalence rates of suicidal ideation and suicide attempts. Identifying major risk factors, such as depression and dysphoria, underscores the need for integrating tailored mental health interventions into routine oncological care.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-05511-7.

Introduction

The prevalence of suicidal ideation (SI) and suicide attempts (SA) among patients with oral cancer, which includes malignancies of the lip, and oral cavity, is a critical yet underexplored aspect of the disease’s psychological impact. Oral cancer remains a significant global public health challenge, with high morbidity and mortality rates despite advances in diagnostic and therapeutic approaches [1]. In 2020, oral cancer accounted for over 377,000 new cases and approximately 177,000 deaths worldwide, with a disproportionate burden on low- and middle-income countries [2]. Beyond the physical health implications, the psychological toll on patients is substantial and often underappreciated in clinical practice. The visible disfigurement, functional impairments, chronic pain, and stigma associated with oral cancer contribute to profound psychological distress, manifesting in conditions such as anxiety, depression, and, critically, suicidal ideation and attempts [3].

The risk of suicide among cancer patients is a well-documented concern, with evidence suggesting that cancer patients are more likely to die by suicide compared to the general population. However, the specific risks of suicidal ideation and attempts in oral cancer patients—who are particularly vulnerable due to the visible and socially stigmatizing nature of their disease—have not been adequately explored [4]. Previous research on the psychological impact of cancer often aggregates data across different cancer types, potentially obscuring the unique challenges faced by oral cancer patients [5–7]. While existing literature suggests that patients with oral cancer may experience higher levels of psychological distress compared to those with other cancers, the extent to which this distress translates into suicidal thoughts and behaviours is not well characterized [8].

Understanding the prevalence and risk factors for SI and SA in oral cancer patients is essential for informing targeted interventions to mitigate these serious complications. This systematic review and meta-analysis aim to address these gaps by synthesizing the available evidence on the prevalence of suicidal ideation, suicide attempts, and associated risk factors among patients with oral cancer. The findings of this review will have significant implications for clinical practice, research, and policy, emphasizing the need for a holistic approach to oral cancer care that includes robust mental health support.

Methods

The PRISMA-2020 guidelines (Table S1) was followed in conducting this systematic review and meta-analysis [9]. The study protocol was registered in the PROSPERO database (Registration No. CRD42024575488), ensuring methodological transparency and adherence to predefined objectives and procedures.

Eligibility criteria

Eligibility criteria were established to include observational studies that reported on the prevalence or determinants of SI or SA among patients diagnosed with oral cancer, including malignancies of the lip, and oral cavity. Studies were eligible if they were published in English. Only observational studies, such as cross-sectional, cohort, and case-control designs, were included, whereas case series and case reports were excluded due to their limited generalizability.

Database search

A comprehensive search strategy was developed to retrieve relevant studies from PubMed, Embase, and Web of Science databases. The search terms included combinations of keywords and Medical Subject Headings (MeSH) related to “oral cancer,” “suicidal ideation,” and “suicide attempts,” with the final search conducted from the inception of databased to August 13, 2024, and subsequently updated on December 19, 2024. Additionally, a manual search was conducted using Google Scholar, and citation tracking of included studies was performed to ensure all relevant articles were identified. The search results were then imported into the Nested Knowledge software for further processing (Table S2).

Screening of articles

A rigorous two-step screening process was employed using Nested Knowledge to ensure the quality and relevance of the included studies. Initially, two independent reviewers assessed the titles and abstracts to eliminate non-relevant articles. Subsequently, the same reviewers conducted independent evaluations of the full-text articles to determine their eligibility based on predefined criteria. Any discrepancies between the reviewers were resolved through discussion, with the involvement of a third reviewer as needed to reach consensus.

Data extraction

The data extraction was performed using the Nested Knowledge software, with two independent reviewers systematically extracting data on study characteristics, population details, and outcomes. Specific data points included the prevalence of SI and SA, demographic and clinical characteristics of the study populations, and any identified risk factors associated with SI or SA. The extracted data were compared between the two reviewers, and discrepancies were resolved through discussion.

Quality assessment

Using a modified Newcastle-Ottawa Scale (NOS), which assesses the choice of research groups, group comparability, and outcome determination, the quality of the included studies was evaluated. All of these domains were used to rate the studies, and lesser risk of bias was indicated by higher ratings. The overall quality of the evidence was evaluated using the evaluation’s findings [10].

Evidence synthesis

Statistical analyses were conducted using R version 4.4, with a random-effects model (REM) employed to estimate the pooled prevalence of SI and SA, accounting for between-study heterogeneity [11]. Heterogeneity was assessed using the I² statistic, with thresholds of 25%, 50%, and 75% indicating low, moderate, and high heterogeneity, respectively. Where data allowed, subgroup analyses were conducted to explore the impact of demographic, clinical, and psychosocial factors on the outcomes [12, 13]. A leave-one-out sensitivity analysis was performed to assess the robustness of the findings and Doi plot and Luis Furuya-Kanamori (LFK) index were used to assess publication bias. The Doi plot helps visualize the symmetry of meta-analysis data, aiding in the detection of publication bias. The LFK index quantifies this symmetry, where a value of 0 indicates no bias. Deviations from zero, either positive or negative, suggest increasing levels of bias, with values near ± 1 indicating mild bias and values beyond these indicating substantial bias. These tools are essential for maintaining the integrity of our findings by quantitatively assessing the impact of publication bias [14, 15].

Results

Literature search

A total of 5,145 records were identified through database searches, including 3,103 records from Embase, 855 from PubMed, and 1,165 from Web of Science (Fig. 1). After removing 811 duplicate records, 4,334 records were screened. Of these, 4,289 records were excluded based on title and abstract screening, leaving 45 reports for full-text retrieval. Ultimately, 33 full-text articles were excluded for various reasons: reviews (n = 13), editorial or opinion pieces (n = 10), lack of relevant population (n = 4), or outcomes of interest (n = 6). An additional study found through citation search brought the total to 13 studies that met the eligibility criteria for inclusion in the systematic review and meta-analysis. Of these, 9 studies reported data on suicide attempts, while the remaining 4 focused on suicidal ideation.

Fig. 1.

PRISMA-2020 flow diagram showing the study selection process

Summary characteristics included studies

The review included 13 studies—2 from Taiwan, 1 from South Korea, 1 from Germany, 2 from Canada, and 7 from the United States—examining suicidal ideation and suicide attempts among patients with oral cancers (Table 1). The studies employed various designs, including retrospective longitudinal, cross-sectional, and cohort studies, with sample sizes ranging from 126 to 415,268 participants. The mean age of participants, when reported, ranged from 52.9 to 67 years, with diverse gender distributions. Five studies focused on suicidal ideation, while six studies investigated suicide attempts. Psychiatric assessments varied, with some studies utilizing structured clinical interviews, such as the Structured Clinical Interview for DSM-IV-TR Axis I Disorders (SCID-I), and others employing scales like the EORTC QLQ-C30 and the Demoralization Scale; several studies did not specify the tools used. The overall quality of the studies included is moderate to high (Table S3).

Table 1.

Summary characteristics of included studies

Study	Country	Study design	Type of cancer	Age (mean)	Sample Size	Suicidal ideation/ attempted	Events	Risk factors
Misono et al., 2008 [28]	USA	Retrospective cohort study	Oral cavity	NA	51,807	Suicide attempted	191	NA
Kam et al., 2015 [29]	USA	Retrospective cohort study	Oral cancer	NA	120,193	Suicide attempted	396 (61.78 per 100,000 person-years; SMR, 5.23; 95% CI: 3.93–6.54)	NA
Henry et al., 2018 [30]	Canada	Prospective longitudinal study	Oral cancer	NA	126	Suicide ideation	19	NA
Klaassen et al., 2019 [31]	Canada	Retrospective cohort study	Oral cancer	67 [57–75]	10,780	Suicide attempted	74	NA
Chang et al., 2019 [6]	Taiwan	Retrospective Longitudinal Study	Oral cancer	54.1 ± 10.0	224	Suicide ideation	19	Depression (odds ratio [OR]: 42.83, 95% [CI]: 11.62-157.81), dysphoria (OR: 8.06, 95%CI: 2.55–25.54), hypopharyngeal cancer (OR: 4.61, 95% CI:1.14–16.56), previous cancer history (OR: 4.53, 95%CI: 1.30-15.83
Choi et al., 2021 [7]	South Korea	Retrospective cohort study	Lip, Oral Cavity	NA	2,054	Suicide attempted	11	NA
Osazuwa-Peters et al., 2021 [26]	USA	Cross-sectional study	Oral cavity	57.7 [10.3] years	3,803	Suicide attempted	18	NA
Yang et al., 2021 [32]	USA	Population-based retrospective study	Oral cavity	NA	7,020	Suicide attempted	25	NA
Chang et al., 2022 [5]	Taiwan	Cross-sectional study	Oral cancer	52.9 (45.4–59.2)	155	Suicide ideation	18	NA
Yu et al., 2022 [33]	USA	Retrospective study	Oral cancer	NA	415,268	Suicide attempted	224	NA
Stanbouly et al., 2023 [34]	Germany	Cross-sectional study	Oral cancer	NA	139	Suicide ideation	26	NA
Hu et al., 2023 [35]	USA	Retrospective cohort study	Oral cavity	NA	409,033	Suicide attempted	1,157	NA
Wang et al., 2024 [16]	USA	Population-based retrospective study	Oral cancer	≥ 20 years	50,011	Suicide attempted	203	Male: HR: 4.21, 95% CI: 2.61–6.80, White: HR: 5.85, 95% CI: 1.45–23.63, Chemotherapy (Yes): HR: 0.35, 95% CI: 0.13–0.93, Single Primary Tumor: HR: 1.53, 95% CI: 1.13–2.06

^{NA: Not Available; EORTC QLQ−C30: European Organization for Research and Treatment of Cancer Quality of Life Questionnaire; SMR: Standardized Mortality Ratio; 95% CI: 95% Confidence Interval; SCID−I: Structured Clinical Interview for DSM−IV−TR Axis I Disorders}

Meta-analysis

Prevalence of suicidal ideation among patients with oral cancers

The meta-analysis included four studies assessing the prevalence of suicidal ideation among 664 patients with oral cancers (Fig. 2). Using a random-effects model, the pooled analysis indicated an overall prevalence of suicidal ideation of 13.08% (95% CI, 7.41-22.03%). The analysis demonstrated moderate heterogeneity (I² = 65%), suggesting some variability across the studies. The prediction interval ranged from 2.86 to 43.44%, reflecting potential variations in the prevalence estimates when applying these findings to different settings or populations.

Fig. 2.

Forest plot depicting the pooled prevalence of suicidal ideation among oral cancer patients

Prevalence of suicide attempts among patients with oral cancers

The meta-analysis included nine studies examining the prevalence of suicide attempts among patients with oral cancers, involving a total of 1,069,969 participants, with 2,299 recorded suicide attempts (Fig. 3). The pooled prevalence of suicide attempts, calculated using a random-effects model, was 0.33% (95% CI, 0.18-0.57%). The prediction interval ranged from 0.05 to 2%, indicating significant heterogeneity (I² = 99%) among the studies. This high level of heterogeneity suggests considerable variability in the reported rates of suicide attempts across different studies.

Fig. 3.

Forest plot illustrating the pooled prevalence of suicide attempts among oral cancer patients

Associated risk factors

Chang et al., 2019 (6) identified several significant risk factors associated with suicide in oral cancer patients. Depression was the most potent risk factor, with an odds ratio (OR) of 42.83 (95% CI, 11.62-157.81), strongly linking it to suicidal behaviour. Other significant factors included dysphoria, with an OR of 8.06 (95% CI, 2.55–25.54), and hypopharyngeal cancer, with an OR of 4.61 (95% CI, 1.14–16.56). A history of previous cancer also increased suicide risk, with an OR of 4.53 (95% CI, 1.30-15.83). Demographically, individuals in urban and metropolitan areas exhibited a lower suicide risk (ORs of 0.52 and 0.55, respectively) compared to those in rural areas. Non-Hispanic Black, Hispanic, and other non-Hispanic races showed significantly lower risks (ORs of 0.19, 0.36, and 0.53, respectively) compared to Non-Hispanic Whites. Additionally, females had a substantially lower risk of suicide than males, with an OR of 0.23 (95% CI, 0.15–0.33).

Similarly, Wang et al., 2024 [16] reported significant findings concerning the demographic and clinical factors influencing suicide risk among oral cancer patients. Males showed a significantly higher risk with a hazard ratio (HR) of 4.21 (95% CI: 2.61–6.80), while whites faced an elevated risk with an HR of 5.85 (95% CI: 1.45–23.63). Chemotherapy was associated with a reduced risk, marked by an HR of 0.35 (95% CI: 0.13–0.93). Additionally, patients with a single primary tumor had a higher risk (HR: 1.53, 95% CI: 1.13–2.06) compared to those with multiple primaries. Concerning surgical treatment, patients with an unknown surgical status had a lower risk (HR: 0.36, 95% CI: 0.15–0.87) relative to those who did not undergo surgery.

Sensitivity analysis

A leave-one-out sensitivity analysis was conducted to evaluate the robustness of the pooled prevalence estimates for suicidal ideation and suicide attempts among oral cancer patients. For suicidal ideation, the pooled prevalence was 13.08% (95% CI, 7.41–22.03) with moderate heterogeneity (I² = 65%). Omitting individual studies did not significantly alter the overall prevalence, though removing Chan et al., 2019, reduced heterogeneity to I² = 30% and yielded a prevalence of 15.13% (95% CI, 8.11–26.48) (Figure S1). For suicide attempts, the pooled prevalence was 0.3% (95% CI, 0.1–0.7) with high heterogeneity (I² = 99%). Excluding Yu et al., 2022, it reduced heterogeneity to I² = 93% and increased the prevalence to 0.4% (95% CI, 0.3–0.6) (Figure S2). These findings demonstrate the stability of the prevalence estimates while highlighting the variability in heterogeneity across studies.

Publication bias

The Doi plot with an LFK index of -1.22 for the meta-analysis on suicidal ideation indicates mild asymmetry, suggesting a low likelihood of significant publication bias (Fig. 4). However, the Doi plot with an LFK index of 2.58 for the meta-analysis on suicide attempts demonstrates major asymmetry, indicating substantial publication bias.

Fig. 4.

Doi plot representing the publication bias in results of suicidal ideation and suicide attempts

Discussion

This systematic review and meta-analysis assessed the prevalence and psychosocial risk factors associated with SI and SA among oral cancer patients. The findings revealed that 13.08% of patients reported SI and 0.33% attempted suicide, highlighting the considerable psychological distress within this group. The analysis demonstrated a notably high prevalence of SI among oral cancer patients, consistent with broader literature on psychological distress in cancer patients. The observed heterogeneity in SI prevalence (I² = 65%) suggests that the risk of SI varies significantly across different demographic, clinical, and contextual factors. This variability indicates the need for context-specific mental health interventions. Although the prevalence of SA is lower at 0.33%, it remains a critical concern, highlighting the importance of timely and effective suicide prevention strategies. The high heterogeneity in SA (I² = 99%) reflects substantial variability in suicide risk, likely due to differences in study design, population characteristics, and healthcare infrastructure. The critical risk factors identified include depression (OR = 42.83), dysphoria (OR = 8.06), history of previous cancer (OR = 4.53), and hypopharyngeal cancer (OR = 4.61). Demographically, males and rural residents exhibited higher risks, whereas Non-Hispanic Blacks, Hispanics, and other non-Hispanic races showed lower risks compared to Non-Hispanic Whites. Notably, chemotherapy was associated with reduced risk (HR = 0.35), and patients with a single primary tumor showed increased risk (HR = 1.53). Additionally, an unknown surgical status correlated with a lower risk (HR = 0.36), suggesting the importance of comprehensive follow-up care.

The findings of this systematic review align with existing literature that underscores a broad spectrum of psychological and demographic factors contributing to suicidal ideation across various cancer types. For instance, a systematic review by Chen et al., 2023 [17] identified key risk factors for suicidal ideation in cancer patients, including marital status, living alone, PTSD, panic disorder, education level, psychiatric history, social functioning, childhood adversity, financial problems, pain, depression, demoralization, vomiting, residence, and anxiety. Similarly, Jimenez-Labaig et al., 2024 [18] documented significant emotional distress in head and neck cancer patients, highlighting a notable prevalence of mental health concerns across all domains, despite a lower incidence of clinically diagnosed mental health disorders. Tuan et al., 2024 [17] reported an 8.2% prevalence of suicidal ideation among lung cancer patients in Vietnam, with pain, depression, and anxiety markedly associated with increased suicidal thoughts, emphasizing the urgent need for targeted mental health assessments and interventions. These studies collectively emphasize the necessity for comprehensive screening and proactive mental health interventions tailored to the needs of high-risk cancer patients, particularly during the initial months post-diagnosis—a critical period identified by Henson et al. 2018 [19] for heightened suicide risk. Moreover, the integration of targeted mental health services into oncological care is crucial, as findings suggest higher suicide risks associated with specific cancer types such as oral, lung, and pancreatic cancers [20]. Additionally, this meta-analysis highlights the significant role of psychosocial factors in suicide risk, resonating with Filiberti et al., 2001 [21], who noted the loss of autonomy and the fear of becoming a burden as significant contributors to suicide among terminal cancer patients. The identification of depression, demoralization, and dysphoria as key risk factors enrich our understanding of how psychological distress influences suicidal behaviours in cancer patients, advocating for strengthened collaboration between oncologists and mental health professionals to effectively mitigate these risks.The study by Chang et al., 2022 [5] introduces demoralization as a predictive factor for SI among oral cancer patients, marking a pivotal exploration into how demoralization impacts suicide risk. This research reveals that high levels of demoralization significantly increase the likelihood of experiencing SI, thereby extending prior understanding of demoralization syndrome in cancer contexts and highlighting the potential of interventions focused on enhancing spiritual and existential well-being to mitigate suicide risk [22]. Conversely, findings on the impact of cancer type on SI risk show variability. For example, while Chang et al., 2019 [6] noted hypopharyngeal cancer as a significant risk factor, other studies did not replicate this association, suggesting that variations in study populations or the differential manifestation of psychological distress across cancer types could contribute to these inconsistent results.

Unexpectedly, the meta-analysis revealed a lower prevalence of SA compared to SI, which, while initially counterintuitive, aligns with broader cancer research suggesting that many patients may experience suicidal thoughts, but fewer progress to making attempts [23]. This could reflect effective protective factors, such as robust social support systems and timely mental health care [24]. Alternatively, it may indicate underreporting of SA, particularly in retrospective studies where self-harm incidents might not be fully documented [25]. The analysis also uncovered intriguing demographic influences on suicide risk, with rural residence emerging as a significant risk factor for SA among oral cancer patients, as noted by Osazuwa-Peters et al. 2021 [26]. This finding contrasts with the generally lower suicide rates observed in rural populations for other conditions, suggesting that the unique challenges of living with cancer in rural areas—such as limited access to healthcare and increased social isolation—may exacerbate suicide risk. Additionally, the observed decrease in suicide risk over time, likely due to better integration of psychosocial care in oncology, highlights the effectiveness of comprehensive cancer care models in mitigating suicide risk [27].

This meta-analysis is robust, incorporating studies from diverse geographical regions, which enhances the generalizability of the findings. The systematic approach to study identification, screening, and inclusion, coupled with rigorous statistical methods for pooling prevalence rates and assessing heterogeneity, provides reliable estimates of SI and SA among oral cancer patients. The risk of bias was systematically evaluated using a modified Newcastle-Ottawa Scale, ensuring that only studies meeting moderate to high-quality standards were included in the analysis. Sensitivity analysis, particularly the leave-one-out approach, demonstrated that the overall pooled prevalence estimates for suicidal ideation and attempts were robust across the removal of individual studies, although slight variations in heterogeneity were observed. These analyses are essential to validate the stability and reliability of our findings, affirming that the reported prevalence rates are reflective of the data despite the potential methodological constraints.

However, this study is not without limitations that warrant consideration. The inherent heterogeneity among the included studies, indicated by an I² statistic of 99% for suicide attempts, points to substantial variability in study designs, populations, and measurement tools. This diversity likely reflects differing regional practices, diagnostic criteria, and patient demographics, which could affect the generalizability of our findings. Additionally, a meta-analysis for specific risk factors could not be performed due to limited data availability, making it impractical to visually represent these findings in a forest plot. Nonetheless, the identification of key psychosocial risk factors provides valuable insights for targeted clinical interventions. Doi plots and LFK index values, revealed substantial Publication bias. This suggests a potential overrepresentation of studies with significant findings, which could skew the overall conclusions of the meta-analysis. Moreover, the observational nature of the studies included may introduce some bias, and significant asymmetry in the Doi plots indicates that studies reporting non-significant outcomes might be underrepresented, potentially inflating the observed prevalence rates. This requires cautious interpretation of the meta-analytic results. Furthermore, the reliance on observational study designs limits the ability to establish causality between identified risk factors and suicidal behaviours. Despite rigorous statistical adjustments, residual confounding might still affect the outcomes.

The findings of this study have significant public health implications, particularly in the realm of oncological and mental health care integration. Recognizing the high prevalence of suicidal ideation and attempts among oral cancer patients, it becomes imperative to implement comprehensive screening protocols for psychological distress early in the cancer care pathway. Such measures should not only assess depression and anxiety but also consider broader psychosocial factors like demoralization, which have been shown to significantly influence suicide risk. Moreover, these findings underscore the need for targeted mental health interventions that are tailored to the specific psychosocial profiles of different cancer types. Enhancing the collaboration between oncologists and mental health professionals can facilitate the development of more effective, holistic care strategies that address both the physical and psychological needs of cancer patients. Ultimately, these approaches aim to improve patient outcomes and quality of life, demonstrating the critical intersection of mental health and cancer care in public health policy.

Conclusion

This systematic review and meta-analysis revealed substantial psychological distress experienced by oral cancer patients, with 13.08% reporting suicidal ideation and 0.33% attempting suicide. The study identifies key risk factors, including depression, dysphoria, and certain cancer types, emphasizing the need for comprehensive mental health support within cancer care. Early identification and tailored interventions addressing these factors are essential to improving patient outcomes. Strengthened collaboration between oncology and mental health professionals is critical to ensuring holistic care that addresses both physical and psychological needs, ultimately improving the quality of life for oral cancer patients.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: G.B., M.N.K., A.M.G.; Data curation: M.P., T.V., E.M.; Formal analysis: M.P.S., A.S.C., S.S.; Investigation: S.B., P.B., A.A.; Methodology: M.R.K., M.S., B.S.T.; Project administration: M.N.K., B.S.T., G.B.; Resources: A.M.G., T.V., P.B.; Software: E.M., M.S., A.S.C.; Supervision: G.B., M.N.K., B.S.T.; Validation: S.S., M.P., A.M.G.; Visualization: M.R.K., E.M., T.V.; Writing – original draft: P.B., S.B., A.S.C.; Writing – review & editing: M.N.K., M.R.K., G.B.

Funding

This study received no funding.

Data availability

All data generated or analyzed during this study are included in this published article (and its Supplementary information files).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.