Improving access to care through transportation and internet use for patients with atypical parkinsonism: A pilot study

Highlights

• •Transportation challenges present a barrier to care access for atypical parkinsonian disorders.
• •Providing transportation can increase access to care for patients with PSP, CBS and MSA.
• •Scheduling complexities and disease progression can impact effectiveness of transportation interventions.

Abstract

Introduction

This quality improvement study aimed to assess whether arrangement of transportation and/or distribution of internet-enabled tablets could improve access to care in a multi-disciplinary Atypical Parkinsonism clinic for patients with Progressive Supranuclear Palsy (PSP), corticobasal syndrome (CBS), and multiple system atrophy (MSA) who 1) lack the ability to arrange transportation to our clinics; 2) lack reliable Internet access/devices for telehealth, or 3) both 1) and 2).

Methods

Patients and care partners completed digitally distributed surveys and were given the option to complete hard copy surveys. The primary outcome variable was change in patient-reported Accessibility survey responses at 6 months, and secondary outcome variables included change in patient-reported assessments (PSP-QoL, CBFS, MSA-QoL) and Zarit Caregiver Burden Scale scores at 6 months.

Results

While limited access to resources was a less frequently reported factor at 6-month follow-up, there was an increase in factors contributing to worse quality of life (QoL) for PSP patients. There was no effect on transportation barriers at 6-month follow-up.

Conclusions

Integrating transportation support into routine care processes can help improve access to care for patients with PSP, CBS, and MSA.

1. Introduction

Neurodegenerative diseases (NDDs) are a group of debilitating and currently incurable illnesses affecting >30 million individuals worldwide with severe negative consequences for patients and their families. NDDs that affect both movement and non-motor function include PSP, CBS, and MSA. Due to disease progression and lack of access to local specialists familiar with these diagnoses, travel to in-person appointments becomes stressful and laborious for patients and caregivers over time [1]. These individuals are at a higher risk of becoming homebound, isolated from specialty care, and are more susceptible to adverse illness-related effects leading to emergency room visits and hospitalizations [2].

In 2017, an estimated six million people in the United States (U.S.) reported barriers to transportation that hindered them from receiving needed medical care [3]. As a result of transportation issues, there became an increased need for telemedicine to deliver specialty care to individuals with Parkinson’s Disease (PD) and related disorders and their caregivers [4]. In particular, previous research has shown that patients with atypical parkinsonian disorders were satisfied with tele-palliative care; however, participants also emphasized the continued need for in-person visits to allow for comprehensive physical examinations and other essential services [5]. This underscores the need to overcome these barriers and facilitate access to expert care. Based on patient feedback, it was determined that there was lack of adequate transportation and lack of reliable high-speed Internet access for some of our PSP, CBS, and MSA patients.

Given the rapid progression and complex care needs in these conditions, access to specialized care may directly influence patient-reported outcomes, including quality of life. QoL assessment in APDs incorporates disease-specific instruments (e.g., PSP-QoL, CBFS, MSA-QoL) alongside broader health-related measures to capture the multidimensional impact of motor and non-motor disease burden [6], [7], [8].

To our knowledge, this is the first study to prospectively evaluate system-level interventions—specifically whether providing transportation arrangements and/or distributing internet-enabled tablets could improve access to care for patients with PSP, CBS, and MSA who: 1) lack the ability to arrange transportation to our clinics; 2) lack reliable internet access or appropriate devices for telehealth; or 3) experience both limitations.

2. Materials and methods

2.1. Participants

The Institutional Review Board (IRB) of the Johns Hopkins Hospital (IRB# 00369111) approved the study protocol, and all participants provided written informed consent.

Inclusion Criteria

• •Men and women
• •Clinically diagnosed possible or probable PSP [9], CBS [10] or MSA [11] who visited the JHU CurePSP Centers of Care clinics
• •Native English speakers or English proficiency
• •A life expectancy > 1 year, so that follow-up assessments were more likely to be completed
• •A dedicated care partner able to attend clinic visits and help with questionnaire completion.
• •The patient must be able to understand the study regimen, its requirements, risks, and discomforts, and was able and willing to sign an informed consent form.

Exclusion Criteria include any of the following:

• •Lack of English proficiency
• •Life expectancy <1 year
• •Lack of dedicated care partner who was able to attend clinic visits and help with questionnaire completion

Inclusion of Women and Minorities. Both men and women of all races and ethnic groups are eligible for this trial.

2.2. Study design

This was a multi-center, quality improvement project conducted at the Johns Hopkins University School of Medicine. This pilot study was designed to assess whether patient access to a multi-disciplinary Atypical Parkinsonism clinic can be improved by providing patients with transportation services and Internet enabled devices. The overall study duration was approximately one year and seven months.

2.3. Data collection

A REDCap accessibility survey was sent to all patients with PSP, CBS and MSA seen in the Johns Hopkins Atypical Parkinsonism Center clinics in person between October 2023 and August 2024… Completed REDCap accessibility surveys were screened, and patients were contacted for an opportunity to participate in the project and given the option of completing in-person or virtual consent. Consented participants completed baseline assessments either in person or virtually, per their preference. Baseline assessments (REALM-SF, Newest Vital Sign, disease-specific Quality of Life/functional assessments, and Zarit Caregiver Burden assessment) were completed within 2 weeks of participants’ scheduled in-person or virtual multi-disciplinary clinic visit (email or printed REDCap survey). Intervention (transportation arrangements for in-person visits or provision of Internet-enabled tablets for telehealth visits) type was determined on the basis of participants’ accessibility survey responses. Participants who completed the baseline assessments received a 6-month (180 days +/– 2 weeks) reassessment (accessibility survey, Zarit Caregiver Burden assessment, disease-specific Quality of Life/functional assessments), collected via email or printed REDCap survey. Cognitive screening in this study was limited to the Montreal Cognitive Assessment (MoCA), which was administered as part of routine clinical evaluation.

2.4. Outcome measures

The Rapid Estimate of Adult Literacy in Medicine (REALM-SF) is a validated seven-item word literacy test used by providers to assess patient health literacy. Scores range from 0 to 7. Lower scores mean lower health literacy below a third-grade level, while higher scores reflect higher health literacy above a high school level [12].

The Newest Vital Sign assessment is a six-question health literacy validated screening tool (Appendix B). Scores range from 0 to 6. A score of 0 to 1 correct answer indicates people are more likely to have limited health literacy. A score of 2 to 3 correct answers indicates someone may have a possibility of limited health literacy. A score of 4 to 6 almost always indicates that health literacy is accurate [13].

Health literacy was assessed using the REALM-SF and the Newest Vital Sign because low health literacy is independently associated with reduced ability to navigate transportation systems to access healthcare, decreased likelihood of successfully using internet-enabled devices and telehealth platforms, and greater need for assistance performing online tasks—factors that directly compound the transportation and digital access barriers targeted by this intervention [14].

The Progressive Supranuclear Palsy-Quality of Life (PSP-QoL) is a patient-reported outcome scale that measures quality of life in patients with PSP. The PSP-QoL is made up of 28 items in six categories: daily activities, behavior, bulbar, ocular motor, limb motor and gait. Scores range from 0, which is normal, to 100. Each item on this scale is graded 0–2 (six items) or 0–4 (22 items) [6].

The Cortical Basal ganglia Functional Scale (CBFS) is a patient/caregiver-reported outcome scale which includes questions for patients with CBS and 4-repeat tauopathies on fourteen items of motor experiences of daily living and seventeen items of non-motor experiences of daily living. These items are rated on a Likert 5-point scale rating function from 0 to 4, where 0 indicates normal or no problems and 4 indicates severe problems [7].

The Multiple System Atrophy Quality of Life (MSA-QoL) is a validated patient-reported health related quality of life outcome scale for patients with MSA. It is a 40-item questionnaire and is composed of three different subscales: motor (14 items), non-motor (12 items) and emotional/social (14 items). The response options for each question range from 0 (no problem) to 4 (extreme problem) with higher total scores reflecting greater quality of life impairment. The MSA-QoL also includes a visual analogue scale (VAS) of how satisfied patients feel (range 0–100, with lower scores indicating lower quality of life) [8].

Zarit Caregiver Burden Scale is a 22-item caregiver-reported questionnaire that measures caregiver burden in patients with dementia. It is based on a 5-point Likert scale that ranges from 0 (never) to 4 (nearly always) with the sum of scores ranging between 0 and 88. Higher scores reflect greater caregiver burden. A score of 17 or more indicates high caregiver burden [15].

2.4.1. Primary outcomes

Primary outcome variables included the participant scores on health literacy measures (REALM-SF, Newest Vital Sign assessment) and change in patient-reported Accessibility survey at 6 months.

2.4.2. Secondary outcomes

Secondary outcome variables included the change in patient-reported assessments (PSP-QoL, CBFS, MSA-QoL) and Zarit Caregiver Burden Scale scores at 6 months.

Given the rarity and heterogeneity of atypical parkinsonian disorders, particularly CBS, this pilot exploratory study was not powered for disease-specific analyses. The sample size was determined by the consecutively enrolled patient cohort within the timeframe of our grant (CARES grant, CurePSP) and was intended to evaluate feasibility and inform future study design.

2.5. Data analysis

The analytic approach was selected to generate robust descriptive summaries and effect size estimates for future larger-scale studies. We focused on an intention to treat analyses and included data for all patients who completed their baseline visit, rather than doing a per-protocol analysis, which would be more limited due to smaller sample size. The primary outcome was the change from baseline to 6 months in REALM-SF, Newest Vital Sign assessment, and Accessibility survey scores, with secondary outcome (PSP-QoL, CBFS, MSA-QoL, and Zarit Caregiver Burden Scale) analyzed as continuous variables. Outcomes were summarized using descriptive statistics. Given small sample sizes and a high attrition rate (61%), along with non-normal distributions, medians and interquartile ranges (IQRs) were reported for each variable to characterize central tendency and dispersion. Baseline MoCA and literacy measure were also reported with medians and IQR to provide Context on cohort characteristics.

To assess changes over time, Hedges’ g was calculated as a measure of effect size for within-group differences between baseline and follow-up. Hedges’ g was selected instead of Cohen’s d to correct for small sample bias. Although medians were used to describe the data, effect sizes were calculated using available continuous raw scores, assuming approximate interval scaling.

Effect sizes were interpreted based on standard benchmarks: small (|g| = 0.2), medium (|g| = 0.5), and large (|g| ≥ 0.8). All analyses were conducted using R Studio version 4.3.2 with the “effsize” package.

3. Results

3.1. Participant characteristics

A total of 248 potential participants were invited to complete the accessibility survey. Of these, 54 individuals (21.8%) completed the survey. Among respondents, 30 of 54 (55.6%) reported transportation barriers with a self-reported impact score of 2 or higher, indicating at least a moderate effect on their ability to participate in research activities (Fig. 1). These 30 respondents were invited to take part in the project. Only 18 of 30 (60%) completed a baseline visit because 12 of those participants who reported transportation barriers on the accessibility survey declined follow-up for access to care. All 18 participants received transportation (door to door, to the clinic and back to the patient’s home) for baseline in-person visits, arranged by study staff through a local transportation agency (FreedomCar). At 6-month follow-up, 16 participants remained eligible (1 declined participation, 1 relocated). Of these 16 participants, 7 completed follow-up visits with transportation provided as above. Nine participants did not complete follow-up due to relocation (n = 4), death (n = 1), hospitalization (n = 1), hospice placement (n = 1), and withdrawal (n = 2) (Fig. 1).

Fig. 1.

Participant flow diagram. Participant flow diagram illustrating the number of individuals screened, excluded, enrolled, and included in the final analysis.

No participants reported a lack of reliable Internet access; consequently, no Internet-enabled devices were provided as part of this study. This is consistent with recent findings that, while access to technology can be a barrier in some populations, it was not a limiting factor in our cohort [16].

Eighteen participants [PSP = 11 (PSP-Richardson syndrome = 6, PSP-Parkinsonism = 2, PSP-Primary gait freezing = 2, PSP-CBS = 1), CBS = 3, MSA = 4 (MSA-C = 3 and MSA-P = 1), median age = 70 [IQR 62–76]; 8 Females) enrolled in the study, completed in-person consent, and underwent baseline assessments. Of these, 15 were in the multi-disciplinary clinic group and 3 had individual visits. At baseline, the median MoCA score was 21/30 (IQR = 16–24), indicating that mild cognitive impairment (as defined by a MoCA score between 18–25) was common in the cohort (Table 1).

Table 1.

Demographic breakdown for participants who were selected for the CARES project.

Diagnosis	PSP	CBS	MSA
	11 (61%)	3 (17%)	4 (22%)
Sex	Female	Male
	8 (44%)	10 (56%)
Age	Median	IQR
	70 years old	62–76 years old
Race	Caucasian	Black or African American
	16 (89%)	2 (11%)
Ethnicity	Hispanic or Latino	Non-Hispanic or Latino
	0 (0%)	18 (100%)
Education	Less than high school completion	High school degree or equivalent	Some college	4-year college degree	Professional degree
	1 (6%)	2 (11%)	3 (16%)	2 (11%)	10 (56%)
MoCA	Median	IQR
	21	16–24
Employment	Employed full-time	Employed part-time	Unemployed	Disabled	Retired
	2 (11%)	1 (6%)	1 (6%)	3 (16%)	11 (61%)
Native language	English	Russian	Farsi
	16 (88%)	1 (6%)	1 (6%)
Combined family income	Less than $9,999	$10,000-$19,999	$20,000-$49,000	$50,000-$99,999	$100,000-$149,000	More than $150,000	Chose not to answer
	1 (6%)	0 (0%)	6 (33%)	5 (28%)	3 (16%)	2 (11%)	1 (6%)
REALM-SF	Median	IQR
Number of words known/7	7	7
Newest Vital Sign	Median	IQR
Number correct/6	2	2–5

Note: REALM-SF: Rapid Estimate of Adult Literacy in Medicine − Short Form.

3.2. Primary outcome

The median REALM-SF score at baseline was 7/7 (IQR = 7) while the median Newest Vital Sign score was 2/6 (IQR = 2–5) indicating high literacy in medical term recognition but low comprehension (Table 1).

In terms of the accessibility survey scores, at the 6-month follow-up, participants reported minimal change in transportation barriers, with a median score of 3 (IQR = 2–5) at baseline (n = 18) and 3.5 (IQR = 2.25–4) at follow-up (n = 10), corresponding to a negligible effect size (Hedges’ g = –0.10). Reported access to resources slightly improved, with baseline median = 3 (IQR = 2–4) and follow-up median = 2.5 (IQR = 2–3), yielding a small effect (Hedges’ g = 0.26) (Table 1).

3.3. Secondary outcomes

Following the baseline visit, 2 of 18 participants (11.1%) withdrew prior to scheduling follow-up (1 declined follow-up, the other relocated to another state), resulting in 16 participants eligible for 6-month screening. 7/18 (38.9%) participants withdrew after scheduling a follow-up appointment but before completing a 6-month follow-up survey. 4 of these 7 patients relocated to another state, 1 of 7 patients passed away, 1 of 7 patients was hospitalized for the duration of the follow-up window, and 1 of 7 patients was placed in hospice care. The 6-month survey was completed by 9 of 16 participants (56.3%). Subsequently, 2 additional participants withdrew after completing their 6-month follow-up survey but before their 6-month follow-up appointments, resulting in 7 of 18 original participants (38.9%) completing the 6-month follow-up (Fig. 1 and Table 2).

Table 2.

Median and IQR values for variables of interest collected from CARES project participants at baseline and 6-month follow-up intervals.

	Baseline			6-month follow-up
	n	Median	IQR	n	Median	IQR	Effect size (Hedge’s g)
Transportation barriers	18	3	2–5	10	3.5	2.25–4	−0.10
Lack of access to resources	18	3	2–4	10	2.5	2–3	0.26
MSA Rating Scale Pt II	4	32		1	33
Total MSA QoL	4	80		1	80
Quality of life for MSA	3	66%		1	41%
PSP Rating Scale	9	49	33–72	4	61	41–75	−0.17
Total PSP QoL	11	70	51–105	4	79	67–97	−0.16
Quality of life for PSP	11	70%	52–79%	4	41%	24–55%	0.69
Total CBFS	3	33	27–44	1	22
Quality of life for CBS	3	37		1	22
Caregiver burden	15	35 Mild to moderate burden	20–46	6	41 Moderate to severe burden	22–55	−0.06

In participants with PSP, the PSP Rating Scale total scores showed an increase in impairment from a median of 49/100 (IQR = 33–72, n = 9) to 61/100 (IQR = 41–75, n = 4), with a small negative effect (Hedges’ g = –0.17). As expected for the rapidly progressive disease, PSP-specific quality of life ratings decreased markedly from 70% (IQR = 52–79%, n = 11) to 41% (IQR = 24–55%, n = 4), corresponding to a moderate-to-large effect size (Hedges’ g = 0.69). One outlier was noted in the PSP cohort, whose PSP Quality of Life score changed from 90% to 0% and PSP Rating Scale changed from 31/100 to 49/100. While changes for this patient were more significant than for others in the PSP cohort, their data were not removed from analysis as these changes could be reasonably observed with the rapidly progressive nature of PSP. Quality of life scale results for the CBS and MSA cohorts are listed in Table 2. Overall caregiver burden increased slightly from a median of 35 (mild-to-moderate; IQR = 20–46, n = 15) to 41 (moderate-to-severe; IQR = 22–55, n = 6), with a near-zero effect size (Hedges’ g = –0.06).

4. Discussion

There was substantial attrition between initial enrollment and 6-month follow-up, consistent with previously reported challenges in retention for Parkinsonism research, particularly in the context of transportation barriers and the need for in-person visits [17]. Notably, the absence of technology barriers in this cohort suggests that future studies may benefit from leveraging telehealth modalities to mitigate transportation-related attrition, as supported by recent literature on telemedicine adoption in PD research. Our finding aligns with prior literature identifying transportation as a significant barrier to clinical trial participation in Parkinsonian populations [18]. REALM-SF and Newest Vital Sign scores reflected that our cohort had high literacy in medical term recognition but low comprehension, respectively. Cognitive impairment (as reflected in subnormal MoCA scores) likely limited participants’ comprehension during Newest Vital Sign assessment and future studies with larger cohorts should formally explore the relationship between health literacy and limitations to care access in patients with atypical parkinsonism.

The higher-than-anticipated dropout rate can in part be attributed to the rapid progression and high morbidity associated with atypical parkinsonian syndromes. This observation is consistent with prior longitudinal and interventional studies, which have documented substantial attrition due to clinical deterioration, increased care needs, and mortality in these populations [18]. In particular, the need for patients to relocate for in-home care—often moving back in with family—reflects the profound loss of functional autonomy and the escalating care requirements that characterize the natural history of these disorders [19]. We acknowledge that by choosing to perform an intention to treat analysis rather than a per protocol analysis, we may have biased the effect of our intervention toward the null. However, we note that there were no significant differences in Transportation Barriers rating scores or Lack of Access to Resources rating scores between the 8 participants who did not complete follow up and the 10 who did.

The small sample size in our study, especially CBS and MSA, limited the statistical power of the study and the generalizability of the findings. This is a recognized challenge in single-center research on atypical parkinsonism, where low prevalence and rapid progression frequently constrain recruitment and retention, as highlighted in recent cohort and trial reports [18]. The clinicopathological heterogeneity of CBS and the clinical variability of MSA and PSP (multiple variants existing for each condition) further complicates case ascertainment and underscores the need for improved diagnostic biomarkers and multicenter collaboration to achieve adequate sample sizes [20].

Despite these limitations, the study identified a small effect size for the reduction in lack of access to resources at six-month follow-up, suggesting that targeted interventions or increased awareness may have modestly improved patients' ability to access care. This aligns with emerging evidence that multidisciplinary and palliative care models can facilitate resource navigation and address unmet needs in parkinsonian syndromes, though persistent gaps remain [21]. The literature also emphasizes that transportation, cost, and geographic barriers continue to impede access, particularly for those in rural or underserved areas, and that telehealth and community-based interventions may partially mitigate these challenges [21].

A medium effect size for decline in quality of life (QoL) was observed in the PSP cohort over six months, consistent with the known trajectory of rapid functional and symptomatic deterioration in this disorder [22]. Prior longitudinal studies have demonstrated that both motor and non-motor symptom burden, as well as neuropsychiatric disturbances, drive QoL decline in PSP and related syndromes, and that these changes often outpace those seen in idiopathic PD [23]. The multidimensional nature of QoL impairment in MSA and PSP, including motor, non-motor, emotional, and oropharyngeal domains, further supports the need for comprehensive outcome assessment and early palliative interventions [24]. The lack of decline in QoL in MSA and CBS is likely multifactorial as certain domains may be impaired early and remain relatively stable over time. Neuropsychiatric symptoms are pervasive and may not correlate directly with motor progression. Furthermore, patients may adapt to their disabilities over time, leading to a recalibration of expectations and coping strategies. Finally, the symptomatic management in our multi-disciplinary clinic setting (where most patients were evaluated) is focused on maintaining function and comfort, which may help stabilize QoL even as the underlying disease progresses [25], [26]. A comprehensive neuropsychological testing was not performed, as the primary focus of this pilot study was feasibility and access to care rather than detailed cognitive phenotyping.

Given the high attrition and small sample size in our study, the findings must be interpreted as preliminary. Limited statistical power and potential attrition bias may have contributed to the modest effect sizes observed, thereby restricting interpretability and generalizability. Larger, multicenter studies with robust retention strategies are needed to confirm these trends and to refine study design. Approaches such as proactive caregiver engagement, flexible follow-up modalities (including telehealth), and integration of palliative care principles may enhance participation and better capture the lived experience of patients with atypical parkinsonism [21]. Future research should also prioritize the development and validation of sensitive, disease-specific outcome measures and explore interventions that address both access to care and quality of life in this vulnerable population.

In terms of diagnosis, serum neuronal exosome biomarkers, particularly α-synuclein and tau aggregates, demonstrate robust diagnostic accuracy for differentiating Parkinson’s disease from atypical parkinsonism, with area under the curve values of 0.86 to 0.98 [27]. Digital health platforms incorporating speech analysis further enhance remote diagnostic capabilities, achieving 93% AUC in distinguishing progressive supranuclear palsy from Parkinson’s disease [28], [29]. Integrating these technologies into clinical practice may substantially improve diagnostic access for underserved populations.

5. Conclusion

In summary, early identification of transportation needs and the integration of transportation solutions into care coordination are essential components of comprehensive atypical parkinsonism management. Although the observed effect sizes in our study were quite modest, this pilot study demonstrates the feasibility of integrating transportation support into routine care for patients with atypical parkinsonian disorders. These findings provide preliminary insight into structural access barriers and will help inform larger, multisite evaluations.

While logistical barriers—including scheduling complexities, variable reliability, and limited patient awareness of available services—can constrain the effectiveness of transportation interventions, these challenges underscore the necessity of proactive, individualized approaches within multidisciplinary care frameworks [26]. Expanding access to telehealth services and embedding transportation support into routine care processes for atypical parkinsonian disorders will likely enhance access to healthcare services, improve quality of life for both patients and caregivers, and mitigate disparities in care delivery. Future research should focus on optimizing implementation strategies and evaluating the impact of transportation integration on clinical outcomes and health equity in atypical parkinsonian populations.

Transportation barriers significantly limit access to specialized care for patients with atypical parkinsonism, highlighting the urgent need for innovative care delivery models [26]. Telemedicine represents a promising solution by enabling remote specialist consultations while reducing travel burden, though successful implementation requires individualization based on patient technological literacy and preferences.

Funding/support

This study was supported by a Collaborative Approaches to Resources, Education and Support (CARES) grant from CurePSP, Inc., for projects aimed at improving care, access, and education for patients with PSP, CBD, and MSA.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work the author(s) used Open Evidence AI in order to improve language and readability. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

CRediT authorship contribution statement

Maria Schmidt: Writing – original draft, Project administration, Methodology, Investigation, Funding acquisition, Data curation. Kyurim Kang: Writing – review & editing, Methodology, Formal analysis, Data curation. Hannah Jackson: Writing – review & editing, Methodology, Formal analysis, Data curation. Claudia Waddell: Writing – review & editing, Methodology, Formal analysis, Data curation. Anna Hall: Writing – review & editing, Methodology, Formal analysis, Data curation. Alexander Pantelyat: Writing – original draft, Supervision, Conceptualization.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: This study was supported by a Collaborative Approaches to Resources, Education and Support (CARES) grant from CurePSP, Incorporated, for projects aimed at improving care, access, and education for patients with progressive supranuclear palsy, corticobasal degeneration, and multiple system atrophy. The funder had no role in the study design, data collection, analysis, interpretation, manuscript preparation, or decision to submit for publication. All authors declare that this financial support is disclosed in accordance with current editorial and reporting standards, which require explicit identification of all sources of funding and their roles in the research process. No other financial or non-financial conflicts of interest relevant to this work are declared.

Appendices A and B. Supplementary data

The following are the Supplementary data to this article: