Exploring the relationship between self-awareness of driving efficacy and that of a proxy when determining fitness to drive after stroke

Authors


  • Tadhg Stapleton MSc, BSc, PG Dip Stats; HRB Research Fellow. Deirdre Connolly PhD, MSc, PG Dip Stats, Dip COT; Lecturer. Desmond O’Neill MA, MD, FRCPI, AGSF, FRCP (Glasg); Associate Professor.

Tadhg Stapleton, Discipline of Occupational Therapy, Trinity Centre for Health Sciences, St. James’s Hospital, James’s Street, Dublin 8, Ireland. Email: tadhg.stapleton@tcd.ie

Abstract

Background/aim:  The inclusion of a driving specific self-awareness measure may assist the clinical screening process to determine fitness to drive after stroke. This article reports on the use of the Adelaide Driving Self-Efficacy Scale (ADSES) and a proxy ADSES for completion by a significant other in assessment of fitness to drive post-stroke.

Method:  A prospective study among a clinical sample of stroke patients was conducted incorporating an off-road occupational therapy assessment, an on-road assessment and a six-month follow-up. Self and proxy driver efficacy ratings were compared with each other at off-road assessment and at six-month follow-up, both ratings were compared with structured on-road ratings of driving performance.

Results:  Forty-six stroke patients (37 men), mean age 63.5 years, were recruited to the study. Thirty-five participants successfully completed the on-road test. ADSES and proxy ratings were high and a ceiling effect was noted. Self and proxy ratings were significantly correlated with each other and both correlated with the on-road assessment ratings. The ADSES ratings were sensitive to the final driving outcome with scores of the restricted driving group significantly lower than the unrestricted group. Proxy ratings showed a decrease at six-month follow-up.

Conclusions:  The ADSES is an easy to administer tool that warrants further use in stroke rehabilitation. Scores on the ADSES differentiated between restricted and unrestricted driving recommendations post-stroke. These preliminary findings indicate its potential use as a proxy measure to assist in identifying patient who are not ready for formal driving assessment.

Introduction

Deficits of awareness present challenges in rehabilitation as patients with awareness deficits may choose activities that are beyond their capabilities and as such place themselves and others at risk (Toglia & Kirk, 2000). Awareness has been described as ‘the capacity to perceive the “self” in relatively “objective” terms while maintaining a sense of subjectivity’ (Prigatano & Schacter, 1991, p. 13). While awareness refers to self-knowledge about one’s capabilities, self-efficacy refers to one’s belief in the effectiveness of engaging these capabilities to successfully execute tasks and achieve goals (Bandura, 1997). Self-efficacy is reliant on the person’s underlying awareness and accurate appraisal of their abilities and limitations (Bandura; Toglia & Kirk). Both concepts are closely linked and it is often difficult to make a distinction between self-awareness and self-efficacy. Awareness deficit has been highlighted as a significant phenomenon in post-stroke rehabilitation, particularly unawareness of cognitive deficit (Hartman-Maeir, Soroker, Ring & Katz, 2002). The importance of assessing and understanding awareness deficits over the course of stroke rehabilitation given the relationship between awareness and rehabilitation outcomes has been highlighted (Ekstam, Uppgard, Kottorp & Tham, 2007; Simmond & Flemming, 2003).

Examination of these constructs would appear pertinent when considering a high-risk activity such as driving. However, self-efficacy and awareness scales are not routinely included within off-road assessment practices, although the need to include such measures has been recommended (Heikkila, Korpelainen, Turkka, Kallanranta & Summala, 1999; Lundqvist & Alinder, 2007; Patomella, Kottorp & Tham, 2008). It is suggested that the patients who have a realistic awareness of their own capabilities may be better able to adjust their behaviour and respond to driver rehabilitation than those with poor self-awareness. In addition, these patients with good awareness are perhaps more likely to be successful in subsequent on-road driving assessment (Lundqvist & Alinder). As such, the inclusion of some measure of awareness in the off-road assessment may assist therapists screen suitable patients to be referred onwards for on-road assessment and may increase the likelihood of success in the on-road assessment.

It is recognised that people generally over rate their own driving abilities when making comparison with the ‘average driver’ (Groeger & Grande, 1996). Similar findings were reported among older drivers despite having had a previous adverse driving event or being independently judged as having driving difficulties (Marottoli & Richardson, 1998). Interestingly, this phenomenon has even been observed in stroke survivors who had not returned to drive post-stroke (Scott et al., 2009). Discrepancy between the stroke or brain-injured person’s rating of their own driving ability and that of a driving therapist, health-care professional, or significant other, have been reported in the literature (Heikkila et al., 1999; Lundqvist & Alinder, 2007). This discrepancy between the person’s self-rating of driving ability and the rating of a proxy may indicate an underlying deficit of awareness (Marottoli & Richardson, 1998).

Previous research has highlighted that the caregivers’ perception of the patient’s fitness was a significant determinant of whether the person drove or not after brain injury (Coleman et al., 2002). The inclusion of a proxy measure in the off-road assessment may help identify patients with impaired awareness and unrealistic expectations with regard to returning to drive post-stroke. Based on Coleman et al.’s findings, the inclusion of a proxy measure may further strengthen the off-road screening process and selection of suitable patients to be referred for on-road testing.

No single gold standard driver self-awareness scale is evident from the literature. Coleman et al. (2002) used a general awareness scale that included just one question on driving competency. Patomella et al. (2008) modified a generic competency rating scale and made it specific to driving for their study. Other measures that are specific to driving have been reported in the literature but are still under development (Classen et al., 2010), or in need of further testing (Kay, Bundy & Clemson, 2009). The Adelaide Driving Self-Efficacy Scale (ADSES) was specifically developed to rate self-efficacy on 12 typical driving tasks (George, Clarke & Crotty, 2007). The ADSES was developed and trialled primarily on stroke patients and was found to distinguish between stroke and non-stroke groups when rating driving self-efficacy. Scores on the scale were significantly lower for stroke patients who failed their on-road assessment than those who passed indicating a relationship between the scale and final driving outcome (George et al.).

This article reports on the findings of a prospective study including a measure of self-awareness in examining the determination of fitness to drive among a sub-acute clinical sample of stroke patients. The aims of the study were to:

  • 1 Examine the relationship between the stroke patient’s self-rating of driving competence as measured by the ADSES, with that of a proxy rater/family member using an adapted ADSES as a proxy measure (ADSES-P).
  • 2 Explore the relationship between the ADSES and the ADSES-P scores with the outcome of the on-road assessment.

Methodology

A prospective observational study composed of an off-road assessment followed by an on-road assessment and a six-month follow-up was set up. Participants were recruited from an acute stroke unit in a large urban teaching hospital in Dublin, Ireland. The study design was influenced by the international literature in the area (Korner-Bitensky, Gelinas, Man-Son-Hing & Marshall, 2005; Unsworth, Lovell, Terrington & Thomas, 2005), and by the local context of service provision (Stapleton & Connolly, 2010; Tan, O’Driscoll & O’Neill, 2011).

Subjects

As the study was descriptive in nature and aimed to reflect typical clinical practice the inclusion criteria were kept deliberately broad. Stroke patients consecutively referred for driving assessment during the data collection period were recruited to the study. The inclusion criteria were that the person be diagnosed with a stroke, deemed medically fit to undertake the off-road and on-road assessments (no evidence of hemianopia or epilepsy) and each participant had been driving up to the time of stroke onset.

Assessment tools

The ADSES (George et al., 2007) is a self-completed 12-item scale. Each item on the scale describes a typical driving activity such as ‘driving in your local area’, ‘driving at night’, ‘driving around a roundabout’, ‘parallel parking’, etc. The person rates how confident they feel at each driving task on a scale of 0–10, where 0 is not confident and 10 is completely confident. The maximum score on the ADSES is 120; higher scores indicate greater confidence in driving ability. The only change made to the ADSES to facilitate its use as a proxy measure (ADSES-P) was to change the initial question to ‘How confident do you feel your family member can complete the following driving tasks safely?’ Construct and criterion validity of the ADSES have been reported and the scale was found to have good reliability and validity (George et al.).

Typical on-road assessment practices in this country do not routinely include the use of any standardised rating scales, and as each participant was assessed on a local route starting from their own home it was not possible to develop a route-specific rating form. Two generic on-road rating forms were sourced and used in this study, the Jewish Rehabilitation Hospital Road Evaluation Form (JRHREF) as described by Mazer, Korner-Bitensky and Sofer (1998), and the Test Ride for Investigating Practical Fitness to Drive (TRIP – Belgian version) as described by De Raedt and Ponjaert-Kristoffersen (2001).

The JRHREF (Mazer et al., 1998) is a generic driving skill rating form that consists of 34 driving skills graded on a 5-point rating scale; maximum score is 170. Inter-rater reliability of the JRHREF on a limited sample size was found to be good (Mazer et al.). The TRIP – Belgian version (De Raedt & Ponjaert-Kristoffersen, 2001) consists of 53 items, 50 of which are rated on a 4-point scale and three items rated on a 3-point scale. The maximum score is 209. The only alteration made to the TRIP was the subscale ‘turning left’; this was changed to ‘turning right’ to allow for the fact that in Belgium driving is on the right whereas in Ireland driving is on the left. Studies reporting the reliability and validity of the TRIP have reported high validity and moderate to high reliability (Akinwuntan et al., 2003, 2005). On both the TRIP and JRHREF, higher scores indicate better performance; there are no recommended cut-off scores to indicate safe or not safe to drive on either scale.

Procedure

Ethical approval was granted by St. James Hospital/Adelaide and Meath Hospitals Research Ethics prior to commencement of the study. The occupational therapy off-road assessment was completed by the first author and included a measure of driving self-efficacy (ADSES), when available a family member was asked to complete an ADSES-P at the time of the off-road assessment. Each participant was then referred for an on-road assessment which is typically completed separately from the off-road assessment by a driving assessor. For the purpose of this study, the researcher was permitted to attend the on-road assessment and sit in the back seat of the car as an observer. Each participant completed a local area on-road assessment in their own car commencing at the participant’s own home. The route and duration of the on-road assessment was determined by the driving assessor. Upon completion of the on-road assessment the final driving decision was made by the driving assessor. The researcher completed a six-month follow-up with those participants who successfully passed the on-road assessment to establish if the person was still driving; at follow-up, they were asked to complete a final ADSES and the family member to repeat the ADSES-P.

The JRHREF and the TRIP were completed by the researcher immediately following each on-road assessment where he was present. The driving assessor agreed to independently complete TRIP forms on a proportion of the participants to check for inter-rater reliability between the researcher and the driving assessor, and also to identify if there was any bias on the part of the researcher as the participant’s performance on the off-road assessment was known to him and might potentially influence his rating during the on-road assessment.

Data were analysed using SPSS for Windows, version 16 (IBM Corporation, New York, NY, USA). Non-parametric analysis was most appropriate given the nature of the rating scales and the fact that the data could not be assumed to be normally distributed.

Results

Over an 18-month period, 46 stroke patients (37 men, 9 women) were referred for a driving assessment. The mean age of the group was 63.5 years (SD: 13.4 years) and the age range was from 29 to 83 years. The median time of off-road assessment post-onset of stroke was two months (range: 1–23 months). All 46 participants completed the occupational therapy off-road assessment and were referred for on-road driving assessment with a driving assessor, 35 completed the on-road assessment.

Eleven participants did not complete the on-road test. Six refused to take the on-road assessment, although all were adamant that they wanted to return to drive at the time of the off-road assessment. Five of these participants stated that they did not feel ready to return to driving when the on-road assessment was arranged. One participant refused to take the on-road test and stated that his medical doctor had already given him clearance to drive. Five others did not complete the on-road test (moved jurisdiction for further rehabilitation, n = 2; medical deterioration, n = 1; suitable car not available, n = 1; deemed unsafe by driving assessor to take the on-road test, n = 1).

The scoring patterns on ADSES and ADSES-P are provided in Table 1. The mean and median proxy ratings of the 11 participants who did not take the on-road test appear lower than the proxy ratings of the participants who completed the on-road test (see Table 1); however, no significant difference was found between the ADSES or ADSES-P scores of the two groups (Mann–Whitney U-test: ADSES, P = 0.904; ADSES-P, P = 0.295).

Table 1. ADSES and ADSES-P scores at off-road assessment and at six-month follow-up
 Initial off-road assessmentSix-month follow-up
ADSES scoresADSES-P scoresSix-month follow-up ADSES scoresSix-month follow-up ADSES-P scores
Mean (SD)
Median (range)
Mean (SD)
Median (range)
Mean (SD)
Median (range)
Mean (SD)
Median (range)
  1. ADSES, Adelaide Driving Self-Efficacy Scale; ADSES-P, Adelaide Driving Self-Efficacy Scale by Proxy; SD, standard deviation.

ADSES score of the 11 participants who refused to take the on-road testN = 10N = 5N/AN/A
105.6 (16.4)94.8 (25.9)
112 (70–120)106 (56–116)
ADSES score of those who took and passed the on-road testN = 35N = 24N = 25N = 18
107.8 (11.6)105.1 (16.1)107.16 (13.5)105 (15)
111 (80–120)112.5 (72–120)111 (75–120)109 (75–120)

ADSES scores were available for all 35 participants who completed the on-road test. A number of participants did not have a close family member available to complete the ADSES-P, some family members refused to complete, or failed to return the forms, subsequently ADSES-P scores were available for 24 of the participants who completed the on-road test. A ceiling effect was noted on all the individual items on both the ADSES and the ADSES-P. Mann–Whitney U-test revealed no significant differences in the overall ADSES and ADSES-P scores at time of the off-road assessment (P = 0.883; see Fig. 1). ADSES and ADSES-P scores of those who took the on-road test were significantly correlated with each other at the time of the off-road assessment (Spearman’s rho = 0.707, P < 0.001).

Figure 1.

Adelaide Driving Self-Efficacy Scale (ADSES) and ADSES by Proxy (ADSES-P) scores at initial assessment for participants who completed the on-road test (n = 24).

All 35 participants who completed the on-road assessment were cleared to return to drive by the driving assessor. Based on feedback and advice given to the participant by the driving assessor following the on-road test, the researcher stratified the participants into those who were suitable to return to unrestricted driving (n = 27), and those who were advised to restrict their driving, that is, local area driving only (n = 8).

Five on-road assessments were completed without the occupational therapy researcher present and as a result JRHREF and TRIP scores were not recorded for these five participants (JRHREF, n = 30). However, as the TRIP scores of the occupational therapy researcher and the driving assessor were significantly correlated (n = 11; Spearman’s rho = 0.632, P = 0.037), the driving assessor had recorded TRIP scores for two of these five participants and these were included for analysis (TRIP, n = 32).

Both the ADSES and ADSES-P scores had significant positive correlation with the on-road ratings (see Table 2). Subgroup analysis was completed comparing scores of the group who were suitable for unrestricted driving and those who were advised to restrict their driving following on-road assessment. Scores on the ADSES, ADSES-P and both on-road assessments (TRIP and JRHREF) were compared between these two driving classification groups using Mann–Whitney U-test. All four measures were sensitive to the final driving classification with lower mean and median scores on each measure noted among the restricted driving group and the differences in scores were statistically significant (see Table 3).

Table 2. Correlations between ADSES ratings and on-road ratings
TRIPJRHREF
  1. *Significance at P < 0.05 level; **significance at P < 0.01 level. ADSES, Adelaide Driving Self-Efficacy Scale; ADSES-P, Adelaide Driving Self-Efficacy Scale by Proxy; JRHREF, Jewish Rehabilitation Hospital Road Evaluation Form; rho, Spearman’s correlation coefficient; TRIP, Test Ride for Investigating Practical Fitness to Drive.

ADSES
 n = 32n = 30
 rho = 0.433, P = 0.013*rho = 0.497, P = 0.005**
ADSES-P
 n = 21n = 19
 rho = 0.507, P = 0.019*rho = 0.614, P = 0.005**
Table 3. Summary scores for each of the driving outcome subgroups on the ADSES, ADSES-P, TRIP and JRHREF
Final driving outcomeUnrestricted drivingRestricted drivingMann–Whitney U-test
Mean (SD)
Median (range)
Mean (SD)
Median (range)
  1. *Significance at P < 0.05 level; **significance at P < 0.01 level; ***significance at P < 0.001 level. ADSES, Adelaide Driving Self-Efficacy Scale; ADSES-P, Adelaide Driving Self-Efficacy Scale by Proxy; JRHREF, Jewish Rehabilitation Hospital Road Evaluation Form; SD, standard deviation; TRIP, Test Ride for Investigating Practical Fitness to Drive.

ADSES
 Unrestricted, n = 27110.8 (10.4) 97.6 (9.9)P = 0.001**
 Restricted, n = 8113 (80–120)100.5 (84–111) 
ADSES-P
 Unrestricted, n = 18110.3 (11.7) 89.3 (18.1)P = 0.012*
 Restricted, n = 6115 (84–120) 81.5 (72–118) 
TRIP
 Unrestricted, n = 25196.6 (10.9)162.4 (27.6)P = 0.002**
 Restricted, n = 7199 (169–209)158 (132–196) 
JRHREF
 Unrestricted, n = 23155.6 (5.6)129.4 (12.2)P < 0.001***
 Restricted, n = 7156 (114–165)132 (112–146) 

Six-month follow-up

At six-month follow-up, 25 ADSES and 18 ADSES-P forms were returned. Scores on both scales were significantly correlated at follow-up (n = 18, spearman’s rho = 0.927, P < 0.001), and there were no significant differences between the scores on both scales at follow-up (Mann–Whitney U-test, P = 0.728). All 25 stroke participants reported that they were driving at six-month follow-up. There was no significant difference between the ADSES scores at initial assessment and six-month follow-up (Wilcoxon signed ranks, P = 0.273), with 12 participants showing a drop in their ADSES score, 11 participants improving their score and two participants remaining the same.

There was a significant difference between the ADSES-P scores at initial assessment and at six-month follow-up, with 11 (60%) respondents showing a drop in their follow-up score (Wilcoxon signed ranks, P = 0.028). Three ADSES-P ratings were improved and four remained the same at follow-up (the four who remained the same at follow-up rated their significant other at the maximum score on both occasions).

Discussion

The ADSES and ADSES-P scores correlated significantly with each other at initial assessment and at six-month follow-up among our sample of stroke patients who successfully completed an on-road assessment. It is interesting that the mean ADSES score of 107 reported by George et al. (2007) among their sample of stroke patients who passed a driving assessment is identical to the mean ADSES score of 107.8 among the stroke participants in our study. Similarly, all 35 participants in our study were deemed fit to return to drive, although a subgroup of eight participants were advised to restrict their driving. We found the ADSES to be sensitive to the final driving outcome with a statistically significant difference in ADSES scores of those stroke patients who were suitable to return to unrestricted driving and those who were advised to restrict their driving.

A ceiling effect was noted on all the individual items of the ADSES scale when rated by both the person themselves and the proxy rater. It is possible that ADSES ratings at initial assessment were based on the participant’s previous driving experience and ability and the potential impact of stroke on current driving ability was not appreciated. At the time of the off-road assessment the majority of our participants were early post-stroke onset, as such it is possible that they may not have had adequate time to develop ‘intellectual awareness’ of the existence or impact of stroke deficits on driving, and as they had not driven since stroke onset they would not have had opportunity to develop ‘online awareness’ of how driving may be impacted by the stroke (Simmond & Flemming, 2003). This possible lack of awareness coupled with previous research showing that people generally overestimate their driving ability (Groeger & Grande, 1996; Scott et al., 2009) may help explain the high ADSES ratings at the time of the off-road assessment.

At six-month follow-up, during which time ‘online awareness’ should have an opportunity to develop through the active engagement in driving, the mean ADSES score remained stable. The proxy scores on the other hand did show a decrease at follow-up, suggesting that the significant other was less confident in the stroke participant’s driving ability after the period of observation of actual driving. This finding highlights the importance of continuing to include the significant other in the decision to return to drive as in previous research (Coleman et al., 2002; Heikkila et al., 1999), and in the future development of rating scales for use in driver rehabilitation (Classen et al., 2010).

The participants who refused or cancelled the on-road test similarly recorded high ADSES scores at the time of the off-road test. Efficacy belief may waver as the time of actual performance draws nearer, particularly if the performance appears formidable or threatening as may be the case with an on-road driving test (Bandura, 1997). Perhaps as the on-road assessment approached, this group realistically re-evaluated their driving competency and readiness for actual driving by cancelling the on-road assessment, thus negating their previous high self-efficacy rating. Assessment of efficacy beliefs is more accurate when conducted in close temporal proximity to the action (Bandura), this highlights a deficiency in practices where off-road and on-road testing is conducted in isolation, and suggests the necessity of ensuring that the off-road assessment occurs in closer temporal proximity to the on-road testing in future clinical practice.

We found that the ADSES-P ratings tended to be lower for the participants who did not take the on-road test, suggesting a more accurate appraisal of the stroke participant’s driving ability by the proxy rater. However, this finding was not statistically significant perhaps because of the small sample size; more research is needed to further explore the usefulness of the ADSES as a proxy measure. However, our finding supports Coleman et al.’s (2002) suggestion that a proxy rater may have a more realistic view on the person’s readiness to resume driving. Continued inclusion of the ADSES as a proxy rating may assist the clinical screening process in identifying patients who may not yet be ready to take an on-road assessment.

The ADSES scores in our study appeared high and it is possible that participants were over-estimating their post-stroke driving competence, or may be demonstrating a lack of awareness of impact of stroke on driving capability as in other studies (Heikkila et al., 1999; Patomella et al., 2008). However, given the fact that we had a 100% success rate in the on-road test, which in itself was an unexpected outcome, it is possible that the ADSES scores were an accurate reflection of the participant’s ability, particularly as the scores on both the ADSES and ADSES-P were significantly correlated with the scores on the on-road rating scales.

Several factors may explain the unexpected 100% on-road success rate observed in our study and also highlight limitations in our study design, and weaknesses in current driving assessment practices in the Irish context. First, each participant completed a local area on-road assessment and was performing what essentially is an over-learned skill (Coleman et al., 2002; Fox, Bowden & Smith, 1998) in a very familiar environment. Second, as reported in the literature, on-road assessment processes often lack standardization and at times the driving decision following on-road assessment can be more subjective than objective (Fox et al.). It is also possible that current clinical pre-driving assessment processes may be effectively screening out inappropriate referral for driving assessment. This may explain the high ADSES scores recorded among the stroke patients referred for driving assessment in our study, and also the unexpected 100% on-road success rate. Exploration of the clinical decision-making process guiding the referral for driving assessment is the subject of our ongoing research.

Limitations

The study findings are limited by the small sample size, particularly the lower number of proxy raters. The fact that there was a 100% success rate in the on-road assessment was an unexpected finding and limits any comparative analysis of the findings. There is a possibility of bias in the on-road scoring as the ratings were completed by the first author who had also completed the occupational therapy off-road assessment. However, attempts were made to eliminate bias by the fact that the final driving decision was made by the driving assessor and not the researcher, and second, good inter-rater reliability was found between the researcher and driving assessor on independent completion of the TRIP for a proportion of the participants.

Summary

In our study, we attempted to include the person themselves and their significant other in the process of determining fitness to drive following stroke. We used the ADSES to elicit the person’s self-perception of fitness to drive, and a proxy rating from the significant other through the ADSES-P. Scores on the ADSES correlated significantly with on-road ratings and were sensitive to the final driving classification of the stroke participants in our study.

Scoring patterns on the ADSES may indicate the likelihood of successful outcome in the on-road assessment or identify stroke patients who may need to restrict their future driving post-stroke. Using the ADSES as a proxy measure may be of assistance in screening out those who may not yet have adequate awareness and insight to return to drive or take an on-road assessment. Given the fact that the patient must bear the financial cost of undertaking an on-road assessment it would seem beneficial to include a simple screening tool in the clinical setting that could guide the appropriateness and optimum timing of such an undertaking.

Conclusion

The use of the ADSES as a proxy measure needs further investigation given the small numbers in our study. However, our findings suggest that continued use of the ADSES and the ADSES-P may assist therapists when screening stroke patients in the clinical setting to determine readiness for referral for on-road assessment. It would not be the intention to replace the on-road assessment with a clinic-based screening, but use of the ADSES may augment current clinical screening practices. From the limited research available, it would appear that an ADSES score of 107 or above may indicate suitability to return to driving. Increased use of the scale and further research may establish more definite cut-off scores to guide clinical decision making in this area.

The ADSES was easy to administer in the clinical setting, the items were clearly understood and the relevance of the questions to driving were clear to both the participant and the significant other. We would suggest that the ADSES and ADSES-P be trialled further in clinical settings by occupational therapists as a potentially useful component to foster self-awareness of driving capability among people recovering from stroke. Administration of the scales may facilitate active participation of the person themselves and their significant others within the overall discussion, and assessment of fitness to drive that should occur as a routine component of stroke rehabilitation.

Acknowledgement

Mr Tadhg Stapleton is funded through a Health Research Board (HRB) Research Training Fellowship for Healthcare Professionals.

Ancillary