Infants born extremely preterm (EP, <28 weeks’ gestational age) and/or extremely low birthweight (ELBW, <1000 g) are at higher risk of developmental delay and/or disabilities, such as cerebral palsy (CP), compared with infants born at term. Early detection is paramount to ensure that infants who are at the highest risk of developmental delays are identified early to commence intervention and improve functional outcomes. The General Movements Assessment (GMA) involves visual clinical recognition of patterns of spontaneous infant movement from video recordings and is a key assessment for the diagnosis of high-risk CP but is not universally accessible due to resource constraints, such as limited out-patient follow-up services. The Baby Moves smartphone application (app) offers a novel method of allowing clinicians to partner with parents by providing them with a tool to record their infant’s movements via their smartphone to be used for a remote GMA. This thesis explores the feasibility of Baby Moves in a geographical sample of infants born EP/ELBW and at term, and whether there is a relationship between perinatal clinical history, GMA, and motor outcome at 4 months’ corrected age. Furthermore, this thesis investigates the nuances of the GMA in relation to preterm birth. Study 1 is a systematic review of the literature investigated the reported predictive validity of spontaneous infant movements for later CP. Forty-seven studies were identified, with fidgety movements assessed according to the Prechtl GMA found to be the most accurate spontaneous movement for predicting CP. Study 2 is a cohort study, which was conducted within a geographical sample of infants born EP/ELBW and term-matched controls. Of the 226 infants born EP/ELBW and 225 infants born at term who were recruited to the study, at least one video was received from 158 and 188 families of infants born EP/ELBW or at term respectively. This thesis found that families tended to use Baby Moves less if they were of lower sociodemographic status, regardless of whether the infants were born EP/ELBW. Families found Baby Moves easy to use and considered it a secure way to transmit videos. Study 3 analysed the data from infants’ GMA further and the occurrence and trajectories of fidgety movements was explored. Infants tended to have more normal fidgety movements with increasing age within a window of 12-16+6 weeks’ corrected age regardless of birth group (EP/ELBW vs term). More infants born EP/ELBW had absent/abnormal fidgety movements than term-born controls. Study 4 explored the relationships of perinatal variables with absent/abnormal fidgety movements within the EP/ELBW cohort. Brain injury was independently associated with absent/abnormal fidgety movements. Finally, Study 5 assessed 4-month motor outcomes in a sample of 56 infants born EP/ELBW. There was a high rate of motor impairment within this sample. Brain injury was strongly related to poorer 4-month motor outcome and neonatal surgery was independently related to a poorer AIMS score. Absent/abnormal fidgety movements were not associated with 4-month motor outcomes. Findings from this thesis provide insight into the GMA and confirm that smartphone technology can be used with the GMA in a population of infants born EP/ELBW and at term.

Training targeted towards goals that are meaningful to children with cerebral palsy (CP) and their families is needed to improve function and support participation in physical activities in this population. Riding a two-wheel bike is a common goal for ambulant children with CP, yet little specific evidence exists to guide clinicians and families. This thesis developed and tested a task-specific approach to training bike skills in this population through three studies: 1) a systematic review, 2) a practice survey, and 3) a randomised controlled trial (RCT). Each study used the International Classification of Functioning, Health and Disability (ICF) as a framework. While strong evidence exists for task-specific training (TST) for improved upper limb (UL) function in this population, prior to this thesis the literature regarding TST for gross motor skills, including bike riding, in ambulant children with CP had not been systematically appraised. Thus, Study 1 aimed to systematically evaluate the effectiveness of task-specific gross motor skills training for improving activity and participation outcomes in ambulant school-aged children with CP. This review involved 13 studies of low-to-moderate overall quality and found effects of TST were positive for participation-related outcomes, and mixed for specific skill performance and functional skills, while little or negative effects were found for general gross motor skills. This study identified the need for higher quality studies and reporting that enables evidence synthesis. Given the importance of understanding current practice when designing effectiveness studies, Study 2 involved a survey of 95 physiotherapists (PTs) and occupational therapist (OTs) in Australia about their practices when training two-wheel bike skills in children with CP. This study found that while functional approaches to training and goal-based assessment and evaluation were predominant, overall practices appear highly variable. Moreover, the need to develop and test bike-specific measures and interventions in this population was highlighted. The findings from Studies 1 and 2 informed the design of Study 3. This multi-site assessor-blind RCT aimed to determine if a task-specific approach was more effective than a parent-led home program for attaining individualised two-wheel bike riding goals in ambulant children with CP. Sixty-two children were randomly allocated to either the task-specific approach (n=31) or home program (n=31). The primary finding was that the task-specific program was more effective than the home program for goal attainment at one week post-intervention. Greater odds of goal attainment were retained at three months and evidence of better outcomes following the task-specific program were found for some outcomes related to participation in bike riding, physical activity and self-perception. In addition, there was evidence of mixed effects for functional skills, and little difference in bike skills and health-related quality of life. While each of these studies provides an original contribution to the literature, together they form a significant foundation for evidence on training bike skills in ambulant children with CP. Use of the ICF across the thesis meant findings could be synthesised and enhanced the clinical relevance of the research. Given that an effective approach for attaining two-wheel bike riding goals in this population now exists, training for clinicians to optimise knowledge translation should be developed. Future research should seek to understand relationships between bike skills training and a broader range of ICF domains and levels of function in CP, tailor interventions to individuals and determine longer-term outcomes.

Participation in physical activity (PA) is associated with wide ranging health benefits at preschool age and across the lifespan, including favourable cardiometabolic and psychosocial outcomes. Many typically developing preschool age children are not meeting 24-hour movement guidelines in Australia and internationally. Children born very preterm (VP; <32 weeks’ gestation) may be at higher risk of physical inactivity due to myriad comorbidities associated with preterm birth, including adverse respiratory, neurological and motor outcomes. However, studies investigating PA participation in preschool age VP and term-born (born >=37 weeks’ gestation) children are sparse. This thesis used the International Classification of Functioning, Disability and Health children and youth version (ICF-CY) framework to compare PA participation and motor outcomes of children born VP with term-born children at preschool age over four studies. Study one was a systematic review and meta-analysis investigating motor outcomes of 3- to 6-year-old VP and term-born children within the ICF-CY framework. Children born VP had poorer outcomes than their term-born peers within the body structure and function and activity domains. However, no data on participation domain outcomes were identified. The second study examined the agreement between parent-reported and accelerometer-measured 24-hour movement behaviour in 4- to 5-year-old children. Agreement between the two measures of sleep was moderate, but poor for PA and stationary duration. Parents under-reported PA and stationary behaviour relative to the accelerometer, and VP birth, higher social risk and male sex were associated with the difference between the two measures. Study three compared motor outcomes within the ICF-CY body structure and function, activity and participation domains by examining grip strength, motor competence (MC) and PA participation in 98 VP and 84 term-born 4- to 5-year-old children. Children born VP had poorer preferred and non-preferred grip strength than term-born children, but there was little difference in bimanual grip strength between the groups. Motor competence was poorer for children born VP than term-born children as measured by the Movement Assessment Battery for Children, second edition and the Little Developmental Coordination Disorder Questionnaire. Using accelerometer data, children born VP completed less PA and more stationary time, and VP parents reported less unstructured PA and more minutes of screen time per day than parents of term-born children. Although adherence to the Australian 24-hour Movement Guidelines was poor for both groups, a lower proportion of VP children met the PA and screen time recommendations than their term-born peers. Study four examined the relationships between ICF-CY domains by investigating associations between grip strength, MC and PA levels at 4 to 5 years of age, determining if associations differed between VP and term-born children. Irrespective of birth group, more PA and less stationary behaviour were associated with better MC, and better MC was associated with greater grip strength. For children born VP, more PA was associated with better balance skills, and better balance skills were associated with greater grip strength. Physical activity levels did not appear to be associated with grip strength in 4- to 5-year-old children. Overall this thesis demonstrates that preschool age children born VP experience poorer motor outcomes than term-born children within all ICF-CY domains. The studies within this thesis form the foundation for future research of PA participation in VP preschool age children, and contribute to a deeper understanding of the PA participation trajectory in this population. These findings have substantial implications for allied health clinicians involved in the management and developmental follow-up of children born VP, as well as for early educators and the design of PA promotion strategies.

Worldwide lung cancer is the second most frequently diagnosed cancer. In Australia it is the leading cause of cancer burden and death. The overwhelming majority of people are diagnosed once the disease has spread beyond the primary site. Lung cancer is associated with high levels of poorly controlled symptoms, a decline in physical function, low physical activity (PA) levels and poor health-related quality of life (HRQoL) compared to the healthy population. For people with operable non-small cell lung cancer (NSCLC) increased PA levels are associated with improvements in HRQoL, exercise capacity, fatigue, and psychological distress. Increased exercise capacity at diagnosis of inoperable NSCLC is associated with improved survival. Further research regarding the effects of exercise interventions is required in populations with inoperable NSCLC as the majority of studies conducted to date in this population have either been single-group studies or small randomised controlled trials which are frequently subject to high rates of attrition and lack longer-term follow-up of outcomes. The first aim of this thesis was to conduct a systematic review of outcome measures that have been utilised to assess PA levels in lung cancer and report on the psychometric properties of included measures. The second aim of this thesis was to conduct a randomised controlled trial (RCT) to determine the effects of home-based multi-disciplinary rehabilitation both during and following treatment for inoperable NSCLC and report on quantitative and qualitative outcomes. The systematic review identified significant variation in the measurement of PA in lung cancer; 34 articles utilising 21 different outcome measures were included in the review. Seventeen (50%) studies used performance-based measures, such as accelerometers or pedometers, to quantify PA and the remaining studies used patient self-report measures, such as the Godin Leisure Time Exercise (GLTEQ) or the International Physical Activity (IPAQ) questionnaires. Only two studies used both performance-based and self-reported methods of PA measurement. Eighteen (53%) studies reported on the psychometric properties of the outcomes used to measure PA in lung cancer and the quality of those that did was commonly rated as ‘fair’ or ‘poor’. Reflective of the increased research activity and interest in this area, an update of this systematic review performed in 2018 identified an additional 31 articles measuring PA in lung cancer. Updated findings were consistent with the original review with 14 articles (45%) reporting utilisation of performance-based outcome measures and 19 (61%) including patient-reported outcomes. Again, only two articles measured PA using both methods of measurement. For articles retrieved in the updated review, the IPAQ was the most frequently used questionnaire, this is in contrast to the published review where the GLTEQ was most commonly used. The RCT assessed the efficacy of home-based multi-disciplinary rehabilitation both during and following treatment for inoperable NSCLC. Participants randomised to the intervention group received an eight-week rehabilitation package of care, delivered by a combination of home-visits and telephone calls, involving exercise (aerobic and resistance), behaviour change techniques to support increased exercise and PA and early initiation of patient-centred symptom self-management support. Following the initial eight-week program intervention participants received reduced frequency telephone contact to support exercise behaviours until trial completion at six months. Quantitative outcomes were measured at baseline (prior to randomisation), nine weeks and six months post-baseline and included: exercise capacity (six-minute walk distance (6MWD), the primary outcome), PA levels (performance-based using accelerometers and patient self-report), muscle strength (quadriceps and handgrip), patient reported outcomes (symptom severity and distress, HRQoL, mood, exercise motivation, exercise self-efficacy and resilience) and survival. Ninety-two participants were recruited and 78 (all participants who provided data for at least one follow-up measure) were included in modified intention-to-treat analyses. Adherence to the aerobic component of the exercise program was 65%. There were no significant between-group differences for measures of physical function, mood, self-efficacy or resilience at either follow-up time point. However, a significant interaction effect was demonstrated between group allocation and time across the three study timepoints for the 6MWD; indicating that the temporal pattern of 6MWD results was significantly different between the groups. This difference may be in part due to the timing of exercise with respect to treatment for lung cancer, with less decline in 6MWD observed for the intervention group in the long-term (between baseline and six months) than during the initial treatment phase (between baseline and nine weeks). At six-month follow-up statistically and clinically significant between-group differences favouring the intervention group were found for symptom severity levels, HRQoL and exercise motivation. The intervention group survival benefit at censoring for data analyses was not statistically significant, however at a median of 230 days greater than the usual care group, is of likely clinical importance for those with inoperable disease. Following the initial eight weeks of the program 25 intervention group participants completed semi-structured interviews regarding their views and experiences of program involvement. The majority of participants found the program acceptable and reported multiple physical and mental health benefits including improved strength and fitness, motivation and prevention of boredom. Program enablers included: having supportive family and friends; advice and support from expert health professionals; the perception that the exercise program had been individually tailored to be achievable; having ongoing program monitoring and modification as required by the program physiotherapists; and having a program which consisted of exercise that participants found familiar and enjoyable. Barriers to exercise program completion were symptom exacerbations and poor weather. Adherence to exercise was reportedly increased by use of simple activity trackers and exercise diaries and receiving weekday exercise text message reminders. Few participants watched the study DVD of resistance exercises that was provided to them; most feeling it was not required or reporting they did not have the technology to watch the DVD. Recommendations for future research to improve outcomes for people with lung cancer include greater consensus regarding utilisation of a core set of validated outcomes to measure PA. Where possible to employ both patient-reported and performance-based methods of PA measurement. The findings from the RCT support the benefits of the rehabilitation package of care delivered both during and following treatment for inoperable NSCLC. However, the significant interaction between time and group allocation for the 6MWD indicates the need to consider the timing of exercise in relation to lung cancer treatment. Routine self-monitoring of symptoms should be embedded into lung cancer care pathways; electronic self-reported data could be used to trigger an alert for clinician follow-up of symptoms above a given threshold. Future studies should implement strategies to improve exercise adherence and ensure target training intensities are met. This could involve remote monitoring of exercise sessions or the use of online ‘virtual’ exercise groups, formulating alternative indoor exercise plans and implementing flexible program designs incorporating largely home-based exercise with supervised hospital or community-based sessions as needed.

Pubescent females are twice more likely to suffer a non-contact ACL injury than their male counterparts. This disparity has been correlated with multiple concurrent factors, including biomechanical, anatomical and hormonal changes. ACL ruptures require serious and costly surgical interventions, which could be avoided if subjects at higher risk of injury were more carefully monitored and trained. Three-dimensional motion analysis is required to identify individuals at risk of ACL injury. Multi-camera optical systems are the gold standard for 3D motion capture, but they are very expensive and cumbersome. The aim of this thesis was to make motion analysis more accessible, developing an affordable and compact 3D motion tracking methodology, alternative to conventional multi-camera systems. A novel tracking approach was developed using Microsoft Kinect v2, employing custom-made coloured markers and computer vision techniques. This methodology was denoted as Kinect coloured marker tracking (KCMT). The accuracy of KCMT relative to a conventional Vicon motion analysis system was measured performing two Bland-Altman analyses of agreement, the first using single-leg squat (SLS) as benchmark task, the second using treadmill locomotion. The objective of the first study was to determine if KCMT-derived sagittal joint angles of the lower limb were accurate enough to allow discerning individuals at risk of ACL injury from those not at risk. Eleven healthy participants were asked to perform three SLS trials, while three-dimensional marker trajectories were simultaneously recorded using Vicon and KCMT respectively. Joint angles from the two systems were calculated via inverse kinematics using OpenSim. The limits of agreement (LOA) of the joint angles were −16°, 13° for hip flexion, −12°, 0° for knee flexion and −12°, 9° for ankle flexion. These results indicated that the agreement between KCMT and Vicon was joint dependent, and that further work was required for the novel methodology to replace conventional marker-based motion capture systems for the identification of ACL injury risk from SLS data. In the second study, an improved data collection protocol for the KCMT was used. Twenty participants were recruited, and markers placed on bony prominences near hip, knee and ankle. Three-dimensional coordinates of the markers were recorded during treadmill walking and running. The LOA of marker coordinates were narrower than −10 and 10 mm in most conditions, however a negative relationship between accuracy and treadmill speed was observed along Kinect depth direction. LOA of the knee angles measured in the global coordinate system were within −1.8°, 1.7° for flexion in all conditions and −2.9°, 1.7° for adduction during fast walking, suggesting that KCMT may be capable of discerning between subjects at risk of ACL injury and controls. The proposed methodology exhibited good agreement with a marker-based system over a range of gait speeds and, for this reason, may be useful as low-cost motion analysis tool for selected biomechanical applications.

Children and adolescents (“children”) with Charcot-Marie-Tooth disease (CMT) have progressive weakness of the lower limbs causing problems with gait and function. This thesis examined the nature and impact of gait dysfunction in children with CMT, with consideration of the typical everyday environments in which children function. Little is known about the effect of this degenerative peripheral neuropathy on gait and functional ambulation over time in children. Gait dysfunction may contribute to retrospective reports of frequent trips and falls in children with CMT, yet there are no studies of falls in this population. Children with CMT often report reduced physical endurance, yet no studies have investigated physical endurance, and more generally functional ambulation and physical activity. Six studies were designed to address these knowledge gaps, with comparison to typically developing peers (TD) to place gait dysfunction within the context of typical growth and development. A systematic review of gait in paediatric CMT found only a few studies with small sample sizes, examining barefoot gait in clinical and laboratory settings and providing limited comparison to TD databases. Gait was slower, most likely due to shorter stride length, with foot drop, reduced calf push-off power and proximal lower limb compensatory strategies. Two cross-sectional, case controlled studies utilising spatio-temporal gait analysis, one in different footwear and the other during a six-minute walk test, confirmed that gait in 30 children with CMT was slower, with shorter and wider steps, and greater step-to-step variability compared to TD children. Increased barefoot base of support variability was associated with poorer balance. Suboptimal footwear negatively affected gait in all children, irrespective of disease, which has clinical implications for children with CMT who have weak feet and ankles and poor balance. Reduced six-minute walk distance (6MWD) was an indicator of reduced physical endurance, and increased step-to-step variability was moderately associated with reduced 6MWD and increased perceived exertion. A longitudinal study of gait in 27 children with CMT over 12-months highlighted the importance of normalising gait data in children. Over 12-months, children with CMT were found to have reduced walking speed and endurance, and older children (≥ 12 years) exhibited greater disease progression and decline in functional ambulatory capacity than younger children. A novel six-month prospective cohort study of falls identified a markedly higher incidence of falls in children with CMT, with concerningly high numbers of injurious falls compared to TD children. Tripping was the most common mechanism of falls and all children under the age of 7 years fell, irrespective of disease. Age and balance were the strongest predictors of falls. A cross-sectional, case controlled study of functional ambulation and physical activity in 50 children with CMT across two paediatric centres found that functional ambulation was limited on all measures, including capacity, performance in every day environments, and the child’s perception of gait-related disease limitations. Children with CMT were less physically active than their TD peers with greater disease severity associated with lower physical activity. This thesis delivers important information to healthcare providers and families of children with CMT, and highlights the degree to which gait dysfunction impacts everyday activities. Gait dysfunction can be used as a biomarker of disease severity and progression in CMT. These findings will inform management of children with CMT, development of clinical practice guidelines and educational resources, and influence future research, including exercise interventions.

Knee osteoarthritis (OA) is a significant cause of pain and disability among older adults worldwide. Evidence and all guidelines recommend exercise as the cornerstone of non-surgical treatment for all people with knee OA. However, the clinical benefits of exercise among people with knee OA have been demonstrated to be modest, and decline from short- to long-term. Poor adherence to exercise programs as prescribed has been suggested as an important factor impacting the effectiveness of these exercise interventions for people with knee OA. This thesis aimed to gain a better understanding of adherence to exercise among people with knee OA through five related studies. Study One examined the presence of common trajectories of self-reported adherence to home exercise programs over time among a large cohort of people with knee OA. Using latent class growth analysis three distinct trajectory groups were identified: a “Rapidly declining adherence” group, a “Gradually declining adherence” group and a “Poor adherence” group. These findings affirmed the importance of monitoring adherence, and identifying interventions and behaviour change techniques to achieve and maintain adherence to exercise long-term. Study Two clarifies the current understanding of interventions targeting adherence to exercise among older adults with knee/hip OA or chronic low back pain by way of a systematic review. Meta-analysis found moderate quality evidence that booster sessions with a physiotherapist may improve exercise adherence in people with lower limb OA. Findings highlighted the limited number and heterogeneous nature of published randomised controlled trials (RCTs) specifically evaluating interventions aimed at increasing exercise adherence. Study Three explored the perspectives of people with knee OA and physiotherapists who treat people with knee OA regarding theory-derived behaviour change techniques (BCTs) to improve adherence to exercise. Results of the online questionnaire identified a mismatch between the BCTs experienced by people with knee OA and used by physiotherapists, and those perceived to be most likely to be effective. A limitation identified in Studies One, Two and Three was the lack of evaluation of the validity and reliability of commonly used self-reported measures of exercise adherence. Study Four used unique concealed accelerometer technology to examine the concurrent validity of exercise diary completion and a retrospective self-rated adherence scale among a cohort of older adults with chronic knee pain undertaking a home strengthening program. Both self-reported measures showed questionable validity, and the self-rated adherence scale also demonstrated less than acceptable test-retest reliability. Finally, using the accelerometer-measured exercise adherence data Study Five examined the effect of home exercise adherence on changes in patient outcomes of pain, function and quadriceps strength over the 12-week intervention. While a significant decline in adherence, and significant improvements in patient outcomes were observed, the level of home exercise adherence was not significantly associated with changes in these outcomes in linear or non-linear models. Taken together, the findings of these studies provide new knowledge of adherence to exercise specifically among people with knee OA. Furthermore, the results of this work raise a number of research questions worthy of future investigation.

Knee and hip osteoarthritis (OA) is highly prevalent and has a significant burden on both the individual sufferer and society. All current clinical guidelines recommend education and exercise for management of OA, however exercise participation amongst people with OA is sub-optimal. Barriers to exercise uptake and adherence include inequitable access to appropriate healthcare, and difficulties changing behaviour and incorporating exercise into daily life. This thesis explores remote models of service delivery (telerehabilitation) as a potential method of improving exercise participation in people with OA. Specifically, this thesis aimed to explore the perceived acceptability of telerehabilitation services amongst people with OA and physiotherapists, and also investigate physiotherapist training in behaviour change techniques and person-centred care for telephone-delivery. Study 1 involved a survey that investigated the perceptions people with knee and/or hip OA (n=330) have towards the delivery of exercise therapy by a physiotherapist via internet-mediated video and telephone consultations. Participants had overall positive perceptions, acknowledging ease of use and time saving advantages. However, most (>50%) did not agree that they would like the lack of physical contact, that they would be willing to pay for telerehabilitation services, that telephone-delivery would be effective, or that a physiotherapist would be able to adequately monitor their condition via telephone. Study 2 involved a survey that investigated the perceptions of physiotherapists (n=217) towards the delivery of exercise therapy via internet-mediated video and telephone consultations for people with OA. Most physiotherapists agreed such services would save patient’s time and maintain their privacy, but most did not believe they would like the lack of physical contact with patients. In addition, physiotherapists favoured the use of video technologies over telephone, with most feeling uncertain about the safety, effectiveness, usefulness, or acceptability of telephone-delivered care. Study 3 qualitatively explored physiotherapists’ (n=8) perceptions before and after a training program in behaviour change techniques and person-centred principles that was done in preparation for a clinical trial involving the delivery of exercise therapy via telephone. After training, physiotherapists’ perceptions about their role managing patients with OA had changed, increasing their feelings of responsibility to assist their patient with exercise adherence, and they felt confident and prepared to deliver the intervention remotely via telephone. Study 4 evaluated the fidelity of physiotherapists from Study 3 (n=8) to the behaviour change techniques and person-centred principles taught during training, and involved both self- and expert audits of practice telephone consultations. Physiotherapists performed moderately well, but had room for improvement from further practice and/or training. Physiotherapists’ self-ratings of performance generally agreed with expert ratings, however they tended to underestimate their ability to implement some principles and techniques. Study 5 explored the perceptions of people with knee OA (n=20) who participated in a clinical trial involving telephone-delivered exercise therapy by a physiotherapist. Although participants were initially sceptical about receiving care via telephone, they described mostly positive experiences, valuing the sense of undivided focus and attention and feeling confident performing their exercise program without supervision. Study 6 explored how experience delivering exercise therapy via telephone as part of a clinical trial changed physiotherapists’ (n=8) perceptions about such services. Physiotherapists were initially sceptical about the effectiveness of telephone-delivered service models, expressing concern about the lack of physical and visual contact. However, after experience, physiotherapists were pleasantly surprised by the effectiveness of the intervention and by the positive outcomes that they were able to achieve with their participants. Collectively, findings from this thesis suggest that, overall, telerehabilitation is perceived to be an acceptable model of service delivery by people with OA and physiotherapists. Although there was evidence of scepticism about the effectiveness of telephone-delivered care amongst those who are inexperienced, particularly physiotherapists, these perceptions appeared to change with experience. Physiotherapists believed that training in behaviour change techniques and person-centred care helped them communicate effectively via telephone and also changed their perceptions about their role managing patients with OA. Findings from this thesis can be used to inform the future design and implementation of telerehabilitation services and clinician training programs.

Median sternotomy is the most common incision used in cardiac surgery worldwide with more than a million procedures operated annually (Epstein et al, 2011; Go et al, 2014), due to its ease of performance and provision of optimal exposure of the heart (El Ansary et al, 2007c; McGregor et al, 1999; Robicsek et al, 2000; Zeitani et al, 2006; Deb et al, 2013) It remains the standard of care for myocardial revascularization in cardiac surgery, in particular for multi-vessel disease (Cheng & Slaughter, 2013; Deb et al, 2013; Rosenfeldt et al, 2012; Taggart, 2013b). Despite the advantages of a median sternotomy, the incidence of sternal complications has remained relatively unchanged for the last two decades and is reported to be between 1 to 8% worldwide (Balachandran et al, 2016; El-Ansary, 2000b; Ho et al, 2002; ASCTS Data 2013) Sternal complications can range from post-sternotomy pain, skin infections, dehiscence, sternal instability/non-union and mediastinitis (Crabtree et al, 2004; El-Ansary et al, 2008; Cahalin et al, 2011; ASCTS Data 2013). These complications are associated with significant morbidity, and prolonged patient hospitalization which is reported to triple the cost of care (Losanoff et al 2002b; Crabtree et al, 2004; Zeitani et al, 2006; El-Ansary et al, 2008; Baskett et al, 1999; Filsoufi et al, 2009; Joseph et al, 2014; Cahalin et al, 2011; Mekontso et al; 2011; Lazar et al, 2016). In an attempt to facilitate sternal healing and prevent sternal complications, patients who have undergone cardiac surgery via a median sternotomy are routinely asked to follow sternal precautions post-operatively (Balachandran et al, 2014; Tuyl et al, 2012, Cahalin et al, 2011; Overend et al, 2010). These precautions place restrictions on the use of the upper limbs and trunk immediately following surgery (Balachandran et al, 2014; Tuyl et al, 2012, Cahalin et al, 2011; Overend et al, 2010; Brocki et al., 2010). Sternal precautions are applied worldwide for duration of four weeks to three months following surgery despite a paucity of research to support this practice (Balachandran et al, 2014; Tuyl et al, 2012, Cahalin et al, 2011; Overend et al, 2010; Brocki et al, 2010). Recent research has reported that minimal sternal motion takes place between the sternal edges as measured by ultrasound during upper limb and trunk activity (Balachandran, 2015; Balachandran et al, 2017). Furthermore, such sternal micro-motion may constitute part of the normative path to bone healing (Balachandran et al, 2014; Balachandran et al, 2017; Cahalin et al, 2011). Sternal precautions in their current form may be overly restrictive thus delaying recovery and a return to community role healing (Balachandran et al, 2014; Balachandran et al, 2017; Cahalin et al, 2011). This thesis examined (1) whether change to sternal precautions impact upon function and sternal pain following cardiac surgery via a median sternotomy A randomized controlled trial (RCT) was conducted to assess how changes to sternal precautions impact upon function, and sternal pain following cardiac surgery via a median sternotomy Nested within the RCT, a repeated cohort studies were conducted to assess the clinimetric properties of selected physical function tools ie The Short Physical Performance Battery (SPPB) and The Functional Disability Questionnaire (FDQ) used in the cardiac surgery via a median sternotomy The results provide new clinimetric information on outcome measures targeting the cardiac population and inform the post-operative clinical management and rehabilitation after cardiac surgery. This thesis is composed of four studies, where each study is focused on specific parameters that are essential in obtaining comprehensive data and results. In the first study, a protocol for randomised controlled trial was designed to investigate whether changes to sternal precautions impact upon function following cardiac surgery via a median sternotomy The rationale for developing this study protocol was that the routine implementation of sternal precautions worldwide practice following a median sternotomy may delay recovery and be overly restrictive This study is the first randomized controlled trial using an intervention group to modify sternal precautions, and study its effectiveness in improving physical function in this population The intervention was built on foundational evidence that evaluated the effects of upper limb exercise by investigating the effects of modifying sternal precautions to include the safe use of upper limbs and trunk, and assess their impact on patients’ physical following cardiac surgery via median sternotomy in order to optimize functional recovery in this patient population. The second study nested within the study 1 was conducted to determine the clinical applicability of the Short Physical Performance Battery (SPPB), when used in patients post cardiac surgery This study evaluated the MCID of the SPPB, an outcome measure that has been validated in older patients who were classified as cardiovascular stable Importantly this study is the first to determine the MCID of the SPPB for an adult cardiac surgery population The results of this study should be considered preliminary evidence on the application of the SPPB to evaluate treatment effectiveness by detecting a true improvement An increase or decrease in performance greater than the MCID indicates a high likelihood of a meaningful change These measures can be used to document real improvements in physical function through the course of cardiac rehabilitation Therefore, it is recommended that an MCID reference value above one point of the SPPB scores could serve as an explicit therapeutic goal for rehabilitation intervention and monitoring functional progress following cardiac surgery. The third study incorporated a novel outcome assessment of upper limb and trunk function specific to cardiac surgery developed by a team of researchers within the Department of Physiotherapy at the Melbourne University This was a comprehensive clinometric analysis of the FDQ including: the statistical feasibility of a shortened FDQ (FDQ-s), validity, reliability, responsiveness, interpretability, and feasibility The findings of this study established that the FDQ-s has strong clinimetric properties with moderate to excellent results on all domains As such, it is recommended that the FDQ-s be adopted as an outcome measure of physical recovery after cardiac surgery within the acute hospital setting, and in the community to plot the trajectory of recovery overtime Further, the FDQ-s can be utilized in research trials evaluating function and in the clinical setting by health professionals to inform and guide management after cardiac surgery The FDQ-s may be a useful tool in understanding the benefits of physical limitations after cardiac surgery, and thus lead to more finely tailored and individualized health care interventions. The final study in this thesis presents the results from a RCT, the Sternal Management Accelerated Recovery Trial (SMART) that investigated a standard restrictive versus a program of modified sternal precautions following cardiac surgery via a median sternotomy The findings of this study suggest that a program of modified (less-restrictive) sternal precautions for patients following cardiac surgery did not improve physical recovery, pain or enhance health related quality of life (HRQoL) compared to usual care With no adverse event, the results of this RCT suggest that a precautionary approach that is less restrictive with a progression of activity will likely facilitate optimal functional recovery after a median sternotomy Importantly, this result adds further evidence that strict adherence to SP may not be warranted for all patients as it reinforces kinesiophobia which may potentially impact on patient participation exercise and in cardiac rehabilitation It is recommended that a program of sternal precautions based on individual clinical characteristics and risk profile rather than a generic and routine set of SP may result in optimal recovery Findings in this field of research will be of great importance to enhance, develop and evolve SP to provide patients with the optimal care for post-operative physical and sternal pain management. The findings of this thesis supported the need in the development, and implementation of clinical and regulatory guidelines to improve patient and community safety; quality of life and standards of care for individuals following cardiac surgery The thesis addresses the paucity of research and the inconsistent recommendations with respect to sternal precautions and associated restrictions to upper limb and trunk provided to the large number of individuals having open-heart surgery nationwide In particular, this research will inform guidelines for the commencement of upper limb activities in Cardiac Rehabilitation (CR) and standards for sternal precautions following cardiac surgery This thesis further addressed the gaps in the literature pertaining to the outcome measures used following cardiac surgery performed via sternotomy The results of the two outcome measure studies recommended that the SPPB and FDQ be used as measure tools for outcome assessment in research studies, to assess the impact of post-operative rehabilitation management, and better elucidate the process of recovery after cardiac surgery in the acute clinical setting.  

Adolescent girls are susceptible to knee injuries such as non-contact anterior cruciate ligament (ACL) rupture and patellofemoral pain (PFP). Adolescence is synonymous with pubertal development which drives substantial growth and maturation of the musculoskeletal system, and is thought to contribute to poor knee biomechanics associated with both of these injuries. Specifically, higher tri-planar knee moments during puberty (external peak knee abduction moment (KAbM), flexion moment (KFM) and internal rotation moment (KIRM)) are thought to contribute to a higher incidence of these injuries; however, there are still gaps in our understanding of female pubertal biomechanics. For instance, variations in dynamic tasks (i.e., bilateral vs single limb), unreliable pubertal classification methods, small sample sizes, conflicting findings and data normalisation methods (i.e., mixed between body mass or body mass by height) highlight the need for additional, better designed pubertal studies. The role of athletic footwear is also an important consideration, given it may alter tri-planar knee moments relevant to both non-contact ACL rupture or PFP. Specifically, high-support footwear is thought to control excessive foot pronation, which may transfer up the kinetic chain and confer protection at the proximal knee joint by modifying tri-planar knee moments during dynamic tasks. By contrast, low-support shoes do not possess the same stability features, potentially allowing for greater foot pronation that may have a clinically meaningful effect on tri-planar knee moments compared to high-support shoes. Surprisingly, no studies have explored the effects of these shoes during female pubertal development, which is concerning as many adolescent girls are likely wearing these types of shoes during the various sports in which the aforementioned knee injuries occur. To address the current limitations in pubertal and footwear biomechanical research, four cross-sectional studies are reported in this thesis. Ninety-three girls aged between 7-25 years old were categorised into three key stages of puberty: pre-pubertal (n = 31, mean age = 9.4 ± 1.2), early/mid-pubertal (n = 31, mean age = 11.1 ± 1.4) and late/post-pubertal (n = 31, mean age = 19.8 ± 4.0). Tri-planar knee moments normalised to body mass (Nm/kg) and body mass by height (Nm/kg/m) were analysed across landing and running-related tasks in each pubertal group. These were initially observed barefoot, and then subsequently, the effect of high- and low-support footwear was explored across both tasks. The primary aim of Study 1 was to determine whether peak tri-planar knee moments differed between three stages of female pubertal development during a barefoot single limb drop lateral jumping (DLJ) task. The secondary aim was to explore the hip adduction moment (HAM) at time of peak KAbM and the hip flexion moment (HFM) at time of peak KFM between pubertal groups. In the frontal plane, a higher peak KAbM was found for the late/post- compared to the pre-pubertal group when normalised to body mass (95%CI=-0.02 to -0.17 Nm/kg, p=0.015, d=0.61), but not body mass by height (p=0.88). At the hip, neither body mass or body mass by height normalised data revealed between-group differences for HAM at time of peak KAbM (p>0.05). In the sagittal plane, a higher peak body mass-normalised KFM was found for the late/post- (95%CI=0.19 to 0.68 Nm/kg, p=0.001, d=1.12) and the early/mid-pubertal groups compared to the pre-pubertal group (95%CI=0.05 to 0.52 Nm/kg, p=0.017, d=0.59). No significant between-group differences were found for body mass by height-normalised peak KFM (p=0.30) or the HFM at time of peak KFM (p>0.05). Finally, in the transverse plane, a higher peak KIRM in the late/post- compared to both the early/mid- (95% CI = -0.09, -0.01 Nm/kg, p=0.028, d=0.62) and pre-pubertal groups (95% CI= -0.12, -0.03 Nm/kg, p=0.001, d=0.82) was found for body mass, but not body mass by height normalised data. The primary aim of Study 2 was to determine whether peak tri-planar knee moments differed across footwear conditions (i.e., barefoot, high-support and low-support shoes) during the single-limb DLJ amongst late/post-pubertal girls. Based on the findings in Study 1, the late/post-pubertal group was selected as they displayed higher mass normalised tri-planar knee moments compared to early/mid- and pre-pubertal counterparts and may be at higher risk of ACL injury. Results revealed no significant differences for peak KAbM or KIRM regardless of statistical adjustment for FPI (p>0.05). By contrast, peak KFM was higher in the high-support (95% CI= 0.36, 0.53 Nm/kg, p<0.001, d= 1.11) and low-support shoes (95% CI= 0.25, 0.48 Nm/kg, p<0.001, d= 0.85) compared to barefoot; however, no significant differences were observed between shoe conditions. Together, Study 1 and 2 provide novel insights into the effects of female puberty and footwear on the biomechanics of single-limb landing, revealing that increased pubertal-related height (i.e., stature), rather than body mass, is the main contributor to augmented tri-planar knee moments in the latter stages of female pubertal development, which athletic footwear did not ameliorate. The primary aim of Study 3 was to examine tri-planar knee moments (normalised to body mass and body mass by height) across the three pubertal stages while running barefoot. Higher peak body mass-normalised KFM was apparent in the late/post-pubertal (95% CI= 0.18 to 0.63 Nm/kg, p=0.001, d= 1.01) and early/mid-pubertal (95% CI= 0.02, 0.47 Nm/kg, p=0.034, d=0.52) girls compared to the pre-pubertal girls; however, no significant differences were found when KFM was normalised to body mass by height (p>0.05). Furthermore, no significant differences were found for body mass or body mass by height normalised peak KAbM or KIRM (p>0.05). At the hip, a lower body mass normalised HAM at time of peak KFM (i.e., greater hip abduction moments) was found in the late/post- (95% CI= -0.51, -0.11 Nm/kg, p=0.003, d= 0.86) and early/mid-pubertal (95% CI= -0.42, -0.01 Nm/kg, p=0.039, d= 0.53) girls compared to their pre-pubertal counterparts. Likewise, in the sagittal plane a decrease in body mass normalised HFM at time of peak KFM was evident with the late/post-pubertal girls displaying lower HFM (95% CI= 0.65, 0.28 Nm/kg, p<0.001, d=1.27) compared to their early/mid- and pre-pubertal counterparts (95% CI= 0.36, 0.73 Nm/kg, p<0.001, d=1.42). Study 4 determined whether footwear conditions (i.e., barefoot, high-support, low-support) effect running-related peak KFM amongst a pooled sample of early/mid- and late/post-pubertal girls. Based on the findings in Study 3, girls in the early/mid- and late/post-pubertal groups were pooled as no differences in body mass normalised tri-planar knee moments were observed; however, they both displayed higher peak KFM compared to pre-pubertal girls. The secondary aim explored predictors associated with a change in the peak KFM wearing shoes compared to barefoot (i.e., knee-ground reaction force (GRF) lever arm, sagittal plane resultant GRF magnitude and sagittal plane lower limb kinematics) to help elucidate the underlying biomechanical mechanism. A main effect (p<0.001) for peak KFM was found, revealing both high- (95% CI= 0.36, 0.49 Nm/kg, p<0.001, d=1.07) and low-support (95% CI= 0.31, 0.45 Nm/kg, p<0.001, d=0.97) footwear increased peak KFM compared to barefoot, no differences were found between shoes (p>0.05). The regression models identified that only a change in the knee-GRF lever arm in shoes compared to barefoot was associated with a change in peak KFM (F(1, 109)= 93.56, p<0.001), but not the sagittal plane GRF magnitude or any lower limb kinematics (p>0.05). Combined, Study 3 and 4 provide evidence for a developmental increase in sagittal plane but not frontal or transverse plane knee moments that are likely attributed to differences in adolescent height while running. More importantly, wearing shoes increased peak KFM even further regardless of whether they possessed supportive characteristics, and this was partly driven by an increase in the knee-GRF lever arm. Given the repetitive and chronic loading pattern associated with the development of adolescent PFP, both puberty and athletic footwear may influence the manifestation of this condition. Further studies are required to prospectively determine whether higher body mass landing-related tri-planar knee moments and running-related peak KFM in the later stages of puberty are indeed linked to ACL rupture or PFP, respectively. Moreover, pubertal footwear studies may consider modifying footwear features to determine if these higher knee moments can be attenuated.

Falls are a major public health problem. They can have physical and psychological sequelae and may impose an economic burden for the patient, family, carers, staff and the healthcare system. Although much of the focus of falls prevention and its impact is in the community setting, falls can have equally devastating impact when they occur in hospitals. Despite an expansion in research investigating falls in hospitals, there remains uncertainty about the best way to help prevent falls, thereby minimising potential detrimental outcomes, in the acute setting. Prior to commencement of this research, falls prevention within the acute hospital setting was thought to be best addressed using multifactorial falls prevention initiatives, guided by falls prevention screening or assessment. This research sought to develop a unique two-stage falls risk screening and assessment tool, the Western Health Falls Risk Assessment (WHeFRA), purposely designed for the acute hospital setting, and to determine its rater reliability and predictive validity. This tool was then utilised as the foundation for a multifactorial falls prevention program in the acute hospital setting, and evaluation of the immediate and sustained outcomes was undertaken. Investigation of the clinimetric properties of the WHeFRA on acute hospital wards revealed very good intra-rater reliability and good inter-rater reliability. A further prospective investigation of WHeFRA screening tool accuracy in predicting patients at risk of falls found good predictive validity, and also favourable accuracy compared with the “gold standard” screening tool (STRATIFY). A falls prevention program, based on the WHeFRA and utilising Falls Resource Nurses (FRN’s) as key drivers and ward champions, was implemented on 14 acute wards utilising a staggered rollout over three years. Falls rates monitored over ten years increased initially compared with the two-year baseline rate. Rates then gradually decreased for nearly three years post program completion, after which the rate began to increase again, however still remaining lower than baseline. The program was further investigated through audits of WHeFRA compliance and nurses’ surveys at two time points, at completion of program implementation and 2.5 years following this. Sustainability of program achievements also appeared to be maintained initially, however at 2.5 years post completion it was apparent that booster activity and support avenues were required to preserve benefits longer-term. Although growth in investigations based in the acute hospital setting occurred following commencement of this research, debate regarding the most ideal prevention approach and the use of falls screening tools continues today. Knowledge about staff acceptance of falls risk screening tools and falls prevention programs, and sustainability outcomes of falls prevention initiatives, remains limited. In summary, this research spanning ten years in an acute hospital setting: i) developed a unique two-stage, reliable and accurate, falls risk screen and assessment tool, ii) successfully implemented a multifactorial falls prevention program, and iii) identified important elements required to initially achieve practice change and to maintain this culture change over a sustained period, highlighting what is needed to accomplish optimal long-term falls prevention benefits.

Interacting with patients is integral to the practice of physiotherapy. Notably, however, empirically derived knowledge about how physiotherapists interact with their patients is limited, particularly in the private practice setting. In addition, heavily promoted approaches for interacting with patients, such as patient-centred care and the biopsychosocial approach, have been adopted from the medical profession, are not derived from research evidence, and therefore may not adequately reflect how physiotherapists interact with patients in physiotherapy practice. Thus, this qualitative research had two aims: first, to detail how patients and physiotherapists interact in private practice; second, to consider how the research findings related to promoted healthcare interaction approaches. Methodologically, the research incorporated features of both ethnography and grounded theory. Observations of 52 consultations, as well as in-depth interviews with 9 patient and 9 physiotherapist participants, were undertaken. Data comprised field notes and audio-recordings of observations and interviews, and were analyzed iteratively using principles of thematic analysis and grounded theory. The data analysis yielded two central and complementary themes. The theme ‘physiotherapist-led communication’ encapsulates how physiotherapists directed the style and content of communication to achieve clinical goals by providing structure, making decisions, and focussing on biomedical aspects. The second theme, ‘adapting to build rapport’, describes how physiotherapists incorporated adaptive communication such as eye contact, body language, touch, casual conversation, and humour into their interactions with patients. These adaptations were often intuitively enacted, were responsive to individual patient characteristics, and functioned to build rapport. The findings neither clearly correlated to features of patient-centred care nor to the biopsychosocial approach. Rather, the findings portrayed a dynamic integration of clinical and responsive communication that fostered the development of a trusting relationship between patient and physiotherapist. These results extend knowledge of interactions in physiotherapy by providing detailed descriptions of interactional elements that incorporated patient and physiotherapist perspectives. Furthermore, the findings explain how rapport was developed between patient and physiotherapist with trust as an underlying construct. Relationship-centred care and relational notions of trust are discussed as alternative explanations for how patients and physiotherapists interact in private practice. These findings and explanations have the potential to benefit educators, physiotherapists and, by extension, patients, by offering a framework for education and the practice of patient-physiotherapist interactions.