Paediatrics (RCH) - Theses
Permanent URI for this collection
Search Results
1 - 10 of 10
-
ItemUrinary tract infection in children – changing the management paradigm so more children can be treated at home( 2023-07)Urinary tract infections (UTI) are a common reason for paediatric emergency department (ED) presentations. Although most children are successfully treated with oral antibiotics, approximately one-third receive intravenous (IV) antibiotics and hospital admission. However, there is limited evidence supporting these decisions, and consequently an opportunity to approach antibiotic use differently to potentially treat more children out of hospital. Guidelines recommend most patients can be treated with oral antibiotics, but the studies used to support these guidelines excluded several cohorts. The aims of this thesis were to investigate current UTI management, particularly use of antibiotics and Hospital-in-the-Home (HITH), to define which children actually need IV antibiotics, and to explore the feasibility of using shorter IV antibiotic durations. The initial study focused on use of HITH for IV antibiotics for UTI. The study found that 90% of patients successfully completed treatment without readmission to hospital, and none became severely unwell at home. However, HITH was underused, with only 8% of patients on IV antibiotics using this model. The study supports home IV antibiotics for selected UTI management. The next study used a wider lens, examining all current UTI management in ED and found little variation in antibiotic route, choice, and duration for lower UTIs and uncomplicated upper UTIs. However, complicated upper UTIs, where there is less evidence, showed more variation. No single feature reliably predicted use of IV antibiotics, but multiple complicating factors were associated with their initiation, suggesting they represent ‘red flags’ for clinicians. In the following study, clinicians were asked why they used IV antibiotics for UTI and admitted children to hospital. Although they had defined reasons for using IV, objective clinical evidence did not always support their reasons. The study hypothesised that multiple features lead to the decision to prescribe IV antibiotics and that these features may group together. Given the finding that multiple but not always the same features led to IV antibiotic initiation, the next study aimed to develop a clinical score to determine which children need IV antibiotics. The Melbourne RUPERT score was successfully developed and tested on patients with confirmed and probable UTI. Using six clinical features, the score is straightforward, easily attainable, reproducible and aids consistent stratification of patients to determine route of initial antibiotic treatment. Finally, using outcomes and learnings from the preceding studies, a randomised control trial protocol was developed. This trial aims to investigate whether an ultra-short IV duration (1 day) is non-inferior to the standard duration for children requiring IV antibiotics. If a single daily dose of IV antibiotics followed by oral antibiotics proves sufficient, patients could be discharged home from the ED on oral antibiotics, eliminating the need for hospital admission. In questioning whether more children with UTI could be treated out of hospital, this thesis found that indeed they can, through delivering IV antibiotics at home, treating more children with oral antibiotics, and investigating an identified practice of earlier switch from IV to oral. This has the potential for major impacts on patient care and hospital sustainability.
-
ItemHair cortisol as a measure of stress and social adversity in early childhood( 2019)Background: Children raised in families experiencing adversity (e.g. unemployment, financial hardship, family violence, parent mental health difficulties) are at greater risk of poor health outcomes. Physiological stress is one mechanism thought to explain the effects of early adversity on children's health. Understanding this role of physiological stress and how best to measure it, particularly in population cohorts of young children, is limited. My PhD used child hair cortisol as a measure of physiological stress to address this evidence gap by examining associations between adversity, physiological stress and health in a community-based cohort of young children experiencing adversity. Aims: To investigate 1) whether indicators of adversity are associated with children's physiological stress at 2 years; 2) the role that maternal physiological stress and parenting behaviours play in explaining children's physiological stress response to adversity at 2 years; and 3) whether social adversity is associated with children's health at 3 years, and whether physiological stress is a mechanism which mediates these effects. Methods: This was a prospective longitudinal cohort study nested within the ‘right@home’ randomised controlled trial of nurse home visiting, comprising 722 Australian women recruited for their experience of adversity while pregnant and their subsequent children. Child hair cortisol was collected at ages 2 and 3 years. Social adversity was measured as maternal reported indicators of adversity according to sociodemographic and psychosocial characteristics, collected at pregnancy and child ages 1, 2 and 3 years. Maternal physiological stress (hair cortisol) and parenting were measured at 2 years. At 3 years, child health outcomes of mental health, wellbeing and weight status were assessed. Analyses for Aims 1 and 3 used linear regression models; for Aim 2 used structural equation modelling. Results: Hair cortisol data were available for 319/603 (53%) participating children at 2 years and 297/500 (59%) at 3 years. Aim 1 showed limited evidence of associations between 18 indicators of adversity and child hair cortisol at 2 years across different types, timing and persistence of adversity. Aim 2 showed that maternal and child hair cortisol were associated with one another at 2 years. There was no evidence that adversity or maternal parenting were associated with child hair cortisol; as such there were no pathways by which maternal stress or parenting explained effects of adversity on children’s physiological stress. Aim 3 showed children experiencing greater adversity had higher externalising behaviour problems and poorer physical wellbeing at 3 years, and higher hair cortisol was associated with higher externalising problems; however, there was no evidence that hair cortisol mediated any effects of adversity on health. Conclusions: At ages 2 and 3 years, social adversity was not consistently associated with children’s physiological stress, measured using hair cortisol, nor was there any evidence that physiological stress mediated the effects of adversity on young children’s health. These findings suggest that hair cortisol may be limited as a measure of stress in young children or that measuring one aspect of the physiological stress pathway at the population level provides limited insight into the complex mechanisms underlying children’s health inequalities.
-
ItemEarly-life predictors of vascular phenotypes : in the Child Health CheckPoint and Longitudinal Study of Australian Children( 2018)Background: Cardiovascular disease (CVD) is a leading cause of global mortality and morbidity in the 21st century. The main pathological mechanism in CVD, atherosclerosis, begins early in life. Effective CVD prevention strategies require better understanding of the early-life patterns of CVD risk. Aims: In a cohort of Australian 11- to 12-year-old children and their parents, I sought to describe, in the early-life period, patterns in CVD risk factors and vascular phenotypes indicative of atherosclerosis. Specifically, I sought: I. To investigate, through systematic review, whether early-life SEP influences adult inflammation II. To examine whether early-life SEP is associated with child vascular phenotypes III. To examine whether traditional risk factors (BP, body mass index (BMI), glucose, cholesterol, smoking, diet and physical exercise) in the Ideal Cardiovascular Health (ICVH) score, are associated with child vascular phenotypes IV. To determine whether inflammation, measured by glycoprotein acetyls (GlycA), is associated with child risk factors, and to what extent inflammation also predicts child vascular phenotypes Methods: To address the first aim, I conducted a systematic review and meta-analysis of cross-sectional and longitudinal studies examining child SEP and adult inflammation, in particular C-reactive protein (CRP). The next three aims utilised data from the Child Health CheckPoint (n=1874 child-parent dyads, 49% girls, 88% mothers), a biophysical markers cross-sectional study nested within the Longitudinal Study of Australian Children. It was conducted in 2015 sequentially across 15 cities and towns around Australia. Associations were evaluated with multivariable linear regression, supplemented with survey weights and methods and multiple imputation. Adult associations were examined in parallel when equivalent data were available. Results: In addressing the first aim, meta-analysis showed individuals with parents in the lowest SEP bracket had 25% higher CRP levels than those in the highest SEP bracket in age- and sex-adjusted analyses. The inclusion of adult BMI into models attenuated this association. In the following CheckPoint analyses, both family and neighbourhood socioeconomic disadvantage were associated with structural and functional vascular phenotypes at age 11-12 years, in age- and sex-adjusted linear regression analyses. Structural vascular phenotype associations were independent of traditional CVD risk factors, while functional vascular phenotype associations largely attenuated when adjusted for BMI. The absence of adequate grandparental SEP data precluded parallel adult analyses. Further, each additional CVD risk factor at age 11-12 years was cross-sectionally associated with adverse functional, but not structural, vascular phenotypes. Adult associations were larger in magnitude and associations existed for both structural and functional vascular phenotypes. Associations with BP and BMI were largest in magnitude and independent of other CVD risk factors. Finally, in addressing the last aim, I observed GlycA at age 11-12 years was associated with structural and functional vascular phenotypes in age- and sex-adjusted models. Associations attenuated with inclusion of traditional CVD risk factors, specifically BMI and BP. Adult associations were similar, however GlycA remained associated with PWV independent of CVD risk factors. Conclusion: Socioeconomic gradients in childhood were associated with vascular phenotypes and CVD risk factors. These risk factors were themselves associated with functional vascular phenotypes in children. In adults, CVD risk factors were associated with both vascular structure and function. These findings broadly support CVD prevention strategies starting earlier in life. Further investigation of socioeconomic influences on both novel and traditional CVD risk factors may afford opportunities to improve population cardiovascular health.
-
ItemNovel technology for the measurement of newborn and infant heart rate( 2017)Background: Monitoring heart rate in newborns and infants is crucially important in guiding resuscitation and medical care. Established methods for heart rate assessment of these children have inherent drawbacks. In recent years, novel methods for assessing neonatal and infant heart rate have been developed, with varying levels of evaluation conducted. Digital stethoscopes may provide a better means of heart rate assessment for newborns and infants. Aim: The aim of this thesis was to comprehensively review existing established and novel technologies used to monitor newborn and infant heart rate, and compare new digital stethoscope technology with the gold standard, electrocardiogram (ECG). Methods: This thesis (a) outlines the definition and importance of heart rate in medicine, presented in the context of a review of cardiac anatomy and physiology relevant to understanding this vital sign and aspects of its measurement in neonates and infants; (b) presents a narrative review of established methods for monitoring heart rate; (c) expands the scope of this review from established to emerging methods for monitoring heart rate with a systematic literature review of novel methods for newborn and infant heart rate assessment; (d) describes original research using a prototype digital stethoscope attached to a smart device containing software for detecting and displaying heart rate in real-time that was conducted on infants in the neonatal intensive and special care setting, as well in the delivery room setting using an improved version of the device and software. Results: A review of the literature analysing methods of assessing neonatal and infant heart rate found strengths as well as significant weaknesses in the various methods in clinical use or in development. In the neonatal unit, a prototype digital stethoscope and smartphone device for assessing heart rate had a mean difference (±2 standard deviations) of 7.4 (48.5) beats per minute (bpm) when compared to the gold standard of electrocardiography. The mean (interquartile range) time to first digital stethoscope heart rate display was 4.8 (1 to 7) seconds, and the device failed in 12.3% of use attempts. Repeating the comparison in the delivery room setting using an updated algorithm and new hardware, Bland-Altman analysis revealed a smaller mean difference (±2 standard deviations) between the digital stethoscope and electrocardiography of 0.2 (-18 to +18) bpm including crying periods (Figure 23), and 1.0 (-11 to +12) bpm excluding crying periods. The improved digital stethoscope took a median (interquartile range) of 7 (5 to 11.5) seconds after application to display a heart rate. It failed to detect heart rate in 37% of cases, all of which were in crying infants. Conclusion: A digital stethoscope and smart device with software can rapidly detect neonatal and infant heart rate. In the delivery room, device failure primarily occurred during infant crying, with improved accuracy during non-crying periods.
-
ItemImproving pain and distress management of children in the emergency department( 2015)This thesis represents a compilation of peer-reviewed papers addressing the care for children in pain or undergoing painful or distressing procedures in the emergency department. After an introductory chapter, papers related to one of several themes are briefly introduced and placed in context with each other. The papers, and their compilation in this thesis, aim to provide guidance for paediatricians and emergency physicians caring for children on how to prepare for and reduce procedural and non-procedural pain and distress. Many aspects of this work, including certain research findings, evolving approaches to pain and sedation management and evidence-based teaching materials, are being used by, or have been implemented in, the Emergency Department at Royal Children’s Hospital and other emergency departments across Victoria, Australia, as well as interstate and overseas.
-
ItemA prospective evaluation of symptom based screening for child contact screening and management of tuberculosis in Yogyakarta, Indonesia( 2013)Children who are in close contact with a case of tuberculosis (TB) have a high prevalence of infection with Mycobacterium tuberculosis (M. tuberculosis).Infected children, and in particular young infants, are at high risk of developing severe disseminated TB such as tuberculous meningitis or miliary TB, which is associated with high mortality rates.Screening and management of children who are in close contact with a case of pulmonary TB has the potential to increase TB case finding and prevent disease. This is universally recommended but rarely implemented in TB endemic settings. The WHO recommends a symptom based screening approach that may improve implementation at the primary care level. The main aim of the study was to prospectively evaluate the performance of symptom based screening approach for child contact screening and management. A prospective cohort study of children who were close contacts of a pulmonary TB patient in Yogyakarta, Indonesia, was conducted from August 2010 to December 2012. Clinical assessment, tuberculin skin test and chest X ray (CXR) were performed on all eligible children irrespective of symptoms at baseline. Sputum smear and culture or other appropriate investigations were performed only in children with suspected TB. Following assessment, children were assessed as having TB disease, latent TB infection (LTBI) or TB exposed only. They were treated appropriately as TB disease or with isoniazid preventive therapy (IPT) if they did not have TB and were less than 5 years of age, as per WHO guidelines. All children were followed for 12 months to evaluate response to therapy if diagnosed as TB at baseline evaluation, or to monitor for the development of TB among those without disease at baseline. The effectiveness and cost of the symptom-based screening approach was compared to other possible screening approaches. Data were further analysed to determine factors associated with the risk for infection with M. tuberculosis and with TB disease, to evaluate adherence to IPT, to describe CXR findings and to measure inter-observer agreement in the interpretation of CXR findings in the context of child contact screening and management. At baseline evaluation, of 269 child contacts, 8% had TB disease, 38% had LTBI and 64% were TB exposed only. The results of further evaluations and follow up showed that symptom-based screening provides a simple, effective and low-cost approach to child TB contact management that can be implemented at the primary health care setting. An important strength of this study was that it observed outcomes in the cohort for 1 year with a very low rate of lost to follow up. We identified that children who had close contact with a TB case that was sputum smear-positive case or female were most likely to be infected with M. tuberculosis. We observed that none of 99 children who were provided IPT at baseline developed TB in the first year of observation. Incomplete adherence to IPT was common, and problems related to health care service were identified as possible barriers to adherence. The majority of CXRs of child contacts were normal but there was substantial disagreement between reviewers on the quality and the findings of the CXR. This study provides original data from a prospective and comprehensive evaluation of the symptom-based screening approach. The findings support implementation of contact screening and management at a primary care level, with implications for global efforts to detect and prevent TB in children.
-
ItemNovel approaches to the immunodiagnosis of childhood tuberculosis( 2012)Approximately one-third of the global population is infected with Mycobacterium tuberculosis (MTB). Despite a number of advances, tuberculosis (TB) remains one of the diagnostically most challenging diseases. For almost a century, the tuberculin skin test (TST) was the only method available to identify latent TB infection (LTBI). However, the TST has variable sensitivity and limited specificity. In 2002, a new generation of immune-based tests, interferon-γ release assays (IGRA), were licensed for clinical use. IGRA rely on the detection of interferon-γ produced by sensitised T cells in response to stimulation with relatively MTB-specific antigens. Since these are absent from all BCG vaccine strains and most non-tuberculous mycobacteria, it is assumed that IGRA have greater specificity than the TST. However, increasing data cast doubt on the performance of IGRA, especially in the paediatric setting. Two constellations are particularly common in children: (a) discordance between TST and IGRA results (ie TST+/IGRA- or-TST-/IGRA+), and (b) indeterminate IGRA results. The underlying basis for discordance remains unknown and it is uncertain which of the test results accurately reflects the true TB infection status. Both indeterminate and discordant IGRA results present a considerable management dilemma in clinical practice. The principle hypothesis of this thesis was that a detailed analysis of mycobacteria-specific cellular immune responses would provide insights into the origins of discordance, and provide important information to enable the development of improved immunoassays for the diagnosis of TB. This thesis begins with a systematic review of the literature on the performance of IGRA in children, highlighting the lack of evidence supporting the hypothesis that prior BCG vaccination and exposure to non-tuberculous mycobacteria are the primary cause of common discordance (TST+/IGRA-). The next chapter describes the development of whole blood assays that enable detailed analyses of mycobacteria-specific immune responses. It also describes the development of a complex multi-colour flow cytometric procedure for simultaneous analysis of major T cell surface markers (CD3, CD4, CD8), T cell memory markers (CD45RA, CCR7) and intracellular cytokines (IFN-γ, IL-2, TNF-α, IL-17), as well as the optimisation of Luminex bead-based cytokine assays used to determine mycobacteria-specific cytokine responses in supernatants. The subsequent chapters describe the results of a clinical study in children at risk of TB infection that used these methods. The data show that individuals with common discordance are a heterogeneous group, comprising both TB-uninfected and TB-infected cases. A number of mycobacteria-specific cytokine responses that have the potential to improve diagnostic assays for TB were identified. Importantly, the results indicate that certain cytokine responses can identify TB infection, and also discriminate between LTBI and active TB. Furthermore, the results show that the proportion of mycobacteria-specific single-positive TNF-α+ CD4+ T cells differs significantly between uninfected children, those with LTBI and those with active TB. The discovery of biomarkers that can simultaneously identify TB infection and determine disease stage is an important breakthrough, as existing immune-based TB tests are unable to distinguish between disease stages. This represents a significant advance in TB diagnostics, with the potential to be developed for use in low-resource settings where better TB tests are needed most.
-
ItemCerebral palsy in Victoria: a population-based study( 2012)BACKGROUND AND AIMS: This thesis reports an exploration of the epidemiology of cerebral palsy in Victoria and a comparison of Victorian data with data from other populations. Each of six broad aims related to a particular area of investigation and thesis chapter. The aims were to describe 1) the frequency of cerebral palsy in Victoria, 2) overall temporal trends and by gestational age group, 3) clinical characteristics, including motor type, topographical pattern, and motor severity, 4) accompanying disorders and impairments, 5) neuroimaging patterns and their relationship to perinatal variables and clinical characteristics, and 6) survival. METHODS: The Victorian Cerebral Palsy Register was used to extract data on individuals born in Victoria with non-postneonatally acquired cerebral palsy based on specific criteria for each chapter. A systematic review of the literature was also undertaken to enable comparison of classification, definitions, and epidemiological data with the Victorian data. RESULTS: The decade-specific prevalence of cerebral palsy in Victoria was 1.6 per 1000 live births for the 1980s, 1.8 for the 1990s and 1.7 for 2000-2004. In comparison, the range of prevalence estimates from 33 surveillance systems was 1.8-4.0 per 1000 for 1990s birth cohorts and 1.2-3.0 per 1000 for 1980s cohorts; no methodological or population factors explained the wide variability. The rate of cerebral palsy in Victoria increased during the 1980s and early 1990s, particularly for extremely preterm infants, but also for infants born at term. The data suggest a reversal of this trend from the mid-1990s for both term and extremely preterm infants, a finding observed in other populations. In general, the proportion of cerebral palsy cases with each motor type, topographical pattern, motor severity level, and associated impairment varied widely between surveillance systems, and there was a lack of consensus on definitions and classifications. Victorian data showed a relative increase in mild motor impairment between 1970 and 2003 and an increase in the proportion of cerebral palsy cases with spastic hemiplegia and diplegia relative to spastic quadriplegia; however, no improvement in survival was seen, even for the most severely affected. The most common imaging finding was periventricular white matter injury (PWMI), a pattern associated with spastic diplegia and hemiplegia, and mild-moderate motor impairment. While PWMI was the primary pattern in children born preterm, it was also common in children born at term. Grey matter injuries and malformations were predominantly associated with term birth, and bilateral brain injuries were associated with poor gross motor function. The majority of focal vascular insults and unilateral malformations resulted in spastic hemiplegia. CONCLUSIONS: Through a series of systematic reviews, this thesis has contributed to a better understanding of issues relating to the definition and classification of cerebral palsy and, in particular, emphasised the potential for these to affect estimates of prevalence and distribution of clinical subtypes. In addition, it has resulted in a deeper understanding of clinical, temporal and neuropathological patterns in cerebral palsy and exemplifies the potential benefits of inclusion of MRI data in epidemiological research into a condition with diverse causal networks and clinical manifestations.
-
ItemHealth and development in the 12 months following arterial ischaemic stroke in children( 2012)Background: The breadth of consequences of stroke in children is poorly understood, impacting the targeting of health resources and appropriate measurement of intervention outcomes. Clinical experience suggests children experience functionally disabling consequences following a stroke that change over time. There is limited understanding of the clinical course following diagnosis and factors that impact long-term functional outcome. Aim: This study aimed to: 1) Prospectively measure and describe neurological, language, cognitive and motor abilities of children from acute presentation to 12 months following diagnosis of arterial ischaemic stroke 2) Measure the children’s activity limitations in daily life activities, quality of life, and parent health at 6 and 12 months 3) Explore the relationship between brain capacity markers, cognitive capacity markers, age at injury and environmental factors over a 12 month period 4) Identify factors predictive of functional outcome at 12 months. Method: A prospective single-centre study was undertaken to evaluate the clinical course of 27 children following diagnosis of first arterial ischaemic stroke. The children were evaluated during the acute phase, and at 1, 3, 6 and 12 months following diagnosis across domains of health including neurological function, language, motor, cognitive, social-emotional, adaptive behaviours and quality of life. Family functioning and parent health was also measured and the concerns of the children and parents elicited. Outcomes were explored for the neonatal group (n=6) compared to the older children (aged > 30 days at diagnosis), and the cohort overall. Findings: Neonates had few neurological or detectable health consequences in the first 6 months following diagnosis, however motor impairments and limitations in functional daily abilities were evident by 12 months. In the older age group, a wider range of sequelae was evident. Motor impairments were the most prevalent consequence, and made the most rapid recovery in the first 6 months. Fine motor difficulties emerged over time. Social, behavioural and communication limitations were evident by 12 months. None of the children aged older than 30 days at diagnosis returned to their pre-morbid levels of functional daily ability by 12 months. The health-related quality of life of the older children was impacted, particularly psychosocial health. Parent health was impaired, and burden on the family was sustained through the first 12 months following diagnosis. Functional outcome at 12 months was predicted by the presence or absence of a speech deficit at the time of symptom onset, and one month fine motor or cognitive abilities. Socio-economic status also contributed to functional outcome. Conclusions: The consequences of stroke in children are diverse, can appear over the months following diagnosis, and may continue to emerge beyond the 12 month surveillance period of this study. Measures of impairment only detect a limited range of sequelae. Activity limitations, and participation and quality of life restrictions are prevalent. The impact of the stroke can also extend to the broader family. Health resources need to be targeted to long-term surveillance of this population, to address changing needs in daily life (including home, school and community activities) and parent and family health. Further prospective multi-centre studies of larger populations will enable prognostication of outcome, and the exploration of the individual and environmental mechanisms that contribute to variations in clinical course and outcome.
-
ItemThe ‘Kids Sleep Study': effects of an infant sleep intervention at 6 years of age: follow-up of a randomised controlled trial( 2011)BACKGROUND: Behavioural sleep problems in infancy are highly prevalent. They increase the risk of maternal depression symptoms and predict later child sleep problems, which in turn are associated with adverse mental and general health outcomes for children and parents. Controlled trials demonstrate that behavioural sleep techniques effectively reduce infant sleep problems and associated maternal depression symptoms in the short- to medium-term. However, theoretical concerns remain that these techniques harm child-parent attachment and lead to later child emotional and behaviour problems. Although they are yet to be empirically tested, these concerns are potentially causing harm, as are some practitioners who hold these concerns, by discouraging the use of effective behavioural management strategies. The capacity for behavioural techniques to have long-lasting benefits on child, child-parent and maternal outcomes to school-entry age is also unknown, because studies are yet to examine whether early child sleep problems predict these later outcomes. AIMS: Part 1 aimed to determine whether an infant behavioural sleep program caused lasting harms or benefits at child age 6 years (school-entry) to: (1) attachment-related, (2) sleep and (3) maternal mental health outcomes. Part 2 aimed to determine whether early child sleep problems (4, 12 and 24 months) predicted (4) child sleep, and (5) secondary child, child-parent, and maternal outcomes at child age 6 years, after accounting for concurrent child sleep problems. METHODS: This study drew on a pre-existing population-based cluster controlled trial of a behavioural sleep intervention (n=328, ISRCTN48752250), which recruited infants with parent-reported sleep problems at 7 months from six socioeconomically diverse Melbourne (Australia) local government areas. The original study compared an intervention comprising behavioural sleep strategies delivered by well-child nurses at 1-3 individual consultations at child age 8-10 months, with usual care. The current study revisited children and their mothers at child age 6 years with a parent questionnaire and direct assessment in the family home. Part 1 analysed data as a randomised controlled trial follow-up. Based on the results of Part 1, Part 2 pooled trial arms to analyse data as an epidemiological cohort. Aims were analysed using unadjusted and adjusted regression modelling (linear for continuous outcomes, logistic for binary outcomes) controlling for known confounders selected a priori, where possible. RESULTS: Families retained at 6 years (n=225, 69%) were more likely to speak English than those lost to follow-up (84% versus 74%, p=0.03). There was no evidence of attrition bias between groups. Part 1: There was no evidence of a difference between groups for any attachment-related, sleep or maternal mental health outcome. Although power was limited to detect effect sizes ≥0.4 standard deviations between groups, 95% confidence intervals tended to centre on equivalence and there was no pattern of effect across the sample estimates toward harms or benefits. Part 2: Early sleep problems at individual time points (4, 12, 24 months) rarely predicted outcomes at 6 years, but, when measured as an ordinal persistent count (1, 2, 3 or 4 sleep problems at these ages, taking a single sleep problem at 7 months as the reference category), demonstrated an increasing dose-response relationship with later child, child-parent and maternal outcomes, with evidence strongest for child outcomes. CONCLUSION: This brief behavioural sleep intervention was effective in the short- to medium-term and safe to use in the long-term, at least to five years post-intervention. Parents and health professionals can feel confident using and offering behavioural strategies to manage infant sleep in the second six months of life. In a sample selected for infant sleep problems at 7 months, recurring or persistent early child sleep problems predisposed children and their mothers to poorer outcomes at 6 years. Health professionals need to offer, and parents need to be able to access, strategies that adapt to the changing nature of child sleep problems as the child develops.