The Good Schools Study consists of a cluster randomised controlled trial, a qualitative study, an economic evaluation and a nested cohort study. We are conducting a process evaluation but are seeking funds for that elsewhere. Trial. The trial is a 2-arm, unblinded, cluster randomised controlled trial with parallel assignment (NCT01678846, clinicaltrials.gov). Clusters are 42 primary schools registered in Luwero District, Uganda. Within each primary school, all school staff and a random sample of up to 130 children in Primary 5, 6 and 7 are eligible to participate. Two cross-sectional surveys, one completed in July 2012, and one to be conducted in June-July 2014, are being used to assess children's and schools staffs' experiences/use of violence, and to compare the randomised groups. Our main analysis will be a cross-sectional comparison at follow-up. Qualitative study. The main aims of the qualitative study relate to exploring possible causal pathways through which the intervention is operating and how different aspects of context may affect scalability of the intervention. A range of qualitative methods of data collection will be used with education officials, headteachers, school management committees (which include parents), school staff and learners. We will draw on a range of analytic techniques, including techniques from Grounded Theory, such as constant comparison and seeking deviant cases. Economic evaluation. An economic evaluation of the Good School Toolkit will be performed in order to assess the cost effectiveness of implementing the intervention compared to a 'do nothing' alternative. We will do a full economic costing from the provider's perspective. Nested cohort study. A nested cohort study of all school staff and children in Primary 5 at baseline is also being included to examine changes in educational outcomes and mental health within individuals over time. We will also be able to follow school staff longitudinally. Violence against children in schools is common practice in many countries, and research into prevention and treatment hasbeen outlined as a priority in the World Report on Violence against Children. In most countries, children spend more timeat school than anywhere else besides their family home, and can suffer violence from teachers and other school staff, andother children. Despite this, in most countries evidence is lacking from rigorously conducted studies on the prevalence,epidemiology and consequences of this violence. Existing studies tend to come from North America, where corporalpunishment from teachers is far less common.In Uganda, no rigorous, representative prevalence data exist, but one survey indicated more than 80% of children haveexperienced physical punishments such as caning and slapping by teachers. More research exists on sexual violence inschools suffered by girls in African schools, and qualitative studies indicate girls in Ugandan secondary schools experiencesexual violence and harassment from teachers and fellow students but are not able to report it for fear of reprisals.The Good School Toolkit has been developed and refined for 6 years in Uganda by Raising Voices. The Toolkit takes asystemic approach, involving an entire school in a process of change to reduce violence and improve teaching techniques.The Toolkit draws on the Transtheoretical Model and incorporates standard behaviour change techniques such as setting agoal and making an action plan, which are effective in modifying behaviour. This study aims to determine whether use ofthe Good School Toolkit reduces children's experience of violence by school staff. We will also examine the effects of theToolkit on children's mental health, well-being, and educational outcomes.The study will include a trial, which will be complemented by a qualitative study, a process evaluation, and an economicevaluation. We will also follow a sub-group of trial participants over time. We will conduct a cluster randomised controlled trial in 42 primary schools in Luwero District, Uganda. Half of the schoolswill receive the Toolkit, and other half will be put on a waiting list to receive the Toolkit at the end of the study if it is shownto be effective. School staff, and children in Primary 5, 6 and 7 will be surveyed (aged about 11-14 years) at the beginningand the end of the study, and schools which received the Toolkit will be compared with those which did not.A qualitative study will also be conducted to explore mechanisms by which the Toolkit might be affecting violence, mentalhealth and educational outcomes. In-depth work will focus on how school staff and children have experienced the Toolkitintervention, and what aspects of it may be refined to be more effective.The Toolkit is specifically designed to be implemented at very low cost, appropriate for low income settings. An economicevaluation will be performed to explore the economic and financial costs of this intervention, with the aim of informingpossible scale-up of the Toolkit.All students in Primary 5 at baseline will be followed longitudinally until follow-up, when most will be in Primary 7. This willenable exploration of trajectories of change in mental health and educational outcomes over time, and how violenceexperience impacts this. We will also be able to follow school staff over time.The results from this evaluation will be used to brief policy-makers within the Ministry of Education and Sports involved indeveloping country-wide policy and practice around violence against children in schools.

Bacterial infections contribute significantly to high child mortality rates in Africa, and non-typhoidal Salmonellae (NTS) are consistently amongst the most common invasive isolates. Whilst NTS rarely cause severe illness in the western world, they are a frequent cause of major illness in the tropics. Understanding of how NTS infections are acquired and estimations of NTS disease burden in these settings are limited and strategies for NTS control are non-existent. Goals: 1) establish signifi cant risk factors for NTS acquisition in the community in Nairobi 2) determine estimates for NTS carriage in the general population. I propose a case-control study of children with microbiologically-proven NTS disease with controls from hospital out-patients to investigate risk factors for acquisition of NTS. I will measure the duration of stool carriage and estimate population NTS carriage prevalence. The study will take place in an urban African setting in conjunction with a well-establish ed microbiology laboratory inside a Wellcome Trust MOP. The strain structure of isolates obtained will be explored by MLST. I will prepare for this project by studying the LSHTM MSc in epidemiology. My work will be supervised by two experienced epidemiologists and a Kenyan NTS microbiology expert.

Whooping cough, known as pertussis, is a serious infection, especially in infants. Two types of pertussis vaccine exist: whole-cell (wP) and acellular (aP). We now think that aP vaccines may not be as long-lasting or effective in preventing infection as wP vaccines, but the cause remains unknown. It may be due to differences in specific groups of cells in the immune system called T-helper cells. The type of T-helper cells found in animals and humans who have received wP- compared to aP-vaccines is better at preventing bacteria from being passed on. We still do not know enough about all the important T-helper cells involved and there is no data in very young children. My project is set within a larger study in The Gambia that is vaccinating babies with aP or wP vaccine. I will investigate T-helper cells activated by these vaccines and see how they may help other components of the infant’s immune system. As well as taking blood samples, I will collect and analyse fluid from the infant’s nose (where pertussis bacteria initially infect), to provide further insight into how these vaccines differ in their action. My findings will help to design better pertussis vaccines in the future.

Group A Streptococcus (GAS) is a bacteria which causes severe infections and leads to deadly diseases such as rheumatic heart disease which kills over 300,000 people a year globally, particularly in low-income countries. We do not know how GAS is spread between people, how often people carry GAS in their throat or on their skin without having symptoms, or what factors increase the chance of this occurring. It is important to understand these factors in order to know how to reduce GAS-related disease. This study will follow 444 people in The Gambia, over 12 months, taking samples from the throats and skin of people living in the same households, and asking questions about themselves and their behaviour, at regular intervals. By taking samples over time, we can understand how common it is to carry GAS without having symptoms, how GAS is spread between people, and whether carrying GAS leads to more GAS infections in people or their household members. We will use state-of-the-art techniques to look at the DNA of GAS bacteria that we find, and combine this with a mathematical model to investigate how different strains spread to people within and between households in the community.

The UK has a low proportion of Intensive Care (ICU) beds and a correspondingly high sepsis-related mortality. Appropriate identification of deteriorating ward patients for ICU admission is a national priority area. Current methods use severity-based scores derived from routine ward observations (e.g. pulse, blood pressure). However, the current biological paradigm of sepsis emphasises its time-dependent nature progressing from a pro- to an anti-inflammatory cytokine milieu. Piggybacking a cohort study onto a national ICU audit, this project will firstly evaluate how delay to ICU admission affects outcome (hypothesising that the relationship will rest on these time-dependent sepsis phenotypes). Accordingly, delay should not simply lead to greater severity, but to different clinical states with different outcomes. Secondly, in a detailed prospective cohort of patients with severe pneumonia, daily organ dysfunction will be used to plot timing of ICU admission and organ support aga inst the duration and trajectory of illness. Correlation with clinical phenotypes defined by duration and trajectory will be sought using biological (cytokine and endocrine) markers. Using these longitudinal descriptions of illness, a multi-state model will be developed to assess the effect of altering criteria for, and timing of, ICU admission.