>
>Institute of Child Health, University College London - Great Ormond Street
>Hospital NHS Trust
>
>PhD studentships
>
>The joint institutions are offering a number of three-year research PhD
>studentships to start in the academic year 2006.
>Full details and how to apply can be found at
>http://www.ich.ucl.ac.uk/ich/humanresources
>
>For further information about the ICH see our website,
>http://www.ich.ucl.ac.uk
>Applications are invited from committed individuals wishing to do research
>in a clinical context,
>and who expect to graduate with a UK 1st class or upper 2nd class honours
>degree or equivalent.
>
>
>Three projects are ideally suited to individuals with a strong background
>in Mathematics and Statistics:
>
>1. Applications of functional data analysis in physiotherapy and life
>course epidemiology.
> Supervisors: Dr Mario Cortina Borja and Professor Tim Cole
>
>2. Creating age-related centiles with smaller sample sizes.
> Supervisors: Dr Angie Wade and Professor Tim Cole
>
>3. Infection in neonatal intensive care units: health care
>determinants and outcomes.
> Supervisors: Dr Ruth Gilbert and Dr Mario Cortina Borja
>
>The PhD students will be based within the Centre for Paediatric
>Epidemiology and Biostatistics at the Institute of Child Health.
>More information about the Centre is available from:
>http://www.ich.ucl.ac.uk/ich/html/academicunits/paed_epid/paed_ep_unit.html
>
>Informal enquiries may be made to the supervisors.
>Applications should include a CV and the names and email addresses of two
>academic referees.
>
>Eligibility: Full studentships are available to UK applicants.
>Other EU applicants may apply for a fees-only award. Non-EU residents are
>not available for funding.
>
>Closing date for applications: 6th January 2006.
>Interview dates: 30th and 31st January 2006.
>
>If you wish to visit the ICH to discuss these opportunities, there is an
>Open Day on 23rd November 2005 from 2.00pm.
>This will include a display of posters by PhD students which demonstrates
>the range of ICH research.
>
>Project abstracts
>
>1. Applications of functional data analysis in physiotherapy and life
>course epidemiology
>
>Hypothesis: Functional Data Analysis provides a better understanding of
>intrinsic aspects of the data regarded as curves
>than conventional statistical methods.
>
>Aims and methods: Functional data (FD) consist of samples of curves,
>images, or other types of function (1,2).
>These forms of data occur often in clinical and epidemiological studies:
>examples include a child's height measured repeatedly,
>ECG traces, angles formed by the hip and the knee during a child's gait
>cycles, and age-related patterns of viral load
>in an HIV infected child. In all these examples the data change according
>to another variable, usually time,
>and can be plotted as a curve; the aim of FD analysis (FDA) is to model
>characteristics intrinsic to the curves
>(e.g. cyclical patterns or changes in the function's derivatives). The
>curves can also be multivariate: for instance,
>we may be interested in the joint description of height and weight over time.
>
>There are three types of analysis involving FD. In the simplest case we
>seek to quantify the ways in which individual
>curves vary among themselves, for instance differences in growth patterns
>among children. The second case is a regression
>model with a scalar response variable and FD explanatory variables. For
>instance we could model the mucous expulsion from
>the airways as a function of the force-time profile of a physiotherapist's
>hand pressing on a child's chest wall, recorded
>on a mat during physiotherapy manoeuvres (3); this can be plotted as a
>curve versus time and is thus FD. Other examples
>refer to the effect of variables measured in early life, e.g. early growth
>patterns, on the life course e.g. later chronic
>disease (4-6). In the third case the response variable is itself
>functional, with scalar and possibly FD explanatory variables.
>Children with brain injury walk along an electronic walkway which provides
>co-ordinates of point of force/pressure under the
>foot over time so each footfall provides a trace showing the path of the
>centre of pressure (7). The FD response variable is
>the pressure pattern, and this can be related to the child's status (e.g.
>brain injury or control) and other factors relating
>to their severity of injury and/or degree of recovery.
>
>We have access to datasets provided courtesy of Dr Eleanor Main from the
>Portex and Physiotherapy Department at GOSH,
>and to data sets from the Centre for Paediatric Epidemiology and
>Biostatistics relating to life course research (8).
>
>The student will: apply advanced programming in R, review applications of
>FDA in clinical and epidemiological contexts,
>discuss general inference procedures in FDA, and develop models for the
>datasets available.
>
>
>References:
>1. Ramsay J, Silverman BW. Functional data analysis (2nd ed.). New
>York: Springer; 2005.
>2. Ramsay JO, Silverman BW. Applied functional data analysis. New
>York: Springer; 2002.
>3. Gregson RK, Petley GW, Browne M, Pickering RM, Warner JO. A new
>method to quantify manual paediatric chest physiotherapy techniques.
> Physiotherapy 2003;89:611-2
>4. Cole TJ. Modeling postnatal exposures and their interactions with
>birth size. J Nutr 2004;134:201-4.
>5. Singhal A, Fewtrell M, Cole TJ, Lucas A. Low nutrient intake and
>early growth for later insulin resistance in adolescents born preterm.
> Lancet 2003;361:1089-97.
>6. Singhal A, Cole TJ, Fewtrell M, Deanfield J, Lucas A. Is slower
>early growth beneficial for long-term cardiovascular health?
> Circulation 2004;109:1108-13.
>7. Alderson LM, Peters J. Dynamic balance in children with
>coordination problems: the usefulness of the 'Gaitrite' mat.
> In: 6th International Conference on Children with Developmental
> Coordination Disorder; 2005; Trieste; 2005.
>8. de Stavola BL, Nitsch D, dos Santos Silva I, McCormack V, Hardy R,
>Mann V, Cole TJ, Morton S, Leon DA.
> Statistical issues in life course epidemiology. Am J Epidemiol
> 2005;(in press).
>
>Contact: [log in to unmask]
>
>2. Creating age-related centiles with smaller sample sizes.
>
>Hypothesis: The joint modelling of several correlated outcomes to create
>age-related centiles will require fewer measurements
>to achieve the same precision than when each outcome is modelled
>separately. The development of this methodology will be of
>widespread clinical usage.
>
>Aims and Methods: Population reference centiles are used within clinical
>practice to contrast a measurement for a single subject
>with the values seen in a control population. In paediatric applications,
>it is often necessary to adjust reference centiles
>for the age of the child. The statistical aspects of constructing
>age-related centile curves have developed greatly over the last
>10-15 years and have been an area of particular interest to the project
>supervisors (1-10). One area that has not been investigated
>is the joint modeling of correlated outcomes, an approach which will lead
>to greater precision for each univariate set of centiles.
>This development will be particularly important where the collection of
>samples and/or measurements from normal individuals is problematic.
>Each of 3 locally available datasets involving correlated outcomes have
>previously been the subject of developments in the field:
>
>1. Ratings of emotion recognition (10): There are six6 types of emotion to
>be recognised (fear, surprise, anger, happiness,
>sadness and disgust). These data are by nature multinomial although, and
>previous analyses have treated each emotion as
>a separate ordinal outcome (10).
>2. Assessment of visual acuity (5,9): For each child, 4 related
>measurements are made: 2 with both eyes open and 1 for each
>of the eyes separately (other eye patched). If vision is bad in one eye
>then this may manifest itself additionally when
>the eyes are tested jointly.
>3. Immunological measurements of CD4, CD8 and lymphocytes (1,4,7): CD4
>and CD8 are conventionally analysed as percentages of
>all lymphocytes. The percentages of CD4 and CD8 cells are thus inversely
>related to each other and also to CD4, CD8 and absolute
>lymphocyte counts.
>
>The student will develop multivariate methods and quantify how this
>approach affects the sample sizes necessary to obtain
>a specified centile precision. The student will also investigate
>multivariate assessment of individuals i.e. assignment of a
>single composite centile score to an individual dependent on their
>measurement portfolio and contrast the usefulness and
>applicability of composite versus separate scores.
>
>References:
>1. Wade AM, Ades AE, Dunn DT, Newell M-L, Peckham CS with De Maria A
>[The European Collaborative Study].
> Age-related standards for T lymphocyte subsets based on
> uninfected children born to human immunodeficiency
> virus 1-infected women. Pediatr Infect Dis J 1992; 11): 1018-1026.
>2. Cole TJ. Fitting smoothed centile curves to reference data (with
>discussion). J Roy Statist Soc A 1988; 151: 385-418.
>3. Cole TJ, Green PJ. Smoothing reference centile curves: the LMS
>method and penalized likelihood. Stat Med 1992; 11: 1305-1319.
>4. Wade AM, Ades AE. Age-related reference ranges: significance tests
>for models and confidence intervals for centiles.
> Stat Med 1994; 13: 2359-2367.
>5. Wade AM, Ades AE, Salt AT, Jayatunga R, Sonksen PM. Age-related
>standards for ordinal data: modelling the changes
> in visual acuity from 2 to 9 years of age. Stat Med 1995; 14:
> 257-266.
>6. Cole TJ, Freeman JV, Preece MA. British 1990 growth reference
>centiles for weight, height, body mass index and
> head circumference fitted by maximum penalized likelihood. Stat
> Med 1998; 17: 407-429.
>7. Wade AM, Ades AE. Incorporating correlations between measurements
>into the estimation of age-related reference ranges.
> Stat Med 1998; 17:1989-2002.
>8. Pan H, Cole TJ. A comparison of goodness of fit tests for
>age-related reference ranges. Stat Med 2004; 23: 1749-1765.
>9. Wade AM, Salt AT, Proffitt RV, Heavens SJ, Sonksen PM.
>Likelihood-based modelling of age-related normal ranges for ordinal
> measurements: changes in visual acuity through early childhood.
> Stat Med 2004; 23: 3623-3640.
>10. Wade A, Lawrence K, Mandy W, Skuse D. Charting the development of
>emotion recognition from 6 years of age.
> Journal of Applied Statistics. In Press. August 2005.
>
>Contact: [log in to unmask]
>
>
>3. Infection in neonatal intensive care units: health care determinants
>and outcomes.
>
>Hypothesis: Better understanding of the determinants and consequences of
>hospital acquired infection in neonatal intensive care units
>would lead to the development of reliable measures of infection control.
>
>Aims and Methods: The long term aim of the study is to reduce serious
>bacterial infections in neonatal intensive care units (NICUs)
>through better understanding of the reasons for variation in the incidence
>of infection and its consequences. Specific objectives of the
>project are to compare the incidence of bacteraemia in babies admitted to
>NICUs, adjusted for routinely collected risk factors. The study
>will develop analytic and data collection methods for ongoing surveillance
>to measure the effect of interventions to reduce serious
>bacterial infection and evaluate variation between units. A secondary aim
>is to examine the association between bacteraemia and subsequent
>health outcomes during the first few years of life.
>
>Newborn babies admitted for neonatal intensive care rank among the
>populations at highest risk of bacteraemia (bloodstream infection) (1)
>due partly to the relative immune deficiency of premature newborns and the
>use of highly invasive and prolonged supportive care.
>Approximately, 75% of NICU inpatients receive antibiotics, and
>increasingly antibiotics are used intrapartum (2).
>Between 10% and 20% of babies admitted to NICU experience at least one
>episode of bacteraemia (3).
>
>In the UK, information is lacking about how the types of organisms causing
>bacteraemia vary between NICUs, whether these have changed over
>time, and what the consequences of infection are for subsequent
>health.3 In North America, coagulase negative staphylococcal bacteraemia
>has been used as a marker of the quality of infection control and an
>outcome measure for changing practices.4 Analytic methods for monitoring
>risk stratified incidence rates of bacteraemia in NICU need to be
>developed and tested using NHS datasets.
>
>Three datasets (total 6000 patients) are available, ethics approved, and
>cleaned ready for analysis. More sites can join during the project.
>We will examine potential biases in comparing rates over time and between
>units, and use different modelling approaches for risk stratification,
>calculation of incidence rates, and for measuring divergent performance.5
>The aim will be to develop and validate an audit template for monitoring
>bacteraemia rates more widely in the NICUs. This studentship would be
>appropriate for someone with experience of programming and data manipulation,
>and statistical and/or epidemiological expertise.
>
>References:
>1. Nosocomial Infection National Surveillance Service. Surveillance
>of hospital-acquired Bacteraemia in English Hospitals.
> 2002. London, Public Health Laboratory Service.
>2. Gilbert, RE, Pike K, Kenyon SL, Tarnow-Mordi W, Taylor DJ. The
>effect of pre-partum antibiotics on the type of
> neonatal bacteraemia: insights from the MRC ORACLE trials. BJOG
> 2005;112(6):830-832.
>3. Gilbert RE Prenatal screening for group B streptococcal infection:
>gaps in the evidence. IJE 2004;33(1):2-8
>4. Kilbride HW, Wirtschafter DD, Powers RJ, Sheehan MB Implementation
>of evidence-based potentially better practices to
> decrease nosocomial infections. Pediatrics. 2003 Apr;111(4 Pt
> 2):e519-33.
>5. Spiegelhalter, DJ. Monitoring clinical performance: a commentary.
>J Thorac Cardiovasc Surg. 2004 Dec;128(6):820-2.
>
>Contact: [log in to unmask]
>
>
>The UCL Institute of Child Health (ICH) is a postgraduate research
>institution which, together with its partner,
>Great Ormond Street Hospital for Children NHS Trust, contains the largest
>concentration of research expertise in
>the scientific basis of child health in Europe. Its standing is reflected
>in the Grade 5*A awarded in the 2001 HEFCE
>Research Assessment Exercise. The Institute is committed to high quality
>postgraduate education and has a strong
>track record of training and support for its postgraduates. R&D is
>organised into a number of themes operating across
>the site which facilitate a high level of interaction between basic
>scientists, clinicians and population health scientists.
>
>The Institute is offering a number of research studentships for the
>academic year 2006-07 supported by the
>Child Health Research Appeal Trust (CHRAT) and MRC (Medical Research
>Council). Applicants selected for CHRAT or MRC
>funding will be able to choose from the proposed PhD projects
>
>Please note, applications that are submitted without following the correct
>procedure may not be considered.
>4-year PhD studentships funded by the Medical Research Council are also
>tenable at ICH, in the project areas
>listed above. To apply for a UCL 4 year MRC DTA PhD studentship, please
>see: http://www.ucl.ac.uk/mrc-dta
>.
>
>
>
|