World Health Organisation Collaborating Centre

1. Undertaking your first research project David Goldberg Institute of Psychiatry Thursday 24th May 2012, 16.30 pm “New directions in psychiatry” World Health Organisation Collaborating Centre

2. RUMBA! R Relevant U Understandable M Measurable B Behaviour should be influenced A Attainable KISS = Keep it simple, stupid

3. This goes though stages: • Thinking of an idea • Reading round the subject • Deciding on the method • Research protocol • Finalising the procedures: pilot study • Doing the fieldwork • Processing the results • Writing up

4. Thinking of an idea: • Read a journal • Help an experienced investigator • Carry out a systematic review • think of your daily work • Adapt or develop an instrument? • experiments of opportunity

5. Reading round the subject: • Medline, Psychlit, Pubmed, Web of Knowledge • Decide on keywords • Limit your search • Follow Key papers How to make notes! - note FULL reference right away

6. THE RESEARCH PROTOCOL: Your name Supervisor’s name • Title • Aim (disproving the null hypothesis) • Background • Method • Power calculation • Statistical treatment of results

7. A Time budget Start with today’s date, end with time research must be handed in Time for instrument preparation Pilot study Main field work Processing your results Writing up Time for supervisor to read it Time for you to make corrections andINJURY TIME!

8. What kinds of research designs are unsuitable for single investigators? Assessment of treatment studies Prospective studies of risk – case/control Most studies of process of psychotherapy Evaluations of psychiatric services Epidemiological studies & Cohort studies

9. SAMPLE SIZE POWER: is the ability of a test to show that a relationship exists, when it DOES exist. Also called “false negative”, or Type 2 error.Is sample size big enough? SIGNIFICANCE: is the probability we shall make a false claim, and say a relationship exists which did so by chance. (Also called “false positive”, or Type 1 error). Usually set power at 0.80 (giving an 80% chance of showing a relationship), with significance at 0.05 (giving a 5% chance of a false claim)

10. For example - How big a difference would you be impressed by? (The smaller the difference, the larger the sample!) With power at 0.80, and significance at 0.05 Difference Sample size 30% 95 15% 160 10% 360 Can be looked up in nomograms!

11. Joining an existing research group If their interests are also yours, this is the best solution to your problems. Both in the UK and here in Italy, it is the strategy most often done. You learn research methodology, and get a paper in a high impact journal

12. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 2012 Dorsolateral prefrontal cortex volume in patients with deficit or nondeficit schizophrenia Volpe U, Mucci A, Quarantelli M Galderisi S, Maj M The aim of the present study was to investigate whether patients with deficit schizophrenia differed from those without deficits…In the present study a 3D-T1w MR imaging procedure and an extensive clinical assessment was carried out in 18 patients with schizophrenia (10 with deficits) …. We found a significant reduction of dorso-lateral prefrontal cortex in the entire schizophrenia group, with respect to healthy subjects, but the reduction was greatest in those without deficits….

13. BRITISH JOURNAL OF PSYCHIATRY (2007), 191, 113 - 11 9. • Cortical white-mattermicrostructure in schizophrenia. Diffusion imaging study • N. ANDREONE, M. TANSELLA, R. CERINI, A. VERSACE, G. RAMBALDELLI, • C. PERLINI, N. DUSI, L. PELIZZA, M. BALESTRIERI, C. BARBUI, M. NOSE, • A. GASPARINI and P. BRAMBILLA • Aim To investigate cortical white matter microstructure with • diffusion-weighted imaging in a large community-based sample of • people with schizophrenia. • 68 people with schizophrenia and 64 healthy controls underwent a session of diffusion-weighted imaging to obtain the apparent diffusion coefficient of white-matter water molecules. • Our findings confirm the presence of cortical white-matter • microstructure disruption in frontal and temporo-occipital lobes in • the largest sample of people with schizophrenia thus far studied • with this technique.

14. Working on an existing data-set Find out if there is more work to be done analysing a large data-set that has already been collected. This can be done at home, or even abroad! For this, you need to be in contact with the Principal Investigator!

15. Acta Psychiatrica Scandinavica 2011: 1–8 Patient characteristics and symptoms associated with perceived coercion during hospital treatment Fiorillo A,plus 17 others, including Mario Maj & Stefan Priebe This study aimed to identify the patients’ clinical characteristics associated with compulsory admission or perceived coercion and changes in symptoms and global functioning. Three thousand and ninety three in-patients who were involuntarily admitted or felt coerced to hospital treatment despite a legally voluntary admission were recruited in the European evaluation of coercion in psychiatry and harmonization of best clinical practice – EUNOMIA project in 11 European countries. Involuntary admission, female gender, poorer global functioning and more positive symptoms were associated with higher levels of perceived coercion at admission. Perceived coercion significantly decreased over time, and the improvements in global functioning and positive symptoms were associated with reduction in perceived coercion.

16. General Hospital Psychiatry 28 (2006) 503– 508 Self-reported thyroid disease and mental disorder prevalence in the general population Patten SB, Williams JV, Esposito E, Beck CA Objective: The objective of the current study was to determine whether self-reported thyroid disease is associated with elevated mental disorder prevalence in the general population. Method: Data from the Canadian Community Health Survey Mental Health and Well-being were used. Results: Twelve-month and lifetime mental disorder prevalence was higher in subjects with thyroid disease than in subjects reporting no chronic conditions. … After adjustment for age, sex and other chronic conditions, only social phobia was found to be associated with thyroid disease. Conclusions: People with thyroid disease are not a particularly high-need group for mental disorder screening or intervention, at least not in the community population.

17. The Canadian Journal of Psychiatry, 2007, 52, 780 -789 Frequency and Adequacy of Depression Treatment in a Canadian Population Sample Esposito E, Wang JL…[7 others]...Barbui C, Patten SB The objective of this study was to describe the pattern of treatment for major depression (MD) in Alberta. Telephone survey methods were employed. Random digit dialing was used to select a sample of 3345 household residents aged 18 to 64 years in Alberta. A computer-assisted telephone interview…The antidepressants were taken at therapeutic dosages 87.4% of the time. Most (80.7%) of those taking ADs reported taking them for more than 1 year. The frequency of receiving psychotherapy, or talk therapy was ….14.3%. in respondents with MD. However, most of these subjects were unable to name the type of counselling they were receiving. Conclusions: When compared with previous estimates, these results suggest continued progress in the delivery of evidence-based care to the population….

18. Carrying out a systematic review You may have to think up a subject for one yourself. But compared with most clinical projects, the invasion of your free time is fairly small. You may even go abroad!

19. Schizophrenia Bulletin 2005, 31, 613–617 Association Between the Neuregulin 1 Gene and Schizophrenia: A Systematic Review Sarah Tosato, Paola Dazzan, and David Collier Institute of Psychiatry, Kings College, London Chromosome 8p22–p11 has been identified as a locus for schizophrenia in several genome-wide scans, which has been confirmed by meta-analysis of published linkage data…. …..Not only are these alternative explanations for the linkage seen between chromosome 8p and schizophrenia, but it is equally possible that there is more than 1 susceptibility gene at this locus.

20. Journal of Clinical Psychopharmacology _ 2004, 24, 126-130 ‘‘Wish Bias’’ in Antidepressant Drug Trials? Corrado Barbui, Andrea Cipriani, Paolo Brambilla, and Matthew Hotopf Abstract: The present study investigated whether the outcome of randomized clinical trials studying fluoxetine favored fluoxetine, where this was the experimental agent, and favored comparator antidepressants in trials where fluoxetine was the reference agent. A systematic reviewof all double-blind, randomized clinical trials comparing fluoxetine with any other antidepressant drug in patients suffering from depression was carried out. Thirty-seven studies meeting the inclusion criteria were analyzed. … The evidence that the outcome of fluoxetine trials varied according to whether this drug was used as a new compound or a reference one suggests the presence of bias.

21. Psychological Medicine, 2003, 33, 1149–1160 How often do patients with psychosis fail to adhere to treatment programmes? A systematic review M. NOSĖ, C. BARBUI & M. TANSELLA The purpose of this paper is to provide an overall estimate of treatment non-adherence by psychotic patients in community psychiatric services,…. A total of 103 studies were included in this systematic review. Eighty-six of these studies were suitable for data re-analysis. The overall weighted mean rate of non-adherence, calculated in a sample of 23,796 patients, was 25.78%..... Factors associated with poor compliance included: lack of insight; positive symptoms; younger age; male gender; history of substance abuse; unemployment; and low social functioning. Conclusions. Approximately one in four patients with psychosis fails to adhere with treatment programmes. Preventive evidence-based clinical interventions should be routinely implemented in community settings to reduce patient non-adherence.

22. What other kinds of research designs might you use? A descriptive study of a group of patients that interests you

23. Descriptive studies: • DON’T just look at level of mental morbidity! • Look for an internal comparison: measure mental status, AND other characteristics (eg physical features, eg - extent of diarrhoea, extent of eczema; social features - quality of parenting, social deprivation)

24. What other kinds of research designs might you use? A descriptive study of a group of patients that interests you A study of retrospective risk – a case control study

25. Retrospective studies of risk: Retrospective study, backwards in time a/c b/d Measure with Odds ratio

26. What other kinds of research designs might you use? A descriptive study of a group of patients that interests you A study of retrospective risk – a case control study A validity study of a psychometric test in an Italian setting

27. Two stage designs Suppose you want to find whether a particular screening test works as well in Italy as it is claimed it did in another country… You naturally want as many as possible of those with high scores, and a sample of those with low scores…..

28. Two stage designs Consecutive Selected Non-Cases True cases Attenders for interview HIGH SCORES 178 102 13 89 LOW SCORES 375 98 94 4

29. Two stage designs Consecutive Selected Non-Cases True cases Attenders for interview HIGH SCORES 178 102 13 89 LOW SCORES 375 98 94 4 TOTAL 553 200 107 93 But, those with high scores were under-sampled by 178/102 = 1.75 And those with low scores were under-sampled by 375/98 = 3.80 Multiply last 2 columns by these numbers to re-create the consecutive patients

30. Two stage designs Consecutive Selected Non-Cases True cases Attenders for interview HIGH SCORES 178 102 13 22.7 89 155.3 LOW SCORES 375 98 94 359.4 4 15.3 TOTAL 553 200 107 382.4 93 170.6 Specificity = True negatives / Total non-cases = 359.4/382.4 = 93.6% Sensitivity = True positives / Total Cases = 155.3/170.6 = 91.0% 382.4 + 170.6 = 553!

31. What other kinds of research designs might you use? A descriptive study of a group of patients that interests you A study of retrospective risk – a case control study A validity study of a psychometric test in an Italian setting The assessment of a training course

32. OK, now you can start? • Are you now trained in the clinical interview, or experimental procedure? • Are the measuring instruments appropriate in your setting? • Are the patients available, will the nurses co-operate? • Have you obtained ethical committee approval? • YES: Then try it all out, noting date in your project log. If you have to change ANYTHING, call it a pilot study • Arrange a meeting with your supervisor very quickly, to report that all is now ready