HOW TO TEACH BASICS IN OCCUPATIONAL EPIDEMIOLOGY Guido

HOW TO TEACH BASICS IN OCCUPATIONAL EPIDEMIOLOGY Guido F. MOENS • Dept. Occupational & Environm. Medicine, Catholic University of Leuven, Leuven (Belgium) • Dept. Research & Development, Occupational Health Services IDEWE, Leuven (Belgium) EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

1. TEACHING FRAMEWORK 1. 1. Trainees/students: • Post-academic level students : during the 2 -year theoretical training in OM for Ma. Med graduates (former MD) • Other groups : during the training for agreed O. Hygienists • Previous knowledge of epidemiology : = poor or none = BASICS EASOM Summerschool, Bari (IT), 1 -9 -2006

1. TEACHING FRAMEWORK 1. 2. Teacher: Professional situation / expertise : • • • Epidemiologic researcher since 30 years University teacher in OE since 15 years Epidemiologist in OHS since 18 years Director R&D in OHS since 6 years Experience with / access to practice-linked studies/data e. g. : • Standard registration data • (collected during the periodical medical examinations) Surveys by OPh e. g. stress at work, MSD, … EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

2. TEACHING OBJECTIVES Trainees should be able to : 1. Assess the role of E. for OM and for OM-practice: Know the = Objectives of OE = Strengths and limitations in solving workplace health problems = Key features of OE, distinguishing it from other fields of E. EASOM Summerschool, Bari (IT), 1 -9 -2006

2. TEACHING OBJECTIVES Objectives of OE: 1. Describe the occurrence & characteristics of (work 2. 3. -related) health problems among workers Identify hazardous workplace exposures to prevent O. disease (examples of successful prevention) Provide information for • • Risk assessment Prevention in the general population e. g. health effect of low exp. levels of pesticides --> estimation from exposed workers 4. Assess the effect of workplace interventions e. g. replacement of asbestos by MMF EASOM Summerschool, Bari (IT), 1 -9 -2006

2. TEACHING OBJECTIVES Trainees should be able to : 2. Design and perform small OE-study • • • Write a study protocol (draft) Discuss E. research questions with professionals and lay people from OH-practice Collect and analyse data properly Report OE-findings (communication) Interpret and communicate findings from OE-literature Cooperate with research centre in larger OE-study EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

3. TEACHING METHOD & FORMAT 3. 1. METHOD • Practice-oriented / problem-based approach = application of E. methods in a real setting = practical problems encountered in setting up and performing an E. study = interactive teaching EASOM Summerschool, Bari (IT), 1 -9 -2006

3. TEACHING METHOD & FORMAT 3. 1. METHOD • Individual activity of students + group activity e. g. drafting a study protocol (about problem from one’s own practice) = preparation of Master-dissertation e. g. critical reading of E-papers • But, no time left for presentation/ discussion of protocols e. g with course coordinator with other participants with users (company, union) EASOM Summerschool, Bari (IT), 1 -9 -2006

3. TEACHING METHOD & FORMAT 3. 2. FORMAT • Incorporated into one course with • • • Medical statistics Medical data handling (informatics) Theoretical lectures and practical exercises (individual/group) spread over two years (13 hours + 13 hours = 26 hours) • Course notes (containing questions and answers) EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 1. General introduction to E. methodology 4. 2. Setting up an E. study : • Importance of the study protocol • Design of a draft study protocol using a practical example : Prevalence study on low back pain among hospital workers EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 3. Sources of error in E. studies • • Random error Bias - Selection bias - Confounding - Observation bias 4. 4. Characteristics and design of questionnaires EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 5. Examples of types of E. studies • • Short introduction of each study type Real world example for each study type using questions to be answered, from setting up the study, analyzing the data, to the interpretation of the findings. (partly based on the K. Steenland book : Case studies in OE) EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 5. Types of E. studies : 1. Prevalence study 2. (Retrospective) cohort study 3. Case-control study 4. Disease clustering EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 5. Types of E. studies 1. Prevalence study : “Low back pain among hospital workers” (most common study type in OH practice) 2. (Retrospective) cohort study : “Risk of spontaneous abortion among VDT-users” 3. Case-control study “Occupational exposure and end stage renal disease” Calculation of association-, and impact-parameters, sources of bias and error, and the control for these are discussed and illustrated working through the examples. EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 5. Types of E. studies 4. Disease clustering • How to investigate a disease cluster using a stepwise investigation method A cluster is usually a trigger / motivation for a formal E. study To avoid expensive, large, unnecessary (and mostly negative) E. studies --> FIRST: Cluster investigation through 3 steps / 3 questions on Outcome / Association / Exposure EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. 5. Types of E. studies 4. Disease clustering Cluster investigation through 3 steps / questions on Outcome / Association / Exposure 1. Orientation 2. Quantification 3. Quantitative analysis (aggregated level + control group) If positive : Formal E. study design EASOM Summerschool, Bari (IT), 1 -9 -2006

4. TEACHING CONTENT 4. Disease clustering • How to communicate about risk (notion of risk perception) • Two examples : • • Cluster of cancer cases among young nurses in a hospital ward Cluster of urogenital disorders among production workers exposed to chromium 4. 6. Assessment of E. papers Critical assessment of two or three papers from peer reviewed journals (+ research report) (to be prepared by students) EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 1. Political and ethical issues • Controversies : - Strengths and weaknesses of E. methodology - Role in causal/risk assessment - Role of social, economic, political influences (e. g. asbestos debate, EM-fields-debate) • Warning : (Undue) emphasis on criticism of publications --> desillusion on the value of OE versus practical problems in setting up a study (writing of protocol) EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES Practical problems in setting up a study (writing of protocol) • Practical compromises • No study can be perfect / definitive • Contribution to the pool of knowledge BALANCE OF EVIDENCE • Pressure from interested parties to - set up a study quickly, or - delay a study or action e. g. “further studies needed” ≠ delay of prevention • Disclosing conflict of interest / sources of funding EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 1. Healthy Worker Effect (HWE): How to handle? 2. Exposure data: Problem of historical exposure Markers of internal dose /biological effect 3. Confounding 4. Specific study designs (see Content) 5. Use of routine data EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

6. ASSESSMENT / EVALUATION / TIMETABLE 6. 1. Assessment of the students • Year 1 : Classical examination (oral with written preparation) • Year 2 : Discussion with course coordinator on a draft of a study protocol • From 2006 -2007 on : also self teaching exercises in year 2 EASOM Summerschool, Bari (IT), 1 -9 -2006

6. ASSESSMENT / EVALUATION / TIMETABLE 6. 2. Assessment of the course content / format (quality assessment) Each year: anonymous questionnaire filled in by students on - content - format - course notes EASOM Summerschool, Bari (IT), 1 -9 -2006

Overview 1. 2. 3. 4. 5. Teaching framework Teaching objectives Teaching method & format Teaching content Specific content issues • • Political & ethical issues Major methodological issues 6. Evaluation / timetable 7. Limitations EASOM Summerschool, Bari (IT), 1 -9 -2006

7. LIMITATIONS • Very poor basic knowledge of E. or statistics among starters (Ma. Med) • More time needed : e. g. • • Public presentations / discussions e. g. of protocol More exercises EASOM Summerschool, Bari (IT), 1 -9 -2006

7. LIMITATIONS • Co-ordination / timing with other subcourses (statistics, informatics) is not always easy in practice e. g. the statistics course has to be finished before the analysis aspects of E. studies e. g. discussion about the content of medical informatics subcourse EASOM Summerschool, Bari (IT), 1 -9 -2006

REFERENCES • Olsen J, Saracci R, Trichopoulos D. Teaching epidemiology: Chpt 11. IEA, New York : Oxford Univ. Press, 2000. • Steenland K. Case studies in occupational epidemiology. New York : Oxford Univ. Press, 1993. EASOM Summerschool, Bari (IT), 1 -9 -2006

THANK YOU FOR YOUR ATTENTION! ANY QUESTIONS? EASOM Summerschool, Bari (IT), 1 -9 -2006

EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 1. Healthy Worker Effect (HWE) = type of selection bias in comparing workers outcome with general population outcome because: • Workers are healthier at hiring • Selection of unhealthy workers out of the workforce. Mostly discussed in the context of cohort (follow-up) studies, but also present in other types (e. g. prevalence studies) EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues How to handle the HWE? • Use of appropriate comparison groups (within the cohort = internal comparison) but does not affect those lost to the workforce. • Limitation of the extrapolation of the findings. EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 2. Exposure data • • Estimation of historical exposure = difficult Personal information/ interview/ records on work history, causes: - Recall bias - Other forms of reporting bias Job exposure matrices (but also based on personal job information) Real life examples to show the problems involved EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 2. Exposure data • Molecular E. methods = markers of internal dose / biological effect but major limitations for historical exposure (only last weeks or months) Choice of most appropriate method = varies from study to study EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 3. Confounding • Could be present in all E. studies but the association with exposure and outcome must be strong in OE e. g. exposed (manual) workers usually will have the same distribution of lifestyle factors (e. g. smoking) as non-exposed (manual) workers • Complex mixtures (e. g. welding fumes) --> identification problem EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 4. Study designs : Specific methodological issues and characteristics e. g. due to the OH-situation, cohort studies are more common (most occupational exposures are rare in general population). 5. Use of routine data • Mortality data : - Occupation is not reliably registered and coded on • • the Belgian death certificate - Only the last occupation Registers : occupation is mostly not registered Thus : standardization techniques are not in the course EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 6. Cohort studies • Relatively frequent : rare exposure + historical personal hygiene records --> retrospective / historical cohort studies • Calculation of : - Descriptive parameters • Incidence risk • Incidence rate (person-time) - Associative parameter • Relative risk = I risk ratio or I rate ratio - Impact parameter • Attributable risk (etiologic fraction) EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 7. Case-control studies • Nested within a cohort • More efficient than analysis of the whole cohort • Combination of • • • Rare exposure (cohort) Rare outcomes (case-control) For - common exposure - dispersed groups (e. g. farmers) Population-based case-control studies EASOM Summerschool, Bari (IT), 1 -9 -2006

5. SPECIFIC CONTENT ISSUES 5. 2. Major methodological issues 8. Prevalence studies • Frequent use in practice (description of a health problem) • Mostly for non-fatal O. disorders (asthma, MSD, …) • HWE is of concern (prevalent cases = selected sample) EASOM Summerschool, Bari (IT), 1 -9 -2006

STEPS IN CLUSTER INVESTIGATION EASOM Summerschool, Bari (IT), 1 -9 -2006