INTRODUCTION TO CLINICAL RESEARCH July 2011 David H. Rubin, MD Chairman, Department of Pediatrics, St. Barnabas Hospital Professor of Clinical Pediatrics.

INTRODUCTION TO CLINICAL

RESEARCH July 2011

David H. Rubin, MDDavid H. Rubin, MDChairman , Department of Chairman , Department of

Pediatrics, St. Barnabas HospitalPediatrics, St. Barnabas HospitalProfessor of Clinical Pediatrics Professor of Clinical Pediatrics

Albert Einstein College of Albert Einstein College of MedicineMedicine

OBJECTIVES• Discuss framework of clinical researchDiscuss framework of clinical research

• Development of hypothesis, research Development of hypothesis, research question, methods, analysisquestion, methods, analysis

• Development of project within Development of project within residency - advantages and residency - advantages and disadvantagesdisadvantages

• Prepare for ILPPrepare for ILP• Achieve competency in Achieve competency in practice based practice based

learninglearning

THE PROCESS OF THE PROCESS OF RESEARCHRESEARCH

• Phase I (pre-operational)Phase I (pre-operational)• Period of creativityPeriod of creativity• Laying the groundworkLaying the groundwork• Asking the right questionAsking the right question

• Phase II (operational)Phase II (operational)• IRB application and presentation IRB application and presentation

(with mentor); deadline Jan 1, (with mentor); deadline Jan 1, 2012 2012

• Initiation of study planInitiation of study plan

THE PROCESS OF THE PROCESS OF RESEARCHRESEARCH

•Phase III Phase III •Data analysisData analysis•Presentation – Resident Presentation – Resident Research Day, Local and Research Day, Local and National MeetingsNational Meetings

•ManuscriptManuscript

PHASE IPHASE I • Develop Develop hypothesishypothesis and and research question research question • Complete literature reviewComplete literature review

• National Library of Medicine/pubmed National Library of Medicine/pubmed

• ““has it been done before?”has it been done before?”• Determine methods and statisticsDetermine methods and statistics

• Sample size – are there enough Sample size – are there enough patients?patients?

• Independent and dependent variablesIndependent and dependent variables• Confounding variablesConfounding variables• Any unusual problems/costs related to Any unusual problems/costs related to

your project?your project?

PHASE IIPHASE II • Presentation and development Presentation and development

of ideas with peers and facultyof ideas with peers and faculty• Finalize methods, analysis, Finalize methods, analysis,

sample sizesample size• Submit IRB applicationSubmit IRB application• Pilot instrument/surveyPilot instrument/survey• Prepare data collection formsPrepare data collection forms• Enroll subjectsEnroll subjects

PHASE IIIPHASE III • Data entry and cleaning Data entry and cleaning

(statistical package: SPSS, (statistical package: SPSS, sysstat, SAS)sysstat, SAS)

• Data analysisData analysis• Prepare abstractPrepare abstract• Present of project Present of project • Prepare of manuscriptPrepare of manuscript

RESEARCH PROJECT: RESEARCH PROJECT: Practical Practical

ConsiderationsConsiderations• ““Do-able” in 3 years?Do-able” in 3 years?• Funding required?Funding required?• Research assistant required?Research assistant required?• Interesting question?Interesting question?• Do I have enough passion to Do I have enough passion to

spend the time necessary to spend the time necessary to complete project?complete project?

INSTITUTIONAL REVIEW INSTITUTIONAL REVIEW BOARDBOARD

• IRB approval required prior to contact IRB approval required prior to contact with medical records or study with medical records or study subjectssubjects• Approval also required for abstract Approval also required for abstract

submission, presentation, and submission, presentation, and publicationpublication

• Protection of study subjectsProtection of study subjects• Importance of consent form – English Importance of consent form – English

and Spanishand Spanish• May take several months for approvalMay take several months for approval

RESEARCH PROJECT: RESEARCH PROJECT: Potential TopicsPotential Topics

• Case study and review of the Case study and review of the literatureliterature

• SurveySurvey• Cross-sectional studyCross-sectional study• Case-control studyCase-control study• Retrospective chart reviewRetrospective chart review• Prospective studyProspective study

TIMELINETIMELINE• Year 1 (July 11-June 12):Year 1 (July 11-June 12):

• July-December: determine question July-December: determine question and methods; complete literature and methods; complete literature search (National Library of Medicine, search (National Library of Medicine, etc) and faculty/colleague critiqueetc) and faculty/colleague critique

• January 1, 2012: submit application to January 1, 2012: submit application to IRB (with faculty mentor)IRB (with faculty mentor)

• January-July: initiate projectJanuary-July: initiate project• Year 2 (July 12-June 13): Year 2 (July 12-June 13):

• July-July: collect data, analyze dataJuly-July: collect data, analyze data

TIMELINETIMELINE• Year 3 (July 13-June 14):Year 3 (July 13-June 14):

• July-December: prepare July-December: prepare abstract for Spring 2014 abstract for Spring 2014 presentationpresentation

• May: prepare poster for SBH May: prepare poster for SBH Resident Research DayResident Research Day

• June: presentation at Grand June: presentation at Grand RoundsRounds

LITERATURE SEARCH • St. Barnabas Hospital librarySt. Barnabas Hospital library• National Library of MedicineNational Library of Medicine

• Pubmed Pubmed • Google Google • Topic, authorTopic, author• Read/critique all pertinent articlesRead/critique all pertinent articles

• Similar ideas in the literature?Similar ideas in the literature?• Methodology problems? Methodology problems? • Can you do it better?Can you do it better?

• If journal not available, order through If journal not available, order through PMID numberPMID number

OUTLINE OF STUDY OUTLINE OF STUDY PROTOCOLPROTOCOL

Research questionResearch question (objective (objective of the study, must be focused)of the study, must be focused)

What question(s) does What question(s) does the study address?the study address?

SignificanceSignificance (review prior (review prior research and state its research and state its problems; proposed research problems; proposed research may help resolve problems)may help resolve problems)

Why is the research Why is the research question important?question important?

DesignDesign (time frame and (time frame and epidemiologic approach)epidemiologic approach)

What is the structure of What is the structure of the study?the study?

SubjectsSubjects (selection and (selection and sampling)sampling)

Who are the subjects and Who are the subjects and how will they be how will they be selected?selected?

VariablesVariables (independent, (independent, dependant, confounding)dependant, confounding)

What measurements will What measurements will be made?be made?

Statistical issues Statistical issues (hypotheses, (hypotheses, sample size, approach to sample size, approach to analysis)analysis)

How large is the study; How large is the study; what is the analysis?what is the analysis?

STUDY OUTLINETITLETITLERESEARCH RESEARCH QUESTION/HYPOTHESISQUESTION/HYPOTHESISSIGNIFICANCE (REVIEW OF SIGNIFICANCE (REVIEW OF LITERATURE)LITERATURE)DESIGNDESIGNSUBJECTS-ENTRY CRITERIASUBJECTS-ENTRY CRITERIA

SUBJECTS-RECRUITMENTSUBJECTS-RECRUITMENTVARIABLES – PREDICTOR VARIABLES – PREDICTOR (INDEPENDENT)(INDEPENDENT)VARIABLES – OUTCOME VARIABLES – OUTCOME (DEPENDENT)(DEPENDENT)SAMPLE SIZE, POWER, SAMPLE SIZE, POWER, α,α,ß, ß, STATISTICAL STRATEGY STATISTICAL STRATEGY

ASKING THE RIGHT QUESTION

(Eng, 2004)• State the question in writingState the question in writing• Question should be important, novel, Question should be important, novel,

and answerableand answerable• Question should provide useful Question should provide useful

informationinformation• Question should be significant – ask Question should be significant – ask

colleagues if it iscolleagues if it is• Interesting• Novel• Ethical• Relevant

CHOOSING THE RIGHT PROJECT

• What makes a research project What makes a research project outstandingoutstanding??• Logical flowLogical flow of ideasof ideas

• Hypothesis/aim -> methods -> Hypothesis/aim -> methods -> analysis -> conclusion based on analysis -> conclusion based on data -> impact of study data -> impact of study (?new (?new way of thinking about subject?)way of thinking about subject?)

• Every detailEvery detail reviewed – can it be reviewed – can it be improved?improved?

PICKING A RESEARCH PROJECT

(Kahn, 1994)

• Anticipate results Anticipate results beforebefore the study the study • Choose area on the basis of interest Choose area on the basis of interest

of the outcome to the scientific of the outcome to the scientific communitycommunity

• Look for “underoccupied niche” with Look for “underoccupied niche” with potentialpotential

• Attend lectures and read papers Attend lectures and read papers outside of your area of interestoutside of your area of interest

• Build on a themeBuild on a theme

PRACTICAL ISSUESPRACTICAL ISSUES• Are questionnaire and/or instruments

sensitive enough to detect differences in the major outcome variables?

• Are there enough subjects available?• Expand inclusion criteria, lengthen

enrollment period• Too many subjects excluded refusing to

participate, lost to follow-up? • Reduce exclusion criteria

• Do you have and/or need a lot of time and funding?

• Should you consider a pilot study first?

PRACTICAL ISSUES

• If considering a retrospective If considering a retrospective design, watch out for selection bias design, watch out for selection bias (e.g. asthma treatment at a (e.g. asthma treatment at a community hospital)community hospital)

• Describe study populationDescribe study population• Collect information on those who Collect information on those who

declined to participate or “dropped declined to participate or “dropped out”out”

• Define “positive, negative, no Define “positive, negative, no change” change”

POTENTIAL PROBLEMS POTENTIAL PROBLEMS AND SOLUTIONSAND SOLUTIONS

Potential Problems SolutionsResearch question too Research question too broadbroad

Specify smaller set of variables, Specify smaller set of variables, narrow the questionnarrow the question

Not enough subjectsNot enough subjects Expand inclusion criteria, Expand inclusion criteria, modify exclusion criteria, add modify exclusion criteria, add other sources for subjects, other sources for subjects, lengthen entry time into study, lengthen entry time into study, decrease sample sizedecrease sample size

Methods beyond Methods beyond investigator’s skillsinvestigator’s skills

Collaborate with other Collaborate with other colleagues, review literaturecolleagues, review literature

Too expensiveToo expensive Consider less costly study Consider less costly study designs, fewer subjects, designs, fewer subjects, measurements, follow-up visitsmeasurements, follow-up visits

Not interesting or Not interesting or vaguevague

Modify question, specify Modify question, specify outcome, independent and outcome, independent and dependent variablesdependent variables

VOCABULARY

VARIABLES• DimensionalDimensional

• Age, scores, serum NaAge, scores, serum Na• CategoricalCategorical

• Gender (male, female), age (0-Gender (male, female), age (0-10, 10, ≥ 10-20, ≥20-30), ethnic ≥ 10-20, ≥20-30), ethnic (white, black, asian, hispanic)(white, black, asian, hispanic)

• Independent – how does this Independent – how does this variable affect outcome (under variable affect outcome (under researcher’s control) researcher’s control)

• Dependant – outcome variables Dependant – outcome variables (not under researcher’s control)(not under researcher’s control)

VARIABLECATEGORICAL(QUALITATIVE)

NUMERICAL(QUANTITATIVE)

NominalCategories are mutually exclusive & unordered; gender, blood group

OrdinalCategories are mutually exclusive & ordered; social class, disease stage

CountsInteger values; sick days per year, ED visits for asthma in 6 months

Measured (continuous)Any value in a range of values; birthweight (kg), age (years), scores on a testCampbell, 2007

NULL HYPOTHESIS• There is There is nono association between the association between the

independent and dependant variablesindependent and dependant variables• Assuming no association, statistical Assuming no association, statistical

tests estimate the probability that an tests estimate the probability that an association is due to chance (p<.05, association is due to chance (p<.05, 1/20)1/20)

• If there IS an association (p<.05, If there IS an association (p<.05, p<.01), we reject the null hypothesisp<.01), we reject the null hypothesis

HOW ARE THESE RELATED?HYPOTHESISHYPOTHESIS

SAMPLE SIZESAMPLE SIZE

POWERPOWER

RELATIONSHIP• The The hypothesishypothesis determines the type determines the type

of study of study • Risk of Reyes syndrome and aspirinRisk of Reyes syndrome and aspirin• Drug A v Drug B and asthmaDrug A v Drug B and asthma

• Avoidance of type I and II errors Avoidance of type I and II errors needs to be assured by adequate needs to be assured by adequate sample size sample size so study is adequately so study is adequately poweredpowered to show a difference to show a difference

SAMPLE SIZE CALCULATIONS

(Maggard et al, Surgery 2003;134:275)• Identified articles in 3 major surgical Identified articles in 3 major surgical

journals from 1999-2002 (journals from 1999-2002 (Annals of Annals of Surgery, Archives of Surgery, Surgery)Surgery, Archives of Surgery, Surgery)

• Was there 80% power to detect Was there 80% power to detect treatment group differences – large treatment group differences – large (50%) and small (20%), one-sided, (50%) and small (20%), one-sided, =.05=.05

• If underpowered, how many more If underpowered, how many more patients needed?patients needed?

SAMPLE SIZE CALCULATIONS (Maggard , 2003)

• 127 RCT identified; 48 (38%) reported 127 RCT identified; 48 (38%) reported sample size calculationssample size calculations

• 86 (68%) reported positive treatment 86 (68%) reported positive treatment effecteffect

• 41 (32%) found negative treatment effect41 (32%) found negative treatment effect• 63 (50%) of studies appropriately 63 (50%) of studies appropriately

powered to detect 50% effect changepowered to detect 50% effect change• 24 (19%) had power to detect 19% 24 (19%) had power to detect 19%

differencedifference• Of underpowered studies: >50% needed Of underpowered studies: >50% needed

to increase sample size 10 Xto increase sample size 10 X

COMMON ERRORS•Sample size estimates Sample size estimates

subjects to be subjects to be followedfollowed not not subjects subjects enrolledenrolled ( (beware beware of dropouts and problems of dropouts and problems in enrollmentin enrollment))

•Don’t estimate sample size Don’t estimate sample size late in the studylate in the study

and P VALUE• Significance level = Significance level = (Type I error) (Type I error)• QuestionQuestion: What is the association of : What is the association of

watching TV and developing asthma?watching TV and developing asthma?• Set Set to .05 to .05• 5% is maximum chance of incorrectly 5% is maximum chance of incorrectly

inferring TV and asthma are related inferring TV and asthma are related when they are not related when they are not related

• If P value < If P value < , null hypothesis rejected , null hypothesis rejected – conclusion: TV is related to asthma– conclusion: TV is related to asthma

• If P value > If P value > , null hypothesis accepted , null hypothesis accepted – conclusion: TV not related to asthma– conclusion: TV not related to asthma

β and POWER• ββ: probability of Type II error: probability of Type II error• Type II errorType II error: incorrectly : incorrectly

assuming no difference exists assuming no difference exists between 2 groupsbetween 2 groups

• Small differences require large Small differences require large sample sizessample sizes

TYPE I AND II ERRORS• Type I (false positive)Type I (false positive)

• Investigator Investigator rejectsrejects the null the null hypothesis (no association hypothesis (no association between groups) that is actually between groups) that is actually true in the population true in the population

• Effect sizeEffect size: size of association : size of association detectable in population sample detectable in population sample of clinical importanceof clinical importance

TYPE I AND II ERRORS• Type II (false negative)Type II (false negative)

• Investigator Investigator fails to rejectfails to reject the null hypothesis that is the null hypothesis that is actually not trueactually not true• Sample size too small to Sample size too small to detect difference in detect difference in comparison groupscomparison groups

POWER PROBLEMS• Low PowerLow Power

• Too little dataToo little data• Meaningful effect size Meaningful effect size

difficult to determinedifficult to determine• High PowerHigh Power

• Too much dataToo much data• Trivial effect sizes detectedTrivial effect sizes detected

EFFECT SIZE• What is the What is the magnitudemagnitude of the of the

association between independent and association between independent and dependant variables?dependant variables?• Large: Large: easy to detecteasy to detect• MediumMedium• Small: Small: difficult to detectdifficult to detect

• Decide a priori what is important Decide a priori what is important clinicallyclinically

• Should be units of a response – not %Should be units of a response – not %• Use effect size for the most important Use effect size for the most important

hypothesis for sample size planninghypothesis for sample size planning

NUMBER NEEDED TO TREAT

• Usually seen in results of clinical trialUsually seen in results of clinical trial• PPexpexp = number of subjects having = number of subjects having

success in experimental groupsuccess in experimental group• PPcontrolcontrol = number of subjects having = number of subjects having

success in control groupsuccess in control group• With With nn patients treated in both patients treated in both

groups, then groups, then nnPPexpexp and and nnPPcontrolcontrol are the are the number of patients with success in number of patients with success in each groupeach group

NUMBER NEEDED TO TREAT

• If there was 1 extra success in the If there was 1 extra success in the experimental group, thenexperimental group, then• nnPPexpexp – – nnPPcontrol control = 1= 1

• Thus, the number needed to treat Thus, the number needed to treat in each group in order to obtainin each group in order to obtain one extra success isone extra success is• N = 1/(N = 1/(PPexpexp – P – Pcontrol )control )• NNT = 1/ NNT = 1/ PPexpexp – P – Pcontrolcontrol

NUMBER NEEDED TO TREAT

(Campbell 2007)• Tremendous impact of baseline Tremendous impact of baseline

incidence incidence (Sackett 1997)(Sackett 1997)• Use of antihypertensive drugs to Use of antihypertensive drugs to

prevent death, stroke, or MI prevent death, stroke, or MI • Over 1.5 years with diastolic Over 1.5 years with diastolic

115-129mmHg; NNT = 3115-129mmHg; NNT = 3• Over 5.5 years with diastolic Over 5.5 years with diastolic

90-109mmHg; NNT = 12890-109mmHg; NNT = 128

DIAGNOSTIC TESTS DISEASE DISEASE

++DISEASE DISEASE --

TEST TEST ++ A (TP)A (TP) B (FP)B (FP)

TEST -TEST - C (FN)C (FN) D (TN)D (TN)•Sensitivity: A/A+C•Specificity: D/D+B•PPV: A/A+B•NPV: D/D+C

PREVALENCE/INCIDENCEPREVALENCE/INCIDENCE• PrevalencePrevalence

• Pre-existing + NEW cases in time Pre-existing + NEW cases in time period/population at riskperiod/population at risk

• Has Has allall the cases NEW + old! the cases NEW + old!• Prevalence=Incidence x durationPrevalence=Incidence x duration

• IncidenceIncidence• NEW cases in fixed time NEW cases in fixed time

period/population at riskperiod/population at risk• NEW cases only!NEW cases only!

RELATIVE RISKRELATIVE RISK• Incidence rate of disease in exposed Incidence rate of disease in exposed

group/incidence rate of disease in group/incidence rate of disease in non-exposed groupnon-exposed group• RR=1, risk the same RR=1, risk the same • RR<1, risk RR<1, risk in not exposed group in not exposed group• RR>1, risk RR>1, risk in exposed group in exposed group

• ExampleExample: Among children with : Among children with asthma, there is a 1.5 fold increase asthma, there is a 1.5 fold increase in mortality during the past 5 yearsin mortality during the past 5 years

ODDS AND ODDS ODDS AND ODDS RATIORATIO

• Similar to RR, but is used Similar to RR, but is used primarily in case control studies primarily in case control studies where no true incidence exists where no true incidence exists (need entire population)(need entire population)• OR=1, risk the same OR=1, risk the same • OR<1, risk OR<1, risk in not exposed group in not exposed group• OR>1, risk OR>1, risk in exposed group in exposed group

CONFIDENCE INTERVALCONFIDENCE INTERVAL• Statistical precision of a Statistical precision of a

specific which isspecific which is usually 95% usually 95% around the point estimatearound the point estimate• If CI narrow, If CI narrow, certainty about true certainty about true

effect sizeeffect size• If study unbiased, 95% chance that If study unbiased, 95% chance that

interval includes true effect sizeinterval includes true effect size

CONFIDENCE INTERVALCONFIDENCE INTERVAL• If value corresponding to NO effect (eg If value corresponding to NO effect (eg

RR=1) falls outside the 95% CI, then RR=1) falls outside the 95% CI, then unlikely that results are significant at unlikely that results are significant at the .05 levelthe .05 level

• IF CI barely includes value of no effect IF CI barely includes value of no effect and is wide, significance may have and is wide, significance may have been reached if the study had more been reached if the study had more powerpower

• Advantage of CIAdvantage of CI: can see : can see rangerange of of accepted values and compare with accepted values and compare with what is clinically significantwhat is clinically significant

CONFIDENCE INTERVAL – CONFIDENCE INTERVAL – Clinical ExamplesClinical Examples

• Risk for intracranial bleed after Risk for intracranial bleed after serious head trauma is 8.22, 95% serious head trauma is 8.22, 95% CI=6.25,10.21CI=6.25,10.21• Actual risk could be between 6.25-10.22Actual risk could be between 6.25-10.22• If risk was 1.0, this would indicate no risk If risk was 1.0, this would indicate no risk

between exposed and non exposed between exposed and non exposed groupsgroups

• Sensitivity of clinical exam for Sensitivity of clinical exam for splenectomy is 27% (95% CI 19-36%)splenectomy is 27% (95% CI 19-36%)

PARAMETRIC/PARAMETRIC/NONPARAMETRICNONPARAMETRIC

• Parametric DataParametric Data• Data for which descriptive data are known Data for which descriptive data are known

(usually mean, SD) (usually mean, SD) • Frequency distribution of data defined as Frequency distribution of data defined as

“normal”“normal”• Examples of parametric tests Examples of parametric tests

• T- Test T- Test • Pearson Correlation CoefficientPearson Correlation Coefficient

PARAMETRIC/PARAMETRIC/NONPARAMETRICNONPARAMETRIC

• Parametric DataParametric Data

PARAMETRIC/PARAMETRIC/NONPARAMETRICNONPARAMETRIC

• Nonparametric DataNonparametric Data• Data for which descriptive data cannot be Data for which descriptive data cannot be

obtained due to no measurement scaleobtained due to no measurement scale• No assumption regarding the underlying No assumption regarding the underlying

frequency of the data; only certainty is frequency of the data; only certainty is rank orderrank order

• Examples of nonparametric testsExamples of nonparametric tests• Sign test Sign test • Wilcoxon matched pairs test Wilcoxon matched pairs test • Mann Whitney U TestMann Whitney U Test

PARAMETRIC/PARAMETRIC/NONPARAMETRICNONPARAMETRIC

• Nonparametric DataNonparametric Data

COMMONLY USED STATISTICAL TESTS

PARAMETRIC TESTPARAMETRIC TESTCORRESPONDING CORRESPONDING NONPARAMETRIC NONPARAMETRIC

TESTTESTPURPOSE OF TESTPURPOSE OF TEST

t test for independent

samples

Mann-Whitney U test; Wilcoxon rank-sum test

Compares two independent

samples

Paired t testWilcoxon matched pairs signed-rank

testExamines a set of

differences

Pearson correlation coefficient

Spearman rank correlation coefficient

Assesses linear association

between two variables

One way analysis of variance (F

test)

Kruskal-Wallis analysis of

variance by ranksCompares three or

more groups

Two way analysis of variance

Friedman Two way analysis of variance

Compares groups classified by two different factors

BIASBIAS(Altzema 2004)(Altzema 2004)

• Selection BiasSelection Bias• Selection of subjects systematically Selection of subjects systematically

distorted and may predetermine outcomedistorted and may predetermine outcome• Example: Example: hospital hospital study of diarrhea will study of diarrhea will

overestimate severity of diseaseoverestimate severity of disease• Measurement/information BiasMeasurement/information Bias

• Bias in classifying disease, exposure, or Bias in classifying disease, exposure, or bothboth

• Example:Example: knowing too much about disease knowing too much about disease may influence exposuremay influence exposure

BIAS BIAS (Altzema 2004)(Altzema 2004)

• Confounding VariablesConfounding Variables• A factor that may influence the relationship A factor that may influence the relationship

between dependent and independent variablesbetween dependent and independent variables• Example: Example: Risk of morbidity from hypertension Risk of morbidity from hypertension

should control for age, gender, race, etc should control for age, gender, race, etc • Verification BiasVerification Bias

• Patients with positive or negative test result Patients with positive or negative test result preferentially selected for testing – other preferentially selected for testing – other patients may have been missed for testing with patients may have been missed for testing with milder form of the disease milder form of the disease

• Example: Example: Morbidity and childhood asthmaMorbidity and childhood asthma

STUDY DESIGNS

STUDY STUDY DESIGNDESIGN

FEATUREFEATURE EXAMPLEEXAMPLE

Descriptive Descriptive ReportsReports

Recognize Recognize new/atypical new/atypical characteristic of characteristic of diseasedisease

Case report – first Case report – first case(s) of pediatric case(s) of pediatric lyme disease lyme disease

CohortCohort 1 group followed 1 group followed over timeover time

Infants followed for Infants followed for effects of smoke effects of smoke exposure for 2 exposure for 2 yearsyears

Cross-Cross-SectionalSectional

A group examined A group examined at 1 point in timeat 1 point in time

Psychometric Psychometric testing in homeless testing in homeless vs. nonhomeless vs. nonhomeless childrenchildren

Case-ControlCase-Control Two groups, based Two groups, based on outcomeon outcome

Aspirin and Reyes Aspirin and Reyes SyndromeSyndrome

Randomized Randomized TrialTrial

Two groups, Two groups, randomly created, randomly created, blinded blinded interventionintervention

Effect of Effect of educational educational intervention on intervention on asthma morbidityasthma morbidity

DESCRIPTIVE REPORTSDESCRIPTIVE REPORTS• Description of a new aspect or new Description of a new aspect or new

diseasedisease• No comparison group neededNo comparison group needed• Description is usually a basic statistic Description is usually a basic statistic

summary or profile of the group of summary or profile of the group of casescases• Mean, SD, range, confidence intervals, Mean, SD, range, confidence intervals,

correlation between variablescorrelation between variables

Ann Neurol. 2010 Jan 20;68(1):92-101. [Epub ahead of print]Pediatric moyamoya disease: An analysis of 410 consecutive cases.Kim SK, Cho BK, Phi JH, Lee JY, Chae JH, Kim KJ, Hwang YS, Kim IO, Lee DS, Lee J, Wang KC.Division of Pediatric Neurosurgery, Seoul National University Children's Hospital, Seoul National University College of Medicine, Seoul, Republic of Korea

COHORT STUDYCOHORT STUDY

T0 T1

•Population followed forward over time•Baseline: acute pharyngitis•Outcome: Prevention of rheumatic fever or glomerulonephritis •Admission Criteria?: Evidence of ß-hemolytic streptococcus vs pharyngeal inflammation

CROSS SECTIONAL CROSS SECTIONAL STUDYSTUDY

•Collect data on 2 groups at 1 point in time•Compare group differences•Cholesterol levels in athletes vs. non athletes at a midwest university

T0 T1

CASE CONTROL STUDYCASE CONTROL STUDYCASE

CONTROL

•Risk factors in both cases and controls are compared for a condition – especially rare diseases•Important methodology regarding choice of cases, controls

RANDOMIZED CONTROL RANDOMIZED CONTROL TRIALTRIAL

CONTROL

EXPERIMENTALRANDOMIZATION

TIME 0; BASELINE T1; FOLLOWUP

ENROLL SUBJECTS

SUMMARYSUMMARY• Acquire knowledge of research Acquire knowledge of research

process and initiate process nowprocess and initiate process now• Acquire basic knowledge of Acquire basic knowledge of

epidemiology and research epidemiology and research methodsmethods

• Achieve satisfaction in Achieve satisfaction in production/completion of research production/completion of research projectproject

• RE: ILP; establish method of RE: ILP; establish method of criticism of what you do and what criticism of what you do and what is in the literatureis in the literature

CONCLUSIONCONCLUSION• What are the aims of the What are the aims of the

pediatric residency pediatric residency research program?research program?

• What are the elements of What are the elements of the program?the program?

• What is the timeline?What is the timeline?• How do I start?How do I start?

REFERENCES• Kahn CR. NEJM 1994;330:1530Kahn CR. NEJM 1994;330:1530• J Gen Intern Med. 2009 May;24(5):642-8. Epub 2009 J Gen Intern Med. 2009 May;24(5):642-8. Epub 2009

Feb 27.Feb 27.• Tips for teachers of evidence-based medicine: Tips for teachers of evidence-based medicine:

making sense of decision analysis using a decision making sense of decision analysis using a decision tree.tree.

• Lee A, , Joynt GM, , Ho AM, , Keitz S, , McGinn T, , Wyer PC; ; EBM Teaching Scripts Working Group..

• Collaborators (24) Collaborators (24) Wyer PC, , Cook D, , Guyatt G, , Haines T, , Jaeschke R, , Hatala R, , Hayward R, , Fisher B, , Keitz S, , Barratt A, , Dans AL, , Kennedy C, , Montori VM, , Kleinbart J, Lee A, Ho A, Joynt GM, Leipzig R, McGinn , Lee A, Ho A, Joynt GM, Leipzig R, McGinn T, Moyer V, Newman TB, Prasad K, Richardson WS, T, Moyer V, Newman TB, Prasad K, Richardson WS, Wilson MC.Department of Anaesthesia and Intensive Wilson MC.Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, NT, Hong Kong, China.Wales Hospital, Shatin, NT, Hong Kong, China.

REFERENCES• Maggard et al, Surgery 2003;134:275Maggard et al, Surgery 2003;134:275• Altzema C, Ann Emerg Med 2004;44:169-174Altzema C, Ann Emerg Med 2004;44:169-174