cross sectional study. 2. 1. define the problem 2. specify the objectives 3. select design or type...
TRANSCRIPT
CROSS SECTIONAL STUDY
2
1. Define the problem 2. Specify the objectives
3. Select design or type of study
4. Select study population
5. Collect data
6. Analyze data
7. Determine conclusions
Anatomy of Research
3
Study Design: Definition
• The procedures and methods, predetermined by an investigator, to be adhered to in conducting a research project
• Methods used to obtain valid data to answer a research question (or prove/refute a hypothesis)
5
Relative strength of various study designs (based on level of evidence for a cause & effect relationship)
Strength Design Strong Clinical trial
Cohort study Case control study Cross sectional Case series
Weak Case report
Ecological studiesCross-sectional studiesRetrospective cohort studyProspective cohort studyCase control studyRandomized controlled trial S
tatis
tical
ly s
tron
ger
Mor
e lim
ited
answ
ers
Sta
tistic
ally
wea
ker
Bro
ader
ans
wer
s
Hierarchy of Evidence
A cross-sectional studies
A cross-sectional studies– a type of observational study – the investigator has no control
over the exposure of interest (e.q. diet).It involves
– identifying a defined population at a particular point in time–
measuring a range of variables on an individual basis e.g. include past and current dietary intake
– At the same time measuring outcome of intereste. g. obesity
Measurement of exposure of interest andoutcome of interest is carried out at the same time (e.g.
Obesity and Hypertension)
There is no in-built directionality as both exposure and outcome are present in the study subject for quite some time
Deals with the situation existing at a given time (or during a given period) in a group or populationThese may be concerned with:
– The presence of disorders such as diseases, disabilities and symptoms of ill health
– Dimensions of positive health, such as physical fitness– Other attributes relevant to health such as blood
pressure and body measurements– Factors a/w health & disease such as exposure to
specific environmental exposure or defined social & behavioral attributes and demographic attributes
– Determining the workload of personnel in a health
program as given by prevalence
May be– Descriptive– Analytical or– Both • At descriptive level, it yields information about
a single variable, or about each of number of separate variables in a study population
• At analytic level, it provides information about
the presence and strength of associations between variables, permitting testing of hypothesis
Essential feature of cross-sectional studies
-They collect information relating to a single specified time
•But, often extended to include historical
information which leads to demonstration of statistical associations with past experience e.g. investigation of an epidemic
Temporal association
Steps of cross–sectional study
Choose the problem & analyse it Important steps:– Problem identification– Prioritize the problem– Analyze the problem to convert it in “Research
Question”• Specific• Measurable• Realistic• Time bound • Questions to ask:– What is the problem?
– Why should it be studied?
Literature review What information is already available? • Helps you understand and analyze the
problem – Is it the same thing which is bothering me? – Uncertainty about a health issue that the
investigator wants to resolve • Helps you to frame SMART research question
FINER RQFeasible– Adequate number of subjects– Adequate technical expertise– Adequate resources (time & money) • Interesting to investigator • Novel –Confirms or refutes previous findings – Extends previous findings Provides new findings • Ethical • Relevant – For scientific knowledge – For policy implications – For future research directions
Research methodologyQuestions to be asked:– What data do we need to meet our objectives?– How will I get this?– How will it be collected?• Elements:– Study population– Study subjects – Sampling & Sample size• Variables– Data collection instruments & techniques & plan– Data management – data processing & analysis– Ethical clearance– Piloting
Choosing the study subjectGood choice of study subjects serves the vital
purpose of assuring that the findings in the study accurately represent what is going on in the population
– Sample of subjects which are affordable in time
& money, – yet it is large enough to control random error in
generalizing the study findings to the population – and representative enough to control systematic
error in these inferences
Sampling methodsProbability sampling– Simple random sampling– Systematic sampling– Stratified random sampling– Cluster sampling • Non-probability sampling – Consecutive sampling – Convenience sampling – Purposive (Judgmental) sampling
One sample situation:A. Proportion
Estimating a population proportion with specified precision– Absolute– Relative• Hypothesis test for population proportionB. Mean
Estimating a population mean with specified precisionEstimating sample size with unknown meanHypothesis test for population mean
Two sample situationA. Proportions• Estimating difference between two population proportions with specified
precision• Hypothesis test for two population proportionsB. Means• Estimating difference between two population means with specified
precision• Hypothesis test for two population means
Sample sizeAbsolute N=Z2p(1-p)/d2 • Relative– N=Z2p(1-p)/e2p• Hypothesis test– N={Z1-α* sqrt[p0(1-p0)+ Z1-β*
sqrt[pa(1-pa)]}2/(p0-pa)2 Note – Replace α by α/2 for two tailed
hypothesis
variablesType of variable
characteristic
example
Appropriate statistics
Information content & power
Nominal Unordered categories
Sex, blood group
Counts rate proportion, RR, chi square
low
Ordinal Ordered categories with interval
Degree of pain
Above & median rank correlation
Intermediate
Continuous or discrete
Ranked spectrum with quantifiable intervals
Weight , number , cigarettes /day
Mean, SD t-test, ANNOVA
high
Data collection Data collection instrument Data collection plan
Quality check plan
Data collection instrument / Questionire /interview schedule General: – Brief description of purpose of study – Instructions specifying how to fill – Group the questions concerning major subject
area under a short heading – Warm-up questions• Open-ended & close-ended questions• Instrument format – Format should make it as easy as possible for
filling and avoiding data entry confusions• Wording – Clarity, simplicity, neutrality, double-barreled
questions, time frame• Codes, scores and scales
Steps in designing questionireMake a list of variables•Borrow from other instruments•Write a draft•Revise•Pretest•Shorten and revise again•Precode
Precision & accuracyPrecision Accuracy
Defn The degree to which the variable has same value when measured several times
the degree to which a variable actually represents what is supposed to represent
Best way to assess Comparing among repeated measures
Comparison with a reference standard
Value of study increase power to detect effects
Increase validity of conclusion
Threatened by Random error Systemic error
Sources of error Systematic error (bias):– Confounding bias:• Lack of comparability between the exposed & unexposed with regards to
other factors that affect the risk of developing the disease– Misclassification bias:• Errors in the classification of subjects according to exposure or disease –
interviewer bias, response bias, recall bias– Selection bias:• Selection of subjects or their participation in the study is influenced by the
disease under study– Sample bias – non-representative sample selection– Non-response bias– Non-participant bias– Berkson’s bias– Membership bias• Random error (chance):• Uncertainty introduced by small number of observations
Strategies in dealing with systemic error Confounding bias:– Restriction– Matching– Stratified analysis/Multivariate analysis• Misclassification bias:– Blinding– Minimal gap between theoretical and empirical
definition of exposure/disease• Selection bias:– Population should be defined independently of
disease of interest– All information on the subjects should be secured to
avoid selective loss of information– Prevent loss to follow-up
Uses of cross sectional studyThe findings may be used to promote
the health of the population studied i.e. can be used as tool in community health care
• Can contribute to clinical care • Can provide “new knowledge” • The uses are not mutually exclusive
& single study can fulfill more than one purpose
Uses in community health careCommunity diagnosis – Health status – Determinants of health & disease – Association between variables – Identification of groups requiring special care • Surveillance • Community education & community
involvement • Evaluation of community’s health care
Guideline for critical appraisal of prevalence study 1. Are the study design & sampling method appropriate for the
RQ?2. Is the sampling frame appropriate?3. Is the sample size adequate? 4. Are objective, suitable and standard criteria used to measure
the health outcome?5. Is the health outcome measured in unbiased manner? 6. Is the response rate adequate? Are the refusers described? 7. Are the estimates of prevalence given with CI & in detail by
subgroup – if appropriate? 8. Are the study subjects and the setting described in detail and
similar to those of interest to you?
Cross sectional study advantageCheap and quick studies. •Data is frequently available
through current records or statistics.
•Ideal for generating new
hypothesis
Cross sectional study DisadvantageThe importance of the relationship between
the cause and the effect cannot be determined.
• Temporal weakness: – Cannot determine if cause preceded the
effect or the effect was responsible for the cause.
– The rules of contributory cause cannot be
fulfilled.
Choice of strategy
Advantage & disadvantage of different observational study design
Ecological study
Cross sectional
Case control
cohort
Probability of
Selection bias
NA medium High low
Recall bias NA high high low
Loss to follow up
NA NA low high
confounding
HIGH medium medium low
Time required
LOW medium medium high
cost LOW medium medium high
ReferenceOxford Textbook of Public health,
Fourth Edition, oxford university press.
Rajivir Bhalwar Text Book of Public Health and Community Medicine.
Study design options in epidemiological research at MGIMS Sevagram 2011.