Statistical methods in epidemiology.ppt

Statistical methods in epidemiology.-
Epidemiology
Incidence
Prevalence
Relative risk
Attributable risk

1. By:
Ravinandan AP
Department of Pharmacy Practice
Sree Siddaganga College of
Pharmacy, Tumkur
Statistical
Methods In
Epidemiology

2. Presentation Outlines….
1.Epidemiology
2.Incidence
3.Prevalence
4.Relative risk
5.Attributable risk

3. Epidemiology
Itisthestudyofdiseaseoccurrence
&transmissioninahuman
population.
Itfocusonthedistribution&
determinationofdisease.
Thebranchofmedicalsciencedealing
withthetransmissionandcontrolof
disease
Itmayalsobeconsideredasmethod
ofpublichealth-ascientificapproach
tostudyingdisease&health
problems.

4. Example for Epidemiology

5. Itconsistsofresearchmethods&
specificstrategiesforcounting&
calculatingtheoccurrence&riskof
disease.
Therefore,epidemiologicalstudiesof
druguseemploythesemethods&
statisticalmeasurestostudythe
occurrence&distributionofdruguse
&itsassociatedproblems.

6. Risk Assessment
Riskmeasuresareestimatedthat
describetheamountofrisk
associatedwithaparticular
exposureinasamplepopulation.
Riskestimatescanquantitatively
describetheriskassociatedwith
particularexposure & the
developmentorpreventionof
disease,&theycanquantifythe
associationbetweentheexposureto
aparticulartodrug&anADR.

7. Riskestimatedarepartofourdaily
lives.
Measuresofriskarecommunicated
topatientsvianewspapers,TV,and
theinternetdaily.
Riskmeasuresarecommunicatedto
GP’sviastudiespublishedin
medicaljournals.
Theseriskmeasuresbecome
importantintheclinicaldecision
makingprocessforbothpatients&
thepractitioners.

8. Prevalence
Itisdefinedasthenumberof
existingcasesofdisease(orany
outcome,exADR,druguse)ina
populationataparticularpointin
time.
number of existing cases in a
population
P=--------------------------------------------
total number of people in that
population

9. Example:1
OnJanuary1,2018SSCPhad
3000studentsenrolled.OnJanuary
1,2018,300studentsatSSCP
reporteddengue.
number of existing cases in a population
P=--------------------------------------------
total number of people in that population
P=300/3000=0.10or10%
Therefore,onJan1,2018,10%of
thestudentsatSSCPsufferedfrom
Dengue.
Inotherwords,theprevalenceof
dengueonJan1,2018,was10%

10. Prevalence = no. of cases /
population size
Prevalence can be measured in an
closed cohort or in an open
population.
Prevalence in cross-sectional.
“Old” cases and “new” cases are
counted in the numerator.
Can be measured at a particular
point (point prevalence) or over a
period (period prevalence).
Normally, when we say prevalence
we mean “point prevalence.”

12. Incidence
Itisdefinedasthenumberofnew
casesofdiseasethatdevelopina
populationatriskoveraspecified
timeperiod.
Itisusedtodeterminehowoftenthe
diseaseisoccurring.
Itistypicallydescribedaseither
cumulativeincidence(CI)oran
incidencerate(IR).

13. Cumulative Incidence (CI)
Cumulativeincidenceassumesthat
allofthesubjectswerefollowedfor
theentirestudyperiod.
CIdoesnotreflectstudydropoutsor
lossestofollow-up.
number of new cases disease during
given time
CI=--------------------------------------------
total population at risk

14. UsingSSCPcollege,inthefirstex-1
thetimeframeofJan1,2018,
throughDec31,2018,30students
developeddengue.
CI=30/3000-300=0.011or1.1%per
year.
ForthetimeperiodfromJan1,2018,
throughDec31,2018,1.1%ofstudents
atSSCPdevelopeddengue.Incidenceof
dengueinSSCPwas1.1%

15. Incidencemeasurestheriskof
developingthediseaseoroutcome
ofinterestinapopulationatrisk;
therefore,anysubjectswhocurrently
havethediseaseoroutcomeof
interestattheonsetofstudyperiod
are subtractedfrom the
denominator.

16. Incidence rate (IR)
Alsoreferredtoasincidencedensity
(ID).
Itisamoreaccuratemeansof
measuringdiseaseoccurrence.
IRtakesintoaccounttheactual
observationtimeofeachsubject
duringthestudyperiod,ratherthan
assumingthatallsubjectswere
followedfortheentireperiodoftime.

17. number of new cases
IR=---------------------------------------
total person –time of observation
in population at risk

18. Atoolusedtocalculateriskestimatesis
contingencytable.
Thesetablesareusedtoanalyzediscrete
data.
Incidencesandprevalenceareoften
reportedwithapopulationmultipliersuch
“permpeople”or“permperson-years.”
Toconvertarateorproportionto“perm
people,”simplymultiplyingbym.
Forexample,anincidencerateof
0.00877perperson-year=0.008770×
100,000=877per100,000person-years.

19. Definition of terms
Probability:istheproportions(%)of
timesaneventwouldoccurifan
observationwasrepeatedmany
times.
Risk:istheprobabilityofanevent
amongthoseexperiencingtheevent
dividedbythenumberwhocould
experienceit(atrisk)
Odds:probabilityofaneventdivided
bytheprobabilityoftheeventnot
happening

20. Relative Risk (RR)
RRisthelikelihoodof
developingdiseaseinthe
exposedgrouprelativetothe
unexposedgroup;itisa
measureofassociationbetween
theexposureandthedisease.
RR=CIexposed/
CIunexposed

21. RR can be estimated from a 2 X
2 contingency table.
Interpretation of RRRR Association
betweenexposure
&outcome
1 No association
<1 Negative association
/ Decreased risk
>1 Positive association
/
Increased risk

22. Attributable Risk (AR)
Itisalsocalledriskdifference(RD).
Isanothermeasureofriskusedin
studies.
Itprovidesinformationonthe
absoluteeffectoftheexposure.
Itdescribestheexcessriskof
diseaseintheexposedcompared
withthosewhowereunexposed.

23. ABSOLUTE RISK (AR)
Itistheprobabilitythatan
individualwillexperiencethe
specifiedoutcomeduringa
specifiedperiod.
Itliesintherange0to1,or0%
to100%.
Incontrasttocommonusage,
thewordriskmayreferto
adverseevents(suchas
myocardialinfarction),or
desirableevents(suchascure)

24. Absolute risk reduction (ARR)
Itistheabsolutedifference
inriskbetween the
experimentalandcontrol
groupsinatrial.
Itisusedwhentheriskin
thecontrolgroupexceeds
therisk

25. AR=CIexposed–
CIunexposed
Itallowstheinvestigatorto
determinehowmorbidity&
mortalityareaffectedby
removingexposure.

26. Attributable Risk percent
Theattributableriskcanthenbe
convertedtothe(AR%).
Itmakeseasiertointerpret.
AR%=
CIexposed-CIunexposedX100
CIexposed

27. THANK YOU