why to know statistics

Why to know statisticsWhy to know statistics

To understand dataTo understand data

ExampleExample

One of your colleague is an oncology One of your colleague is an oncology surgeonsurgeon

60% of his cases died60% of his cases died Does this mean that he is a looser!!Does this mean that he is a looser!!

We should ask what are the results of We should ask what are the results of his colleagues in his colleagues in similarsimilar patients patients

How many patients he operated How many patients he operated upon e.g 2/3!!!!upon e.g 2/3!!!!

To summarize dataTo summarize data

ExampleExample

Diastolic Blood pressureDiastolic Blood pressure 80,70,65,90,74,80,60,90,60,75,80,9080,70,65,90,74,80,60,90,60,75,80,90

,100,100,100,95,100,100,100,95 AgeAge 24,30,26,40,28,21,26,31,32,36,27,4524,30,26,40,28,21,26,31,32,36,27,45

,62,58,52,50,60,62,58,52,50,60

Vital for researchVital for research

Without the use of statistics it would Without the use of statistics it would be very difficult to make decisions be very difficult to make decisions based on the data collected from a based on the data collected from a research project research project

Statistical steps in researchStatistical steps in research

Collect dataCollect data Organise dataOrganise data Analyse dataAnalyse data Interpret the dataInterpret the data Present the dataPresent the data

How to read the resultsHow to read the results

An understanding of basic statistics An understanding of basic statistics will provide you with the will provide you with the fundamental skills necessary to read fundamental skills necessary to read and evaluate results section in and evaluate results section in published paperspublished papers

Are groups comparableAre groups comparable!!!!!!

the baseline characteristics of the the baseline characteristics of the groups being studied should be groups being studied should be comparablecomparable

If not, they should be adjusted for If not, they should be adjusted for differencesdifferences

statistical testsstatistical tests

Are they frequently used tests!!Are they frequently used tests!! If not, why!!If not, why!!

Are the data analysed according Are the data analysed according to the original protocol?to the original protocol?

Was follow- up complete?Was follow- up complete?Patients lost to follow-upPatients lost to follow-up………………loss of loss of

subjects biassubjects bias

> 10% - 15 % > 10% - 15 % ……………………………………..invalid results..invalid results

P valueP value

A P value of <0.05 means that this A P value of <0.05 means that this result would have arisen by chance result would have arisen by chance on less than one occasion in 20 on less than one occasion in 20

Standardization of measures of outcomeStandardization of measures of outcome::

Odds and odds ratioOdds and odds ratio

The odds is the number of patients who fulfil The odds is the number of patients who fulfil

the criteria for a given endpoint divided by the criteria for a given endpoint divided by

the number of patients who do not.the number of patients who do not.

For exampleFor example

the odds of diarrhoea during treatment with an the odds of diarrhoea during treatment with an

antibiotic in a group of 10 patients may be 4 antibiotic in a group of 10 patients may be 4

to 6 (4 with diarrhoea divided by 6 without, to 6 (4 with diarrhoea divided by 6 without,

0.66);0.66);

in a control group the odds may be 1 to 9 in a control group the odds may be 1 to 9

(0.11). The odds ratio of treatment to control (0.11). The odds ratio of treatment to control

group would be 6 (0.66÷0.11).group would be 6 (0.66÷0.11).

Risk and relative riskRisk and relative risk

The risk is the number of patients who The risk is the number of patients who

fulfil the criteria for a given end point fulfil the criteria for a given end point

divided by the total number of patients. divided by the total number of patients.

For exampleFor example,,the risk of diarrhoea during treatment the risk of diarrhoea during treatment

with an antibiotic in a group of 10 with an antibiotic in a group of 10

patients may be 4 to 10; in the control patients may be 4 to 10; in the control

group the risks may be 1 to 10. The group the risks may be 1 to 10. The

relative risk of treatment to control relative risk of treatment to control

group would be 4 (0.4÷0.1).group would be 4 (0.4÷0.1).

C.IC.I

The confidence interval around a The confidence interval around a result in a clinical trial indicates the result in a clinical trial indicates the limits within which the "real" limits within which the "real" difference between the treatments is difference between the treatments is likely to lie, likely to lie,

hence the strength of the inference hence the strength of the inference that can be drawn from the result that can be drawn from the result

Example:Example:

95 % CI for RRR 25 % :95 % CI for RRR 25 % :

sample size 100sample size 100………………………….= -- 38 % to 59 %.= -- 38 % to 59 %

Sample size 1000Sample size 1000…………………….= 9 % to 41 % .= 9 % to 41 %

The larger the sample size , the narrower and The larger the sample size , the narrower and more precise the CI , and the greater our more precise the CI , and the greater our confidence that the true RRR is closer to what confidence that the true RRR is closer to what we have observedwe have observed. .

OR = 0.34, 95% CI 0.23 - 0.52

• Odds Ratio < 1 decreased risk• Confidence Interval does not cross 1 statistically significant

A statistically significant result may A statistically significant result may not be clinically significant. not be clinically significant.

The results of intervention trials The results of intervention trials should be expressed in terms of the should be expressed in terms of the likely benefit an individual could likely benefit an individual could expect (for example, the absolute expect (for example, the absolute risk reduction) risk reduction)

How large was the treatment How large was the treatment effecteffect? ?

Treatment effectTreatment effect………………………….. Adverse .. Adverse outcomeoutcome

e.g.;e.g.;

Risk of outcome without therapy Risk of outcome without therapy ( baseline risk ) ( baseline risk ) XX ( = 20 % or 0.20 ) ( = 20 % or 0.20 )

Risk of outcome with therapy Risk of outcome with therapy Y Y ( = 15% ( = 15% 0r 0.15) 0r 0.15)

Absolute risk reduction= X -- YAbsolute risk reduction= X -- Y

0.20-0.15= 0.050.20-0.15= 0.05

Relative risk ( RR )= Y / X = 0.15 /0.20 = 0.75Relative risk ( RR )= Y / X = 0.15 /0.20 = 0.75

Relative risk reduction ( RRR ) =Relative risk reduction ( RRR ) =

{ X -- Y/ X } x 100 % = 0.05 / 0.2 x 100%= { X -- Y/ X } x 100 % = 0.05 / 0.2 x 100%= 25 % i.e : 25 % i.e :

therapy reduced the risk of the outcome by 25 therapy reduced the risk of the outcome by 25 % relative to that occurring among the % relative to that occurring among the controlscontrols

the greater the RRR, the more the greater the RRR, the more effective the therapy. effective the therapy.

Example: M.IExample: M.I

patients receiving medical treatment patients receiving medical treatment have a chance of 404/1324=0.305 or have a chance of 404/1324=0.305 or 30.5% of being dead at 10 years. 30.5% of being dead at 10 years.

Let us call this risk Let us call this risk xx. Patients . Patients randomised to coronary artery randomised to coronary artery bypass grafting have a chance of bypass grafting have a chance of 350/1325=0.264 or 26.4% of being 350/1325=0.264 or 26.4% of being dead at 10 years. Let us call this risk dead at 10 years. Let us call this risk yy. .

RRRR

The relative risk of deathThe relative risk of death—— that is, the risk in surgically treated that is, the risk in surgically treated

patients compared with medically patients compared with medically treated controlstreated controls——is is

y/xy/x or 0.264/0.305=0.87 (87%). or 0.264/0.305=0.87 (87%).

RRRRRR

The relative risk reductionThe relative risk reduction——that is, that is, the amount by which the risk of the amount by which the risk of death is reduced by the surgerydeath is reduced by the surgery——is is 100%-87% (1-100%-87% (1-yy//xx)=13%. )=13%.

ARRARR

The absolute risk reduction (or risk The absolute risk reduction (or risk difference)difference)——that is, the absolute that is, the absolute amount by which surgical treatment amount by which surgical treatment reduces the risk of death at 10 yearsreduces the risk of death at 10 years——is 30.5%-26.4%=4.1% (0.041). is 30.5%-26.4%=4.1% (0.041).

NNTNNT

The number needed to treatThe number needed to treat——how how many patients need coronary artery many patients need coronary artery bypass grafting in order to prevent, bypass grafting in order to prevent, on average, one death after 10 yearson average, one death after 10 years——is the reciprocal of the absolute is the reciprocal of the absolute risk reduction: 1/ARR=1/0.041=24. risk reduction: 1/ARR=1/0.041=24.

ConclusionConclusion

to be able to effectively conduct to be able to effectively conduct researchresearch

to be able to read and evaluate to be able to read and evaluate journal articlesjournal articles

to further develop critical thinking to further develop critical thinking and analytic skillsand analytic skills

to know when you need to hire to know when you need to hire outside statistical help outside statistical help