research methodology and biostatistics

[email protected] 03336422555

Dr. Raham BachaUltrasound specialistMD, PhD ScholarMSc Ultrasound (Gold Medalist) Lecturer Afro-Asian Institute Lahore


Research Methodology &

Biostatistics.Respected readers it is mandatory for every researcher to know somewhat about research and biostatistics. But unfortunately we have been teach by biostatisticians in

our universities and medical schools. They want to make us biostatistician and teach us thoroughly. I

therefore felt a need to prepare a lecture in condense form which make enable us to know those terms which

are frequently used in research results.The purpose of the lecture is to read and understand some research and even analyze your research data and make inferences with the help of SPSS [email protected]

03336422555

What is Research?• Research is composed of two worlds.“Re” mean again

And Search mean to find new facts

It means to search again new facts or To modify older ones, in any branch of knowledge.



Webster international Dictionary proposed a very inclusive definition.“A careful inquiry or examination in seeking facts or principles, diligent investigation in order to ascertain something.”Crawford Says: “Research is simply a systematic and refined technique of thinking, employing specialized tools, instruments & procedures in order to obtain a more adequate solution of a problem than would be possible under ordinary means.”


It starts with a problem, collects data (facts and figures) analyses these critically based decisions of actual evidence. It evolves original work instead of mere exercise of personal opinion (theory). It is quantitative and measurements is therefore a central feature of it.


• Don’t appreciate the research of other and yourself.

• Develop critical thinking.• Specialized tools and instruments

(Material i.e. data sheets)In fact research is treated as advancement in knowledge acquired through scientific methods


Research provides:1. Analytic framework for subject

matter investigation2.Establishes relationship between

different variables e.g. cause and effect relationship, Co-relationship between variables.


Important points• Research should not be treated as a

place of compilation work. It is not simply the compilation of material taken from different books and sources. It is always expected to something original that advances some knowledge.

• Research should be very clear in Researcher mind that he is going to do.

• The domain of Research problem must be clear.


STATISTICSnumerical facts in any field of study. It concerns with collection, organization, summarization, analysis and drawing inferences from a data set. the science comprising methods, which are used in collection, presentation, analysis, and interpretation of numerical data (information).


Bio-statistics

Bio-statistics is the branch of statistics that concerns

with the applications of statistical methods to

medical and biological data.


In medical field, statistical methods enable us to

check the effectiveness of different testing

modalities in medicines. Recently, it has been

found that applications of statistical methods in

medical data are very effective. Testing of

hypothesis, analysis of variance, chi-square, non-

parametric methods, regression and correlation,

logistic regression etc. are frequently used in the

analysis of health and medical data.


Being medical students often we need to

study research paper or we need to present

our research work in statistical terms. It is

therefor needed to know some of the

fundamental terms frequently used in

research papers.

Significance of Bio-Statistics in Medical.


Central Tendency

Central tendency is a characteristic of a data set

which relates to its average value.

Most commonly used measures are,

Mean

Median and mode

Quartiles

Deciles

Percentiles


MEAN

The most commonly used measure of central tendency. It is

the sum of all the observation divided by the number of

observation is mean.

nsobservatioofnumbertotalsetdataofnsobservatiotheallofsum

Mean


Median

The median of a set of data arranged in order of

magnitude is the middle most value.

Mode

Mode is the most frequently occurring number in the data

set.


Rates

The number of events in a population in specific time.

Suppose, in a specified population “n” events occur

during a fixed period of time n(A). If n(A) of these events

possess some characteristic, say A, then rate of the event

having the characteristic A is given by


Prevalence Rate (P.R.)

Prevalence rate of an attribute or event in any group, is the

proportion of individuals in the groups having that

attribute at one point in time. This is also known a

prevalence ratio.

Kgroupinsindividualofnumbertotal

timegivenaatdiseaseorattributewithsindividualofnumberRP .


Incidence Rate (I.R.)

The risk of developing the disease over a period of time is

called incidence rate and is calculated as:

K

groupinsindividualofnumbertotaltimegivenaatofdiseasecasesnewofnumberRI .


Measures of Dispersion

The measure of spread of data/observations from its

averages. Statistical measures of dispersion are used to

measure the extent to which individual observations

disperse or cluster around the average. two commonly

used measures are;

range and

standard deviation.


Range

Range is the difference between maximum and minimum

values of data set

Standard Deviation

The most widely used and stable measure of dispersion is

the standard deviation. This is a square root of variance.


The most pleasurable moment of my life, when I was awarded with Gold medal


Share feelings with media.

Dr. Raham Bacha Ultrasound Specilist

The Statistical Profile

25

The Statistical Profile

These include;

1. sensitivity,

2. specificity,

3. positive predictive value, and

4. negative predictive value.

5. accuracy,

6. reliability

Dr. Raham Bacha Ultrasound Specilist 26

These terms are defined as follows:

1. Sensitivity (True Positive Ratio or Rate):

The ability of a test to find disease when disease is present.

2. Specificity (True Negative Ratio or Rate):

The ability of a test to indicate no disease when there is no disease

present (i.e. to exclude disease when a patient is normal.

3. Positive Predictive Value (PPV):

PPV is a measure of how likely it is that disease is actually present

when the test result is positive. In other words, it is the proportion

of people with a positive test who actually have the disease.


These terms are defined as follows:4. Negative Predictive Value (NPV):

NPV is a measure of how likely it is that disease is actually

absent when the test result is negative. In other words, it is the

proportion of people with a negative test who are free of disease.

5. Accuracy:

The ability of a test to give the correct answer.

6. Reliability:

Consistency in obtaining similar results in similar

circumstances.


The Chi Square Test

This test brings together the statistical terms discussed

above. In the Chi Square, data from any test is compared

to data obtained using the gold standard. In graphical

form, the Chi square looks like this:


The Chi Square Test

Gold Standard result

Our testing modality result

present +

absent +

Positive +

A B

Negative -

C DDr. Raham Bacha Ultrasound Specilist 30

The Chi Square Test

Formulas for statistical terms:

1. Accuracy = (A+D)/ A+B+C+D

2. Sensitivity = A/ (A+C)

3. Specificity = D/ (B+D)

4. Positive Predictive Value = A / (A+B)

5. Negative predictive value = D/ (C+D)


Chi Square Example

Let’s take a few moments to work through an example using

the Chi square.

Suppose 100 non-invasive vascular tests were performed in

the vascular lab on patients who also had an arteriogram

(considered the gold standard in this case). The non-invasive

test yielded 55 positive results and 45 negative results. Of the

55 positive tests the arteriogram found 5 to be negative and of

the 45 negative tests, the arteriogram found 10 to be positive.


Chi Square Example

Let’s create a Chi square to illustrate the data above.

NON- INVASIVE VASCULAR TEST

Gold Standard present

+absent + Total

Positive + 50 5 55

Negative - 10 35 45

Total 60 40 100


Chi Square Example

The chart above illustrates the results of the 100 tests performed on our

patients.

Note that the 55 total positive exams by the non-invasive test are broken

down to indicate the ones with which the gold standard agreed and

disagreed.

Also note that the total of all four squares must equal the total number

of tests performed (i.e. 100 in our example.

Now let’s perform the calculations required for each of the six statistical

formulas discussed previously.

(Multiply the results by 100 in order to arrive at a percent.)


Chi Square Example

Accuracy = (A+D)/A+B+C+D

= (50+35)/50+5+10+35

= 0.85 x100

= 85%


Chi Square Example

Sensitivity = A / (A+C)

= 50/(50+10)

=50/60

= 0.83

= 0.83x100

= 83%


Chi Square Example

Specificity = D/ (B+D)

= 35/ (5+ 35)

= 35 / 40

=0.88

= 88%


Chi Square Example

Positive predictive value = A/ (A+B)

=50/ (50+5)

= 50/55

=0.91

= 0.91 X 100

= 91%


Chi Square Example

Negative Predictive Value = D/ (C+D)

= 35/ (10+35)

= 35/45

= 0.78

= 78%


Test Goals

The purpose of any of the non-invasive tests discussed in this

course is to identify disease when it is present and exclude it when

it is absent. Whether we are in Search of atherosclerosis or venous

thrombosis, we strive to produce the most accurate result within

our capabilities. In order to confirm or dispute the results of non-

invasive tests, our results are often compared with an established

diagnostic test known as the gold standard. In cerebrovascular

diagnosis, the current gold standard is the angiogram. In the study

of vascular disease, the venogram is still the gold standard.


Test Goals

The predictive value of a positive or negative test result is

perhaps the most useful expression of accuracy as far as

the referring physician is concerned. It is essential that we

remember that the skill of the vascular technologist may be

crucial in attaining accurate test results. I am sure that is

why we all study so hard to become excellent in the

performance of our examinations.



Who knew that statistics could be this much fun!

Dr. Raham Bacha 43

research methodology and biostatistics

Health & Medicine