Biostatistics

Major Lutphun Nahar

Anaesthesiology
• Statistics
Statistics may be defined as the science and art of
collection,organization,presentation,analysis and
interpretation of numerical data
Components
• Descriptive statistics
Study design
Data collection
Data compilation
Data presentation
Analysis
• Inferential statistics
Test of hypothesis
Inference
Biostatistics
It is a special branch of statistics which deals with different
types of data pertaining to biological science

Medical statistics
Statistics applied to the medical science

Components
1)Vital statistics
2)Health statistics
 Variables
• Variable is a characteristic or attribute of a person,object or
phenomenon that can take different values

Types(Based on relationship with each other)

1)Independent
2)Dependant
3)Intervening
4)Confounding
Types(Based on characteristics)
• Quantitative data
a)Continuous
b)Discrete

• Qualitative data
a)Nominal
b)Ordinal
 Other terms associated with data

• Primary data
• Secondary data
• Derived data
• Binary data
• Univarient data
• Bivarient data
• Multivarient data
• Outliers
 Methods of data collection
• Document review
• Observation
• Experiment
• Questionnaire
• Interviewing
 Methods of Data presentation
1)Textual method
2)Tabular method
Frequency distribution table
Contingency table
Cross table
3)Graphical method
Graphs(For quantitative data)
Presentation of continuous,numerical or measured data are through graphs
Examples
• Histogram
• Frequency polygon
• Frequency curve
• Dot or scattered diagram
• Line chart
• Diagram (For qualitative data)
Presentation of qualitative or discrete or categorical or counted data are through
diagram.

Examples
Bar diagram
Simple
Compound
Multiple
Pie diagram
Pictodiagram
Map diagram
Population
• Population is the entire group of people or person or
study elements or things or mesurements

It depends
• Area of research or study design
• Area of interest
Types
• Homogenous
• Heterogenous
• Static
• Dynaemic
Another types
• Finite
• Infinite
 SAMPLE AND SAMPLING
• Sample
Sample is the subgroup of population selected for an
investigation which is usually represent the whole group

• Sample size
The number of sampling units selected from the study
of population
Study or survey population
• The entire population of sampling units from which the
sample is drawn

Target population
• The population to which findings of the survey are to be
extrapolated
 SAMPLING
Sampling
• Sampling is the technique of collecting information
about the whole group by examining the sample
Sampling unit
• Each member of the sample is the sample unit or unit
of analysis
Sampling frame
• All listed units or items constituting the study
population
 TYPES OF SAMPLING
1)Random or non purposive or unbiased or
probability or representative sampling
• Simple random sampling
• Systematic random sampling
• Stratified random sampling
• Multistaged sampling
• Multiphased sampling
• Cluster sampling
2)Non random or purposive or biased or non
probability or purposive sampling
• Convenient or accidental sampling
• Purposive or judgement sampling
• Quota sampling
• Snowball sampling
• Consecutive
 Measures of location
1)Measures of central tendency
2)Percentiles
3)Deciles
4)Quartiles
 Criteria to measure central tendency
• 1)All measures of central tendency are used for numerical data
• 2)Mean is appropriate for numerical and normally distributed data
• 3)Median is used for ordinal and numerical skewed data
• 4)Mode is used for data with bimodal distribution
 CENTRAL TENDENCY
• Mean
Arithmetic mean
Geometric mean
Harmonic mean
Quadratic mean
• Median
• Mode
 MEAN
• Arithmetic mean
It is the real average of the number of the item
• Geometric mean
It is the n th root of the product of values in the series
• Harmonic mean
It is the reciprocal of the arithmetic mean of values
• Quadratic mean
It is used to determine the standard deviation utilizing it
 MEDIAN
• It is the central or middle value of series
• If n is odd,the value in the position is the median
• If n is even,the arithmetic mean of two values in the
and +1 th positions are the median
 MODE
• Mode is the value that most frequently occurs in the series

Series may be
• Unimodal
• Bimodal
• Trimodal
• Without mode
Percentile
• Value of a data set arranged in ascending order and divided into 100
equal parts by 99 imaginary lines called percentiles(P1-P99)
• K th percentile=k(n+1)/100
• 5th percentile means,5% data are below and 95% data are above that
value
Decile
• Values of a data set arranged in ascending order and divided into 10
equal parts by 9 imaginary lines called Decile(D1-D9)
• Kth decile=k(n+1)/10
• 5 th decile means 50% data are below and 50% data are above the value
Quartiles
• Values of a data set arranged in ascending order and divided into 4
equal parts by 3 imaginary lines is called quartiles
• 1th quartile means 25% data are below and 75% data are above that
value
• IQR=Q3-Q1.It contains central 50% values
 MEASURES OF DISPERSION
• Measures of dispersion is the measures of extent of
deviation of individual value from the central value
Types
a)Absolute measures
• Range
• Mean deviation
• Standard deviation
• Standard error
• Variance
Relative measures
• Co efficient of variance
• Co efficient of standard deviation
• Range
Range is the absolute difference between the highest
value and the lowest value in a series of observations.
Range =Mean±SD
• Mean deviation
Mean deviation is the average or mean of the absolute
deviations of the individual values from the central values
of the series
• Variance
It is the mean squared deviation(SD²)
• Standard deviation(SD)
It is a measure of dispersion from the mean
SD=
• Standard error(SE)
It is a measure of inaccuracy of dispersion from the mean
SE=
• Co efficient of variation
Standard deviation expressed as percentage of mean is called co
efficient of variation
• CV=
• Co efficient of mean deviation=
 PROBABILITY DISTRIBUTION
• Distribution
Distribution is the shape and pattern of location or data in a set of
value or data
• Frequency
Frequency is the number of individual in each group or the count of
individuals having a particular quality called frequency
• Frequency distribution
It is the arrangement of data into class intervals showing the
frequency of each class
Requirements of construction of frequency distribution
a) Range of lowest value and highest value
b) Class value
c) Whole set of data
d) Tally mark
• Types of distribution or probability distribution
1)Normal or Gaussian distribution
2)Asymmetric or Skewed distribution
3)Log normal distribution
4)Binomial distribution
5)Poisson distribution
• Characteristics of a normal curve
• It is bell shaped
• Mean,median,mode lie centrally
• It is symmetrical
• It has two inflections
• Area of curve will be 1
• Value of mean,median,mode=0
• Specified by its mean and
standard deviation
• 68%
• =95%
• =99%
• In a Skewed distribution
• Mean is always on the long tail
• Mode is always on the short tail
• Median is somewhere between two.
 PRBABILITY OR CHANCES
• Probability is a measure of chances of occurance of an event by chance
• P value:Quantitative estimate of probability
• Probability =
• Probability scale ranges from 0 to 1
• P=0 means no chance of an event to occur
• P=1 means 100% chance of an events to occur
• P↑ → ↓sampling error by chance→real→result significant
• P ↓→↑sampling error by chance→not real→result is not significant
• P value 0.05 means
Level of significance 5%
Probability by chance is 5 out of 100 events
 HYPOTHESIS
• A statement of belief about the population parameter
• Statistical hypothesis
Research hypothesis can be done in two different ways
1)Null hypothesis
2)Alternative hypothesis
 SAMPLING ERROR
• Sampling error may be due to systematic or random errors
Systematic errors
1)Selection bias
2)Measurement bias
3)Confounding bias
Random errors
1)Type 1(α)error:False positive
2)Type 2(β)error:True negative
 TEST OF SIGNIFICANCE
Test of significance is performed to accept null hypothesis or to reject null hypothesis
and accept alternative hypothesis

Test of significance
1.According to the purpose of study
a)Test of association between two or more variables eg X² test
b)Test of difference between two or more variables eg t test
2.According to the type of data and their distributional assumption
a)Parametric test
b)Non parametric test
3.According to the nature of research hypothesis
1)One tail test
2)Two tail test
Parametric test(for quantitative data)
• Students t test
• F test or Fishers test or analysis of variance test or ANOVA
• R test or pearsons correlation co efficient test
• Z test
• Regression test
4)Non parametric test(for qualitative data)
• Chi square test
• Fisher exact probability test
• Mann whitney U test
• Wilcoxon rank sum test
• Wilcoxon signed rank test
• Kruskal Wallis test
• Fried man test
• Spearman rank correlation test
 STUDENTS t TEST
• T test is applied to find out the significance of difference
between the two mean
• Criteria of applying t test
1)Parametric test
2)Used in quantitative test
3)Random sample
4)Sample size less than 30
5)Variables must be normally distributed
6)Applied to find out the difference between two means of
two groups
 t test are of two types
• Paired t test
• Unpaired t test
 Z test of proportion test
• Pre requisites of applying Z test
1)Data must be quantitative or numerical data
2)Samples must be randomly selected
3)Sample size must be large(n>30)
4)Values must be normally distributed
5)Unpaired design
6)To assess the difference between two means of two
groups
 Chi square test(X²)
Criteria
• Non parametric test
• First developed by karl pearson
• Unpaired design
• Used for qualitative or discrete data
• No mean,median,mode
• Applied to measure the probability of association
between two qualitative or discrete attributes
 ANOVA
Pre requisites for ANOVA
• Quantitative data
• Data should be normally distributed
• Sample selected from population randomly
• Unpaired design
• Comparison among the means of more than 2 groups
• Variance of each group must be same even when the
means are different
 MEASURES OF MORBIDITY
Incidence :It can be defined as the number of new cases occurring in
a defined population during a specified period of time
• Incidence
rate=
Uses
• To control disease
• For research into aetiology,pathogenesis,distribution of diseases and
efficacy of preventive and therapeutic measures
Incidence rate refers
• Only to new cases
• During a given period(usually one yr)
• In a specified population or population at risk
• Refers to new spell or episodes of disease
 PREVALENCE
It is defined as all current cases(old and new)existing at a given point of
time or over a period of time in a given population
Types
• Point prevalence
• Period prevalence
• Point prevalence:it is defined as the all current cases(old and new) of
a disease at one point in time in relation to a defined population
• Point prevalence=
• Period prevalence
• It measures the frequency of all current cases exixting during a
defined period of time expressed in relation to a defined population
• Period prevalence=
Uses
1)It helps to estimate the magnitude of disease problem
2)It is useful for administrative and planning purpose
• Relationship between prevalence and incidence
• P=ID=Incidence Mean duration
 Measures of Association
1)Pearson correlation co efficient
2)Spearman rank order correlation co efficient
3)Chi square test
4)Relative risk and Odds ratio
 Measures of Association
3)Risk ratio or relative risk(RR)
• It is calculated by dividing the number of events by the number of
people at risk
• If risk ratio=1,it means there is no difference in risk between the
exposed and non exposed groups
• If risk ratio is >1,the rate of that event is increased compared to
controls.
• If <1,the rate of that event is reduced
4)Odds ratio
• It is calculated by dividing the number of times an event happens by
the number of times it does not happen
• If the odds ratio of an event is >1,the rate of that events is increasedin
patient who has been exposed to risk factors
• If <1,the rate of that event is redued
 STUDY DESIGN
• Study design
Scientific method of research
Types
1)Observational study(non interventional)
a)Descriptive study(without comparison group)
Case report
Case series
Pilot study
Community survey
b)Analytical study(with comparison group)
Case control study
Cohort study
2)Experimental study(interventional)
Randomized control trial
Field trial
Community interventional trial
Clinical trial
 SCREENING TEST
It has 5 components
• Sensitivity
• Specificity
• Predictive value
• Likelihood ratio
• Reproducibility
• Sensitivity
The ability of the test to identify correctly those who have the disease

=
• A high sensitivity implies few false negatives which is important for
very rare or lethal disease
• Specificity
The ability of the test to identify correctly those who don’t have the
disease

=
• A high specificity implies few false positives which is important for
common disease
 Predictive value
Its of 2 types
a)Positive predictive value
• The proportion of patients who test positive who actually
have the disease
• PPV=
b)Negative predictive value
• The proportion of the patients who test negative who are
actually free of disease
• NPV=