Biostatistics: Definition, Types, and Methods

• Biostatistics involves the definition, uses, concept of variables and data, and measurements in Public Health.

Definition of Statistics

• Statistics is the science of data, involving collection, analysis/summarization, interpretation, and presentation.
• It uses data to estimate the level of associations and test hypotheses.
• Statistics is concerned with the treatment of numerical data derived from groups of individuals.
• It deals with variation in data through collection, classification, and analysis to obtain reliable results.

Types of Statistics

• Statistics is divided into descriptive and inferential types.
• Descriptive statistics deals with the description of characteristics of a finite population.
• Inferential statistics makes deduction from a sample of a population to the population.

Methods of Descriptive Statistics

• Descriptive statistics summarize patterns in the responses of people in a sample using frequency tables, diagrams (graphs/charts), and summary indices.

Inferential Statistics

• Statistical inference generalizes from a sample to a population with a calculated degree of certainty.
• The two traditional forms of statistical inference are estimation and hypothesis testing.
• Estimation predicts the most likely location of a parameter.
• Hypothesis testing ("significance" testing) provides an answer to a statistical question.

Examples of Inferential Statistics

• Measures of Association (Chi-square)
• T-Test (Paired and Independent)
• One-Way Analysis of Variance
• Multi-way ANOVA
• Regression Analysis
• Correlation Analysis
• Factor analysis

Medical Statistics/Biostatistics

• It provides appropriate methods for collecting, organizing, analyzing, interpreting, and presenting medical and health data.

Benefits of Statistics

• Statistics has a central role in biomedical investigations.
• It offers a better way of organizing empirical evidence on a wider and more formal basis than anecdotes and personal experience.
• It accounts for the intrinsic variation inherent in most biological processes such as blood pressure.

Benefits of Statistics in Health

• Enables measurement of population health.
• Allows for comparisons of the state of health between different periods or locations.
• Statistical presentations/reports provide the basis for health planning.
• Actions on health issues are usually dependent on relevant health information.

Biostatistics in Public Health History

• John Graunt was the founder of vital statistics.
• In 1662, Graunt's publication quantified patterns of disease/mortality in a population.
• Graunt’s publication was based on “Bills of mortality,” a count of people who died weekly since 1592.
• Graunt observed excess of male births, high infant mortality, proportional mortality, seasonal variation, and urban-rural variation in mortality.
• William Farr compiled statistical abstracts in Britain from 1839-1880.
• Farr compiled annual counts of births, marriages, and deaths.
• He used these counts as numerators and census data as population data to obtain crude rates.
• He examined the effects of altitude, location (densely & sparsely populated areas), and marital status.
• John Snow, in 1854, formulated and tested hypotheses concerning the origins of Cholera epidemics in London.
• This happened 20 years before the microscope was discovered.
• Snow is described as the father of field epidemiology.
• Snow used spot maps to show case distribution and determined that contaminated water was the source of the outbreak, removing the handle of pump to control.

Pertinent definitions in statistics

• A POPULATION is a set of units (usually People, Objects, Transactions, or Events) of interest in a study.
• A SAMPLE is a subset of the units of the population.
• A VARIABLE is a characteristic or property of an individual population unit that varies or changes.
• DATA refers to values obtained when measuring one or more variables on the units in a sample.
• Data is usually collected via a sample from the larger population/universe.
• Samples are not of interest on their own, but for what they can tell about the population.
• Statistics allows the use of the sample to make inferences about the population from which it was drawn.
• Different samples from the same population may give different results due to chance or sampling variation.
• Samples have to be drawn in a representative manner.

Types of Variables

• Numerical (Quantitative) variables are measured on a naturally occurring numerical scale.
  • Continuous variables are measured on a continuous scale and include fractions and decimals, e.g., age, weight, and height.
  • Discrete variables can only take limited numbers of values, usually whole numbers, e.g., episodes of diarrhea in a child or number of men in a village.
• Categorical (Qualitative) variables are non-numerical, e.g., place of birth, ethnic group, social class, and gender.
  • Dichotomous (binary) variables have only 2 possible outcomes, e.g., sex (M/F) or survival status (A/D).
  • Ordered categorical variables have a natural ordering but not in a numerical sense, e.g., social class (I, II, III, IV, V).

Elements of Statistical Inference

• Population
• Sample
• One or more variables of interest
• Reliability

Reliability

• Reliability measures how good an inference is.
• Using a sample introduces an element of uncertainty into inferences.
• Determining and reporting the reliability of each inference made is important.
• An inference is incomplete.
• A measure of reliability accompanying an inference separates statistics from fortune-telling.
• The measure states the degree of uncertainty associated with a statistical inference, e.g., a 95% confidence interval.

Nature of Statistical Data

• Primary data is originally collected in the process of any statistical inquiry.
• Secondary data is collected by other individuals/people/organizations.
• The primary source of data is preferred to a secondary source.

Sources of Health Data

• Census
• Vital Registration systems
• Institutions (school health, hospitals, health centers, Veterinary Clinics)
• Notification centers/Epidemiological surveillance (infectious diseases, cancer registries etc).
• Surveys

Census

• National census enumerates the entire populace in a country, usually every ten years.
• Nigeria's last census was in 2006, with a population of 140,003,542.
• The North West zone was the most populous at 35,786,944, followed by the South West at 27,581,992.
• Osun state in Nigeria has a population of 3,423,535, representing 2.45% of Nigeria's population.
• Census enables the calculation of crude or total population rates.

Vital statistics

• Vital events records (births, deaths, marriages & divorces) are obtained by registration.
• These are used for generating birth and mortality rates for populations or subgroups.
• "Vital Registration systems" do not exist or are ineffective in Nigeria.
• This leads to a lack of relevant up-to-date health and demographic information.
• Data on indicators of development e.g. IMR, U-5MR & MMR, are estimated rather than directly measured.

Institution Based Records

• School health records
• Pre-employment screening for occupational health
• Hospital-based records

Surveys

• Surveys can be ad-hoc, but may be routine.
• Popular national surveys include:
  • NDHS (National Demographic and Health Survey)
  • HIV Sero prevalence Sentinel study among pregnant women
  • NARHS (National HIV/AIDS and Reproductive Health Survey).
• Surveys are conducted by individual or group of researchers, organizations, or governments.

National Demographic Health Survey 1999-2013

• The NDHS is designed to provide up-to-date information on respondents' background, fertility levels, nuptiality, sexual activity, fertility preferences.
• It also provides information on: awareness and use of family planning methods, breastfeeding practices, nutritional status of mothers and young children.
• Included as part of the NDHS are: early childhood mortality and maternal mortality maternal and child health with awareness of HIV/AIDS and other STIs.
• Target groups encompass women aged 15-49 and men aged 15-59 selected in randomly selected households across Nigeria.
• Information about children aged 0-5, including weight and height, was also part of the survey.

Measurement

• Variables measurement method is important.
• Measurement is the assignment of numbers to variables.
• Measurement determines the choice of relevant statistical method.

Scales of Measurement

• Nominal (non-numerical/qualitative)
• Ordinal (non-numerical/qualitative)
• Interval (numerical/quantitative)
• Ratio (numerical/quantitative)
• Nominal scale is the lowest level of measurement.
• It classifies the measure into mutually unordered categories without numerical magnitude, e.g., gender (male, female), blood group (A, B, AB, O).
• It classifies persons or objects into two or more categories where members of a category have at least one common characteristic.
• It is not possible to quantify or rank-order the categories.
• Nominal variables use numbers for identification purposes.
• The scale values have no 'numeric' meaning in the way that is usually thought about numbers.

Examples of Nominal Variables

• Ordinal scale has the ability to rank or order phenomenon along with its nominal property.
• It is defined by related categories, e.g., grades of pain (mild, moderate, severe), social class (I, II, III, IV, V).
• Interval scale measurements are expressed in numbers except that the starting point is arbitrary and depending on the units of measurement.
• Meanings can be physically attached to the difference between 2 measurements on this scale, but not to their ratios.
• Ratio scale has nominal, ordinal, and interval scale properties and a true zero point.
• The ratio of any 2 measurements on the scale is physically meaningful, e.g., height or weight.
• Arithmetic operations of addition and multiplication are not possible on the nominal or ordinal scales.
• Only addition (subtraction) is possible on the interval scale while all operations are possible on the ratio scale.
• The exclusive nature of the scales, requires transforming data from one scale to another to remove properties that may invalidate statistical theories.