Biostatistics Review

Introduction to Research Methodology & Research Ethics

• Research defined: a systematic approach to answer questions, solve issues, and create knowledge through information collection, organization, and analysis to aid decision-making.
• Research steps:
  • Asking a question
  • Understanding the problem's nature
  • Reviewing literature
  • Collecting data through observation, measurement, and recording
  • Analyzing collected data to find significance and generalize
  • Report writing is vital to convey information to the readers.
• Key difference: a research problem is an issue or gap in knowledge, while a research question aims to study and explore the research topic.
• Research Types: dependent on data type.
  • Qualitative research: uses words
  • Quantitative research: uses numbers
  • Aims of epidemiological quantitative research:
    • Measuring disease frequency and determinants
    • Evaluating screening tests
    • Measuring morbidity and mortality
    • Determining prognostic factors
    • Testing new treatments or vaccines
  • Aims of qualitative behavioral/socioeconomic/cultural research: understand behavior for a phenomena.
  • Action research/health system research: focus on priorities and to evaluate policies and programs that provide greatest benefits with resources.
• Data sources of research:
  • Primary data: collect original data
  • Secondary data: using previously collected data
• Objective of research:
  • Observational: measures something as it is
  • Experimental: performs an experiment
• Application research:
  • Basic research: related to new discoveries and molecular basic science such as genome studies
  • Applied research: improves health or quality of life

Research Ethics

• Defined: moral principles for research conduct, including protection of human and animal subjects.
• Key components that make research ethical:
  • Social value to improve health
  • Scientific validity
  • Fair subject selection
  • Favorable risk-benefit ratio
  • Informed consent
  • Qualified investigators
• Informed Consent: Patient agrees with the medical activities.
  • For children or incapacitated persons, consent is given by parent/guardian
• Content of Informed consent:
  • Name of Institution
  • Title of Project
  • Principal Investigator
  • Other Investigators
  • Participant's Names
  • Purpose of Research
  • Procedures (main & alternatives)
  • Time and Duration of Procedures and Study
  • Discomforts and Risks
  • Potential Benefits
  • Statement of Confidentiality
  • Costs for Participation
  • Compensation for Participation
  • Research Funding
  • Voluntary Participation
  • Contact Information for Questions or Concerns
• Principles of Biomedical Ethics:
  • Respect for autonomy: Respect individual decisions about health and future
  • Beneficence: Actions to benefit the patient by care in different medical fields that prevent health problems in the community.
  • Non-maleficence: Designed to not harm the patient and for others
  • Justice: Fair and just to the community in terms of action
  • Procedural Justice: Distribution by established rules (first come, first served)
  • Distributive Justice: Distributes limited resources
  • Compensatory Justice: Is punishment to combat discrimination
  • Fidelity: Duty to observe the action made by the profession.
  • Truthfulness: Tell the truth to decide w/o paternalism
  • Confidentiality: Patient record access is legal but limited; use fake names in conferences.

Sampling

• Definition: a method to gain information about only a part of a population; a good sample represents the population well.
• Sampling Advantages:
  • Lower cost and saves time
  • Providing more intensive and accurate investigations
  • Elimination of bias
• Target population: the group of individuals from which the sample might be drawn.
• A sample is a subset of a population used to gain information about the entire population.
• Sampling Precautions:
  • Well chosen
  • Sufficiently large to limit variation
  • Adequate coverage to avoid bias
• Methods of sampling:
  • Non-probability samples:
    • Convenience sampling: collects samples conveniently by location or Internet service in order to get the feel for a situation.
    • Example: first 10 patients in the clinic, students in the library
    • Purposive sampling: for a specific purpose.
    • Ex: to know more about disabled students at your university, purposefully select students with different support needs
    • Quota samples: a non-probability sample used in sampling public opinions
    • Snow-ball sampling: also known as chain/sequential samples, adopted when identifying members in a sample becomes difficult by identifying potential people through friends or relatives
  • Probability sample/Random sample: all units have an equal chance of selection for generalization.
  • Types of Random Samples:
    • Simple Random Sample:
      • Population uniform and homogenous
      • Equal and independent chance of selection
      • List all (sampling frame) and then select
      • Use a random # table to select
      • Simple to conduct
      • Require sampling frame, and be unsuitable for heterogeneous populations.
    • Systematic Sampling:
      • Systematic selection from sampling frame; start by picking some random point in the list and then select every nth element until the desired number is secured. -Easy to conduct
      • Requires sampling frame, and only the 1st subject is selected randomly.
    • Stratified random sampling:
      • Used when population is heterogeneous
      • Divide heterogeneous population into homogeneous strata
      • Proportional allocation is respected
      • Most suitable procedure to ensure representativeness
      • More expensive and time consuming.
    • Cluster Sample:
      • Sample unit is a group not an individual (family, school class, ....)
      • Selected randomly from all of the same type
      • All members of the selected group will be included in the study
      • Field is concentrated, so is simple & cheaper
      • Can cause errors if the studied disease or variable is clustered in the population.
    • Multistage Sample:
      • Sampling in 2 or more stages
      • 1st stage: a random sample of schools.
      • 2nd stage: random sample of children from each selected school -Can have select town, districts, streets & finally houses
      • Advantages: -Concentrate resources/Requires no sampling for whole population -Only a list of 1st-stage units is required; frame is needed for selected last-stage units.

Descriptive Studies

• Health: a state of holistic wellbeing and the capacity to live a productive life.
• Epidemiology: the study of the distribution and determinants of health-related states or events in specified populations to control health issues.
• Uses of Epidemiology:
  • Identify disease patterns
  • Identify disease risk determinants
  • Determine relative importance of causes of illness, disability, and death
  • Find segments of the population with the greatest risk from specific causes of ill health
  • Evaluate the effectiveness of health programs and services in improving overall population health

Epidemiologic Research Methodology

• Epidemiological studies are classified as either observational or experimental.
• Observational: studies where the investigator measures but does not intervene, and includes descriptive or analytical studies.
  • Descriptive studies: describing disease incidents in a population
  • Analytical studies: goes further by analyzing relationships between health status and other variables.
• Experimental: studies that involve a change in a disease determinant with experiment design in other sciences
• Studies according to Evidence Based Medicine range from weak to strong:
  • Descriptive Studies: Case report-case series/Ecological studies - Cross sectional studies
  • Analytical Studies: Case control studies- Cohort studies
  • Clinical Trials
  • Systematic Reviews
  • Met analysis Studies
• Descriptive Studies:
  • First phase in the epidemiological investigation, which describes disease distribution in the population by time, place, and person.
• Types:
  • Case Reports:
    • Presentation of newly or uniquely found single case, report new therapeutic effect
  • Case Series: Characterizes a number of patients with a disease to describe unusual variations and creates hypothesis
  • Ecological studies: generates hypotheses by looking at groups of people instead of individuals, comparing populations in different places
  • Cross-Sectional Studies: is a study of exposures & disease status at a single point in time so you cannot determine exposure preceded disease
    • Measures the prevalence of health in a population within one point of time
    • Advantages: used to study conditions; conditions that are frequent with long duration- generates hypotheses, assesses estimates, and takes the first step to be a causal association
    • Disadvantages: is not useful for studying acute diseases with seasonal issues- Impractical for rare diseases that do not allow one to determine quickly the disease.
    • Examples: investigate the presence of diabetes depending on obesity, or musculoskeletal problems between radiologists

Prevalence Rate

• Prevalence is the rate the study cases happen in time
• Survey study: a field study to investigate community issues, with classifications by target population or purpose
• Objectives:
  • Case finding by screening: for pre-clinical phases that include symptoms and signs, and have a advanced pathological damage
  • Determine problem magnitude and features according to host, agent and environment to help plan and assess prevention and control programs.
• Steps:
  • Preliminary by clarifying purpose and ethical consideration to formulate the topic
  • Planning by defining objectives, type, target and timetable
  • Preparation of requirements by team, and then approach community
  • Collection, tabulate and writing the data to include results

Analytical Studies:

• Case Control study: investigating exposure and outcome of disease.
  • Used to test an etiologic hypothesis - Steps to conduct a case control study: - Identify the study group: define criteria for diagnosis and the criteria for inclusion and exclusion of cases. - Identify controls: controls must be free from the studied disease considering proper matching with cases (confounding factors) equal to or more by 2-4 times as much as the cases - Sources of controls: - General population - Hospital controls
    • Calculated Measures: -P1 proportion of the exposed among cases P₁= a/a + c -P2 proportion of the exposed among controls P2= b/b + d -Relative: (OR) can identify risk factor more consuming than a normal control -Advantages: -Easy/Quick/Allows study of risk/used on low latency disease Can estimate risk of odd -Disadvantages: -Can not calculate rate-biased can come from recall- estimate times
• cohort study: studies that exposure of group people is free of what effect the cause of it -measure of best observational study - ensure exposure with the risk to developing the disease. - Incidence is measured and provides information - steps to conduct: -exclude cases of disease we want to study - the free cohort, divided into 2 groups is exposed to 1 factor, the other has no exposed -both groups are followed up over sufficient amount of time

Medical Statistics I

• Medical Statistics: definition encompasses the methods of collecting and summarizing (Descriptive Statistics), analyzing and taking statistics, and field of medicine-
• Importance :
• facts/ figures/comparison/ administration/plan -policies to health standards
• -Populations: people or things that need to be analyzed- ex. Data
• • raw-the is data
• -parameters- value summarizing population, while statistic is the data of sampling -Types:
  • Definition (quant is measured- person/ person while var- value can vary
  • Categories:
    • Quantitative: Numerically Continuous and have discreet
• Continuous : value between 2 limits
• Discreet: value between 2 limits
• Categories: Qualitative: fall into a number of separate categories of classes
• Nominal- no order, can order like married/status
• Ordinal: order in this way but you it is not
• Steps: collection and impact which also includes
• population/survey, and vital statistics and providers
• diseases to show inc/prevalence Summarizing of data of what you want in your
• Can be table form, diagrams, or statistics.
• Tables- Has titles, number, and graphs with all units to blood pressure

Graphs pie Charts- simple/understood and has each portion

• But will. have each set of the values for the variable, to look at others Bar has vertical and horizontal which also bar to show the frequency Component- by stacked of how other bars are being stacked like a variable
• Histrogram and the frequency from a continious one bar to another with out the gaps in between the bars.

Medical statistics II

• III Data and analyses
  • Data are described as frequency and proportion.
• Tendency: -Central : single value - Arithmetic/or (mean or -is usually reported for continutious data
• Median is a data set, divided by half is that. For - odd- (cent/2) - used for data Data are spread with following values, Range, which is the amount of spread out Mean- the number of spread it out Percintile-value has % to all values

Factors To show that is used in values, it is very small. With what averages

Definiton: has number, values, all that

Questions covering appropriate statistical tests for different data types and distributions, correlation measures, and the role of data analysis in research. Includes qualitative research and problem vs question definitions.