Epidemiologic Principles and Methods (Ch. 5-6)

Epidemiology

the study of the distribution and determinants of disease frequency in human populations

Disease or Health Outcome

must have a clear definition of what is to be counted or studied

Disease Frequency

incidence rate, prevalence rate, mortality rate

Incidence Rate

rate of new cases appearing in a defined population in a given period of time e.g. new cases/100,000 population

Prevalence Rate

frequency in population at given time e.g. total cases/100,000

Epidemic Curve

graphs the when of an outbreak, number of cases day by day, calculate time of incubation (infectious window), comparison of incidence in population at convention and population, number of cases subsides as transmission shows

Experimental Study Designs

evaluating the association of risk factor and disease, cohort study, case control, intervention study

Cohort Study

exposure is greater than outcome (rate), relative risk: exposed/unexposed, assemble healthy participants, longitudinal and prospective

Case Control

compare groups with and without disease, probability of previous exposure, odds ratio

Intervention Study

test effect of intervention: experiment, control vs treated

Example: Lung Cancer Risk

• increasing mortality rate of lung cancer since early 1900s

• possible factors: increased tobacco smoking, increased air pollution, studies began in 1940s

Trends in Smoking

peak in 1960s U.S.

Current Trends in Smoking

6% high school students smoke cigarettes, 27% report using vape devices mainly Juul

Smoking Policy Changes to Discourage Use

tobacco 21 legislation (Federal law), federal minimum age 21 for purchase, ban on flavored cigarettes/e-cig, taxation

Reye’s Syndrome

1980s, deadly disease of children after viral infection such as chicken pox

Case-Control Study: Reye’s Syndrome

begin with outcome, targeted and retrospective study, participants already have condition, control is closely paired but without condition, shorter duration

Measures of Association

link between exposure and disease, strength of association, relative risk (cohort study), odds ratio (Case control)

Cohort Study General

comprehensive data, long follow up and time commitment, multiple determinants, directly calculates rate so larger numbers needed, change in circumstances during time of study

Case Control General

relatively quick and inexpensive, small sample size using existing records, rare disease with long latency, relies on recall, difficult to control confounding factors, cannot directly calculate rates

Experimental Study

intervention study, clinical trials, treated group vs placebo group, randomization, matched controls, double blinded, therapeutic clinical trial (all patients have disease)

Human Experimentation Ethical Standards

e.g. Tuskegee Syphilis Study, the conduct of research involving human subjects has to go through institutional review boards, informed consent, full and timely reporting of adverse events, same standards should apply to research conducted outside U.S.

Francis Trial 1954: Intervention Study

400,000 schoolchildren, Salk vaccine vs placebo, measured incidence of polio

Physician’s Health Study: Intervention Study

22,000 American physicians, Aspirin: betacarotene, measured incidence of heart disease and cancer

Kingston-Newburgh Study: 1945

fluoridation of water and tooth decay, Kingston water had fluoride Newburgh didn’t, dental exams on school children to measure the prevalence of tooth decay in children after 10 years, ½ level of mission or filled teeth in Kingston children, no evidence of adverse effects of fluoridation

From the Lab to the Pharmacy

drug discovery, translational research, pre-clinical testing, clinical trials (phase 1,2,3), FDA approval, post-marketing surveillance

Drug Discovery

basic research

Translational Research

transition from basic to applied pharmaceutical research

Flieger, K: Testing Drugs in People 1995

tests usually carried out by manufacturers: phase 1, 2, 3, new drug application reviewed by FDA to get approved for marketing, post marketing surveillance recalls if adverse events accumulate

Limitations of Clinical Trials

trials testing efficacy may be too small, monitoring might not be sensitive enough, trial duration too short, enrollment criteria excludes susceptible groups, not enough funding, head-to-head comparisons between manufacturers, trials to test without treatment is unethical to detect adverse events

Drugs Recalls: Fen-phen (Fenfluramine/Phentermine)

increased heart valve problems

Drug Recalls: Vioxx

increased risk of heart attack, investigations about suppression of negative data

Drug Recalls: Conflicts of Interest

selective data reporting by pharmaceutical companies, contflicts of interest influence results, registration of all clinical trials: open access to database of all results

Sources of Error

sample size too small, random variation, multiple determinants, confounding variable is greater than the failure to control for other factors, selection bias, observer bias, recall or reporting bias

Cause and Effect

strengthening the association of risk factor with disease, magnitude of relative risk, increasing number of subjects, dose-response relationship, correspondence to possible biological mechanism, consistency across multiple studies

Conflicting Results: Hormone Replacement Therapy 1960s

administration of female hormones to menopausal women, lower CV disease risk, osteoporosis risk, menopausal symptoms, improve memory

Conflicting Results: Nurses’ Health Study 1976

cohort study that noted many benefits, observational study

NIH Clinical Trials of HRT

Bernadine Healy, head of NIH funded clinical trial in 1991, 2002: dramatic differences among groups

Women’s Health Initiative (WHI) Clinical Trial

HRT group lower risk of fractures and colorectal cancer, higher risk of heart attack, stroke, blood clots. trial terminated early due to selection bias in Nurses’ Study

Controversial Issues

clinical trials where no treatment is available e.g. AIDS drugs, bone marrow transplantation for breast cancer shows little evidence of being effective, conflicts of interest, evaluating risks, bias

Problems and Remedies

full disclosure of results, publication bias, placebos can be misleading, researchers can have financial interests

Publication Bias

release of data controlled by company

2007 FDA Revitalization Act

registration of trials and submission of complete data to public database at National Library of Medicine