Epidemiology (midterm) Flashcards
What is Hippocrates credited with?
The concept of “healthy body, healthy mind” (460 BC)
What is John Graunt credited with?
Being the father of demographics, quantifying births, deaths, and diseases in London (1662)
What is James Lind credited with?
Treating scurvy among sailors with fresh fruit (lemons) (1747)
What is William Farr credited with?
Applying vital statistics to the evaluation of health problems (1839)
What is John Snow credited with?
Establishing that cholera is a waterborne disease and identifying the origin of a cholera epidemic in London (1854)
What is Alexander Louis credited with?
The systematized application of quantitative reasoning and clinical trials (1872)
What is Bradford Hill credited with?
Suggesting criteria for establishing causation (1937)
What is public health?
The science and art of
- Preventing disease,
- Prolonging life, and
- Promoting health and efficiency
- Through organized community effort
What is the definition of health?
A state of complete mental, physical, and social well-being, and not merely the absence of disease or infirmity
What is disease?
A physiological or psychological dysfunction
What is illness?
A subjective state of not being well (the state of a person who feels aware of not being well)
What is sickness?
A state of social dysfunction (the role that an individual assumes when ill)
What is the (latest) definition of epidemiology?
The study of the distribution and determinants of health-related states or events in specified populations, and the application of the study to the control of health problems
How has the definition of epidemiology changed over time?
- The science of the phenomenon of infectious diseases and their natural history (1927)
- The science of infectious diseases, what causes and propagates them, and how to prevent them (1931)
- The study of distribution and determinants of all health-related events in specified populations, and the application of this study to their control (1988)
What does “distribution” refer to in the definition of epidemiology?
The frequency and pattern of health events in a population
How can the frequency of health-related events be assessed?
- Incidence (risk)
- Prevalence (distribution)
- Death rates (mortality, case-fatality rate)
(etc.)
What is the pattern of health-related events concerned with?
Person, place, and time
What questions does descriptive epidemiology answer?
Who, where, what, and when?
What is an important outcome of descriptive epidemiology?
(Impacts analytical epidemiology as well)
The formulation of an etiological hypothesis
How can the person distribution of health-related events be characterized?
- Age
- Gender
- Ethnicity
- Occupation
- Marital status
- Habits
- Social class
(The host factors)
How can the place distribution of health-related events be characterized?
- Geographical pathology: international, national, urban/rural differences
- Relating the geographical distribution to population density, social class, difference in health services, sanitation, education, environmental factors, etc.
What are the types of time distribution?
Short-term fluctuation
- Single exposure: one incubation period and one peak (e.g. single food poisoning event)
- Multiple/continuous exposure: continuous incubation periods and peaks (e.g. a cholera-contaminated water well, Minamata disease)
Periodic fluctuation
- Seasonal: e.g. GI infections in the summer, influenza in the winter, West Nile virus infections in Aug–Sep
- Cyclic: e.g. novel human coronaviruses every 7–10 years
Long-term/secular trend
- Chronic diseases: e.g. cardiovascular disease, lung cancer
What does “determinants” refer to in the definition of epidemiology?
The factors whose presence or absence affects the occurrence and level of a health-related event (i.e. risk factors)
What questions does analytical epidemiology answer?
How/why?
How are hypotheses of the determinants of health-related events proven or disproven?
Through analytical studies (case–control and cohort studies), experimental studies, and the use of statistical hypothesis testing throughout
What is the use of analytical epidemiology?
Developing scientifically sound health programs, interventions, and policies
What does “health-related states and events” refer to in the definition of epidemiology?
Not only the health of patients as individuals, but anything in the environment that may affect their health and well-being in any way
What does “specified population” signify in the definition of epidemiology?
That the unit of study is a population, not individuals
What do the terms “endemic,” “epidemic,” and “pandemic” mean?
- Endemic: the habitual presence or usual prevalence of a given disease in an area
- Epidemic: the occurrence of a group of illnesses in an area, clearly in excess of normal expectancy (outbreak)
- Pandemic: a worldwide epidemic
What are some sources of information in epidemiology?
- Registration of births, deaths, and diseases
- Population censuses
- Routine health information systems
- Surveillance
- Investigation of epidemics
- Sample surveys
What are some of the historical theories of disease causation?
- Supernatural theories: a curse, the evil force of a demon
- Hippocratic theory
- Miasmatic theory
- Theory of contagion
- Germ theory (Henle–Koch postulates)
- Classic epidemiologic theory (agent–host–environment theory)
- Multicausality and webs of causation
What are the factors of classic epidemiologic theory (the epidemiologic triad)?
- Agent of disease
- Susceptible host
- External environment
What are the definitions of “infectivity,” “pathogenicity,” and “virulence”?
- Infectivity: proportion of exposed persons who become infected (i.e. when the agent multiplies in the host)
- Pathogenicity: proportion of infected persons who develop clinically apparent disease
- Virulence: proportion of clinically apparent cases that are severe or fatal
What are some of the parameters of infectious agents?
- Infectivity
- Pathogenicity
- Virulence
What is the definition of “agent” in classic epidemiologic theory?
An infectious microorganism or pathogen, as well as chemical and physical causes (e.g. poisons, smoke, repetitive mechanical stress in carpal tunnel)
What is the definition of “host” in classic epidemiologic theory?
The human who can get the disease/health-related event
What are some of the factors related to host?
- Exposure, i.e. practices, personal choices
- Susceptibility, e.g. genetic composition, psychological status, age, sex
- Response to the causative agent
What is the definition of “environment” in classical epidemiologic theory?
The extrinsic factors that affect the agent, host, and opportunity for exposure
What are examples of environmental factors?
- Physical factors, e.g. geology, climate
- Biological factors, e.g. insect vectors
- Socioeconomic factors, e.g. crowding, sanitation, availability of health services
What is the definition of etiology?
The sum of all factors that contribute to the occurrence of a disease (agent factors + host factors + environmental factors)
What is the model of disease causation most suitable for chronic and noninfectious diseases?
Multicausal theory/web of causation
What are the types of causal relationships?
- Direct: A causes B without intermediates (very rare)
- Indirect: A causes B with intermediates (i.e. by the effect of C, D, etc.)
What are the types of disease-causing factors?
- Sufficient factors: the occurrence of these factors inevitably produces disease, but these factors may not be present in every occurrence
- Necessary factors: disease cannot occur without these factor, but these factors alone may not be enough to cause the disease
What types of factors (necessary and/or sufficient) are present in most infectious diseases (e.g. HIV)?
Neccessary, but not sufficient
E.g. AIDS will never occur in the absence of exposure to HIV, but not every person exposed to HIV will develop AIDS
What type of factors (necessary and/or sufficient) describes radiation, cigarette smoke, and genetic predisposition in the development of lung cancer?
Sufficient, but not necessary
Each of the factors in the question is enough to cause lung cancer on its own, but not every case of lung cancer will include exposure to any of the three
What type of factors (necessary and/or sufficient) best apply to chronic and noninfectious diseases?
Neither sufficient nor necessary
Factors implicated in chronic diseases are often neither sufficient nor necessary. What is the implication of this fact in public health?
- Public health action does not depend on the identification of every cause, as no single cause is enough to cause the disease, nor will a single cause be found in all cases
- These diseases may be preventable by blocking a single factor out of the many factors
What is the definition of natural history?
The progression of a disease process in an individual over time in the absence of treatment
E.g. the natural history of HIV/AIDS includes some clinical presentation within the first few months (acute retroviral disease) but then latency for around 10 years until AIDS-proper develops, and usually terminates with death
What is the term used to refer to persons who are infectious but are asymptomatic or have subclinical disease?
Carriers
What separates chronic carriers from ordinary carriers?
Chronic carriers remain infectious after recovery from the clinical illness, or may have never exhibited symptoms
What are examples of diseases that produce chronic carriers?
- Hepatitis B
- Typhoid fever
What is the majority clinical outcome of infection with tubercle bacilli (agents causing tuberculosis), without treatment?
(Unapparent, mild, moderate, severe, or fatal)
Unapparent
What is the majority clinical outcome of infection with the measles virus, without treatment?
(Unapparent, mild, moderate, severe, or fatal)
Moderate clinical severity
What is the majority clinical outcome of infection with the rabies virus, without treatment?
(Unapparent, mild, moderate, severe, or fatal)
Fatality
In disease natural history, what is a reservoir?
The habitat in which the agent normally lives, grows, and multiplies, be it humans, animals, or the environment. This may or may not be the source of infection.
E.g. for Clostridium botulinum (the bacterium causing botulism), the source of infection is usually improperly canned food containing spores, but the reservoir is soil
What is the implication of a disease having only a human reservoir?
The naturally occuring disease may be eradicated, as with smallpox in 1977
What is the term that refers to an infectious disease transmissible from vertebrate animals to humans?
Zoonosis
What are examples of (proven) zoonoses?
- Brucellosis (from cows and pigs)
- Anthrax (from sheep)
- Plague (from rodents)
- Trichinellosis/trichinosis (from pigs)
- Rabies (from bats, raccoons, dogs)
What are examples of diseases that are hypothesized to be zoonotic?
- HIV/AIDS
- Ebola
- SARS and COVID-19
What is a portal of exit?
The path by which a pathogen leaves its host, usually corresponding to where the pathogen is localized in the host
E.g. respiratory infections like Mycobacterium tuberculosis exit via the respiratory tract. Other agents exit by the blood, whether through direct exposure (e.g. HIV, hepatitis B) or by a vector (e.g. malaria)
What are examples of direct transmission?
- Skin-to-skin contact
- Kissing/saliva
- Sexual contact
- Droplets produced by sneezing, coughing, talking, etc.
What are examples of indirect transmission?
- Airborne: via dust or aerosols (droplet nuclei)
- Vehicleborne: via food (e.g. botulism), water (e.g. cholera), biologic products like blood, or fomites
- Vectorborne (using another organism): the organism may carry the infectious agent by purely mechanical means (e.g. fleas, lice, ticks) or may support biological changes in the agent (e.g. mosquitoes in malaria)
What is a portal of entry?
The manner through which the pathogen enters a susceptible host and gains access to tissues in which it can multiply or act
What are examples of portals of entry?
- Respiratory tract (e.g. cholera)
- GI tract and the fecal–oral route
- Skin (e.g. hookworm)
- Mucous membranes like the vagina (e.g. syphilis)
- Blood (e.g. hepatitis B)
What is the principle of herd immunity
Developing specific immunity in a certain percentage of the population (whether from prior infection or immunization) to limit the number of susceptible hosts that the agent can infect
What are the requirements for herd immunity?
- The disease must be restricted to a single host species (e.g. smallpox in humans)
- There must be relatively direct transmission
- Infections/immunization must produce solid, specific immunity
How is relative risk/risk ratio (RR) calculated?
RR = risk in exposed/risk in not exposed
i.e. RR = [a/(a+b)]/[c/(c+d)]
What are the commonly used measures of association?
- Chi-squared (χ2) test
- Relative risk/risk ratio
- Odds ratio
- Attributable risk
How is the relative risk/risk ratio evaluated?
≥ 3: high association
1.5–2.9: moderate association
1.2–1.49: weak association
1: no association
< 1: negative association (protective effect)
What does relative risk/risk ratio measure? What does a RR of 5 mean?
RR measures the risk of the outcome in those exposed compared to those not exposed.
A RR of 5 means those exposed are 5 times more likely to develop the outcome compared to those not exposed.
What kinds of studies can produce a relative risk/risk ratio?
Cohort studies
What does the chi-squared (χ2) test measure?
Whether there is an association between two categorical variables, and the probability that this association is purely due to chance (P value)
How is a chi-squared (χ2) test evaluated?
If the P-value ≤ α-value (where α is usually 0.05), there is a significant association
i.e. if the test statistic is greater than the critical value
What is a drawback of the chi-squared (χ2) test?
It only tells us if there is an association, not how strong this association is
What types of studies are suitable for a chi-squared (χ2) test?
Cross-sectional studies
How is the odds ratio (OR) calculated?
OR = (a∙d)/(b∙c)
What does the odds ratio measure? What does an OR of 5 mean?
OR measures the probability that those with the disease were exposed, compared to those who do not have the disease
An OR of 5 means that those with the disease are 5 times more likely to have been exposed compared to those without the disease
What types of studies produce an odds ratio?
Case–control studies
Under which circumstances can the odds ratio be used to estimate the relative risk/risk ratio?
- The selected controls are representative of the general population
- The selected cases are representative of all cases
- The disease is rare
How is the attributable risk (AR) calculated?
Give both the absolute value and the percentage value
AR = a/(a+b) – c/(c+d)
AR% = [a/(a+b) – c/(c+d)] ∙ 100/[a/(a+b)]
∴ AR% = (AR ∙ 100)/[a/(a+b)]
How is attributable risk evaluated?
﹥0: positive association
= 0: no association
< 0: negative association (protective effect)
What does (absolute) attributable risk measure? What does an AR of 3 mean?
(Absolute, i.e. not the percentage)
AR measures the risk of the outcome in the exposed due to exposure alone, not due to other factors
An AR of 3 means those exposed are 3 times more likely to develop the outcome due to this exposure alone, removing the risk due to other factors
E.g. for smoking and lung cancer, those who smoke are 3 times more likely to develop lung cancer due to the fact that they smoke alone
What does the attributable risk percentage measure? What does an AR% of 90% mean?
AR% measures how much of the outcome among those exposed is directly attributable to the exposure, not other factors
An AR% of 90% means that 90% of the outcome among exposed is due to the exposure alone
E.g. for smoking and lung cancer, 90% of the lung cancer cases are due to the individuals smoking, not due to other factors
What is the implication of attributable risk?
The AR values suggest the amount of the disease that could be eliminated if the exposure is controlled or eliminated.
E.g. for smoking and lung cancer, an AR of 90% means that 90% of cases of lung cancer among smokers could have been prevented if they did not smoke. The remaining 10% are due to other factors and may still have occurred if the individuals did not smoke
Are all risk factors and preventitive factors manipulable?
No, e.g. age
What is the definition of association?
The concurrence of two variables more often than would be expected by chance
What are the types of associations?
- Spurious (false) associations
- Indirect (confounding) association
- Direct (causal) associations: one-to-one or multifactorial
What are the steps in establishing a cause–effect or exposure–outcome relationship?
- Evaluate the results of the study and ensure that they are accurate: evaluate the method, validity, reliability, and bias
- Perform statistical analysis on the results (calculate P-value, confidence interval). If the P-value is not significant, this may be due to a small sample size (low power)
- If the P-value is significant, evaluate whether the relationship is indirect—the presence of a confounder
- If the possibility of confounding is excluded, apply the criteria of causation
What kind of error is bias?
Systematic—once it is introduced to the study, it cannot be corrected
What are the types of bias?
Selection bias: distortion in the selection of participants, either:
- A difference in those selected to participate and those who were not, or
- A difference in those selected for the treatment group and those selected for the control group
Information bias: distortion in the way the data were obtained from the study groups (e.g. recall bias)
What is recall bias?
Those in the “outcome” group are more likely to remember possible exposure than those in the “non-outcome” group
What are the criteria for a confounding variable?
- Must be a risk factor for the disease independently
- Must be associated with the exposure under study
- Must not lie on the causal pathway between the exposure and the disease
How can confounding be adjusted for?
- At the design stage: restriction, matching, randomization
- At the analysis stage: stratification, multivariate analysis, standardization
What are Hill’s criteria for causation?
Most important criteria
* Temporality: cause precedes effect
* Strength of association: large RR or AR
* Consistency: the exposure–outcome repeatedly observed by different people in different places under different circumstances at different times
Other criteria
* Biological gradient (dose–response): increasing the exposure is associated with higher rates of disease, and vice versa
* Biological plausability: causation makes sense according to biological knowledge of the time
* Experimental evidence
Additional, non-necessary criteria
* Analogy: similar cause–effect relationships established for related diseases
* Specificity: one cause leads to one effect
* Coherence: relationship doesn’t seriously conflict with known facts of the disease and its biology
What is the definition of research?
A systematic investigation to develop or contribute to generalizable knowledge
What is the definition of study design?
A specific plan or protocol for conducting the study, which allows the investigator to translate the conceptual hypothesis into an operational one
What are the types of uncontrolled-assignment studies?
- Descriptive: cross-sectional, case report, case series, ecological studies, surveys
- Analytical: case–control, cohort studies
What are the types of controlled-assignment studies?
- Community trial (community assignment)
- Clinical trial (individual assignment)
What is a prospective study?
One that watches for outcomes based on specific exposure, e.g. cohort studies
What is a retrospective study?
One that looks backwards to examine exposures based on an identified outcome
What is the most common type of study published in the medical literature?
Case report
What is the most common type of analytical study?
Case–control study
What are limitations of case reports and case series?
- Cannot estimate prevalence or incidence (except if the outcome is surveilled, e.g. breast cancer)
- There is no control group for comparison
What is the ecological fallacy?
Drawing unfounded conclusions about exposure–outcome association on the individual level from data on the population level
E.g. in Country A, 72% of people eat 5+ pieces of fruit a day, compared to 24% in Country B. In Country A, the rate of obesity is 34%, compared to 66% in Country B. We cannot say that eating 5+ pieces of fruit a day decreases your chances of being obese, as those in Country A who eat 5+ pieces of fruit a day may not be those with the low BMIs
What is the unit of analysis in ecological studies?
Populations/aggregates of individuals
How are ecological studies typically analyzed?
Looking for associations and correlations (correlation requires that both variables are continuous)
- Pearson’s product–moment correlation coefficient (r)
- Spearman’s rank correlation coefficient (ρ)
What can be calculated from cross-sectional studies?
- Disease prevalence
- Prevalence rate ratio (PRR) = [a/(a+b)]/[c/(c+d)] {similar to risk ratio but for prevalence, not incidence}
Can cause–effect relationships be determined from cross-sectional relationships?
No, as temporality cannot be determined
What kinds of diseases are best studied using case–control studies?
- Rare diseases
- Diseases with a long latency period
Can incidence be calculated from case–control studies?
No
What is the ideal characteristic of the control group in a case–control study?
Similar to the case group in all respects except having the disease
What can be calculated from case–control studies?
Odds ratio (OR)
What errors can be present in case–control studies due to their retrospective nature?
- Recall bias
- Confounding variables
What type of observational study is most suitable for assessing cause–effect relationships for a rare disease?
Case–control studies
What type of observational study is most suitable for assessing cause–effect relationships for a rare exposure?
Cohort studies
What can be calculated from a cohort study?
- Incidence
- Relative risk/risk ratio (RR)
What are the types of cohort studies?
- Prospective cohort
- Retrospective cohort
- Ambidirectional cohort
What is the “placebo” treatment if a treatment for the disease already exists?
The best standard of care. Giving an inert placebo in this case is considered unethical
What is the ascending order of strength for types of evidence?
- Background information and expert opinions
- Case reports and case series
- Case–control studies
- Cohort series
- Randomized controlled trials
- Critically appraised topics
- Systematic reviews
- Meta-analyses
In the hierarchy of evidence, what are the types of unfiltered information?
- Case reports/case series
- Case–control studies
- Cohort studies
- Randomized controlled trials
In the hierarchy of evidence, what are the types of filtered information?
- Criticaly appraised topics
- Systematic reviews
- Meta-analyses
What is a critically appraised topic (CAT)?
A short summary of the evidence on a topic of interest, usually focused on a clinical question
To what type of evidence does a clinical practice guidline belong?
Critically appraised topic
What is a drawback of CATs?
They have a short useful lifetime if they are not updated with new information
What is a systematic review?
A literature review that considers the entirety of available literature to provide a comprehensive overview of a subject and minimize bias
Where are systematic reviews typically found?
In research databases like Cochrane or PubMed
What is a meta-analysis?
A systematic review that summarizes information from all studies on a specific clinical question, using quantitative, statistical methods to determine whether findings are reliable and applicable to different populations
What is the difference between filtered and unfiltered information?
- Unfiltered information: shaped by the researcher’s approach—the researcher is the primary source of data collection
- Filtered information: the researcher applies quantitative or qualitative methods to filter information from unfiltered sources
What is Trip?
A meta-search engine and database for searching through both filtered and unfiltered medical research (similar to Google Scholar)
Where do blogs and sources like Wikipedia lie in the evidence hierarchy?
The lower part of the pyramid—they are less reliable due to lack of rigorous peer review and a systematic methodology
What is surveillance?
The systematic ongoing collection, collation, analysis, and interpretation of health data; the evaluation of public health practice; and the timely dissemination of these data to those who need to know
(Combined WHO and US CDC definitions)
What are the general objectives of surveillance?
- Following trends in the health status of a population over time
- Establishing health care and public health priorities
- Ensuring those with the greatest need are prioritized
- Detecting and responding to epidemics
- Evaluating the effectiveness of programs and services
What are the specific objectives of surveillance?
- Giving early warning of changes in incidence (e.g. the recent increases in tuberculosis cases in many countries)
- Early detection of outbreaks
- Evaluating the effectiveness of interventions (e.g. a new vaccine)
What are examples of events that may require surveillance?
- Epidemic diseases (e.g. measles, meningococcal meningitis)
- Malnutrition
- Animal reservoirs and vectors of communicable diseases
- Environmental pollution, especially water pollution
- Demographic events, e.g. births and deaths
A public health authority sets a goal of decreasing deaths due to malaria by 20% in the coming 5 years. What is the role of surveillance in achieving this goal?
- There must be a baseline figure on deaths due to malaria
- There must be a way of monitoring the trends in deaths due to malaria over time
What are the types of surveillance?
- Passive surveillance
- Active surveillance
- Negative surveillance
- Sentinel surveillance
- Mortality surveillance
What is passive surveillance?
- Data are generated without intervention or contact by the surveilling agency
- Other agencies will initiate reporting
- E.g. for notifiable diseases, like anthrax, cancers, COVID-19
What is active surveillance?
- The surveilling agency initiates procedures to obtain reports
- The data collator ensures that the reporting agency is in fact collecting data to the fullest degree
What is negative surveillance?
- The surveilling agency requires the collector to report even the absence of cases
- Ensures that a “nil report” truly reflects zero incidence, not a failure to report
- E.g. for uncommon disorders, pediatric conditions
What is sentinel surveillance?
- The surveilling agency selects (either randomly or intentionally) a small group of health workers from whom to gather data
- Often produces more detailed data on cases of illness because the participating health workers may receive incentives
- E.g. analyzing the nasopharyngeal swabs from all patients at only 15 COVID-19 swabbing sites to analyze and identify the strain of the virus
What is mortality surveillance?
- Present in most countries—the requirement that every death in the population be reported or registered
- Yields a death rate
- Countries differ in the quality and completeness of their death registers
Why is mortality surveillance near-universal?
Death is the common final outcome of severe diseases. Changes in death rates can be an indicator of changing health status of a population
What are the components of the surveillance arc?
- Problem
- Data gathering
- Data analysis/integration
- Intervention
(loops back to step 1)
What are the steps of surveillance?
- Data collection
- Data collation
- Data analysis
- Dissemination (often with resulting action)
What are the levels of prevention?
- Primary: preventing disease when risk factors are present (e.g. by immunization, lifestyle modification)
- Secondary: preventing complications when the disease is present (e.g. by early diagnosis, sustained care)
- Tertiary: prevention of disability or death (e.g. by rehabilitation)
- Quaternary: prevention of over-diagnosis and over-treatment from interventions which are too invasive or ethically unacceptable (using evidence-based medicine)
What is screening?
The systematic application of a test or inquiry to identify individuals at sufficient risk of a specific disorder to benefit from further investigation or direct prevention among persons who have not sough medical attention from symptoms of that disorder
(Wald, 2004)
What are the aims of screening?
- Better prognosis for individuals
- Protecting the public from communicable diseases
- Rational allocation of resources
- Research into the natural history of disease
What is opportunistic screening (case finding)?
Screening someone when they come into contact with the health system for another reason
E.g.
- Checking the lipid profile for an overweight or obese patient when they come to your clinic (screening for diabetes)
- Referring all women at your clinic within a certain age range for cervical or breast cancer screening
What is the difference between screening and diagonsis?
Early detection/diagnosis is based on symptoms and signs, while screening is performed for people who are usually asymptomatic
What are the criteria for the necessity of screening?
- The disease is an important, well-defined health problem with known epidemiology, well-understood natural history, and prevalent undiagnosed cases
- The presence of a presymptomatic or early stage, where screening would be actually beneficial
What are the characteristics of an ideal screening test?
- Safe
- Inexpensive
- Acceptable
- Reliable
- Valid
- No, or minimal, adverse effects (e.g. pain, radiation exposure)
What are the parameters of screening test validity?
- Sensitivity
- Specificity
- False positive rate
- Positive predictive value (PPV)
- Negative predictive value (NPV)
What is sensitivity? How is it calculated?
The ability of a screening test to correctly identify those with the disease
= TP/(TP + FN)
What is specificity? How is it calculated?
The ability of a screening test to correctly identify those without the disease
= TN/(TN + FP)
What is the false positive rate? How is it calculated?
The proportion of “positive” results that are erroneous—incorrect identification of disease
= FP/(FP + TN)
= 1 – specificity
What are the implications of the false positive rate?
Estimates the magnitude of:
- the economic effects of a false positive (further testing, investing health care resources)
- the psychological distress of a false positive
What is the PPV? How is it calculated?
The likelihood that someone with a “positive” result actually has the disease
= TP/(TP+FP)
What is the NPV? How is it calculated?
The likelihood that someone with a “negative” result actually does not have the disease
= TN/(TN+FN)
How can prevalence be calculated from a two-by-two table of a screening test?
= (TP + FN)/(N),
where N is the total sample size
What is the importance of a pilot project for a public health screening program?
- Evaluating the economics of the program
- Setting the cut-off age based on local data
- Improving performance from the experience gained
- Comprehensive assessment of the screening process
- Quality assessment of staff
- Assessing the barriers to screening
What is the lead time bias?
A perceived longer survival time in those who underwent screening as they were detected at a stage too early for effective treatment
E.g. a person who underwent early brain cancer screening survived 8 years, compared to 3 years for a person diagnosed from symptoms. If the person who underwent screening were left to be diagnosed from symptoms, they may too have survived for only 3 years—there is a 5 year lead time bias.
What is the length time bias?
Screening tends to detect cases where the disease is progressing slowly due to a longer asymptomatic phase, inflating the survival rate
Those with severe, short disease have a very short asymptomatic phase, and so wouldn’t be detected from screening anyway
What is the volunteer bias?
Those who volunteer for screening tend to be:
- of a higher socioeconomic class,
- more health-conscious, and
- comply with prescribed advice better.
Therefore the percieved benefits of screening may only be the benefits of volunteerism
