Non-Communicable Disease Epidemiology Final Exam

Natural history (def)

Typical progression of a disease in an individual from its beginning until its final outcome without treatment

Stages of disease (3; interventions)

Predisease: health promotion and specific protection

Latent disease: Early diagnosis and treatment

Symptomatic disease: Disability limitation, rehabilitation

Primary prevention (methods)

3 E's

Education

Environmental change

Enforcement/legislation

Population strategies (def, key factors)

Bell curve -> shift risk factor distribution in the population

Move everyone down 5lbs or lower chol

Radical

Benefit-to-risk ratio low

High-risk strategies (def, key factors)

Screening and aggressive treatment

Behaviorally inappropriate

Difficult to identify

Risk scores: Framingham, SCORE

Secondary prevention (methods)

Early detection

Screening

Tertiary prevention (methods)

Prevent and delay development of acute/chronic complications

COPD (cases, def, types)

210 million cases/year

Chronic bronchitis/emphysema

Decreased elasticity of the lung

Asthma (cases, prevalence, why deaths)

300 million cases/year

Increasing in prevalence

Deaths related to lack of proper treatment

Occupational lung diseases (examples)

Pneumoconiosis

Occupational asthma

Work-aggravated asthma

Sleep apnea (%, sequelae)

25% men, 10% women

Untreated -> hypertension, stroke, cardiac failure

Hospital-based cancer registry (limitation)

Only the people visiting that hospital

Population-based registry (US, when, who, what)

USA: Surveillance, Epidemiology and End Result (SEER) program

started 1973

36% of population

"tracks the incidence of persons diagnosed with cancer during the year and collects follow-up information on all previously diagnosed patients until their death."

Age standardization (def, use)

Apply rates to a relevant population standard

Take away impact of age on epi

Account for differences in age structure of the population/s being considered

Geographical variation (factors)

- Genetics

- Lifestyle

- Environmental factors

- Screening

(migrant studies)

Migrant studies (def/use)

Determine if causes are lifestyle or environmental

Lung cancer (cases)

2.48 million cases, 1.82 million deaths (g)

2,500 cases, 1.9k deaths (I)

Most common overall

Lowest 5-year survival rate (1.7/10)

90% from smoking

Lung cancer (trends)

-Higher in men, but decreasing

-Increasing in women, effects from smoking in 60/70s

Lung cancer (risk factors)

-Age

-M>F

-SEG: lower>upper

-Smoking

-Dose-response relationship, no safe level

Doll and Hill (type, years, focus, results)

Cohort study of doctors

1971-91

Cases with lung cancer, controls without

Smoking & lung cancer link

Lung cancer (prevention)

Primary: stop smoking

Secondary: screening?

Breast cancer (cases)

2.31 million

3.6K (i)

Second most common after lung

Breast cancer (incidence/mortality)

Highest incidence in the West

Higher deaths in Africa/LMICs

Breast cancer (risk factors)

-increasing age

-SEG incidence higher in upper, mortality higher in lower

-age at menarche/menopause

-age at first pregnancy (over 30 >risk, under 20

Breast cancer (prevention)

Modify lifestyle factors

Awareness and screening

Treatment and specialist centers

Colorectal (cases)

1.9m (g)

2.5k (i)

Third most common

Second leading cause of cancer deaths (900k/year)

Colorectal (risk factors)

-Family history

-Increasing age

-Gender no

-SEG no

-Ulcerative colitis 2-8.2

-Lifestyle

-Diet

Colorectal (prevention)

Diet

Exercise

Screening

Prostate cancer (cases)

1.4m (g)

4.1k (i)

Second-most common diagnosed among men

Mort low in West, high in Africa

Prostate cancer (risk factors)

-Increasing age

-Race (AA men higher, Asian men lower)

-Development of prostate

-Family history

-Diet

-Smoking

-Older paternal age

Prostate cancer (prevention)

Uncertain

General health health diet

Cervical cancer (cases)

662k (g)

250 (i)

350k death g

Cervical cancer (geography, progression)

High in Africa

10-15 year progression from CIN to CC

CC (risk factors)

-High risk HPV infection (12 strains, especially 16/18)

-HIV

-Smoking

-Oral contra

-High parity

-Low SEG

-Immunodeficiency

-Diethylstiboestrol during preg

Framingham study (type, location, when, goal)

Longitudinal CV cohort study

Massachusetts

1948-now

Identify risk factors for CVD

Nurse's health study (when, what)

1976

Risk factors for major chronic disease in women

Doll and Hill (what, when, who)

Link between smoking and cancer/diseases

1950s

British doctors

Skin cancer (cases)

1.5m (g most common)

11k (i most common, 90% non-melanoma skin cancers, remainder melanoma)

Skin cancer (types)

Squamous cell carcinoma

Basal cell carcinoma

- Both chronic ulcers scars, burns; sun exposure

Malignant melanoma

- Naevi/moles, childhood sunburn/intermittent sun

Skin cancer (trends)

Trending up in incidence and mort

Often caught early

High 5 year survival and rising since 90s

Skin cancer (risk factors)

-Age

-SEG lower>upper

-Ethnicity

-Family

-Phenotype

-Pigmentation disorders

Skin cancer (prevention)

Primary: hats, suncream, protect children, avoid lamps/beds

Secondary: early diagnosis, awareness

Injury (definition, key stats)

Damage to the body from acute exposure to thermal, mechanical, electrical or chemical energy or from absence of essentials such as o2

Most common cause of death for 1-44 years

35% of all ER visits

General model for injury control

Monitor incidence

Identify risk factors

Intervene

Evaluate

Haddon Phase Factor Matrix (what is, 6 factors)

Phases relevant to the risks posed by the agent, host and environment

Pre-event, Event, Post-event

5 E's of incident prevention

Epidemiology

Education

Enforcement

Engineering

Evaluation

Screening vs diagnostic test

Screening = identifies asymtomatics at risk of disease

Diagnostic = determines presence or absence of disease in symptomatics

Positive predictive value (def, equation)

probability a person with a positive test actually has the disease; true positive

= true positive/total positives

Negative predictive value (def, equation)

Prob person with a negative test is truly disease-free; true negative

= true negatives/total negatives

Sensitivity (def, equ)

test's ability to correctly identify true positives/people with disease

= true positive/(true positive+false negative)

Specificity (def, equ)

Ability of test to correctly identify true negatives

= true negatives/(true negative+false positive)

Lead-time bias (def)

Live with disease longer because it was diagnosed earlier

Length bias (def)

screening tends to detect more slow-growing/indolent cancers; aggressive cancers move faster so less likely to be picked up

Benefits of screening

-Improved prognosis

-Reassurance for ppl with negative tests

-Less radical treatment

Disadvantages of screening

-Longer "illness" if prognosis unchanged

-False reassurance for false negatives

-Anxiety/unnecessary medical intervention

-Cost

Considerations for starting a screening program

-Management

-Population register

-Call-recall system

-Facilities

-Fears/anxiety

-Treatment

-Quality assurance

Junger and Wilson criteria (use, 9)

To introduce a screening program. Must be:

-Relatively common

-Seriously debilitating

-Treatment acceptable to population

-Facilities available for treatment

-Recognized treatment available

-Recognized identifiable preclinical phase

-Natural history understood

-Treatment of preclinical disease more effective than treatment of symptomatic

Wilson & Jungner screening test criteria (7)

-Simple and quick

-Specific

-Sensitive

-Repeatable

-Accurate

-Inexpensive

-Acceptable

Criterion of positivity (def, considerations)

Test value at which screening test outcome is considered positive

Weigh cost of false positives vs negatives, anxiety and costs vs false sense of security

Decreased criterion of positivity =

sensitivity increase, specificity decrease

Increased criterion of positivity =

sensitivity decrease, specificity increase

Breast cancer screening (test, evidence, issues, rec)

Mammography

Reduction of mortality 20% in 50+

Overdiagnosis/treatment

Yes

CC/CIN screening (test, rationale, rec)

Smear and cytology/HPV and reflex cytology (more sensitive)

Smear tests can detect CIN which always precede CC

Yes

Colorectal cancer screening (test, evidence, rec)

Meets ALL disease criteria

Test = fecal immunochemical test

Early detection and removal = reduction in incidence and mort

Yes

Lung cancer (evidence, rec)

Limited evidence, only one RCT showed reduced mort

US rec yes for at risk, EU waiting on more trials

Prostate cancer screening (evidence, tests, rec)

Mixed evidence, no reduction in mortality

Tests = prostate specific antigen (PSA), digital rectal exam

US and EU no

EU running PRAISE-U project to identify cost-effective early detection programmes for prostate cancer

Interval cancer (def, examples)

Cancers diagnosed between routine screening appts

Breast, CC, Colorectal

Interval breast cancer (causes)

Mammographically occult

True interval

Missed abnormality (false neg), subtle signs

Obesity (trends/stats)

3x since 1975

1 in 8 people in 2022

1.18 association with all-cause mort in OB

23% Irish adult OB, 1 in 5 children

BMI (pos/negs)

Pos: quick, simple, inexpensive

Neg: Crude, doesn't take body comp into account, can't be used in children, underestimates prevalence of OW&OB

Metabolically healthy OB (def)

absence of metabolic abnormality in OB person

OB as public health concern (conditions, sequlae)

Metabolic syndrome, T2D and CVD

Higher risk of premature death and all-cause mort

Childhood OB has lifelong impacts

Negative social/econ consequences

Costs to individuals and systems

T2D (def, trends)

Body cannot effectively use insulin, damage to nerves, vessels, eyes, kidneys

4x increase in 18+ with diabetes

Rising faster in LMICs

14% of adults in 2021

90% potentially avoidable

T2D Ireland (stats)

9.5% in 2015

7th in world for diabetes-related health exp. per person

Stable in adults but rising in primary school

Pre-diabetes (def, test, prog)

Abnormal glucose reg but not past threshold

Impaird fasting glucose or impaired glucose tolerance

70% will develop overt T2D

T2D risk factors

-Age

-Family history

-Gestational dia

-OB

-Sendintary

-Smoking

-Very low birth weight

-Depression, antipsychotics

-ART

-Diet

Building a cancer strategy (steps)

FROM PREVENTION TO Palliative andd prediction

1 Population data

2 Diagnosis

3 Treatment

4 Palliative care

Predicting future trends

Building a cancer strat: Population data (considerations)

Primary prevention

Secondary prevention (screening)

Building a cancer strat: diagnosis (considerations)

Facilities available

Building a cancer strat: Treatment (considerations)

Current types and numbers of cancer by age and sex

Personnel required

Hospitals

GP/primary services

Building a cancer strat: Palliative care (considerations)

Cure/survival rates

Where do people like to die

Palliative care specialists

Projections

Testing in series (def, s/s effect)

BOTH screening tests must be pos/abnormal for positive result

Lower sensitivity, higher specificity

Testing in parallel (def, s/s effect)

EITHER test is pos/abnormal = pos screening result

Higher sensitivity, lower specificity

Positive predictive value (PPV) (def, equ)

True positives, person with pos result actually has disease

True positives/total positive on test

Negative predicative value (NPV) (def, equ)

True negatives, person with neg result does not have disease

True negatives/total negative on test

Receiver Operating Characteristic curve (ROC curve) (def, setup, interp)

Graphic determination of test accuracy

Sensitivity on Y vs 1-specificity on X

More area under the curve is GOOD: 1 is perfect, .5 is random, >.5 is model worse than guessing

No discrimination is BAD

Likelihood ratio (def, interp)

Commentary on test quality ;probability of getting a result if the person has the condition vs if they were healthy

Good test has high ratio

LRpos>10 and LRneg<0.1 = convincing evidence

LRpos>5 and LRneg<0.2 = strong evidence

LR+ (def, equ)

Probability of getting a positive result if person is diseased vs healthy

Equ = Likelihood of + test in diseased person / likelihood of + in a healthy person

LR- (def, equ)

Probability of getting a negative result if person is diseased vs healthy

LR neg = (1-sensitivity)/specificity

Cardiovascular disease (CVD) (def, examples)

Conditions of the heart and blood vessels, all stemming from atherosclerosis

-Coronary heart disease (CHD)

-Cerebrovascular disease

-DVT

-Pulmonary embolism

CVD public health significance (stats)

#1 cause of death and premature death globally

19.8m in 2022

32% of all global deaths (75% of which LMIC)

38% of premature deaths

Most can be prevented

CVD and COVID (relationship)

Pre-existing CVD = worse outcomes, greater risk of death from COVID

COVID can induce CVD symptoms

Descriptive epidemiology (def)

Describes distribution and occurrence of a disease

Who, what, where, when

Analytical epi (def)

Analyzing relationship between disease and risk factors

Formulation and testing of etiological hypotheses

Experimental epi/prevention strategies

Descriptive epi CVD (list)

Age: Atherogenesis and lifestyle begins in childhood, increase in m's starts from 30

Gender: mort more common among men, affects women equally but older; women have gender-specific risks

CVD trends (over time, location)

Doubling of cases 1990-2023

Prevalence higher in E EU, C Asia, N Africa, ME, N America, Australia; mort higher in E EU, N Africa, C Asia

CVD deaths leading causes

CHD/ischemic heart disease and stroke

CVD in Ireland (stats)

56% reduction from 1985-2006

48% due to risk factor improvements

56% reduction from 2000-15

60% from improvements in treatment

CVD risk factors

Elevated total or LDL chol

Low HDL

Hypertension

Smoking

Diabetes

Family history

Age

OB

Psychosocial

Diet

Inactivity

HDL (def, implication)

High-density lipoprotein

Confers protection against CVD?