HS300: Measures of Morbidity & Mortality I

Morbidity

preventable disease

Mortality

preventable death

Incidence

how fast new disease is appearing

Prevalence

• How much disease is present right now

• The number of cases — new and existing — of a disease or
  health condition in a population, at or during some designated time

• Provides an indication of the extent of a health problem, a burden measure—how many people have this condition right now

• NOT A MEASURE OF RISK

The epidemiologist’s bathtub

• Factors that affect prevalence

  • Incidence rate: the faucet

  • Recovery/cure rate: the steam

  • Death rate: the drain

  • Transfer (migration): the bucket

Ways to lower prevalence that does not lower incidence

• A cure that shortens duration

• A disease that kills faster

• Out-migration of cases

Quantifying a disease

1. Define the disease

• A case definition, written down before you count anything.

2. Count the individuals affected

• This is your numerator.

3. Determine the population the cases arose from

• This is your denominator.

4. Account for the passage of time

• Over what period did these cases accrue?

Four elements of incidence

1. Numerator: number of new cases

2. Denominator: population at risk

3. Rate base: multiplier (per 100, 1,000, etc.)

4. Time: the period during which cases accrued

Should recurrent events in the same person count as new cases?

• Chronic disease: count first-ever events (etiologic study)

• Infectious disease: recurrence is exactly what you want, anyone can get disease as many times as possible in a year

Population at risk

• People capable of becoming a new case

1. They are at risk of contracting the disease.

2. They are disease-free at the start of the specified time period.

3. Individuals not at risk are excluded from the denominator—this matters most with lifelong diseases and with infections that confer immunity

Who should I exclude from the denominator?

The rate base

• Multiplier, always a power of 10

• Purpose: to express the reported rate as a number equal to or greater than 1, so it can be read at a glance

Cumulative incidence rate

• Number of new cases ÷ population at risk

• Average risk of developing the disease in this population, over a defined period

Incidence rate (incidence density, person-years incidence rate)

• Number of new cases ÷ total person-time of observation

• If period of observation is measured in years: Number of new cases ÷ total person-years

• Measures how fast disease is occurring per unit of time at risk, stays honest even when people drop out

Person-years calculation

1. Total the person-years (Subjects A x Years observed B = Person-years A x B)

2. Calculate the incidence rate

3. Interpret the result in words

• Ex. In a 5-year prospective study of 311 subjects, the incidence rate of colon cancer was 4.9 per 100 person-years.

What incidence data are for

1. Research on the etiology and causality of disease

• Because incidence, unlike prevalence, is uncontaminated by how long people survive.

2. Estimating the risk of developing a disease

• The question a patient actually asks: what are my chances?

Attack rate

• An alternative form of cumulative incidence, used when a disease is observed in a population over a short, sharply bounded period — an outbreak rather than a decade

• Ill ÷ (ill + well) among the exposed group, during the outbreak

• Ex. foodborne outbreak, flu in dorm or cruise ship

Point prevalence

• At one moment in time

• Cases at one point in time ÷ total population at that time

Period prevalence

• Over a stretch of time

• Cases during a time period ÷ average population at the mid-point

Uses of prevalence

1. Describing the burden of a health problem

• How much of this condition is there, in this population, right now?

2. Determining the allocation of health resources

• Facilities, personnel, budgets — prevalence answers the planner’s question.

• How many people need care right now?

Prevalence and etiology

• Prevalence data are less useful for studying the causes of disease because of the survival factor.

• A prevalent case is someone who both developed the disease and
  lived long enough to be counted. Those are two different things, and
  prevalence cannot tell them apart.

The survival trap

1. Suppose a risk factor doubles the incidence of a disease.

2. But it also makes that disease rapidly fatal.\

3. Prevalence among the exposed could be lower than among the unexposed — the exposed cases die before anyone can count them.

• A prevalence study would conclude the exposure is protective. This
  is why etiologic research uses incidence.

Relationship between prevalence and incidence

• P ~ I x D

• The prevalence (P) of a disease is proportional to its incidence rate (I) multiplied by the average duration (D) of the disease.

• The relationship is most accurate when incidence and duration are both relatively stable.

• Higher incidence, or longer duration, produces higher prevalence.

When duration is short

• For many infectious diseases, prevalence becomes similar to incidence.

• Short duration means cases leave the tub quickly: they
  recover rapidly, or they die.

• Ex. cold, ebola

When duration is long

• For chronic disease — diabetes is the standard example — prevalence stays high and keeps climbing even when incidence is flat or falling.

• The result is a large and widening gap between the incidence and the prevalence of the disease.