Measures of Morbidity and Mortality in Populations

Question 1

What are three ways to measure morbidity? (disease frequency)

Answer 1

1) cumulative incidence

2) incidence rate

3) prevalence

Question 2

How do you calculate cumulative incidence?

Answer 2

(number of new cases of the disease occurring in the population during a specific period of time) / (number of persons at risk in the population for the disease at the beginning of the time period)

Notes: Individuals are followed over time. Also, to be able to calculate the cumulative incidence, all individuals must either develop the disease during a specified period of time or be observed for entire period but do not develop the disease. It is a proportion, often expressed as a %. It is a measure of risk.

Question 3

What is risk?

Answer 3

Risk is the probability that an event will occur during a specified time

Question 4

How do you calculate incidence rate?

Answer 4

Incidence rate expressed as number of new cases per person-time “at risk”

= (# new cases) / (total person-time at risk)

Notes: Individuals are followed over time. The value above yields rate of new cases per 1 person-year at risk. Typically used when some individuals who do not develop the disease are not observed for the entire study period (i.e. different people are observed for different lengths of time) and cumulative incidence cannot be calculated directly.

Question 5

How do you calculate prevalence?

Answer 5

Proportion of the population affected by a disease at a given time (est of the existing burden of disease)

= (number of existing cases of disease present in the population at a specified time) / (number of persons in the population at a specified time)

Notes: Individuals are NOT followed over time. This is not a measure of risk of disease, and also ignores whether or not people in the population are “at risk.”

Question 6

How do you calculate proportionate mortality?

Answer 6

What proportion of all deaths are due to a particular disease?

= (# deaths due to particular disease) / (total # deaths)

Question 7

How do you calculate annual all-cause mortality rate? (per 1000)

Answer 7

= 1000* (total number of deaths from all causes in year) / number of persons in the population at mid-year)

Question 8

How do you calculate annual cause-specific mortality rate? (per 1000)

Answer 8

= 1000* (number of deaths from “disease A” in year) / (number of persons in the population at mid-year)

Question 9

How do you calculate proportionate mortality?

Answer 9

What proportion of all deaths are due to a particular disease?

= (# deaths due to particular disease) / (total # deaths)

Question 10

How do you calculate case fatality rate?

Answer 10

What proportion of individuals with a disease die from that disease? How severe/lethal is a disease?

= (# deaths from Disease Y) / (# people with Disease Y

Question 11

What is adjustment and how is it used? What are the two types?

Answer 11

Adjustment is a procedure for adjusting rates (e.g. mortality rates), to minimize the effects of differences in population composition with respect to an extraneous variable (e.g. age) when comparing rates for different populations

two kinds: direct and indirect adjustment

Question 12

Explain direct adjustment.

Answer 12

A standard population is used in order to minimize the effects of any difference in an extraneous variable (e.g. age) between the populations being compared. This allows you to better answer the question: If age distributions of the two populations were similar, would there be any difference in mortality between the two populations being compared?

Direct adjustment can be carried out for any extraneous variable, doesn’t have to be age.

Question 13

EX: How could cMR2 > cMR1 while each of the age-specific mortality rates for popln 1 are greater than the corresponding age-specific rates for popln 2?

Answer 13

The age distributions in the two populations are different (older in popln 2) and older age is associated with higher mortality rates

Question 14

How do you calculate direct age-adjustment?

Answer 14

1) calculate rates for each population overall and by age category (these are the crude and age-specific rates)

2) identify standard population (most often, you will add populations of interest by age categories, may use external population with age categories)

3) “apply” age-specific rates to standard populations to obtain expected number of events for each age category of each population

4) sum the expected number of events for each population

5) for each population, divide the sum of the expected number of events by the total standard population to obtain the age-adjusted rate

Note: the adjusted rates are hypothetical and not observed in the populations. They are almost always used for comparison purposes (i.e. using the ratio of age-adjusted rates to compare one population to another)

Question 15

How do you interpret differences between crude and age-adjusted rates?

Answer 15

When conducting direct age-adjustment, the standard population typically has a different age composition from that of population(s) being compared.

Assuming outcome (i.e. mortality) is associated with older ages… When one of the populations of interest has a higher proportion of older ages (i.e. it’s weighted with older ages) than the standard population, then the age-adjusted rate for that population of interest, age-adjusted rate will be lower than crude rate.

When one of the populations of interest has a lower proportion of older ages (i.e. it’s weighted with younger ages) than the standard population, then the age-adjusted rate for the population of interest will be higher than the crude rate for that population of interest, age-adjusted rate will be higher than crude rate.

Question 16

practice problem

Answer 16

Question 17

When do you use indirect (age) adjustment?

Answer 17

Usually employed when numbers of events (e.g. deaths) for each stratum of the extraneous variable (e.g. age) in the population of interest are not available. It’s used to study mortality in an occupationally exposed population compared to the general population.

Age-specific rates from a standard population are applied to the age strata of the population of interest to obtain the expected number of events in each age strata of the population of interest; compare observed # vs. expected # of events over all age strata. When the event is death, indirect adjustment yields the Standardized Mortality Ratio (SMR)

Question 18

How do you calculate Standardized Mortality Ratio (SMR)?

Answer 18

[(observed number of deaths) / (expected number of deaths)]*100

1) identify the total number of observed events in the population of interest

2) identify age-specific event rates from known comparison (standard) population

3) apply age-specific event rates from comparison (standard) population to age-specific strata of population of interest to obtain expected number of deaths for each age-specific strata of population of interest

4) sum the expected number of events over all age strata for the population of interest

5) divide the number of observed events by the number of expected events and then multiply by 100 to obtain the SMR

An AMR > 100 indicates the observed number of deaths is greater than the expected number of deaths.
An AMR < 100 indicates the observed number of deaths is less than the expected number of deaths.
An AMR = 100 indicates the observed number of events equals the expected number of events

Question 19

Practice Problem

Answer 19

Question 20

What is the Life Table Method?

Answer 20

It describes the pattern of survival in populations. Data is examined in time intervals and cumulative probabilities of survival (or their complement, cumulative mortality) are calculated for each time interval. Time intervals are used because the exact time which events (e.g. deaths) occur are unknown. This can be applied to other outcomes of interest besides death (e.g. time to renal replacement therapy following CKD diagnosis).

Assumptions: No changes have occurred in treatment or survivorship over calendar time of study. For each time interval, the probability of the event (e.g. death) is uniform throughout the interval. Those with incomplete follow-up experience the event of interest with the same probability as those who remain in the study.

Question 21

Practice Problem

Answer 21

Question 22

What is the Kaplan-Meier Method?

Answer 22

• jumps in survival occur when events (e.g. death) occur
• takes advantage of more granular data (o.e. when events occur as opposed to interval when events occur)
• if there are withdrawals between events, they are subtracted from # at risk at the subsequent event time
• probability (P) of event at a given time t = # events at time t/(# under follow-up at time t)
• probability of not having event at a given time t = 1-P
• cumulative probability of not having event by time t = start multiplying as with a life table

Assumptions: No changes have occurred in treatment or survivorship over calendar time of study. Those who are lost to follow-up experience the event of interest with the same probability as those who remain in the study