HEALTHCARE RESEARCH METHODS: Case-Control Studies and Cohort Studies

PHC215

By Dr. Khaled Ouanes Ph.D.

E-mail: [email protected]

Twitter: @khaled_ouanes

INTRODUCTION TO

HEALTHCARE RESEARCH

METHODS

Case-Control Studies

Case-control studies are often used to

identify factors that may contribute to a

medical condition by comparing subjects

who have that disease (the cases) with

other individual patients who do not have

the condition/disease but are otherwise

similar (the controls).

Individual participants in a case-control studyare selected for inclusion in the study based on their disease status

Cases = participants with the disease of interest

Controls = participants without the disease

Both cases and controls are asked the same set of questions about past exposures.

Framework for a Case-Control Study

Use

A case-control study is often the best study

approach when the disease or the

condition is relatively uncommon and

infrequent where a study of the general

population is unlikely to yield more than a

few cases.

Finding Cases

A key initial step is identifying an appropriate

and accessible source of individuals with the

disease of interest

Hospitals, specialty clinics, physicians’ offices,

public health agencies, disease registries, and

disease support groups may be able to assist

researchers in identifying individuals who are

likely to meet the study’s case definition

Case Definition

All cases must have the same disease,

disability, or other health-related condition

A case definition should specify exactly what

characteristics must be present or absent for a

person to be deemed a case

Finding Controls

Controls must be reasonably similar to

cases except for their disease status

The inclusion and exclusion criteria for

cases that do not specifically relate to

the disease should also apply to controls For example, if cases must be males between older than 60 years of

age, controls must also be males in this age group.

Finding Controls

Controls may be recruited from, among other

sources:

Friends and relatives of cases

Hospital or clinic patients without the disease of

interest

The general population

Matching

There are 3 basic options for matching

cases and controls:

no matching

frequency (group) matching

matched-pairs (individual) matching

Many case-control studies use no matching.

They assume that similar inclusion and

exclusion criteria for cases and controls will

result in case and control populations that

have similar distributions according to age

group, gender, socioeconomic status, and

other characteristics.

Matching

The goal of frequency (group) is to recruit a

control population that is similar to the case

population.

Individual cases are not tied to individual

controls during analysis.

Matching

Example: For each hospital case a

researcher identifies select one control from

the hospital registration files who was

admitted the same week as the case, who is

the same gender as the case, and who is

within ± 0-3 years of the age of the case.

Matching

In matched-pairs (individual) matching, each case

is personally linked to a particular individual

control.

This approach is fairly common in genetic studies,

in which a case is linked to a genetic sibling or

other close genetic relative for analysis.

This kind of matched-pairs approach requires a

special type of analysis.

Matching

Limitations of Matching

The variables used as matching criteria

cannot be considered as exposures during

analysis. Example: If cases and controls are matched based age, then the cases

and controls in the study will have the same mean age, even if in the

general population cases tend to be older than non-cases.

It can be difficult to find controls who meet

all of the matching criteria when there are

many matching characteristics.

Misclassification Bias

All participants must be asked questions that confirm whether each is a case, a control, or neither of them

Adhering to strict definitions for what constitutes a case and what constitutes a control minimizes the risk of misclassification bias

Recall Bias

Recall bias occurs when cases and controls systematically have different memories of the

past

Cases may have more vivid memories than

controls of participation or lack of participation

in activities perceived to be risky or beneficial

because they are searching for explanations for

their illnesses

Analysis: Odds

Odds are the most familiar from their

connection with betting

A horse with an equal chance of winning a

race (50% likely to win) or of losing a race

(50% likely to lose) is said to have “even

odds,” or odds of 1 (50%/50%)

A case-control study compares the chance of having

had a particular exposure to not having had it

If 50% of the participants in a study report a history of

exposure and 50% report no exposure history, then the

odds of exposure are 50%/50%, or 1

If 25% report having the exposure and 75% do not, then

the odds are 25%/75%, or 0.33

If 2% report being exposed in the past and 98% report

no exposure, then the odds are 2%/98%, or 0.02

Analysis: Odds

Analysis: Odds

Odds ratio (OR) = the ratio of the odds of exposure in cases to the

odds of exposure in controls

Analysis: Odds

Analysis:

Odds

Ratios

OR = 1: the odds of exposure are the same

for cases and controls

OR > 1: cases have higher odds of exposure

than controls, implying that the exposure

was risky

OR < 1: cases have lower odds of exposure

than controls, implying that the exposure

was protective

Analysis: Odds

Analysis: OR & 95% CI

If the entire 95% confidence interval is less than 1,

then the OR is statistically significant, and the

exposure is deemed protective “Cases had greater odds of exposure than controls”

If the entire 95% confidence interval is greater than

1, then the OR is statistically significant, and the

exposure is deemed risky “Cases had lesser odds of exposure than controls.”

95% confidence interval (95% CI) overlaps

OR = 1 The lower end of the confidence interval is less than 1, suggesting

protection

The higher end of the confidence interval is greater than 1, suggesting risk

Conclusion: The OR is not statistically significant, and the exposure and

disease are deemed to have no association

Analysis: OR & 95% CI

Interpretation of the Odds Ratio Based on Its

95% Confidence Interval

Examples of Odds Ratio Calculations

Matched Case-Control Studies

Individually-matched case-control studies require the calculation of a matched-pairs odds ratio that uses a special kind of 22 table

A ratio of the number of times the case in a pair was exposed and the control was not to the number of times the control in a pair was exposed and the case was not provides an estimate for a special type of odds ratio

Matched-

Pairs Odds

Ratio

Key Characteristics of Case-

Control Studies

Case-Control Studies

Cohort Studies

a cohort is a group of individuals or subjects who have shared a

particular event together during a particular time span.

So, for a research project, a Cohort is going to be any group of similar people followed through time together.

A cohort study follows participants through time to calculate the rate at which new (incident) disease occurs and to identify risk factors for the disease

All cohort studies have at least 2 measurement

times:

An initial survey that determines the baseline

exposure and disease status of all participants

One or more follow-up assessments thatdetermine how many participants have

developed a new (incident) disease since the

initial examination

Framework for

a Cohort Study

Types of Cohort Studies

Cohort studies take many forms.

The 3 most important categories are:

Retrospective cohort

Prospective cohort

Longitudinal cohort

Retrospective & Prospective

Both retrospective and prospective cohort

studies recruit participants based on their exposure status

1st group of participants is recruited because

they are known to have had a given exposure

2nd group is recruited because they are

known not to have been exposed

Retrospective cohort studies use baseline information collected at some point in the pastand follow the cohort to another point in the past or to the present

Prospective cohort studies collect baseline data about exposures and outcomes in the present and follow the cohort to some point in the future


Times of Baseline and Follow-Up

Data Collection for Cohort

Studies

Recruiting based on exposure status makes

retrospective and prospective cohort studies the

optimal study approaches for uncommon exposures.

Indeed, the goal of cohort studies is to examine

incident disease, retrospective and prospective

cohort studies must be able to demonstrate that the

outcome of interest was not present in any members

of the cohort at baseline.


The members of the two comparison groups

for both prospective and retrospective of

studies should be similar except for their exposure status.


Examples:

Recruit industrial workers exposed to a certain

chemical and similar workers in a plant that does

not use that chemical

Recruit children with high blood lead levels and

low blood lead levels who attend the same

elementary school


Longitudinal Studies

Longitudinal cohort studies recruit participants based on their membership in a well-defined source population, then follow them forward in time

Individual participants are assessed at baseline for several exposures and diseases.

Then they are followed forward in time to determine the incidence rate for one or more outcomes of interest.

Examples of populations for a longitudinal study:

All the residents of one town

A representative sample of members of one professional organization

A representative group of students recruited from the same university


In a longitudinal study with a fixed population, all participants start the study at the same time and

no one is allowed to join later

In a study with a dynamic population, participants

are recruited using rolling admission and

replacement of dropouts For dynamic populations, the time to follow up is usually based on individual

participants’ dates of enrollment rather than on a fixed calendar date



Retention

For prospective and longitudinal studies, loss of participants to follow-up before the end of the study period is a major concern.

Researchers must develop strategies that minimize the burden of participation and that maximize interest in continuing to participate.

Information Bias

All participants must complete the same assessments of exposure and disease at

baseline and follow-up to prevent the information bias that might result when

exposed participants are more thoroughly examined for disease than unexposed

participants.

Analysis: Incidence

Incidence rate = the number of new cases of disease in a population during a specified period of time divided by the total number of persons in the population who were at risk during that period

Individuals who already have the disease of interest at the start of the study period are not at risk of getting new disease, so they are removed from the denominator

Analysis: Incidence

Analysis: Person-Years

Some cohort studies use person-time as a denominator

rather than simply counting persons

Person-time is a way of accounting for different

individuals in the study population being observed for

different lengths of time

Example: Over 4 years in a dynamic study, 10

participants may contribute 33 person-years of

observation

Person-Year Analysis

Analysis: Attributable Risk

Excess risk = attributable risk (AR) = the absolute difference in the incidence rate between the exposed population and the unexposed population

Example: If 10% of the unexposed and 15% of the exposed became ill during the study period, then the excess risk in the exposed was 15% – 10% = 5%

This number represents the additional risk of disease in the exposed that can be attributed to the exposure

Analysis: AR%

Attributable risk percent (AR%) = the proportion of incident cases among the exposed that are

due to the exposure

Example: If 10% of the unexposed and 15% of

the exposed became ill during the study

period, then the AR% is 5% 15% = 33%

1/3 of the cases of disease in the exposed

could have been prevented if the exposure

was removed

Attributable

(Excess)

Risk

Analysis: RRsRate ratio (RR) = relative rate = risk ratio = relative risk = ratio of the incidence rate among the exposed to

the incidence rate in the unexposed

Analysis: RRs

RR = 1: the incidence rate was the same in the

exposed and in the unexposed, so the exposure is not

associated with the disease

RR > 1: then the incidence rate was higher in the

exposed than in the unexposed, so the exposure was

risky

RR < 1: the incidence rate was lower in the exposed

than in the unexposed, so the exposure was

protective

Analysis: RR & 95% CI

If the entire 95% confidence interval is less

than 1, then the RR is statistically significant,

and the exposure is protective in the study

population

If the entire 95% confidence interval is

greater than 1, then the RR is statistically

significant, and the exposure is a risk factor

for disease in the study population

95% confidence interval (95% CI) overlaps RR = 1 The lower end of the confidence interval is less than 1, suggesting protection

The higher end of the confidence interval is greater than 1, suggesting risk

Conclusion: The RR is not statistically significant, and the exposure and disease are

deemed to have no association

Analysis: RR & 95% CI

Interpretation of the Rate Ratio Based on

Its 95% Confidence Interval

Examples of

Rate Ratio

Calculations

Key Characteristics of Cohort

Studies

Key Characteristics of Cohort

Studies

PHC215

By Dr. Khaled Ouanes Ph.D.

E-mail: [email protected]

Twitter: @khaled_ouanes

HEALTHCARE RESEARCH METHODS

Based on the textbook of introduction to health research methods – K.H. Jacobsen

HEALTHCARE RESEARCH METHODS: Case-Control Studies and Cohort Studies

Education