Epidemiology: Measures of Association in Cohort and Case-Control Studies

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/22

flashcard set

Earn XP

Description and Tags

Vocabulary practice flashcards covering measures of association in cohort and case-control studies, statistical tests, confidence intervals, and epidemiological examples.

23 Terms

1
New cards

Measures of Association

Statistical indicators used in epidemiological studies (such as cohort studies and case-control studies) to quantify the relationship and strength of association between exposure to a factor and the occurrence of a disease.

2
New cards

Incidence Rate in the Exposed (IeI_e or R1R_1)

The direct rate of new disease cases occurring among exposed subjects in a cohort study, calculated as Ie=aa+b=aN1I_e = \frac{a}{a + b} = \frac{a}{N_1}.

3
New cards

Baseline Risk (IneI_{ne} or R0R_0)

The incidence rate of disease among non-exposed subjects in a cohort study, calculated as Ine=cc+d=cN0I_{ne} = \frac{c}{c + d} = \frac{c}{N_0}.

4
New cards

Attack Rate (Taux d'attaque)

An incidence rate commonly utilized in acute settings such as foodborne outbreak investigations to evaluate the proportion of individuals who become ill after exposure.

5
New cards

Risk Difference (DRDR)

The excess disease risk directly associated with exposure to the factor under investigation, calculated in a cohort study as DR=Ie−Ine=R1−R0DR = I_e - I_{ne} = R_1 - R_0.

6
New cards

Relative Risk (RRRR)

The ratio of the disease incidence in the exposed group to the disease incidence in the non-exposed group (RR=IeIne=R1R0=a/N1c/N0RR = \frac{I_e}{I_{ne}} = \frac{R_1}{R_0} = \frac{a / N_1}{c / N_0}), which measures the strength of the statistical association in a cohort study.

7
New cards

Protective Factor

An exposure status characterized by a Relative Risk or Odds Ratio strictly less than 1 (RR<1RR < 1 or OR<1OR < 1), indicating that exposure reduces the risk or odds of developing the disease.

8
New cards

Absence of Risk

A finding where the Relative Risk is approximately 1 (RR≈1RR \thickapprox 1) or the Odds Ratio equals 1 (OR=1OR = 1), indicating no association between the exposure and the disease.

9
New cards

Risk Factor

An exposure status characterized by a Relative Risk or Odds Ratio strictly greater than 1 (RR>1RR > 1 or OR>1OR > 1), indicating that the exposure increases the risk or odds of disease occurrence.

10
New cards

Mantel-Haenszel Chi-Square Test

A statistical hypothesis test used to verify whether an observed association is statistically significant and not due to chance, computed as [(ad)−(bc)]2×NM1×M0×N1×N0\frac{[(ad) - (bc)]^2 \times N}{M_1 \times M_0 \times N_1 \times N_0}.

11
New cards

Chi-Square Significance Threshold (5% Risk, 1 df)

The critical statistical table value of 3.843.84 at a 5\text{ %} significance level with 1 degree of freedom (ddl)1\text{ degree of freedom (ddl)}; a calculated test statistic exceeding 3.843.84 confirms that the Relative Risk is significantly different from 1.

12
New cards

Miettinen's Method for Relative Risk

The standard epidemiological method used to calculate the 95\text{ %} confidence interval of the Relative Risk: 95\text{ % CI} = RR^{1 \thicksim \frac{1.96}{\backslash sqrt{\backslash chi^2}}}.

13
New cards

Alcohol and Cirrhosis Cohort Study

A worked cohort example where alcohol consumers had an incidence of 35/2500035 / 25000 versus 1/150001 / 15000 in non-consumers, yielding RR=21RR = 21, \chi2=18.5\backslash chi^2 = 18.5, and a 95\text{ % CI} = [5.24 - 84.09].

14
New cards

HIV and Tuberculosis Cohort Example

A 2-year follow-up cohort study showing an incidence of tuberculosis of 3.7\text{ %} (8/2158 / 215) in HIV+ subjects compared to 0.3\text{ %} (1/2981 / 298) in HIV- subjects, yielding RR=11RR = 11.

15
New cards

Odds (Cote)

The ratio between the probability of an event occurring and the probability of that event not occurring (P1−P\frac{P}{1 - P}), such as rolling a 6 on a die: 1/65/6=15=0.20\frac{1/6}{5/6} = \frac{1}{5} = 0.20.

16
New cards

Odds of Exposure in Cases

The ratio of exposed cases to non-exposed cases in a case-control study, calculated as a/(a+c)c/(a+c)=ac\frac{a / (a + c)}{c / (a + c)} = \frac{a}{c}.

17
New cards

Odds of Exposure in Controls

The ratio of exposed controls to non-exposed controls in a case-control study, calculated as b/(b+d)d/(b+d)=bd\frac{b / (b + d)}{d / (b + d)} = \frac{b}{d}.

18
New cards

Odds Ratio (OROR / Rapport de cote)

The principal measure of association in case-control studies, defined as the ratio of exposure odds in cases to exposure odds in controls: OR=a/cb/d=adbcOR = \frac{a / c}{b / d} = \frac{ad}{bc}.

19
New cards

Rare Disease Assumption

The condition in epidemiology where disease prevalence is low (< 5\text{ %}), resulting in a∼a+ba \thicksim a + b and c∼c+dc \thicksim c + d, which mathematically makes the Odds Ratio (OROR) an accurate estimate of the Relative Risk (RRRR).

20
New cards

Mathematical Relationship Between OR and RR

The formula linking the Odds Ratio to Relative Risk and disease prevalence (PP): OR=RR(1−P)1−(RR×P)OR = \frac{RR(1 - P)}{1 - (RR \times P)}.

21
New cards

Saint André du Gaz Outbreak Investigation (1997)

A case-control study of a foodborne infection linked to mayonnaise where cases had exposure odds of 36/236 / 2 and controls had 9/89 / 8, resulting in OR=16OR = 16 and 95\text{ % CI} = [2.4 - 168.2].

22
New cards

Hydrocarbon Exposure and Bladder Cancer Study (USA, 1984)

A case-control study demonstrating a duration-dependent increase in bladder cancer risk: OR=3.6OR = 3.6 for 1-9 years, OR=5.2OR = 5.2 for 10-19 years, and OR=7.2OR = 7.2 for >19> 19 years compared to <1< 1 year.

23
New cards

Asbestos Exposure and Lung Cancer Cohort Study (USA, 1992)

A cohort study showing a dose-response relationship between asbestos exposure and lung cancer rates: RR=2.0RR = 2.0 for 103−10410^3 - 10^4, RR=4.9RR = 4.9 for 104−10510^4 - 10^5, and RR=8.1RR = 8.1 for >105 fibres/(cm3 per day)> 10^5\text{ fibres}/(\text{cm}^3\text{ per day}) compared to <1000< 1000.