25.11.2014 Views

Biostatistics

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

12.7 RELATIVE RISK, ODDS RATIO, AND THE MANTEL–HAENSZEL STATISTIC 645<br />

Odds Ratio and Relative Risk Section<br />

Common Original Iterated Log Odds Relative<br />

Parameter Odds Ratio Odds Ratio Odds Ratio Ratio Risk<br />

Upper 95% C.L.<br />

2.1350<br />

2.2683 0.7585 2.1192<br />

Estimate 1.1260 1.1207 1.1207 0.1140 1.1144<br />

Lower<br />

95% C.L.<br />

0.5883<br />

0.5606 0.5305 0.5896<br />

FIGURE 12.7.1 NCSS output for the data in Example 12.7.1.<br />

These data indicate that the risk of experiencing preterm labor when a woman<br />

exercises heavily is 1.1 times as great as it is among women who do not<br />

exercise at all.<br />

We compute the 95 percent confidence interval for RR as follows. By<br />

Equation 12.4.1, we compute from the data in Table 12.7.2:<br />

X 2 455 22<br />

¼ ½ð<br />

Þð199Þ<br />

ð216Þð18Þ<br />

ð40Þð415Þð238Þð217Þ<br />

Š2<br />

¼ .1274<br />

By Equation 12.7.2, the lower and upper confidence limits are, respectively,<br />

pffiffiffiffiffiffiffiffi<br />

1 1:96= :1274<br />

1:1 ¼ :65 and 1:1 ffiffiffiffiffiffiffiffi<br />

1þ1:96=<br />

p<br />

:1274<br />

¼ 1:86. Since the interval includes<br />

1, we conclude, at the .05 level of significance, that the population risk may<br />

be 1. In other words, we conclude that, in the population, there may not be<br />

an increased risk of experiencing preterm labor when a pregnant woman<br />

exercises extensively.<br />

The data were processed by NCSS. The results are shown in Figure<br />

12.7.1. The relative risk calculation is shown in the column at the far right of<br />

the output, along with the 95% confidence limits. Because of rounding errors,<br />

these values differ slightly from those given in the example.<br />

&<br />

Odds Ratio When the data to be analyzed come from a retrospective study, relative<br />

risk is not a meaningful measure for comparing two groups. As we have seen, a<br />

retrospective study is based on a sample of subjects with the disease (cases) and a separate<br />

sample of subjects without the disease (controls or noncases). We then retrospectively<br />

determine the distribution of the risk factor among the cases and controls. Given the results<br />

of a retrospective study involving two samples of subjects, cases, and controls, we may<br />

display the data in a 2 2 table such as Table 12.7.3, in which subjects are dichotomized<br />

with respect to the presence and absence of the risk factor. Note that the column headings in<br />

Table 12.7.3 differ from those in Table 12.7.1 to emphasize the fact that the data are from a<br />

retrospective study and that the subjects were selected because they were either cases or<br />

controls. When the data from a retrospective study are displayed as in Table 12.7.3,<br />

the ratio a=ða þ bÞ, for example, is not an estimate of the risk of disease for subjects with<br />

the risk factor. The appropriate measure for comparing cases and controls in a retrospective<br />

study is the odds ratio. As noted in Chapter 11, in order to understand the concept of

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!