Common odds ratio

Common Clinical Trial Graphs 200 Scatter Plot 200 Line Plot 201 Bar Chart 202 Box Plot 203 Odds Ratio Plot 203... 

Chapter 3 together with testing hypotheses and the (dreaded ) p-value. Common statistical tests for various data types are developed in Chapter 4 which also covers different ways of measuring treatment effect for binary data, such as the odds ratio and relative risk. 

Benzodiazepines (BZDs) are commonly prescribed in the psychiatric practice to treat anxiety, insomnia, and unpleasant side effects associated with other psychotropic agents. While early case-control studies found that maternal BZD exposure increased the risk of cleft lip and cleft palate, a recent meta-analysis (Dolovich et ah, 1998) examining pooled data from cohort studies published between 1966 and 1998 found no association between antenatal BZD exposure and oral cleft or other major malformations. However, when examining case-control studies published during this period, the authors did find a small, but significant, odds ratio of... 

In the Montreal case-control study carried out by Siemiatycki (1991) (see monograph on dichloromethane in this volume), the investigators estimated the associations between 293 workplace substances and several types of cancer. Isocyanates were one of the substances, and it was stated that the most common form in this study was toluene diisocyanates. The main occupations to which isocyanate exposure was attributed in this study were motor vehicle refinishers, motor vehicle mechanics and foundry workers. Only 0.8% of the study subjects had ever been exposed to isocyanates. For most types of cancer examined (oesophagus, stomach, colon, rectum, pancreas, prostate, bladder, kidney, skin melanoma, lymphoma), there was no indication of an excess risk due to isocyanates. For lung cancer, in the population subgroup of French Canadians (the majority ethnic group in this region), based on 10 cases exposed at any level, the odds ratio was 2.2 (90% CI, 0.9-5.3). [The interpretation of the null results has to take into account the small numbers and presumably low exposure levels. Workers had multiple exposures.]... 

After a study has been conducted, the data are analyzed to see whether an association is observed. This section briefly defines the two major measures of occurrence (incidence and prevalence) and discusses measures of associations in detail, focusing on the most common measures used in the epidemiological study designs described above, such as correlation coefficient, relative risk, and odds ratio. 

First, this chapter will describe a conceptual framework to illustrate the special challenges posed because exposures assessed for epidemiologic studies must be relevant to the health outcome under investigation. Secondly, some of the most commonly applied epidemiological study designs will be introduced, with special emphasis on exposnre assessment issnes associated with the design. Thirdly, some widely applied exposure assessment approaches will be introduced, ranging from qualitative classifications of exposure to quantitative exposure assessment of pesticide concentrations. The influence of measurement error on measures of association between exposure and disease, such as the slopes of exposure-response relationships and risk or odds ratios, will be briefly reviewed. Finally, exposure proxies used in case-control studies of chronic effects of pesticide exposure will be reviewed and the concepts introduced earlier will be applied. 

Risk and odds are two terms used to express, in numbers, the frequency of association of a risk factor with a disease. Risk ratio (relative risk) and odds ratio are two terms that compare the rate of occurrence of a disease in persons exposed and not exposed to a specific risk factor. Odds ratio is commonly used in case-control studies. Risk ratio tends to be used in cohort studies, "Risk" and "risk ratio" are concepts that are easy to grasp, while "odds" and "odds ratio" are somewhat more difficult. 

The odds ratio (OR) is commonly used in epidemiological and clinical studies to express the relationship between two populations [12,13]. It is defined as... 

Odds ratios are one of the most common statistics cited from logistic regression analyses. 

A commonly reported statistic for logistic regression models is the odds ratio. An odds ratio, or the ratio of the odds for x = 1 to the odds for x = 0, may be calculated (14) as... 

In a proportionate mortality or proportionate incidence study, one compares the proportion of deaths or incident cases due to a condition of interest with that expected based on deaths or incident cases in an external, usually the general, population. When the proportions of causes of death are compared, the ratio is known as the proportionate mortality ratio (PMR). The comparison of proportions of incident cases is known as the proportionate incidence ratio (PIR). Proportional measures can be misleading since a decrement in the proportion of deaths or incident cases due to a particular cause will de facto lead to an increase in the proportion due to another cause. The commonly held view with regard to PMRs is that they are good approximations to SMRs from cohort studies when the cohort s all-canses combined SMR is equal to 1.0 (Checkoway et al. 1989). The odds ratio has also been used as a measure of association in PMR or PIR studies and may be a more appropriate measure of association for evalnating proportional measures (WHO 1999). 

In this model, we assume homogeneous variance across all k treatments. For a three-arm trial, having two 5,2, we would further need to consider correlation between them. In this case, we would assume the 8 vector follows a mulfivari-ate normal distribution, typically with common correlation of 0.5 between two log odds ratios (a consequence of the usual assumption of consistency between direct and indirect evidence), as suggested by Lu and Ades (2006). 

Several methods have been found adequate for low event count situations. For odds ratio summary measures, the Mantel-Haenszel, Peto, and exact methods appear to work well. The commonly used method in meta-analysis based on inverse variance weights does not perform well in low event count situations because the weights are not stable with low event counts. The Mantel-Haenszel risk difference appears to work well as well. This method has the added benefit in that unlike the methods for the odds ratio, this... 

The three most commonly used measures to compare two treatment groups are risk difference, risk ratio, and odds ratio. The estimate 0, and the Mantel-Haenszel estimate of Wj can be calculated for each measure as follows (Cochran, 1954 Mantel and Haenszel, 1959 Nurminen, 1981 Tanone, 1981) ... 

A case-control study may be nested within a cohort smdy, i.e. cases and controls are all drawn from a clearly defined cohort. This is an efficient design which is now commonly used in pharmacoepidemiology. Attempts are made in the design and analysis to nimirnise possible biases, and to identify and adjust for confounding factors. Typically a cohort study will measure both absolute and relative risks whereas a case-control study will usually only measure odds ratios which generally approximate to relative risks, hi both cases the data may be summarised in two-by-two tables, as shown in the following examples ... 

---