**[R] epitab** -- Tables for epidemiologists (mhodds)

__Syntax__

**mhodds** *var_case* *expvar* [*vars_adjust*] [*if*] [*in*] [*weight*] [**,**
*mhodds_options*]

*mhodds_options* Description
-------------------------------------------------------------------------
Options
**by(***varlist* [**,** __mis__**sing**]**)** stratify on *varlist*
__b__**inomial(***varname***)** number of subjects variable
__c__**ompare(***v_1***,*** v_2***)** override categories of the control variable
__l__**evel(***#***)** set confidence level; default is **level(95)**
-------------------------------------------------------------------------
**fweight**s are allowed; see weight.

__Menu__

**Statistics > Epidemiology and related > Tables for epidemiologists >**
**Ratio of odds of failure for two categories**

__Description__

**mhodds** is used with case-control and cross-sectional data. It estimates
the ratio of the odds of failure for two categories of *expvar*, controlled
for specified confounding variables, *vars_adjust*, and tests whether this
odds ratio is equal to one. When *expvar* has more than two categories but
none are specified with the **compare()** option, **mhodds** assumes that *expvar*
is a quantitative variable and calculates a 1-degree-of-freedom test for
trend. It also calculates an approximate estimate of the log odds-ratio
for a one-unit increase in *expvar*. This is a one-step Newton-Raphson
approximation to the maximum likelihood estimate calculated as the ratio
of the score statistic, *U*, to its variance, *V* (Clayton and Hills 1993,
103).

__Options__

+---------+
----+ Options +----------------------------------------------------------

**by(***varlist* [**,** **missing**]**)** specifies that the tables be stratified on
*varlist*. Missing categories in *varlist* are omitted from the
stratified analysis, unless option **missing** is specified within **by()**.
Within-stratum statistics are shown and then combined with
Mantel-Haenszel weights.

**binomial(***varname***)** supplies the number of subjects (cases plus controls)
for binomial frequency records. For individual and simple frequency
records, this option is not used.

**compare(***v_1***,***v_2***)** indicates the categories of *expvar* to be compared; *v_1*
defines the numerator and *v_2*, the denominator. When **compare()** is
not specified and there are only two categories, the second is
compared to the first; when there are more than two categories, an
approximate estimate of the odds ratio for a unit increase in *expvar*,
controlled for specified confounding variables, is given.

**level(***#***)** specifies the confidence level, as a percentage, for confidence
intervals. The default is **level(95)** or as set by **set level**.

__Examples__

Setup
**. webuse bdesop**

Calculate the odds ratio for the effect of alcohol controlled for age
**. mhodds case alcohol agegrp [fw=freq]**

Same as above, but perform the calculation by levels of tobacco
consumption
**. mhodds case alcohol agegrp [fw=freq], by(tobacco)**

Calculate the odds ratio for the effect of tobacco controlled for age by
levels of alcohol consumption
**. mhodds case tobacco agegrp [fw=freq], by(alcohol)**

Create a new variable with levels corresponding to all combinations of
alcohol and tobacco consumption
**. egen alctob = group(alcohol tobacco)**

Calculate the odds ratio for the effect of the highest level of alcohol
and tobacco consumption versus the lowest
**. mhodds case alctob [fw=freq], compare(16,1)**

__Stored results__

**mhodds** stores the following in **r()**:

Scalars
**r(p)** two-sided p-value
**r(or)** odds ratio
**r(lb_or)** lower bound of CI for **or**
**r(ub_or)** upper bound of CI for **or**
**r(chi2_hom)** chi-squared test of homogeneity
**r(df_hom)** degrees of freedom for chi-squared test of homogeneity
**r(chi2)** chi-squared

__Reference__

Clayton, D. G., and M. Hills. 1993. *Statistical Models in Epidemiology*.
Oxford: Oxford University Press.