.-
help for ^lrreg^
.-

Diagnostic Likelihood Ratio regression
--------------------------------------

    ^lrreg^ testvar disvar eqlrp [eqlrn], [^if^ exp] [^in^ range]
         [^, e^form ^r^obust ^cl^uster^(^varname^) l^evel^(^#^)^ maximize_options]


^lrreg^ shares the features of all estimation commands; see help @est@.

^lrreg^, typed without arguments, redisplays previous results.


Description
-----------

^lrreg^ estimates maximum-likelihood diagnostic likelihood ratio (DLR) regression
models for binary medical diagnostic tests.
    testvar : the binary test outcome variable where
        testvar==1 ==> a positive test outcome
        testvar==0 ==> a negative outcome
    disvar : the binary disease variable where
        disvar==1 ==> the presence of disease
        disvar==0 ==> the absence of disease
    eqlrp: the name of a previously (user) defined equation
        (see help @eq@) to be estimated in association with the DLR+.
    eqlrn: an equation to be estimated in association with the DLR-.

        If a single equation name is included, the same equation will
        be used for estimation of both the DLR+ and DLR-.
        In this case, subsequent access to coefficient estimates will
        require reference to new equation names LRpos and LRneg.

Options
-------
^eform^ reports the exponentiated estimated coefficients, i.e., exp(b) rather
    than b.  Standard errors and confidence intervals are similarly
    transformed.  This option affects how results are displayed, not
    how they are estimated.  ^eform^ may be specified at estimation or when
    redisplaying previously estimated results.  Note that this transformation
    does not result in predicted DLR's, rather the relative DLR's for
    conditions represented by the associated covariate are obtained.

^robust^ specifies that the Huber/White/sandwich estimator of variance is to
    be used in place of the traditional calculation; see ^[U] 26.10 Obtaining
    robust variance estimates^.  ^robust^ combined with ^cluster()^ allows
    observations which are not independent within cluster (although they may
    be independent between clusters).

^cluster(^varname^)^ specifies that the observations are independent across groups
    (clusters) but not necessarily within groups.  varname specifies to which
    group each observation belongs; e.g., ^cluster(personid)^ in data with
    repeated observations on individuals.  See ^[U] 26.10 Obtaining robust^
    ^variance estimates^.  ^cluster()^ can used with @pweight@s to produce estimates
    for unstratified cluster-sampled data, but see help @svylogit@ for a command
    especially designed for survey data.  Specifying ^cluster()^ implies ^robust^.

^level(^#^)^ specifies the confidence level, in percent, for calculation of
    confidence intervals of the odds ratios.


Remarks
-------

The DLR+ can be thought of as the odds of disease given a positive test outcome
relative to the odds of disease in the absence of test information.  This
regression method allows for specification of different regression models for
the postive and negative DLR's.

At least one equation must be defined and included as an argument to ^lrreg^.
In order to estimate the DLR+ and/or DLR- without covariate effects (i.e.
constant term only), define an empty equation.

    eg.
        .^eq lreqn:^
        .^lrreg test disease lreqn^

Predicted DLR's and associated confidence intervals may be obtained for
particular covariate values using @lincom@ with the ^rr^ option.
Similarly, @predict@ may be used to generate estimated linear
predictor.

Examples
--------
    . ^eq lrpos: age gender^
    . ^eq lrneg: age^
    . ^lrreg test disease lrpos lrneg, e cluster(subj_id)^
    . ^lincom [lrpos]_cons + [lrpos]gender, rr^


Authors
-------

    Wendy Leisenring:   wendy@@fhcrc.org
    Gary Longton:       glongton@@fhcrc.org


References
----------

Leisenring W, and Pepe M. Regression Modelling of Diagnostic
Likelihood Ratios for the Evaluation of Medical Diagnostic Tests.
Biometrics 54:444-52.  1998.


Also see
--------

 Manual:  ^[U] 26 Estimation and post-estimation commands^
          ^[U] 35 Overview of model estimation^
On-line:  help for @est@, @eq@, @predict@, @lrtest@, @lincom@, @vce@