.-
help for ^logitem^                               (SJ5-3: sg139_1; STB-55: sg139)
.-

Logitistic regression when binary outcome is measured with uncertainty 
----------------------------------------------------------------------

	^logitem^ depvar [indepvars] [^if^ exp] [^in^ range]^, sens(^sensvar | #^)^
                 ^spec(^specvar | #^)^ [ ^l^evel^(^#^)^ ^r^obust ^nolog noor ite^rare^(^#^)^ 
				     ^tol^erance^(^#^) ltol^erance^(^#^)^]

The syntax of @predict@ following ^logitem^ is

	^predict^ [type] newvarname [^if^ exp] [^in^ range] [^,^ statistic]

where statistic is

	^p^            predicted probability of a positive outcome; the default
	^xb^           linear prediction
	^stdp^         standard error of the linear prediction
    (*) ^n^umber       sequential number of the covariate pattern
							covariate pattern)

Unstarred statistics are available both in and out of sample; type "^predict^ ...
^if e(sample)^ ..." if wanted only for the estimation sample.  Starred statistics
are calculated for the estimation sample even when "^if e(sample)^" is not
specified.


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

^logitem^ uses an EM algorithm to estimates a maximum-likelihood logit
regression model when the outcome variable is measured with an imperfect test
of known sensitivity and specificity.

The method allows the sensitivity and specificity to vary across observations.

Options
-------

^sens(^varname | #^)^ especifies the value or the sensitivity variable.

^spec(^varname | #^)^ especifies the value or the specificity variable.

^noor^ reports the estimated coefficients instead of odds ratios.  This option
    affects how results are displayed, not how they are estimated.  ^noor^ may be
    specified at estimation or when redisplaying previously estimated results.

^nolog^ supresses the output of the iteration log.

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

^robust^ specifies the Huber/White/sandwich estimator of variance is to be used
    in place of the traditional calculation; see ^[U] 23.11 Obtaining robust^
    ^variance estimates^.  ^robust^ combined with ^cluster()^ allows observations
    which are not independent within cluster (although they must be independent
    between clusters).

^iterate(^#^)^, ^ltolerance(^#^)^, and ^tolerance(^#^)^ specify the definition of
    convergence.  ^iterate(16000) tolerance(1e-6) ltolerance(0)^ is the default.

	Convergence is declared when

                ^mreldif(^b_i+1, b_i^)^ <= ^tolerance()^
           or   ^reldif(^lnL(b_i+1), lnL(b_i)^)^ <= ^ltolerance()^

    (See help functions for information on the ^reldif()^ and ^mreldif()^
    functions.)  In addition, iteration stops when i = ^iterate()^; in that
    case, results along with the message "convergence not achieved" are
    presented.  The return code is still set to 0.


Options for @predict@
-------------------

^p^, the default, calculates the probability of a positive outcome.

^xb^ calculates the linear prediction.

^stdp^ calculates the standard error of the linear prediction.

^number^ creates newvar containing the sequential number of the covariate
    pattern.  Observations with the same covariate pattern have the same
    ^number^.  Observations not used in estimation have the prediction set
    to missing.  This calculation is automatically restricted to the
    estimation sample.

Examples
--------
	. ^logitem testres exposure, sens(.8) spec(.9)^
	. ^logitem low age lwt, sens(.8) spec(.9) nlog^
	. ^logitem low age lwt, sens(.8) spec(.9) nolog noor^
	. ^logitem , nolog noor^


Authors
-------

         Mario Cleves
         Stata Corporation
         email: mcleves@@stata.com

         Alberto Tosetto
         S. Bortolo Hospital, Vicenza, Italy
         email: tosetto@@hemato.ven.it

        
Also see
--------
    STB:  STB-55 sg139
 Manual:  ^[U] 23 Estimation and post-estimation commands^,
	  ^[U] 29 Overview of model estimation in Stata^,
	  ^[R] logistic^
On-line:  help for @est@, @postest@, @logit@