st: RE: RE: Case control study

["Cornelius Nattey" <cornelius.nattey@nioh.nhls.ac.za>]

Dear Paul,
Thank you soo much. That was very insightful and thanks to other contributors.
I will use frequency matching case control approach and adjust for age and gender.


Regards,

Cornelius Nattey

Medical Scientist: Epidemiology and Surveillance
National Health Laboratory Service
National Institute for Occupational Health
Office: 011 712 6438
Fax: 086 604 1214
Cell: 079 631 5857
Email: cornelius.nattey@nioh.nhls.ac.za
www.nhls.ac.za  www.nioh.ac.za




-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Visintainer, Paul
Sent: 27 June 2011 10:06 PM
To: 'statalist@hsphsun2.harvard.edu'
Subject: st: RE: Case control study

Cornelius,

It's not clear whether you intend to conduct "pairwise" matching on age and gender or, rather, a real frequency matching (likely, you're considering both).  In pairwise matching, control selection is contingent on each case characteristic (when a case presents, a control is selected from the pool of controls based on eligibility criteria, matching criteria, and time). Each "pair" has one case and at least one control.  A "pair" ID should generated which uniquely identifies each pair.  Conditional logistic regression is best suited for this approach because it models the "pair" strata.

Frequency matching is much looser. Controls can be "anticipated" based on the expected distribution of the cases.  For example, a clinic may review its cases over a two-year period to see the distribution of cases by age, and gender.  They can develop a grid estimating the expected number of cases by age group and gender that may present each month, say.  They can select controls based on the same grid.  They don't worry about "pairing", or even about insuring that the same number cases and controls are selected each month.  They just want similar (not identical) frequency distributions of cases and controls on age, gender, and time.  Since there isn't any formal "pairing" of cases and controls, no pair ID is generated.  For this approach, unconditional logistic regression controlling for age and gender is fine.  It's important to note that, even though LR will provide coefficients for age and gender, they shouldn't be interpreted, since the distribution of controls was arti!
 ficially modified by the investigator.

- Paul

________________________________________________

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Cornelius Nattey
Sent: Monday, June 27, 2011 9:21 AM
To: statalist@hsphsun2.harvard.edu
Subject: st: Case control study

Hi All,
I am putting together a protocol on lung Cancer and occupational categories a case control study.
I will match by age and gender. I am sure whether to use frequency matched cases control, unconditional logistic analysis or conditional logistics analysis
Can anyone throw more light on the above terminologies and the best approach for this analysis?

Regards,  


Cornelius Nattey

Medical Scientist: Epidemiology and Surveillance
National Health Laboratory Service
National Institute for Occupational Health
Office: 011 712 6438
Fax: 086 604 1214
Cell: 079 631 5857
Email: cornelius.nattey@nioh.nhls.ac.za
www.nhls.ac.za www.nioh.ac.za



N???,{kz?v)?+???j???y^??-???O??^w?a?K~?^j?a???vz-?,z???Z?*'jQ?j[a-??h?$????z?+??&jv?zg????z)????*'?{a????????(???z{bjV?v+"??u?l?W???+??u??A?
z?-??b???j)[??w???????h?,?w???????^?{^????"?'??????rZ,???r?r?x?v+-?&?*'????-???jG???&??^v)?zX??K??b?????!???y?vf?m???_?)l??e??v???-z???Z????????r??s&?v)?z
?*?w?f?????Z??kj{&??^u?^r??'?\???h???z????w?u??b?+az????&??X???z?h???X??Z?????z?^v'??+??b?z+u?Z????!j?z??j?'??+??my?+v????ky?!??"??u?]z?b??b

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

----------------------------------------------------------------------
Please view our annual report at http://baystatehealth.org/annualreport
 

CONFIDENTIALITY NOTICE: This e-mail communication and any attachments may contain confidential and privileged information for the use of the designated recipients named above. If you are not the intended recipient, you are hereby notified that you have received this communication in error and that any review, disclosure, dissemination, distribution or copying of it or its contents is prohibited. If you have received this communication in error, please reply to the sender immediately or by telephone at 413-794-0000 and destroy all copies of this communication and any attachments. For further information regarding Baystate Health's privacy policy, please visit our Internet site at http://baystatehealth.org.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


N���kz�v)����j���y^����Ǭ��^w�~�j���vz-�,z��Z�*'jQ�[a-�譫h�$�����jv�zg���ߢ���*'�{a���������jV�v+"�קu�W��+��u��Ǟ�����)[ɩ��Ǣ����,�������{^�ם��'����rZ,���r�r�x�v+-�&�'���ɧ-������&��v)���Ǩ�������y֧vf�m맕��_ֆ)l��e�欢v�������������&�v)���w����ڝ���kj{&��u���'�\������z�޶��w�퉩b�+az����X�ʊ�h�ț�X��Z�����z�^v'�+��b�z+u�Z����!jׯz����y�v����ky�!���קu����b

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/