Korean J Fam Med Search

CLOSE


Korean J Fam Med > Volume 36(3); 2015 > Article
Park: Comments on Statistical Issues in May 2015
In this section, we explain the FIRTH option in SAS PROC LOGISTIC procedure which used to solve the numerical problems such as non-convergence in estimating regression coefficients, unreasonably large standard errors, and wide ranges of confidence intervals when there are some zero cells in contingency tables, which appeared in the article titled, "Results of an inpatient smoking cessation program: 3-month cessation rate and predictors of success," published in March 2015 by Kim et al.1)

COMPLETE SEPARATIONS

The complete separation occurs when all subjects are completely classified into each response group according to the values of the explanatory variable(s). Consider the following data set (Table 1).2)
The estimated odds ratio (OR) using X=0 as the reference is infinite [OR=(a/b)/(c/d)]=(5/0)/(0/5)=25/0]. Note that if either (b) or (c) is equal to zero, then the OR is undefined. We also obtain an infinite value of standard error for ORSEOR=OR1a+1b+1c+1d=25/0×15+10+10+15=infinite×infinitekjfm-36-154-e001.jpg. Also note that if any one of the four cell frequencies is equal to zero, then SE (OR) is undefined. Gart and Zweifel3) suggested improved estimates of OR and SE (OR) which are calculated after adding 0.5 to each cell. According to their suggestion, we obtain OR'=(5.5/0.5)/(0.5/5.5)=121 and SE (OR')=252.76.

SAS PROC FREQ PROCEDURE

We can obtain the OR and its 95% confidence limits for Table 1 using the following commands in PROC FREQ procedure (here, we do not consider the small sample size problem).
DATA CS; INPUT X Y Z; CARDS;
1 1 5
1 0 0
0 1 0
0 0 5
;
PROC FREQ DATA=CS; TABLES X*Y/ALL; WEIGHT Z; RUN;

<OUTPUT>
Odds ratio = 121.000, 95% confidence limits = 2.0169 ~
7259.1773
Other options, RELRISK or MEASURES in TABLES statement cannot calculate OR and SE (OR) in this case.

SAS PROC LOGISTIC PROCEDURE

We can also obtain the OR and its 95% confidence limits for Table 1 using the following commands in PROC LOGISTIC procedure.
PROC LOGISTIC DESCENDING DATA=CS; MODEL Y=X/
FIRTH; WEIGHT Z; RUN;

<OUTPUT>
Odds ratio = 120.976, 95% confidence limits = 1.364 ~ >999.999
We can see that both procedures give the same ORs, but somewhat different 95% confidence limits. However if we do not use the FIRTH option in MODEL statement, then values of OR and upper 95% confidence limit are infinites.

CONFLICTS OF INTEREST

CONFLICT OF INTEREST: No potential conflict of interest relevant to this article was reported.

References

1. Kim SH, Lee JA, Kim KU, Cho HJ. Results of an inpatient smoking cessation program: 3-month cessation rate and predictors of success. Korean J Fam Med 2015;36:50–59. PMID: 25802686.
crossref pmid pmc
2. Park YG. Comments on statistical issues in July 2013. Korean J Fam Med 2013;34:293–294. PMID: 23904960.
crossref pmid pmc
3. Gart JJ, Zweifel JR. On the bias of various estimators of the logit and its variance with application to quantal bioassay. Biometrika 1967;54:181–187. PMID: 6049534.
crossref pmid
Table 1

A contingency table with a complete separation

kjfm-36-154-i001.jpg


ABOUT
ARTICLE CATEGORY

Browse all articles >

BROWSE ARTICLES
AUTHOR INFORMATION
Editorial Office
Room 2003, Gwanghwamun Officia, 92 Saemunan-ro, Jongno-gu, Seoul 03186, Korea
Tel: +82-2-3210-1537    Tax: +82-2-3210-1538    E-mail: kjfm@kafm.or.kr                

Copyright © 2019 by Korean Academy of Family Medicine. All rights reserved.

Developed in M2community

Close layer
prev next