Cheng YW. the Association Between Persistent. American Journal of Epidemiology 2010

American Journal of Epidemiology ª The Author 2010. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permission s, please e-mail: journals.p ermissions@oxfordjour nals.org. DOI: 10.1093/aje/kwp437 Practice of Epidemiology The Association Between Persistent Fetal Occiput Posterior Position and Perinatal Outcomes: An Example of Propensity Score and Covariate Distance Matching Yvonne W. Cheng*, Alan Hubbard, Aaron B. Caughey, and Ira B. Tager * Correspond ence to Dr. Yvonne W. Cheng, Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Reproduc tive Sciences, School of Medicine , University of California, San Francis co, 505 Parnassus Avenue, Box 0132, San Francisco, CA 94143 (e-mail: yvecheng@hotmail.com). Initially submitted November 17, 2008; accepted for publication October 1, 2009. In a retrospective cohort study of 18,880 full-term, cephalic singletons born in San Francisco, California, during 1976–2001, the authors used multivariable logistic regression (MVLR) and propensity score analysis (PSA) to examine the association between persistent fetal occiput posterior (OP) position and perinatal outcomes. The principles and applications of these techniques are compared and discussed. Pregnancies with OP positions at delivery were compared with those with occiput anterior positions. Perinatal outcomes were examined as adjusted odds ratios determined by MVLR and PSA and as risk differences determined by propensity score matched bootstrapping based on covariate distance. Persistent OP position was associated with operative delivery and maternal morbidity. The odds ratio estimates based on PSA were somewhat larger than those obtained with standard MVLR, and the confidence intervals were narrower. When statistical inference was evaluated with the permutation test, the results were more consistent with the PSA. These analyses demonstrate that PSA is likely to provide more precise estimates of exposure associations and more reliable statistical inferences than MVLR. The authors show that PSA can be extended with Mahalanobis distance matching to obtain estimates of risk difference between exposed and unexposed subjects that avoid violations of the experimental treatment assignment (pos- itivity) assumption that is required for valid causal inference. labor, obstetric; pregnancy outcome; propensity score Abbreviations: C I, confidence interva l; OA, occiput anterior; OP, occip ut posterior; OR, odd s ratio; PSA, propensit y score analysis. Editor’s note: An invited commentary on this article and a related article appears on page 000 , and these authors’ response appears on page 000. Per sist ent occ iput pos teri or (OP) pos itio n is the most common malposition in labor (1) and has an estimated prev- alence ranging from 2% to 10% (2–4). Recent studies sug- gest that persistent OP position is associated with prolonged labor and increased number of operative deliveries (2, 4, 5). Some studies propo se that OP position is a manifes tation of intrapartum malrotation rather than persistent malposition (4, 6–8). In this study , we examin ed factors associated with persist ent OP positio n at deliv ery using diffe rent analytical techniq ues, includ ing tradit ional multiv ariable logistic re- gressio n analys is, propensity score analy sis (PSA), and covariate distance matching. Most studies on OP position and birth outcomes are observational in design, and observational studies may be sub- ject to bias from uncontrolled or residual confounding (7). Thus, consideration of confounding bias is fundamental to the des ign, ana lys is, and int erpret ation of obs erv atio nal studies attempting to estimate causal effects. Common statistical methods for controlling for confounding bias include stratification and multivariable regression analysis (8). Al- though stratification is simple to perform, the trade-off is a paucity of data, which may further contribute to biased or ine ffic ient compar ison s (9) . The acc ura cy of the eff ect 1 American Journal of Epidemiology Advance Access published February 5, 2010 b y g u e s t o n M a y 3 1 , 2 0 1 3 h t t p : / / a j e . o x f o r d j o u r n a l s . o r g / D o w n l o a d e d f r o m

Cheng YW. the Association Between Persistent. American Journal of Epidemiology 2010

Documents

Transcript of Cheng YW. the Association Between Persistent. American Journal of Epidemiology 2010