Amos Grünebaum1, Frank Chervenak, Daniel Skupski. 1. Department of Obstetrics and Gynecology Weill Medical College of Cornell University 525 East 68th Street Suite J-130 New York NY 10065 USA. amg2002@med.cornell.edu
Abstract
BACKGROUND: Raw cesarean section rates are often compared among institutions with different patient populations. The purpose of this study was to develop an easily reproducible mathematical model that allows comparisons of cesarean delivery rates across different patient populations and institutions. STUDY DESIGN: We first calculated three institution's cesarean delivery rates for each of ten groups of patients based on age and parity. The population based adjusted total cesarean delivery rate was then calculated for each institution based on the distribution of patients in the 2004 national birth data (n=4,097,029) but using the three institutions' individual cesarean delivery rates. RESULTS: The adjusted for age and parity cesarean delivery rate was significantly lower from raw cesarean delivery rates in two of the three institutions (A: 28.2% adjusted vs. 36.5% raw; P=0; B: 28.2% adjusted vs. 30.4% raw, P=0.0411; C: 28.7% adjusted vs. 29.7% raw, NS) reflecting the older and more nulliparous patients in these two institutions. CONCLUSIONS: Our study confirms that raw cesarean delivery rate should not be used to compare quality of care within and among different institutions unless they are adjusted for different patient characteristics. We believe that using unadjusted cesarean delivery rates without appropriate adjustments in quality assurance and when comparing data with other institutions and the national rate is erroneous and misleading.
BACKGROUND: Raw cesarean section rates are often compared among institutions with different patient populations. The purpose of this study was to develop an easily reproducible mathematical model that allows comparisons of cesarean delivery rates across different patient populations and institutions. STUDY DESIGN: We first calculated three institution's cesarean delivery rates for each of ten groups of patients based on age and parity. The population based adjusted total cesarean delivery rate was then calculated for each institution based on the distribution of patients in the 2004 national birth data (n=4,097,029) but using the three institutions' individual cesarean delivery rates. RESULTS: The adjusted for age and parity cesarean delivery rate was significantly lower from raw cesarean delivery rates in two of the three institutions (A: 28.2% adjusted vs. 36.5% raw; P=0; B: 28.2% adjusted vs. 30.4% raw, P=0.0411; C: 28.7% adjusted vs. 29.7% raw, NS) reflecting the older and more nulliparous patients in these two institutions. CONCLUSIONS: Our study confirms that raw cesarean delivery rate should not be used to compare quality of care within and among different institutions unless they are adjusted for different patient characteristics. We believe that using unadjusted cesarean delivery rates without appropriate adjustments in quality assurance and when comparing data with other institutions and the national rate is erroneous and misleading.
Authors: J P Souza; A P Betran; A Dumont; B de Mucio; C M Gibbs Pickens; C Deneux-Tharaux; E Ortiz-Panozo; E Sullivan; E Ota; G Togoobaatar; G Carroli; H Knight; J Zhang; J G Cecatti; J P Vogel; K Jayaratne; M C Leal; M Gissler; N Morisaki; N Lack; O T Oladapo; Ö Tunçalp; P Lumbiganon; R Mori; S Quintana; A D Costa Passos; A C Marcolin; A Zongo; B Blondel; B Hernández; C J Hogue; C Prunet; C Landman; C Ochir; C Cuesta; C Pileggi-Castro; D Walker; D Alves; E Abalos; Ecd Moises; E M Vieira; G Duarte; G Perdona; I Gurol-Urganci; K Takahiko; L Moscovici; L Campodonico; L Oliveira-Ciabati; M Laopaiboon; M Danansuriya; M Nakamura-Pereira; M L Costa; M R Torloni; M R Kramer; P Borges; P B Olkhanud; R Pérez-Cuevas; S B Agampodi; S Mittal; S Serruya; V Bataglia; Z Li; M Temmerman; A M Gülmezoglu Journal: BJOG Date: 2015-08-10 Impact factor: 6.531