Cervical effacement, the shortening of the cervical canal as the functional approaches of internal os and external os, is accomplished by the gradual incorporation of the substance of the cervix into the lower uterine segment. Cervical effacement is commonly expressed as a percentage of the presumed length of the uneffaced cervix. However, there is no consensus regarding the length of the uneffaced cervix before labor begins. The 23^rd edition of Williams Obstetrics indicates that the approximate length of the cervix is 2 cm prior to the onset of labor [1]. Yet transvaginal ultrasound studies of the cervix show that the median cervical length ranges from 3.5 to 4.0 cm at 24 and 28 weeks and 3.0 to 3.5 cm after 32 weeks [2]. The aim of our study was to evaluate the accuracy of cervical effacement reported as a percentage by digital cervical exams using cervical length determined by transvaginal ultrasonography as a standard.

This retrospective study was approved by the institutional review board at Stroger Hospital of Cook County, Chicago, Illinois, USA. Data were collected from Stroger Hospital of Cook County Perinatal Network between January 2005 and December 2008. Women who had a digital cervical exam and subsequent transvaginal ultrasound scan for cervical length were included. Digital cervical exams were performed by different examiners within our perinatal network. Cervical effacement was reported as a percentage. A transvaginal ultrasound scan was performed by one of the authors (TN) who did not perform any of the digital exams. All transvaginal ultrasound scans were done at our institution upon patient arrival within 30 minutes of vaginal exams. A Toshiba ECCOCEE Model SSA_340A (Otawara-Shi, TogichiKen, Japan) or Siemens Sonoline Elegra (Issaquah, Washington, USA) ultrasound system with a 7 MHz endovaginal transducer was used to measure cervical length according to a technique described by Iams et al. [3]. The procedure was done with women in the supine position with empty bladders. Care was taken not to put pressure on the cervix. The entire cervical canal was visualized in the sagittal plane. Cervical length was measured from the internal os to the external os and recorded in centimeters. Three measurements were routinely obtained at each study. Intraobserver variability was 3.5%. Average cervical length was calculated and used for data analysis. The following data were also collected for each woman: maternal age, gravidity, parity, gestational age, and indications for cervical evaluation. Data were analyzed with descriptive statistics on SPSS version 17.0 for Windows XP (SPSS Inc., Chicago, IL, USA).

A total of 173 women met the study criteria. Maternal characteristics are summarized in Table 1. Indications for cervical evaluation were threatened preterm labor (55%) and induction of labor (45%). Of the 173 digital exams, 50 (29%), 28 (16%), and 95 (55%) were performed by attending physicians, nurse midwives, and resident physicians, respectively. Vaginal exams were performed by 2^nd, 3^rd and 4^th year residents. Cervical effacements were compared to cervical lengths determined by transvaginal ultrasonography (Figure 1). Cervical effacements and corresponding cervical lengths were grouped according to percentage, and a range of cervical lengths was observed for each effacement group (Table 2). The distributions of cervical lengths are shown in

Figure 2 for 0% effacement and Figure 3 for 50% effacement.