American Journal of Emergency Medicine (2012) 30, 1627-1629

Brief Report

Accuracy of emergency physicians using Ultrasound measurement of crown-rump length to estimate gestational age in pregnant females^?

Caitlin Bailey MD ^a,?, Jennifer Carnell MD ^b, Farnaz Vahidnia MD, PhD ^a,

Sachita Shah MD ^c, Michael Stone MD ^a, Mickeye Adams MD ^d, Arun Nagdev MD ^a

^aDepartment of Emergency Medicine, Alameda County Medical Center, Oakland, CA 94602, USA

^bDepartment of Emergency Medicine, Baylor College of Medicine, Houston, TX, USA

^cDivision of Emergency Medicine, Harborview Hospital, Seattle, WA, USA

^dDepartment of Obstetrics and Gynecology, Alameda County Medical Center, Oakland, CA, USA

Received 2 September 2011; revised 28 November 2011; accepted 2 December 2011

Abstract

Study objective: The objective of this study is to evaluate the accuracy of emergency providers (EPs) of various levels of training in determination of gestational age (GA) in pregnant patients using bedside ultrasound measurement of crown-rump length (CRL).

Methods: We conducted a prospective, cross-sectional, observational study of patients in obstetrical care at an urban county hospital. We enrolled a convenience sample of women at 6 to 14 weeks gestation as estimated by last menstrual period. Emergency providers used ultrasound to measure the CRL. Repeat CRL measurements were performed by either an obstetrical ultrasound technician or senior obstetrician and used as the criterion standard for true GA (TGA).

Results: One hundred five patients were evaluated by 20 providers of various levels of training. The average time required to complete the CRL measurement was 85 seconds. When CRL measurements performed by EPs were compared with the TGAs, the average correlation was 0.935 (0.911-0.959). Using standard accepted variance for CRL measurements at different GAs according to the obstetrics literature (+-3 days for 42-70 days and +-5 days for 70-90 days), correlation between EP ultrasound and measured TGA was 0.947 (0.927-0.967).

Conclusions: Emergency providers can quickly and accurately determine GA in first-trimester pregnancies using bedside ultrasound to calculate the CRL. Emergency providers should consider using ultrasound to calculate the CRL in patients with first-trimester bleeding or pain because this estimated GA may serve as a valuable data point for the future care of that pregnancy.

(C) 2012

^? Equipment used: Sonosite M-turbo, Bothell Washington.

* Corresponding author.

E-mail addresses: [email protected], [email protected] (C. Bailey).

Introduction

Background

Despite the common application of the ultrasonographic crown-rump length (CRL) measurement by obstetricians and

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2011.12.002

1628 C. Bailey et al.

ultrasound technicians to calculate gestational age (GA), there are no published studies assessing the ability of emergency providers (EPs) to accurately measure CRL. Given the rising application of ultrasound in emergency practice, validation of this skill within this field is vital for quality assurance.

Significance

Vaginal bleeding in pregnancy, particularly in the first trimester, is an extremely common chief complaint [1]. In underserved areas with minimal health care access, patients often present to the emergency department (ED) with complaints of vaginal bleeding without any previous prenatal care. In addition, many patients are unsure of the timing of their last menstrual period. Even if it is known, estimation of GA based on last menstrual period dating can be inaccurate [2]. Accurate fetal dating during the initial ED visit may be crucial for the care of that patient because many patients do not return for routine prenatal care until later in their pregnancy. A CRL performed in early pregnancy (up to 12- 14 weeks) is more accurate than sonographic dating methods after 14 weeks. If a pregnant patient subsequently suffers a complication of pregnancy, other medical disease, or trauma later in her pregnancy, knowledge of the GA is critical in determining the safety of early fetal delivery or drug therapy. Accurate dating including the use of ultrasound was also shown to significantly reduce the number of pregnancies considered postterm in 1 large retrospective study [3]. This could result in fewer inductions for postdate pregnancies, with associated cost savings for the hospital. Prior emergency Ultrasound studies of pregnant patients have shown that EPs can accurately identify free fluid in the abdomen of the pregnant patient with blunt trauma [4] and estimate GA in later pregnancy [5]. However, no previous studies have demonstrated the ability of EPs to accurately and rapidly measure GA in the first trimester via CRL.

Goals of investigation

We hypothesized that EPs can rapidly and accurately assess GA in early pregnancy using ultrasonographic calculation of the CRL.

Methods

Study design

We conducted a prospective, blinded observational study of women in their first trimester of pregnancy presenting to our hospital’s obstetrical clinic for a routine sonogram between March 10, 2010, and September 17, 2010. Written informed consent was obtained in all patients. The study was approved by our hospital’s institutional review board.

Setting

The study was performed at a large, urban teaching hospital with an emergency medicine residency and ultrasound fellowship.

Participants

A convenience sample of women older than 18 years old was enrolled on days when study EPs and an ultrasound technician or senior obstetrician (OBGYN) were available. The 20 participating emergency sonographers included senior medical students, emergency medicine residents who had completed a 1-month required rotation in emergency ultrasound, 3 emergency ultrasound fellows, ED physician assistants, and ultrasound-credentialed ED attending physicians. All providers viewed a 5-minute standardized video demonstrating CRL measurements pro- duced by the senior investigator (AN). The participating ultrasound technician was Registered Diagnostic Medical Sonographer (RDMS) certified, and the senior OBGYN had more than 10 years of ultrasound experience.

Study protocol

Each emergency sonographer measured CRL with a standard transabdominal curviLinear probe (Sonosite M- turbo; Sonosite, Bothell, WA). The measurements displayed on the screen of the ultrasound machine were covered while the examinations were performed. The lead author (AN) recorded the measurements and calculated GA immediately after each examination. Time was measured from placement of the transducer on the patient’s abdomen until the point at which each measurement was completed.

Immediately after the EP ultrasound, all study patients underwent comprehensive sonography by our ultrasound technician or senior OBGYN. This measurement served as the true GA. All emergency medicine participants as well as

Fig. 1 Measurement of CRL.

ultrasound measurements“>Emergency physicians crown-rump length accuracy 1629

the obstetrical ultrasound technician and the OBGYN were blinded to each other’s measurements.

Ultrasound measurements

A 2- to 5-MHz curvilinear abdominal transducer was used to measure CRL in the best view possible for the provider. In standard fashion, the yolk sac and limbs were not included in the measurements (Fig. 1). Calipers were placed from the top of the head to the best-estimated region of the rump. The ED sonographers were informed not to include the yolk sac or limbs during the training but were not instructed during the actual measurements. All ultrasounds for the study, includ- ing the comprehensive studies, were performed on a SonoSite M-Turbo ultrasound system.

Results

One hundred five patients were evaluated by 20 EPs, including fourth year medical students, residents (R1-R4), physician assistants, ultrasound fellows, and attending physicians. At least 5 patients were scanned by practitioners at each level of training; medical students performed 27 scans, whereas fellows and attending physicians performed 22 and 11 scans, respectively. The average time required to complete the CRL measurement was 85 seconds. When CRL measurements performed by EPs were compared with the true GA, the average correlation was found to be 0.935 (0.911-0.959). Using standard variance for CRL measure- ments as published (3 days for 42-70 days and 5 days for 70- 90 days [6]), correlation between EP ultrasound and measured true GA was 0.947 (0.927-0.967).

Discussion

Limitations

There are several important limitations of our study. First, our subjects were studied in a controlled clinic environment; this limits our ability to extrapolate this process to the setting of a busy ED. Second, Transvaginal ultrasounds were not performed. For fetuses younger than 8 weeks GA, providers might choose to perform a transvaginal examination for better visualization. Although we might assume that a high degree of correlation would persist with images obtained in this manner, our study does not address this directly. Finally, although we included providers with a range of experience with ultrasound (from medical students to ultrasound fellows and attending physicians), a relatively small number of providers partici- pated at each level of training. Although a small number of

scans were performed by each provider type, we feel that the short standardized training and broad range of experience allows for a moderate level of external validity.

To the best of our knowledge, this is the first study to evaluate the accuracy of EP measurement of CRL to calculate GA in early pregnancy. The high degree of correlation between EP CRL measurement and those of obstetrics personnel suggests that EP measurement of GA in the first trimester may be accurate. In addition, our data suggest that the CRL measurement can be rapidly performed, only minimally prolonging the examination for the busy EP. Emergency provider measurement of the CRL could become a crucial data point in the progression of a pregnancy for patients who present to the ED as their first portal of entry into prenatal care. This is frequently the case in underserved populations who do not have their own physicians when they become pregnant and for patients who do not know they are pregnant at the time of their ED visit. These patients often have difficulty arranging and maintaining timely obstetrics follow-up; by the time they are evaluated by an obstetrician, they may be past the window for the use of the CRL measurement (N14 weeks) and their sonographic dating, therefore, less accurate. This becomes relevant later in pregnancy when complications may arise and require a confident assessment of GA for consideration of delivery or the need for therapeutics such as steroids. In conclusion, our study found that EPs with limited training in sonographic measurement of the CRL were able to determine the CRL with a high degree of accuracy in a short amount of time. This suggests that the EP measurement of the CRL may be used as a reliable estimate of delivery date in subsequent decision making regarding a pregnancy initially assessed in ED.

References

1. Wittels KA, Pelletier AJ, Brown DFM, Camargo Jr CA. United States emergency department visits for Vaginal bleeding during early pregnancy, 1993-2003. Am J Obstet Gynecol 2008;198(5):523.e1-6.
2. Dietz PM, England LJ, Callaghan WM, et al. A comparison of LMP- based and ultrasound-based estimates of gestational age using linked California livebirth and prenatal screening records. Paediatr Perinat Epidemiol 2007;21(Suppl 2):62-71.
3. Mongelli M, Wilcox M, Gardosi J. Estimating the date of confinement: ultrasonographic biometry versus certain menstrual dates. Am J Obstet Gynecol 1996;174(1 Pt 1):278-81.
4. Goodwin H, Holmes JF, Wisner DH. Abdominal ultrasound examina- tion in pregnant blunt trauma patients. J Trauma 2001;50(4):689-93 [discussion 694].
5. Shah S, Teismann N, Zaia B, et al. Accuracy of emergency physicians using ultrasound to determine gestational age in pregnant women. Am J Emerg Med 2010;28(7):834-8.
6. Hadlock FP, Shah YP, Kanon DJ, Lindsey JV. Fetal crown-rump length: reevaluation of relation to menstrual age (5-18 weeks) with high- resolution real-time US. Radiology 1992;182(2):501-5.