Is ultrasound done by emergency physician,a usefull tool in screening for ectopic pregnancy?

Date First Published:

March 10, 2011

Last Updated:

June 23, 2011

Report by:

Dr Muddasar Rasool, ST5 (University Hospital Lewisham NHS Trust)

Search checked by:

Dr Muddasar Rasool, University Hospital Lewisham NHS Trust

Three-Part Question:

In [patients presenting to emergency department with early pregnancy complications] is [ultrasound scan done by emergency physician] useful in [screening for ectopic pregnancy]

Clinical Scenario:

on friday night in a busy ED, a 22 year old single female law student presents with lower abdominal discomfort and per vaginal spotting for 12 hours. She is primigravida, with unplanned pregnancy which happened while she was on oral contraceptive pills. She does not have any other significant medical history. She is afebrile. Pulse of 75/min, B.P is 126/79 with postural drop of 15mm(Hg). Abdominal examination is completely normal. Urine dip shows positive B HCG and blood test are completely normal. The timeliest appointment in early pregnancy unit is not available till Monday. You are concerned about sending this patient home. You have an ultrasound machine available in department. Can this ultrasound be used as a screening tool to risk-stratify this patient with any degree of certainty during first consultation in ED? What does the evidence say?

Search Strategy:

1) The following data bases were searched via the Ovid interface, from 1965 till current.
CINAHL
Medline
Embase
Cochrane Library
2- Further searches were carried out on
Google
Google scholar
Abstract of all the papers identified were checked for relevance to our clinical question.
References cited in original papers were also checked to identify further relevant papers.
Exclusion criteria:
Limit to: Human and adults (Human Age Groups Adult 18 to 64 years).
Studies were excluded where US studies were not done by emergency physicians.

Search Details:

No.tDatabasetSearch termtHits
1 MEDLINE (ectopic AND pregnancy).ti,ab t7472

2tMEDLINE t*PREGNANCY, ECTOPIC/ t7113

3tMEDLINE t1 OR 2 t10874

4tMEDLINE t(ultrasound OR ultrasonography OR sonography).ti,ab t181750

5tMEDLINE t*ULTRASONOGRAPHY/ t34334

6tMEDLINE t4 OR 5 t194724

7tMEDLINE temergency.ti,ab t122477

8tMEDLINE t*EMERGENCIES/ t9200

9tMEDLINE t7 OR 8 t126618

10tMEDLINE t3 AND 6 AND 9 t177

11tMEDLINE t10 [Limit to: (Publication Types Clinical Trial, All)] t1

12tMEDLINE t10 [Limit to: Humans] t168

13tMEDLINE t12 [Limit to: Humans and (Age Groups All Adult 19 plus years)] t110

14tEMBASE t(ectopic AND pregnancy).ti,ab t8319

15tEMBASE t*ECTOPIC PREGNANCY/ t9394

16tEMBASE t14 OR 15 t12722

17tEMBASE t(ultrasound OR ultrasonography OR sonography).ti,ab t222091

18tEMBASE t*ULTRASOUND/ t20923

19tEMBASE t17 OR 18 t229740

20tEMBASE temergency.ti,ab t144702

21tEMBASE t*EMERGENCY/ OR *EMERGENCY HEALTH SERVICE/ OR *EMERGENCY MEDICINE/ OR *EMERGENCY PHYSICIAN/ OR *EMERGENCY WARD/ OR *EVIDENCE BASED EMERGENCY MEDICINE/ t55342

22tEMBASE t20 OR 21 t168788

23tEMBASE t16 AND 19 AND 22 t224

24tEMBASE t23 [Limit to: Human and (Human Age Groups Adult 18 to 64 years)] t135

Outcome:

After reading through abstracts of 135 papers identified by the search,10 papers were identified as being directlty relevant to the clinical question

Relevant Paper(s):

Study Title	Patient Group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
The association between operator confidence and accuracy of ultrasonography performed by novice emergency physicians Davis DP, Campbell CJ, Poste CJ et al. 2005 USA	adults.Evaluating effectiveness of US examinations done by emergency physicians for presence or absence of 6 different conditions. These conditions are 1- peritoneal fluid 2- pericardial fluid 3- gallstones 4- intrauterine fetus 5-hydronephrosis 6- abdominal aorta’s diameter. Data was collected for all eligible exams after one year and ED US results were compared to gold standard investigations. Overall 276 scans were included with sensitivity of 92% and specificity of 86%	This is a single centre prospective study. Evaluating effectiveness of US examinations done by emergency physicians for presence or absence of 6 different conditions. These conditions are 1- peritoneal fluid 2- pericardial fluid 3- gallstones 4- intrauterine fetus 5-hydronephrosis 6- abdominal aorta’s diameter. Data was collected for all eligible exams after one year and ED US results were compared to gold standard investigations. Overall 276 scans were included with sensitivity of 92% and specificity of 86% Primary outcome measure for early pregnancy scans was determination of intra uterine pregnancy. Out of 70 early pregnancy scans done by emergency physicians 52 (74%) positively identified intrauterine pregnancy. When comparing with gold standard, this gives 100% sensitivity and specificity with positive and negative predictive value of 100%.	negative predictive value	100%	, data was collected for effectiveness for 6 different conditions, intra uterine pregnancy being one of them. Thus the study is more relevant for overall effectiveness of scans. Study has been done with no power calculation to determine sample size. Data has been collected retrospectively by the operators themselves, so there is no blinding. The data is presented in percentages not in absolute numbers, confounding the results somewhat
			specificity	100%
Diagnosis and management of ectopic pregnancy using bedside transvaginal ultrasonography in the ED. A 2 year experience Adhikari S, Blaivas M and Lyon M 2007 USA	Pregnant patients with first trimester complications	This is a retrospective single centre study. Only data of positive ED scans (ectopic/ possible ectopic) is being presented, which allows for calculation of positive predictive value only. There is no mention of the final outcome of the patients whose scans were deemed negative in ED. Out of positive scans, patients with small amount of peritoneal fluid with empty uterus but no other findings suggestive of ectopic pregnancy has been excluded. This can lead to purity bias. Results are analysed as per protocol analysis. There is no power calculation, no mention of blinding when reporting results. Study design is poor for a diagnostic study and sensitivity and specificity of the test cannot be calculated as negative results have not been included in the study. Out of 74 scans deemed positive or suspicious for ectopic pregnancy, 47 patients received final diagnosis of ectopic pregnancy. This gives positive predictive value of 63.5%.	positive predictive value	63.50%	Out of positive scans, patients with small amount of peritoneal fluid with empty uterus but no other findings suggestive of ectopic pregnancy has been excluded. This can lead to purity bias. Results are analysed as per protocol analysis. There is no power calculation, no mention of blinding when reporting results. Study design is poor for a diagnostic study and sensitivity and specificity of the test cannot be calculated as negative results have not been included in the study.
Pelvic ultrasound performed by the emergency physicians for detection of ectopic pregnancy in complicated first trimester pregnancies Durham B, Lane B, Burbridge L and Balasubramaniam S 1996 USA	Emergency physicians performed consecutive US scans during 6 months period on all women in first trimester of pregnancy with abdominal pain or vaginal bleeding.	Patients who presented with obvious incomplete abortion have been excluded. Obs and gynae consult was recommended when scan showed EP, intermediate findings or early IUP with gestational age less than 6 weeks. Later recommendation was due to observation that early IUP gestational sac cannot be distinguished from pseudo sac of EP on US alone. Results are promising. And are given in absolute numbers allowing for own calculations to made. Excluding patients with intermediate scan and BHCG <2000 we get Sensitivity =100% Specificity =93% Positive predictive value= 57% Negative predictive value=100%			this is a single centre study. Sample size is small. There is no power calculation to show how sample size is reflective of population. No mention of blinding of operators to final outcome. There is no intention to treat analysis
Effect of emergency physician-performed pelvic sonography on length of stay in emergency department Christina HY Shih 1996 USA	pregnant patients with first trimester complications	Notes were reviewed of 127 patient who underwent pelvic US in emergency department mostly by emergency physicians but also by obs and gynae residents. Primary outcome measure is the length of stay analysis amongst the group that had US done by ED physician compared with the group that had scan performed by gynae resident. LOS for group one is 60 minutes compared to LOS of group B of 180 minutes. In secondary outcome measure 74 scans were done by ED physicians. 47 IUP’s were correctly identified with no false +ve’s. Out of 24 true –ve IUP’s 6 were ectopics which were all positively identified by ED physicians. Plotting this data gives us: Sensitivity =100% Specificity = 72% Positive predictive value = 22% Negative predictive value =100%	negative predictive value	100%	On –ve side it’s a single centre, retrospective study. Study utilizes a convenience sample with no power calculation. It does not give absolute numbers for true and false –ve’s in IUP analysis.
efficacy of transabdominal ultrasound examination in the diagnosis of early pregnancy complications in an emergency department Wong TW, Lau CC, Yeung A et al. 1998 Hongkong, China	all pregnat women within first trimester, presenting with pain and/or bleeding during five months period from February to June 1996.	It’s a prospective single centre trial. Enrolling all women presenting with pain and/or bleeding during five months period from February to June 1996. Haemodynamically unstable patients and patients with previous US scans were excluded. Scans were done by emergency physicians using trans-abdominal probe only. Sensitivity =80% Specificity = 80% Positive predictive value = 12% Negative predictive value = 99%	negative predictive value	99%	Authors used a convenience sample with no power calculation. Had a reasonable inclusion and exclusion criteria. There is no mention of blinding while collating the results. Absolute numbers are not given in the results.
Outcome analysis of a portocol including bedside endovaginal sonography in patients at risk of ectopic pregnancy Mateer JR, Valley TV, Aiman EJ et al 1996 USA	pregnant patients with first trimester complications	Single centre prospective case control trial with historical control. Sample is a convenience sample over period of three years. All eligible were enrolled with no significant exclusions. 314 patients finally enrolled in control group out of which 14 were excluded due to incomplete data. Thus giving us per protocol analysis. Primary outcome measure was number of patients discharged from ED subsequently found to have ruptured ectopic pregnancy. 1/11 in intervention group and 12/24 in control group, who were discharged from ED with ectopic pregnancy, were subsequently found to have ruptured ectopic pregnancy. CER= 50 EER=9.09 ARR= 40.91 RRR=81% NNT=2.4 This is a very significant result considering seriousness of the primary outcome.	RRR	81%	On –ve side study uses a historical control, thus improvement of service, if any, over time may account for some of the relative risk reduction that is observed amongst both groups. It’s a single centre study with relatively small sample size so possibility of type 1 error cannot be ignored.
	pregnant patients with first trimester complications		NNT	2.4
Emergency department sonography by emergency physicians Jehle D, Evan T, Harchelroad F et al 1998 USA	patients have US examaination in ED. Subgroup: patients with early pregnancy complications having ED US	It’s a retrospective single centre trial. Effectiveness of ED sonography is evaluated for cardiac, gynaecological, biliary and abdominal vascular disorders. Study uses a retrospective convenience sample Out of 40 gynaecological exams carried out. 20 showed definitive and 15 probable IUP. Out of 5 scans –ve for IUP, one patient had EP. Sensitivity: 100% Specificity: 89% Positive predictive value: 20% Negative predictive value: 100%			retrospective convnenience sample. no power calculation. no mention of blinding
Transvaginal ultrasonography by emergency physicians decreases patient time in the emergency department Burgher SW, Tandy TK, Dawdy MR 1998 USA	adult pregnant females with early pregnancy complications	Study retrospectively looks at transit time through ED between 2 groups 1- when TVUS was done by ED physicians. 2- When TVUS was done by obs/gynae residents. Study was done at a naval medical centre. Physicians in both groups were not aware of the study at time of exam. An algorithm suggested by mateer at el was used to determine timing of gynae consult according to US findings and B HCG levels. Power calculation was done for primary outcome measure based on p=0.05 and power of 0.80. 46 patients were in ED physician group with mean transit time of 164.70 minutes. 38 patients in obs/gynae group had a mean transit time of 234.79 minutes. Thus there was more rapid transit time in group A by >60 minutes (p=0.0003). In ED physicians group 2 patients were diagnosed as EP. 27 IUPs, 17 intermediate scans. Out of 17 intermediate scans 5 had BHCG>2000, they were admitted to rule out EP. 3 of them turned out to be EP. None of 12 with BHCG <2000 turned out to be EP (excluded from analysis). Sensitivity = 100% Specificity = 93% Positive predictive value = 71% Negative predictive value = 100%			On –ve side, it’s a single centre study done in a naval centre. Retrospective with small sample size. Primarily looking at ED transit times between 2 groups. Study is not powered for efficacy of ED physician’s US scans.
Ultrasonographic examination by emergency physicians of patients at risk for Ectopic pregnancy Mateer JR, Aiman JE, et al 1995 USA	adult pregnant patients with first trimester complications	Prospective single centre study using a convenience sample. Accuracy of scans done by emergency physicians was determined by review of consultant sinologist and also compared with final diagnosis. 152 patients were enrolled. 4 were lost to follow up. 148 patients were included in final analysis. 87 patients had diagnosis of IUP and 17 of abnormal IUP. None of these patients turned out to be ectopic.3 patients were diagnosed as ectopic and diagnosis was confirmed later on. Of 11 patients with non IUP and BHCG of >2000, 5 patients were found to be ectopic. 30 patients with HCG level <2000 are being excluded from final analysis. 8 of them were found to have ectopic. plotting data in 2/2 table gives us Sensitivity = 100% Specificity =94% Positive predictive value = 57% Negative predictive value = 100%			On negative note it’s a single centre study, using convenience sample. No power calculation is done. Results are analysed using per protocol analysis.
Ultrasound availibility in the evaluation of ectopic pregnancy in the ED: Comparison of quality and cost-effectiveness with different approaches Durston WE, Carl M, Guerra W et al. 2000 USA	adult females with first trimester pregnancy complications	Compared effectiveness’ of availability of ED US to groups when ED arranged US through radiology to rule out ectopic pregnancy. Large study divided in three epochs of 2 years time frame. Epoch -1 US imaging available through medical imaging. Epoch – 2: US technician available in normal hours and on call out of hours and weekends. And Epoch – 3: when US were done by ED physicians. Primary outcome measure is number of patients documented to have absent IUP during first ED consultation. And results show increase in from 88% to 96% (p=0.02) During epoch 3 ED physicians performed 996 scans. 75 were intermediate and are excluded. Out of remaining 921 scans 131 are lost to follow up and excluded. 565 had scan +ve for IUP (true positive). 213 are true –ve. One false positive and 11 false –ve. Sensitivity = 98% Specificity = 99% Positive predictive value = 99% Negative predictive value = 95%			This is again a single centre study with historical control groups. Large numbers of patients are excluded from final analysis. No use B HCG levels to further stratify intermediate scans have been done.

Author Commentary:

all papers reviewed are in agreement that ED performed US has a negative predictive value approaching 100% when surveying patients with early pregnancy complications for ectopic pregnancy.

Bottom Line:

US in ED by ED physicians is definitely way forward in screening for ectopic pregnancy.

Level of Evidence:

Level 2: Studies considered were neither 1 or 3

References:

Davis DP, Campbell CJ, Poste CJ et al.. The association between operator confidence and accuracy of ultrasonography performed by novice emergency physicians
Adhikari S, Blaivas M and Lyon M. Diagnosis and management of ectopic pregnancy using bedside transvaginal ultrasonography in the ED. A 2 year experience
Durham B, Lane B, Burbridge L and Balasubramaniam S. Pelvic ultrasound performed by the emergency physicians for detection of ectopic pregnancy in complicated first trimester pregnancies
Christina HY Shih. Effect of emergency physician-performed pelvic sonography on length of stay in emergency department
Wong TW, Lau CC, Yeung A et al.. efficacy of transabdominal ultrasound examination in the diagnosis of early pregnancy complications in an emergency department
Mateer JR, Valley TV, Aiman EJ et al. Outcome analysis of a portocol including bedside endovaginal sonography in patients at risk of ectopic pregnancy
Jehle D, Evan T, Harchelroad F et al. Emergency department sonography by emergency physicians
Burgher SW, Tandy TK, Dawdy MR. Transvaginal ultrasonography by emergency physicians decreases patient time in the emergency department
Mateer JR, Aiman JE, et al. Ultrasonographic examination by emergency physicians of patients at risk for Ectopic pregnancy
Durston WE, Carl M, Guerra W et al.. Ultrasound availibility in the evaluation of ectopic pregnancy in the ED: Comparison of quality and cost-effectiveness with different approaches