June 2021

Authors: Mark Winstanley, Andy Neill, Dave McCreary, Chris Connolly, Becky Maxwell, Ed Abrahamson

/ Codes: / Published: 03/05/2021


Dr Andy Neill
Dr Dave McCreary

Clinical Question

Is a patient-performed vaginal swab sensitive enough for detection of STI?

Title of Paper

[Self-obtained vaginal swabs are not inferior to provider-performed endocervical sampling for emergency department diagnosis of Neisseria Gonorrhoeae and Chlamydia Trachomatis](https://pubmed.ncbi.nlm.nih.gov/33460481/)

Journal and Year

Academic Emergency Medicine. 2021.

Lead Author

Brian Chinnock


– Speculum exam and provider-performed endocervical sampling (PPES) is the textbook approach to female patients with suspected STI in the ED
– These examinations can be uncomfortable and emotionally distressing to patients, and may lead to decreased healthcare engagement for future episodes
– Space to perform said examinations can be difficult to come across in times of access-block, which can delay patient disposition
– In [December 2018](https://www.rcemlearning.co.uk/foamed/december-2018/#1543786461402-00c3f5a7-6482) we covered a paper on the podcast looking at whether the pelvic exam added anything to history when diagnosing cervicitis or PID. They concluded that routine pelvic exam didn’t add anything to the diagnosis.
– Previous studies comparing self-obtained vaginal swabs (SOVS) to PPES have shown them to be highly sensitive and accepted by patients, but these studies were in a variety of non-ED settings for asymptomatic screening rather than assessment in the acute care environment

Study Design

– Prospective observational cohort study comparing two methods of Neisseria Gonorrhoea (NG) / Chlamydia Trachomatis (CT) swab collection

Patients Studied

– Adult females
– Judged by the practitioner to need Gonorrhoea / Chlamydia testing

What They Did

– Patients completed a questionnaire re acceptability of self-obtained vaginal swabs (SOVS), current symptoms, history of STI
– Patients had both SOVS and provider performed endocervical sampling (PPES) with standard speculum examination
– SOVS – patient given one-page information handout on how-to obtain swab
– Both SVOS and PPES were run on hospital’s rapid test for NG/CT


– Self-obtained vaginal swab


– Provider-performed endocervical sampling


– Non-inferior sensitivity of 90% for detection of NG/CT
– Based on point-of-care testing at sexual health clinics priority of 90% sensitivity and a previous ED study looking at urine sampling for same also had 90% sensitivity
– Patient-specific composite measure, not organism-specific
– a true positive = concordant positive test between SOVS and PPES
– If only one organism positive on SOVS but both on PPES, classed as a false negative

Summary of Results

– 515 in final analysis
– 553 enrolled, 15 excluded for incomplete samples, 3 excluded for inderminate results
– 86 (17%) positive for either NG/CT/both
– 34% NG
– 54% CT
– 12% both
– SOVS sensitivity for detection of NG/CT 95% [95%CI 88-99%]
– 3 false neg CT, one NG
– NG sensitivity: 97% [87-100%]
– CT sensitivity: 94% [84-99%]
– Kappa for composite NG/CT 0.93 [0.89-0.98]
– +LR: 83 [34-198]
– -LR: 0.05 [0.02-0.13]

Authors Conclusion

Self-obtained vaginal swab was found to be non-inferior to provider-performed endocervical sampling for the diagnosis of neisseria gonorrhoeae/Chlamydia trachomatis using a rapid nucleic acid amplification test in the ED. This test provides an important ED diagnostic alternative to provider performed endo-cervical sampling for patients in whom a pelvic examination is not possible or declined.

Clinical Bottom Line

I think this study provides good evidence that a self performed HVS is sufficiently sensitive to diagnose NG/CT, so if that’s your only clinical question and you aren’t concerned that the patient has more significant pathology (tubo-ovarian abscess for example) then a HVS may suffice. If the patient is a sniff unwell or has lower abdominal tenderness – they still need the pelvic exam and spec though.

Other #FOAMed Resources / References:

[RCEMLearning Podcast December 2018 – Pelvic Exam for STD](https://www.rcemlearning.co.uk/foamed/december-2018/#1543786461402-00c3f5a7-6482)


Dr Chris Connolly
Dr Becky Maxwell


Dr Andy Neill
Dr Dave McCreary

Clinical Question

– what is the best way to predict stroke following a TIA?

Paper Title

– Prospective validation of Canadian TIA Score and comparison with ABCD2 and ABCD2i for subsequent stroke risk after transient ischaemic attack: multicentre prospective cohort study


– Perry, BMJ, Dec 2020


– TIA is common and clearly offers a potential window for intervention. These interventions are of variable levels of benefit with minor things like aspirin and statins to significant interventions like a carotid endarterectomy.
– it’s a little unclear how urgently to pursue these interventions and in particular whether admission or OPD work up is indicated. This varies of course due to resources from hospital to hospital. Having a means of accurately predicting early risk of stroke following a TIA would be very helpful in risk stratifying patients in terms of the urgency of work up and treatment.
– the ABCD2 score has been around for ages probably exists on your protocol in your department. But it was never really very good. We used it because we didn’t have something much better until recently
– enter Jeff Perry and the Ottawa group, authors of such excellent clinical decision tools ranging from the Ottawa ankle rule, to the Canadian C-spine rule to the more recent SAH tool. They now have one for TIA derived and validated in a separate publication in Stroke in 2018. This time it’s the big prospective validation – a really important step to see if the fancy rule that you derived with logistic regression in one publication applies in the broader context.
– i’ll say up front this is a 13 point score – you will need to MDCalc this rather than hold it all in your head.


– Prospective data, all Canadian including new site not in the original paper
– mix of people doing the assessments in the ED, inc neurologists
– telephone follow up and hospital chart review as gold standard for stroke which is of course not perfect but they had a very low rate of lost to follow up. (0.4%)
– Primary outcome was rate of stroke or carotid end arterectomy within 7 days of the ED visit (which is a very short time period but in reality this is what we’re worried about). These outcomes are clearly not the same with 1st being disease and the second being an intervention to prevent the disease.
– they defined low <1%, medium 1-5% and high >5%. These seem very low but i suppose this is what is expected and it is similar to risk categories for ABCD2.


– 7600 pts recruited (80% of eligible)
– almost perfect follow up
– 1.4% stroke rate at 7 days, 1.1% CEA at 7 days
– 6% admission rate showing which suggests one of two things – they were felt to be incredibly low risk or they have a very good system of follow up.
– 15% low risk, 15% high risk and the rest in the middle.
– because they had prospectively collected all this data they were able to see how well the ABCD2 compared with the Canadian TIA score. Canadian won with an AUC of 0.7 v 0.6 for the ABCD2


– this is high quality research and I think they’ve proven their point that this predicts things fairly well and is probably better than the next leading brand so to speak
– I would be happy to start using this
– however it in no way tells me what to do with these patients. Should I admit them or send them home. And if they go home what is the appropriate time scale
– the other fly in the ointment here is that you need a CT scan to do the score. Certainly over the 17 years I’ve been practicing CT has become more and more available but there is still often a window between about 10pm and 8am where you really need a good reason to get a CT scan. Something like a CT brain for resolved neuro symptoms probably does not fit in that category.


Perry, J. J. et al. A Prospective Cohort Study of Patients With Transient Ischemic Attack to Identify High-Risk Clinical Characteristics. Stroke 45, 92100 (2018).

Perry, J. J. et al. A Prospective Cohort Study of Patients With Transient Ischemic Attack to Identify High-Risk Clinical Characteristics. Stroke 45, 92100 (2018).


Dr Ed Abrahamson Twitter: @ed_abrahamson
Dr Mark Winstanley

Interventions for acute severe asthma attacks in children: an overview of Cochrane Reviews:


Leave a Reply