Table 1.

Study characteristics and extracted data (stratified by referral route) for clinician accuracy^a

Study	Location	Type	Referral source(s)	Mean age, years (SD)	TIA	TIA or stroke	Diagnosis unknown	Total number of cases of suspected TIA, n	PPV for reference Dx of TIA	PPV for reference Dx of TIA and stroke
Banerjee et al, 2009¹¹	London, UK	Prospective cohort^b	GP and ED	68 (13.5)	133	213	2 (15 unavailable)	282	47.2	75.5
Bradley et al, 2013¹²	Dublin, Ireland	Prospective cohort	GP and ED	60 (14.3)	49	56	None	101	48.5	55.4
Cameron et al, 2011⁹	Glasgow, UK	Prospective cohort	GP and ED	65 (13.6)	–	1890	None	3533	–	53.5
Dawson et al, 2009¹³	Glasgow, UK	Prospective cohort	GP and ED	65 (12.8)	–	2358	None	3467	–	68.3
Dutta, 2016¹⁵	Gloucester, UK	Prospective cohort — DOT validation cohort	GP and ED	71 (14.0)	160	236	None	525	30.5	45.0
Dutta et al, 2015¹⁴	Gloucester, UK	Prospective cohort	GP and ED	72^c (IQR: 60–80)	337	529	None	1067	31.6	49.6
Fallon et al, 2006¹⁶	Dublin, Ireland	Prospective cohort	ED (primarily) and other — not specified	75.5 (−)	56	72	18	117	47.9	61.5
Ferro et al, 1996¹⁰	Central & southern Portugal	Prospective cohort	GP	–	10	36	None	52	19.2	69.2
Ferro et al, 1996¹⁰	Central & southern Portugal	Prospective cohort	ED	–	4	18	None	31	12.9	58.1l
Fonseca and Canhão, 2009²⁶	Lisbon, Portugal	Prospective cohort	GP and ED	65 (−)	259 definite TIA	–	109 cases recorded as ‘possible’ TIA	458	−56.6	–
Karunaratne et al, 1999¹⁷	Scottish Borders, UK	Prospective cohort	GP and ED	67 (14)	31	64	7 non-TIA with no clear diagnosis	128	24.2	50.0
Lasserson et al, 2015²⁷	Oxford, UK	Prospective cohort	GP	73 (12.8)	209	–	None	513	40.7	–
Lavallée et al, 2007¹⁸	Paris, France	Prospective cohort	GP and ED	Median ages reported by final diagnosis alone	643 definite TIA	701	144 cases recorded as ‘possible TIA’	1085	72.5	77.9
Lee and Frayne, 2015¹⁹	Melbourne, Australia	Prospective cohort	GP and ED	67 (16.9)	13	18	4 non-TIA with no clear diagnosis	82	15.9	22.0
Magin et al, 2013²⁰	Hunter New England, Australia	Prospective cohort	GP	65 (15)	29	50	13 unclassified	127	22.8	39.4
Magin et al, 2013²⁰	Hunter New England, Australia	Prospective cohort	ED	65 (15)	46	66	9 unclassified	104	44.2	63.5
Martin et al, 1997²⁵	Liverpool, UK	Prospective cohort	GP and ED	62^c (IQR: 23–94)	200	–	Unclear	332	60.2	–
Murray et al, 2007³	Glasgow, UK	Retrospective cohort	GP and ED	Age ranges given	217	283	None	811	26.8	34.9
Palomeras Soler et al, 2015²¹	Barcelona, Spain	Prospective cohort	GP and ED	–	282	310	None	411	68.6	75.4
Sheehan et al, 2009²²	Dublin, Ireland	Prospective cohort	GP and ED	69 (13)	292	337	None	257	49.2	56.7
Walker et al, 2012²³	Leicester, UK	Prospective cohort	GP and ED	–	–	1273	None	2452	–	51.9

↵a Where there is more than one data entry for a single study this reflects that the study provided sufficient information to calculate PPVs by different referral sources.
↵b PPVs only reported for suspected TIAs that were seen at an anterior circulation TIA clinic.
↵c Median not mean reported. Dx = diagnosis. ED = emergency department. IQR = interquartile range. PPV = positive predictive value. SD = standard deviation. TIA = transient ischaemic attack.