| Study Year | Setting Study Type Duration | Author Affiliation | Sample Size and Selection | Peri-operative (30-day) Complications | Quality | Comments |
|---|---|---|---|---|---|---|
| Gasecki 1994 | US and Canada, multicenter Post-hoc subgroup analysis of RCT 1988-1991 | Neurology | Of 328 CEA in patients with 70-99% stenosis, 100 performed in
patients with hemispheric stroke Group A: 42/100 had CEA prior to 30 days Group B: 58/100 had CEA after 30 days (grp B) |
Group A
: 2/42 (4.8%) stroke or death; 0/42 (0%) death Group B: 3/58 (5.2%) stroke or death; 0/58 (0%) death | FAIR. More patients with normal CT in grp A, grp A had overall greater degree of angiographically seen diagnosis. Other clinical/demographic features similar. No statistical analysis of confounders. | Post-hoc analysis of NASCET. Did not find evidence that in patients with abnormal CT, delaying surgery results in less morbidity. |
| Hoffmann 1999 | S Africa, single center Retrospective 1981-1988 | Neurology and surgery | 207/1005 CEA had stroke as indication Group A: 86/207 had CEA within 6 weeks Group B: 121/207 had CEA after 6 weeks study analyzed 86 early (<6 wks, grp A) vs 121 late (>6 wks, grp B) CEA. |
Group A:
3/86 (3.5%) stroke or death; 1/86 (1.2%) death Group B: 7/121 (5.8%) stroke or death; 3/121 (2.5%) death | FAIR. Unclear how patients selected for early surgery. Groups relatively comparable by most demographic/clinical factors. No description of ascertainment technique. No statistical analysis of confounders. Not clear if ascertainment independent. | RR 1.90 (NS) for morbidity/mortality in later surgery group. Rankin score 1.5 across groups. |
| Plotrowski 1990 | US, single center Retrospective 1978-1988 | Surgery | 129/140 CEA had stroke as indication Group A: 82/129 had CEA within 6 weeks Group B: 47/129 had CEA after 6 weeks |
Group A:
1/82 (1.2%) stroke or death; 1/82 (1.2%) death Group B: 2/47 (4.3%) stroke or death; 1/47 (2.1%) death | FAIR. Later surgery group evaluated for CEA after recovery from stroke established. Complications not specified or defined. No description of ascertainment technique. Not clear if ascertainment independent. Groups appear comparable clinically and demographically. No statistical analysis of confounders. | patients had CEA according to protocol, after neurologic stability and soon (generally 24-48 hrs) after angiography. |
| Dosick 1985 | US, 2 centers Retrospective 1980-1983 | Surgery |
Group A:
Of 171 patients with mild-moderate neurologic deficits
> 24 hours, 110 had negative CT scans and positive
angiography and had CEA within 14 days Group B: Of 74 patients with mild-moderate neurologic deficits > 24 hours and positive CT scan, 51 had positive angiography and had CEA at 4-6 weeks |
Group A:
1/110 (0.9%) stroke or death; 0/110 (0%) death Group B: 1/51 (2.0%) stroke or death; 0/51 (0%) death | POOR. No true comparator group. Inadequate data to compare clinical and demographic features between early and late groups. Complications not specified or defined. No description of ascertainment technique. Ascertainment not independent. No statistical analysis of confounders. | RIND defined as symptoms lasting >24 hours, thought ischemic, with normal CT scan. Other comparison groups reported were patients with old stroke (n=56), TIA (n=113; timing of CEA not clear), and asymptomatic (n=41). |
| Eckstein 1998 | US, single center Retrospective 1980-1995 | Surgery, neurology, radiology, anesthesiology |
Group A:
56/2013 patients with TIA/minor stroke had CEA within 4 weeks Group B: No comparison group |
Group A:
2/56 (3.6%) stroke or death; 0/56 (0%) death Group B: No comparison group | POOR. No comparator group. Unclear how patients selected for early CEA. No data to compare clinical and demographic features between early and late groups. No statistical analysis of confounders. Ascertainment not independent. Inadequate description of ascertainment technique. | Outcomes in patients with later (after 4 weeks) CEA not reported. |
| Giordano 1985 | US, single center Retrospective 1973-1983 | Surgery | 49/352 consecutive CEA had stroke as indication Group A: 31/49 had CEA within 6 weeks Group B: 18/49 had CEA after 6 weeks |
Group A:
5/31 (16.1%) stroke or death; 1/31 (3.2%) death Group B: 0/18 (0%) stroke or death; 0/18 (0%) death | POOR. Unclear how patients selected for early surgery. Inadequate description of population. No description of ascertainment technique. No statistical analysis of confounders. Not clear if ascertainment independent. Not clear if followup was 30 days. | For patients with indication other than stroke, 2.3% stroke rate and 0.7% mortality rate. Significant difference in rates of post-op complications when comparing <5 wks with >5 wks CEA. |
| Paty 1997 | US, single center Retrospective 1987-1997 | Surgery | 215/1922 CEA were in patients with stroke <6 months
prior Group A: 149/215 had CEA within 1 month Group B: 66/215 had CEA between 1 and 6 months |
Group A:
3/149 (2.0%) stroke or death; deaths not reported according to timing of CEA Group B: 0/66 (0%) stroke or death; deaths not reported according to timing of CEA | POOR. Unclear how patients selected for early surgery. Unable to compare demographic/clinical factors with group with later surgery. Complications not specified or defined. No description of ascertainment technique. Not clear if ascertainment independent. Unclear duration of followup. No statistical analysis of confounders. | No characterization of stroke severity. |
| Ricotta 1985 | US, single center Retrospective 1978-1984 | Surgery | 181 CEA in patients who had preoperative CT Group A: 27/181 symptomatic patients had CEA within 2 wks Group B: 71/181had positive CT and CEA after 2 weeks |
Group A:
5/27 (18.5%) stroke or death; 1/27 (3.7%) death Group B: 2/71 (2.8%) stroke or death; 0/71 (0%) death | POOR. No true comparator group. Not clear how patients selected for pre-op CT or for early CEA. Inadequate description of pt population. Complications not specified or defined. No description of ascertainment technique. Ascertainment not independent. No statistical analysis of confounders. Unclear duration of followup. | Tabulated results exclude 83 patients with later CEA and negative CT scans because complications for symptomatic patients not reported separately. 4/5 strokes in early CEA occurred in patients with positive CT scans. |
| Rosenthal 1988 | US, single center Retrospective 1977-1987 | Surgery, neurology, radiology | 151/253 CEA in patient with RIND or stroke Group A: 60/151 had CEA within 3 weeks Group B: 91/151 had CEA after 3 weeks |
Group A:
2/60 (3.3%) stroke; deaths not clear Group B: 4/91 (4.4%) stroke; deaths not clear | POOR. Not clear how patients selected for surgery or early surgery. Not clear if ascertainment independent. Unable to compare major clinical and demographic factors in early and later CEA groups. Complications not specified or defined. No description of ascertainment technique. No statistical analysis of confounders. | 2 peri-operative deaths; not clear how many in early CEA or later CEA. |
| Whittemore 1987 | US, single center Retrospective 1979-1986 | Surgery | 651 CEA performed Group A: 44/651 had CEA within 30 days of small, stable stroke with >75% ipsilateral stenosis Group B: 607/651 had CEA after 30 days (severity of symptoms unclear) |
Group A:
1/44 (2.3%) stroke or death; 1/44 (2.3%) death Group B: 12/607 (2.0%) stroke or death; 7/607 (1.2%) death | POOR. Unable to compare major clinical and demographic factors in early CEA and later CEA groups. Complications not specified or defined. No description of ascertainment technique. Not clear if ascertainment independent. No statistical analysis of confounders. | Includes data from Whittemore 1984. |
RCT: randomized controlled trial; CEA: carotid endarterectomy; TIA: transient ischemic attack; MRI: magnetic resonance imaging; TCD: transcranial Doppler; VA: Veterans Affairs
From: Evidence Tables
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.