| Study population | Region | Midyear of study | No of patient years | No of SAH patients | Incidence per 100 000 person years (95% CI) | Additional case finding methods‡ | % of patients with CT | Additional diagnostic criteria†† | Restriction of study population: age (y) |
| Oyabe28 | Japan | 1984 | 492 885 | 124 | 25.2 (20.7 to 29.6) | abj | ‡‡ | ‡‡ | >25 |
| Novosibirsk34 | Russia | 1987 | 971 751 | 64 | 6.6 (5.0 to 8.2) | aehjm | ** | AB | 25–74 |
| Turku15 | Finland | 1987 | 1 249 992 | 278 | 22.2 (19.6 to 24.9) | abeam | ‡‡ | AB | >25 |
| FINMONICA†16 | Finland | 1988 | 3 863 088 | 956 | 24.7 (23.2 to 26.3) | aeh | 84§ | ABC | 25–74 |
| FINSTROKE14 | Finland | 1993 | 1 933 660 | 360 | 18.6 (16.7 to 20.5) | aehk | 86 | AB | 25–74 |
| Arcadia19 | Greece | 1994 | 161 548 | 14 | 8.7 (4.1 to 13.2) | abeh | 82 | AB | >20 |
| Manhattan38 | USA | 1994 | 428 775 | 39 | 9.1 (6.2 to 12.0) | befikl | 99¶ | AC | >20 |
| Innhered32 | Norway | 1995 | 69 295 | 13 | 18.8 (8.6 to 29.0) | abdegk | 88 | BD | >15 |
| ACROSS†6 | Australia, N Zealand | 1997 | 4 916 154 | 400 | 8.1 (7.3 to 8.9) | aefk | 90§ | ABC | >15 |
SAH, subarachnoid haemorrhage.
*Studies listed in ascending order of midyear of data collection and are additional to those in the previous review.
†Studies based primarily on SAH, in contrast with general stroke studies.
‡Case finding methods. For inclusion, involvement of all hospitals in the region necessary and at least a or b. a = death certificates; b = general practitioners; c = rehabilitation; d = nursing homes; e = regular search; f = review radiology requests; g = media attention (campaign/newspaper); h = outpatient clinics, health centres; i = sudden deaths, very early death; j = emergency, ambulance, on call medical services; k = ICD-codes; l = door-to-door, home visit, social services, phone calls; m = autopsy reports.
§Studies providing the proportion of CT use in SAH patients exclusively, in contrast with % of CT in patients with stroke in general.
¶Studies not providing the exact proportion of patients with CT exclusively, but only the proportion of patients investigated with CT, autopsy or MRI.
**CT was available after 1992, and before 1992, all patients were diagnosed with lumbar puncture or autopsy.
††Additional diagnostic criteria, besides CT. For inclusion, at least A or B was necessary in pre-CT era or when CT percentage was below 90%. A = Lumbar puncture; B = autopsy; C = angiography; D = MRI; E = surgery.
‡‡Proportion of patients investigated with CT or diagnostic criteria unknown, but inclusion after discussion among authors of this review.