| 1. |
- Strömsöe, Anneli, et al.
(författare)
-
Validity of reported data in the Swedish Cardiac Arrest Register in selected parts in Sweden.
- 2013
-
Ingår i: Resuscitation. - : Elsevier BV. - 0300-9572 .- 1873-1570. ; 84:7, s. 952-956
-
Tidskriftsartikel (refereegranskat)abstract
- AIM: To describe differences and similarities between reported and non-reported data in the Swedish Cardiac Arrest Register in selected parts in Sweden. METHODS: Prospective and retrospective data for treated OHCA patients in Sweden, 2008-2010, were compared in the Swedish Cardiac Arrest Register. Data were investigated in three Swedish counties, which represented one third of the population. The recording models varied. Prospective data are those reported by the emergency medical service (EMS) crews, while retrospective data are those missed by the EMS crews but discovered afterwards by cross-checking with the local ambulance register. RESULT: In 2008-2010, the number of prospectively (n=2398) and retrospectively (n=800) reported OHCA cases was n=3198, which indicates a 25% missing rate. When comparing the two groups, the mean age was higher in patients who were reported retrospectively (69 years vs. 67 years; p=0.003). There was no difference between groups with regard to gender, time of day and year of OHCA, witnessed status or initial rhythm. Bystander cardiopulmonary resuscitation (CPR) was more frequent among patients who were reported prospectively (65% vs. 60%; p=0.023), whereas survival to one month was higher among patients who were reported retrospectively (9.2% vs. 11.9%; p=0.035). CONCLUSION: Among 3198 cases of OHCA in three counties in Sweden, 800 (25%) were not reported prospectively by the EMS crews but were discovered retrospectively as missing cases. Patients who were reported retrospectively differed from prospectively reported cases by being older, having less frequently received bystander CPR but having a higher survival rate. Our data suggest that reports on OHCA from national quality registers which are based on prospectively recorded data may be influenced by selection bias.
|
|
