Abstract
The intracerebral hemorrhage (ICH) proportion in Romania, Bulgaria, and Hungary is believed to be far higher than the 10–15% typically seen in ethnically similar but more developed Western countries, and comparable to that of developing Asian and African countries. To clarify the proportion of stroke subtypes in Romania, we conducted retrospective and prospective analyses at the Mureş County Emergency Hospital, a major Romanian medical center. This hospital receives patients from a large area of the surrounding Tǎrgu-Mureş city and Mureş county, which are each roughly half Romanian and half Hungarian, with a 2–7% Roma (Gypsy) minority (1). CT scan use for each stroke admission plus subsequent reading by a radiologist for determination of the final diagnosis became standard for the Mureş County Emergency Hospital in mid-2003. ICH vs. cerebral ischemia (CI) were defined in accordance with the WHO MONICA criteria (2).
Paper charts for 60 542 admissions to the emergency department (ED) from January 01, 2004 to June 30, 2006, were obtained and manually examined for a final diagnosis of ICH or CI, and pertinent data were extracted and recorded; consecutive numbering of the 60 542 charts indicated that no records had been lost. The recorded data included the subtype (ICH or CI); age; gender; blood pressure; history of stroke, hypertension, atrial fibrillation, diabetes mellitus, hypercholesterolemia, smoking, aspirin medication, and warfarin medication; and whether the patient died in the ED (30-day case fatality information was not provided in the charts).
This retrospective study was followed by a prospective analysis consisting of 260 consecutive stroke cases from August 08, 2006 until November 30, 2006, which were also classified as ICH or CI by a radiologist's CT scan reading. This study involved obtaining all the data categories of the retrospective study as well as a stroke score at presentation and information concerning each patient's hospital course.
The retrospective study indicated an ICH proportion of 26·8%, and ICH proved to be the far deadlier subtype as it accounted for 63·4% of all ED mortalities. Stated otherwise, the ED mortality proportions were 11·2% for ICH and 2·1% for CI. For the prospective study, the ICH proportion was 21·6%, which represented 69·2% of ED mortalities and 58·8% of mortalities during the entire hospital stay. The ICH mortality proportions were 16·1% in the ED and 17·9% during the hospital course, compared with CI mortality proportions of 2·0% and 3·5%, respectively. ICH patients presented with a mean stroke score of 13·2 (12·1 among survivors and 28·0 among nonsurvivors), while CI patients averaged 8·4 (8·2 among survivors and 16·7 among nonsurvivors).
While the ICH proportions appear to be substantially elevated in both studies, there are several compounding factors concerning health care utilization and delivery that likely skewed the proportions upward. From our experiences talking with patients, we believe that patients in Romania are less likely to seek medical evaluation for a given severity of symptoms than are patients in the West, and ambulance services are certainly less accessible. Because the symptoms of ICH tend to be more severe than those of CI, a relatively greater percentage of patients experiencing CI may not seek treatment in the ED in Romania, thus elevating the proportion of recorded ICH cases. Additionally, as the elapsed time between the onset of symptoms and the ED presentation is likely longer in Romania, many CI cases may have converted into ICH before arrival in the ED and were thus recorded as ICH upon presentation (radiologists can usually differentiate, but often do not do so). Furthermore, the Mureş County Emergency Hospital is the premier medical center in the region, so it may receive a disproportionate number of the more severe cases (which are disproportionately ICH), creating a referral bias that elevates the ICH proportion. The extent of these biases is unknown and was not measured in these studies.
Owing to these unmeasured biases, we can not confidently state that the 26·8% and 21·6% ICH proportions seen in the retrospective and prospective analyses, respectively, indicate that the ICH proportion is indeed higher than the 10–15% seen in Western countries. Nevertheless, it can be reliably stated that the ICH proportion at ED presentation was significantly elevated. Because ICH is far deadlier than CI, this phenomenon is important for resource allocation, emergency care protocols, and triage, especially in overcrowded and underfunded hospitals.