Page 23 - Cardiac abnormalities after aneurysmal subarachnoid hemorrhage

Impact of cardiac complications on outcome after aneurysmal subarachnoid hemorrhage
analysis presented (item 12b), source of funding and role of funders presented (item
13), number of patients at each stage of the study given (14a), reasons for non-
participation given (item 14b), period of recruitment defined (item 14c), baseline
characteristics given (item 15a), completeness of data for each baseline variable given
(item 15b), average and total amount of follow-up given (item 15c), number of
outcome events presented (item 16), association between determinants and outcomes 2 given (item 17a), categories of quantitative variables compared (item 17b), absolute
outcome data given (item 17c), subgroup analysis performed (item 18), key results summarized (item 19), limitations discussed (item 20), external validity of study findings discussed (item 21), overall interpretation of the results given (item 22).
Data Extraction
The three investigators who assessed quality and eligibility reviewed the publications independently. In case of disagreement, the authors reviewed the article in question together until consensus was reached. The following data were extracted: author, year of publication, study design (prospective or retrospective), definition of inclusion and exclusion criteria, number of included patients, gender, mean age, and follow-up period. The neurological condition on admission was dichotomized as good or poor according to the scoring system used in the particular article: Hunt-Hess2, World Federation of Neurosurgical Societies (WFNS)3, Glasgow Coma Scale (GCS)4, or Botterell5. A poor condition on admission was considered when Hunt-Hess ≥3, or WFNS ≥3, or GCS <12, or Botterell ≥3. Additionally, if an article did not use any of these scoring systems or used different criteria for poor outcome, the clinical condition reported by the authors was recorded.
As determinants we extracted the incidence of WMAs, diastolic dysfunction, elevated troponin levels, elevated (NT-pro)-BNP levels, elevated CK-MB levels, atrial fibrillation, tachycardia, bradycardia, wandering P, extrasystoles, peaked P, P mitrale, short PR interval, long PR interval, bundle branch block (BBB), pathological Q waves, ST segment depression, ST segment elevation, T wave changes (inverted or flat), U wave changes (present, inverted or prominent), left ventricular hypertrophy (LVH), and prolonged QT time. Combined ECG criteria were disregarded, because the separate ECG abnormalities may harbor different prognostic value. For outcome measurements we recorded the number of deaths from any cause, the number of patients with poor outcome, and the number of patients with DCI. Poor outcome was defined as death or dependence on activities of daily living, preferably defined by means of a handicap scale such
21