Cardiac tamponade: an initial study in the development of a predictive tool.

This exploratory study was conducted over a 12-month period to describe the relationship between selected clinical variables and the occurrence of postoperative cardiac tamponade. Data were collected on 19 variables. A total of 739 patients were included in the study. A postoperative diagnosis of cardiac tamponade was made in nine of these patients who then comprised the tamponade group. A matched non-tamponade control group of 18 patients was identified for comparison. The following were found to be significantly different in a statistical comparison between variable values in the tamponade and non-tamponade groups: (1) serum creatinine level (p = 0.005), (2) chest tube drainage (p = 0.009), (3) sustained pressure plateau (p = .018), and (4) mediastinal widening (p = 0.039). Two additional variables, pulsus paradoxus and elevated BUN, appeared to be of value in the assessment of late tamponade. Discriminant analysis showed a high predictive accuracy of the above combined variables. Such results warrant the further investigation of these variables in other clinical settings. Both a replication and testing of a tamponade score system would be a fertile and necessary area of research for critical care nurses who provide care for post-cardiac surgery patients. These data suggest that equal weight be assigned to each of the four variables. Thus scores would range from 0 to 4; higher scores would represent greater risk (Table III). With further investigation of this area of clinical practice, the identification of patients at higher risk for tamponade could be greatly facilitated.