Methods

Patient group 

From September 2016 until September 2021, surgery registration databases were retrospectively reviewed for perforation during invasive cardiac procedures requiring surgical exploration. Cardiac perforation was defined as the occurrence of a pericardial effusion during or immediately after the procedure, associated with hemodynamic instability requiring surgical exploration. 25 consecutive patients were selected; 9 underwent primary full sternotomy, 14 underwent VATS, 2 were converted from VATS to sternotomy. Baseline characteristics were gender, body mass index (BMI) and age. Primary outcomes were duration of intensive care unit (ICU) stay, duration of hospital stay, pericardiocentesis, cardiopulmonary resuscitation, 30-day mortality and all-round mortality. Cardiopulmonary resuscitation was defined by active chest compressions perioperatively due to loss of hemodynamic output. Pericardiocentesis was defined as percutaneous placement of any form of pericardial drainage preoperatively.

Surgical method 

Surgical method by sternotomy consisted of urgent full sternotomy followed by opening of the pericardium, central cannulation and surgical exploration followed by closure of the perforation by purse string reinforced by teflon felt.
In the thoracoscopy group, three 5mm trocars were placed in either the left or right axillary region based on the most probable site of the perforation, after consultation with the cardiologist. Surgical exploration and inspection after incision of the pericardium was done. Perforation was closed in the same manner, by purse string reinforced by teflon felt.
Connection to the extracorporeal circulation (ECC) was by central cannulation in the sternotomy-group and by femoral cannulation in the thoracotomy-group. Indications for connection to the ECC were hemodynamical instability, excessive blood loss or perforation not accessible on beating heart or with lungs inflated.

Statistical analysis 

Baseline study population characteristics by outcome group were described using proportions for categorical variables and mean ± SD for continuous variables. Statistical tests were Mann-Whitney-U, Chi-Square and Fisher’s exact test. A significance level of p < 0.05 was used. All tests were 2-sided. Statistical analysis was made by SPSS®.

Results