Treatment of postoperative sternal dehiscence with mediastinitis: Twenty-four-year use of a single method

J Ernesto Molina, Elaine C. Nelson, Robin R.A. Smith

Research output: Contribution to journalArticlepeer-review

30 Scopus citations

Abstract

Objective: Postoperative deep sternal wound infection with dehiscence carries a high mortality rate, a high morbidity rate, and a poor cure rate. We developed a standard protocol of care to treat this complication, achieving primary closure and cure of the infection. Methods: From January 1, 1981, through May 31, 2005 (24 years 5 months), we treated 114 patients with dehiscence and mediastinitis. The diagnosis was made from 4 to 56 days (mean, 14.5 days) after surgery. Mean age was 64.3 years (range, 38-84 years); 79 (69%) were obese, and 48 (42%) had diabetes. Ten had previous attempts (1-4) of repair with other methods. Treatment entailed (1) debridement without removal of bone, (2) bilateral dissection of skin and subcutaneous tissue as one layer, (3) implantation of a staggered double-tube irrigation-suction system posterior and another one anterior to the sternum, (4) lateral reinforcement of the sternum and reclosure with a double wire, and (5) a single-layer closure of the subcutaneous tissue and skin. Results: Of 114 patients, 109 (96%) had mediastinitis, positive for Staphylococcus species in 101 (92.6%). The cure rate was 98% (112/114); hospital stay was 14 days (range, 12-16 days), with no deaths. Conclusions: Use of this standard protocol is effective and highly recommended. It spares the sternum, cures the infection, and leaves the patient physically functional without the use of soft tissue flaps.

Original languageEnglish (US)
Pages (from-to)782-787
Number of pages6
JournalJournal of Thoracic and Cardiovascular Surgery
Volume132
Issue number4
DOIs
StatePublished - Oct 2006

Fingerprint

Dive into the research topics of 'Treatment of postoperative sternal dehiscence with mediastinitis: Twenty-four-year use of a single method'. Together they form a unique fingerprint.

Cite this