The Pittsburgh perforation severity score (PSS) can be used to improve decision making in the management of esophageal perforation, findings from a retrospective, multicenter study have shown.
Dr. Michael Schweigert and his colleagues performed a study of 288 patients with esophageal perforation treated at 11 centers between 1990 and 2014, using them as a completely independent population to validate whether the PSS could be used to stratify such patients into discrete subgroups with differential outcomes.
The PSS was analyzed using logistic regression as a continuous variable and stratified into low, intermediate and high score groups, according to their report published in the Journal of Thoracic and Cardiovascular Surgery (2016 Apr;151:1002-11).
Operative management was more frequent than nonoperative management (200 patients, or 69.4% vs. 30.6%), according to Dr. Schweigert of the Städtisches Klinikum Dresden Friedrichstadt, Germany, and his colleagues. Patients with esophageal cancer (34/43; 79%) and stricture (18/23; 78.3%) mainly were treated operatively. The most common type of surgery was primary repair (83 patients), followed by surgical drainage (38 patients).
Perforation-related morbidity was seen in 180 patients (65%), with sepsis (21%) and pneumonia (19%) being most common. Overall in-hospital mortality was 20%, and the median length of stay was 27 days.
Patients with fatal outcomes had a significantly higher median PSS score (11 vs. 1) and the median PSS was significantly higher in operatively managed cases, compared with nonoperative cases (5 vs. 4, P = .0001). The researchers found that the PSS score predicted morbidity well, with an area under the curve (AUC) of 0.77, as well as mortality (AUC = 0.83). However, prediction of the need for operative management was not as good (AUC = 0.65).
Based upon their analysis, the researchers proposed a treatment decision tree in which group I (low PSS patients) should have a focus of nonoperative management. Group 2 patients (medium PSS) with non–contained leak preferably should be managed by surgery.
They found that the high-risk group (PSS greater than 5) had the worst prognosis and highest mortality, with the odds for mortality being 8 times higher than that the intermediate group and 18 times higher than the low-risk group. “Because these patients are most endangered by esophageal perforation, early and aggressive treatment is mandatory to avoid fatal outcomes,” the authors stated.
They found that nonoperative management was not associated with higher mortality or more unfavorable outcome regarding perforation-related morbidity or length of stay, but they pointed out that nonoperative treatment was only successful in 60% of cases, with 36 out of the 88 nonoperative patients eventually undergoing surgery and 8 undergoing esophagectomy. But patients with a high perforation severity score were 3.37 times more likely to have operative management compared to low-scoring patients. “Better selective criteria for nonoperative management are urgently required,” they stated.
“The Pittsburgh PSS is helpful to assess the severity and potential consequences of esophageal injury and stratifies patients into low-, intermediate-, and high-risk groups with differential morbidity and mortality outcomes. Prospective studies are required to analyze the influence of the Pittsburgh scoring system on the treatment of esophageal perforation,” the researchers concluded.
The authors reported having no disclosures.
A webcast of the AATS Annual Meeting presentation of this paper is available.