Mortality trends and the effects of débridement timing in the management of mediastinitis in the United States, 1998 to 2010

Abstract

Background: The authors examined the relationship between débridement delay and mortality for mediastinitis patients. The authors also assessed mortality trends for mediastinitis patients between 1998 and 2010.

Methods: The authors conducted a retrospective cross-sectional study with data from the Nationwide Inpatient Sample, 1998 to 2010. They studied adult patients, 18 years of age or older, who were surgically treated for mediastinitis. They used a logistic regression model adjusted for patient demographic and clinical characteristics to evaluate the association between timing of first operative débridement and in-hospital mortality. Using their logistic model, they calculated the adjusted probability of in-hospital mortality for each year of the study.

Results: Results showed that initial débridement after the fourth day of admission increased the odds of in-hospital mortality by 50 percent (odds ratio, 1.5; 95 percent confidence interval, 1.0 to 2.1). In addition, the adjusted probability of in-hospital mortality for an average patient treated for mediastinitis decreased from 10.6 percent in 1998 to 3.1 percent in 2010.

Conclusion: There is a survival advantage from timely initial débridement in mediastinitis patients.

Clinical question/level of evidence: Risk, II.

Figure 1

The algorithm used to select the study population. The patients included in the study are a representative nationwide sample of mediastinitis patients treated over the study period. ^ΔWe included these ICD-9 CM diagnosis codes after consultation with coders at our institution confirmed that these diagnosis codes are routinely used for sternal wound dehiscence and mediastinits. We recognize that coding practices differ between institutions. ^†Using the variables available the database, we calculated the “timing of debridement” from the day of admission for mediastinitis. We excluded mediastinitis patients who were never discharged from their index operation because “timing of debridement” would be otherwise erroneously calculated based on the admission date for their index operation.

Figure 2

The declining trend in the probability of in-hospital mortality for an average patient treated for mediastinitis over the study period.

Figure 3

Increasing mean co-morbidity burden (# of co-morbidities/patient) of patients treated for mediastinitis over the study period. The rate of in-hospital mortality declined over the same period in this patient population (Figure 2).