Endoscopy 2019; 51(04): S237-S238
DOI: 10.1055/s-0039-1681883
ESGE Days 2019 ePosters
Friday, April 5, 2019 09:00 – 17:00: Esophagus ePosters
Georg Thieme Verlag KG Stuttgart · New York

EARLY DIAGNOSIS IS ASSOCIATED WITH IMPROVED CLINICAL OUTCOME IN BENIGN ESOPHAGEAL PERFORATIONS: AN INDIVIDUAL PATIENT DATA META-ANALYSIS

B Vermeulen
1   Gastroenterology and Hepatology, Radboud University Medical Center, Nijmegen, Netherlands
,
B van der Leeden
1   Gastroenterology and Hepatology, Radboud University Medical Center, Nijmegen, Netherlands
,
C Rosman
2   Surgery, Radboud University Medical Center, Nijmegen, Netherlands
,
P Siersema
1   Gastroenterology and Hepatology, Radboud University Medical Center, Nijmegen, Netherlands
› Author Affiliations
Further Information

Publication History

Publication Date:
18 March 2019 (online)

Also available at
 
 

    Aims:

    Benign esophageal perforations (BEP) are subdivided in two groups;

    1. spontaneous esophageal perforations, also known as Boerhaave's syndrome, and

    2. iatrogenic esophageal perforations.

    Timing between onset and diagnosis of BEP is regarded as an important risk factor for poor outcome. However, no strong evidence exists to support this finding. We investigated whether timing between onset and diagnosis is associated with clinical outcome of patients with BEP.

    Methods:

    A systematic review (PROSPERO: CRD42018093473) was performed following PRISMA guidelines. Clinical studies of patients treated for BEP were identified from Medline, Embase and Cochrane databases. After study inclusion, corresponding authors were invited to share individual patient data (IPD) and a meta-analysis was performed. Patients were subdivided in two groups; (1) early diagnosis (≤24hours); and (2) late diagnosis (> 24hours), after symptom onset. We used multi-level mixed model logistic regression analysis to compare both groups while correcting for age, gender, etiology, esophageal location and initial treatment strategy. Outcome included mortality, intensive care unit (ICU) admittance and re-interventions.

    Results:

    The systematic search yielded 146 studies eligible for inclusion. If possible, we invited corresponding authors of included studies (n = 115) to share IPD. In total, 25 authors (22%) responded and shared IPD. Out of 960 patients, 672 patients (iatrogenic n = 411, spontaneous n = 261) were included in the IPD meta-analysis. After multivariable logistic regression analysis, late diagnosis (> 24h) was associated with 6% increase in mortality (95% CI 1.20 – 4.10, OR 2.2, p = 0.01), 17% increase in ICU admittance (95% CI 1.21 – 3.34, OR 2.0, p = 0.007), and 19% increase in need for re-intervention (95% CI 1.48 – 3.35, OR 2.2, p < 0.001), when compared with early diagnosis (≤24h) of BEP.

    Conclusions:

    This real-world IPD meta-analysis shows that late (> 24h) diagnosis is associated with higher mortality, ICU admittance and re-intervention rates, when compared with early (≤24h) diagnosed patients with BEP. This association confirms results from previously published small cohort studies and expert opinion.


    #