OBJECTIVES: In patients with high-grade dysplasia (HGD) in Barrett's esophagus (BE), it is incompletely known which factors are associated with developing esophageal adenocarcinoma (EAC). We analyzed prior biopsy and follow-up strategies in a large nationwide population-based cohort of patients with HGD in BE, and identified predictors of EAC progression. METHODS: Prior biopsy records and follow-up evaluations were studied in patients with HGD in BE diagnosed between 1999 and 2008, using PALGA, a nationwide network and registry of histopathology and cytopathology in the Netherlands. Multivariate Cox proportional hazards regression analysis was performed to identify predictors for prevalent (≤ 6 months) and incident (> 6 months) EAC. RESULTS: In total, 827 patients with HGD in BE were included. Follow-up data after HGD diagnosis were available in 699 (85%) patients. In 249 (36%) of these patients, an EAC was detected (14.1 EACs per 100 person-years). The risk of prevalent EAC (n=177) was lower with previous surveillance (hazards ratio 0.7; 95% confidence interval 0.5-0.9), unifocal HGD (0.3;0.2-0.6), diagnosis in a university hospital (0.5;0.3-0.9), endoscopic resection (0.5;0.3-0.7), or ablation (0.0;0.0-0.3); and higher when patients were 65-75 years (1.5;1.04-2.04). After exclusion of prevalent EACs, the progression rate was 4.2 EACs per 100 person-years. The risk of progression to incident EAC (n = 72) was lower with previous surveillance (0.6;0.3-0.9) and ablation (0.2;0.0-0.8), and higher when > 75 years (3.8;2.0-7.2) or with an interval > 6 months between HGD diagnosis and first follow-up (e.g., 7-12 months 2.9;1.3-6.3). CONCLUSIONS: In this cohort of patients with HGD in BE, the EAC detection rate was 14.1 per 100 person-years and 4.2 per 100 person-years after excluding prevalent cases. The risk of both prevalent and incident EAC was reduced with previous surveillance and endoscopic treatment, while it was increased with older age.