Twenty-first century cancer patterns in small island nations: Grenada and the English-speaking Caribbean.

PURPOSE: Grenada is a small island nation of 105,000 in the Caribbean with one single general hospital and pathology laboratory. This study assesses cancer incidence on the island based on existing pathology reports, and compares the cancer mortality burden between Grenada and other Caribbean nations. METHODS: Age-adjusted overall and site-specific cancer "incidence" rates (based on pathology reports) and mortality rates were calculated and compared for 2000-2009. Next, mortality rates for a more recent period, 2007-2013, were calculated for Grenada and a pool of English-speaking, majority African-ancestry Caribbean island nations. Lastly, for direct mortality comparisons by cancer site, mortality rate ratios were computed using negative binomial regression modeling. RESULTS: The pathology reports alone do not suffice to calculate national incidence rates but cancer mortality rates are rapidly increasing in Grenada. The leading causes of cancer mortality were prostate and lung cancers among men, and breast and cervical cancers among women. Overall cancer mortality is significantly higher for both male and female Grenadians than their Caribbean counterparts: RR 1.43 (95% CI 1.32-1.55) and RR 1.26 (95% CI 1.15-1.38), respectively. High prostate and non-Hodgkin's lymphoma rates are concerning. CONCLUSIONS: Given the small existing cancer infrastructure, excessive mortality in Grenada compared to its neighbors may be disproportionately more attributable to low survival than a high cancer risk. Global solutions will be required to meet the cancer control needs of geographically isolated small nations such as Grenada.

An AAST-MITC analysis of pancreatic trauma: Staple or sew? Resect or drain?

INTRODUCTION: Pancreatic trauma results in high morbidity and mortality, in part caused by the delay in diagnosis and subsequent organ dysfunction. Optimal operative management strategies remain unclear. We therefore sought to determine CT accuracy in diagnosing pancreatic injury and the morbidity and mortality associated with varying operative strategies. METHODS: We created a multicenter, pancreatic trauma registry from 18 Level 1 and 2 trauma centers. Adult, blunt or penetrating injured patients from 2005 to 2012 were analyzed. Sensitivity and specificity of CT scan identification of main pancreatic duct injury was calculated against operative findings. Independent predictors for mortality, adult respiratory distress syndrome (ARDS), and pancreatic fistula and/or pseudocyst were identified through multivariate regression analysis. The association between outcomes and operative management was measured. RESULTS: We identified 704 pancreatic injury patients of whom 584 (83%) underwent a pancreas-related procedure. CT grade modestly correlated with OR grade (r 0.39) missing 10 ductal injuries (9 grade III, 1 grade IV) providing 78.7% sensitivity and 61.6% specificity. Independent predictors of mortality were age, Injury Severity Score (ISS), lactate, and number of packed red blood cells transfused. Independent predictors of ARDS were ISS, Glasgow Coma Scale score, and pancreatic fistula (OR 5.2, 2.6-10.1). Among grade III injuries (n = 158, 22.4%), the risk of pancreatic fistula/pseudocyst was reduced when the end of the pancreas was stapled (OR 0.21, 95% CI 0.05-0.9) compared with sewn and was not affected by duct stitch placement. Drainage alone in grades IV (n = 25) and V (n = 24) injuries carried increased risk of pancreatic fistula/pseudocyst (OR 8.3, 95% CI 2.2-32.9). CONCLUSION: CT is insufficiently sensitive to reliably identify pancreatic duct injury. Patients with grade III injuries should have their resection site stapled instead of sewn and a duct stitch is unnecessary. Further study is needed to determine if drainage alone should be employed in grades IV and V injuries. LEVEL OF EVIDENCE: Epidemiologic/Diagnostic study, level III.