Abstract:
Aims: To determine the impact of timing, duration, number, combination and sequence of organ failure (OF) on clinically meaningful outcomes in patients with acute pancreatitis (AP) and the combined impact of certain characteristics of OF and infected pancreatic necrosis on mortality. Methods: This thesis comprises three stages. First, a literature review to identify limitations in existing knowledge regarding OF in AP, was conducted. The second stage was to contact researchers in the field of clinical pancreatology who are currently active and invite them to contribute individual patient data. To be eligible, patients had to meet the criteria of AP, have developed at least one OF during the first week of hospitalisation, and not be enrolled in an intervention study. The final stage was to check, correct, pool and perform statistical analysis of the data. The analysis used was based on a generalised linear mixed-effects logistic regression model (with adjustment for year of enrolment, age, sex, aetiology, and admission APACHE II score) in the statistical programme R 2.15.2 Results: This study found that AP patients with OF for two or more days have significantly increased odds of mortality compared to patients with OF for one day only. The number OFs suffered is also significantly associated with mortality. Interestingly, certain combinations of two OFs (i.e. renal and cardiovascular failures) have a significantly higher mortality than other combinations of two OFs (i.e. renal and respiratory failures). Further the sequence in which these OF occur (i.e. renal followed by cardiovascular failure compared to concurrent OF) have a significant impact on mortality. The presence of infected pancreatic necrosis has a synergistic effect on mortality. Conversely the timing of the first OF, was not significantly associated with mortality. Conclusions: This study comprehensively investigates whether the characteristics of OFs influence mortality outcomes of AP patients. Information derived from this study can be used to optimise routine clinical management and improve clinical care strategies. The study results also validate outcome definitions, and allow comparability of results which will be available to form a more accurate basis for patient allocation in future clinical studies.