Why do complications of gastrointestinal disease and procedures come as serial rather than singular events?

Abstract

It often appears as if complications arising from gastrointestinal disease and interventional procedures come in 'series' and 'cluster' in individual patients. The aim of this study was to analyse the clustering of complications in terms of stochastic modelling and Markov chains. A patient with gastrointestinal disease is simulated to move along either a 'bad track' associated with multiple consecutive complications or a 'good track' free of complications, the transitions within each track and between the two tracks being governed by probability values. Because the occurrence of a single complication increases the risk of further complications, subjects who encounter their first complication are more likely to experience a second, third or even more complications, before they leave the 'bad track' and their personal chain of cascading complications becomes discontinued. The model of a Markov chain explains why the overall number of complications in the total population of patients with digestive disease remains low even if individual patients are expected to encounter more than a single complication. Published complication rates for gastrointestinal procedures underestimate the true risk to the patient, because they do not consider the added vulnerability to cascading complications after the first event.