Capsule endoscopy of the small bowel in the clinical practice: outpatient management is feasible and cheaper.

AIM: Four Italian regions have cost coding for outpatient capsule-endoscopy. Elsewhere it is performed in ordinary hospital admission. To identify, in a cohort of patients of a Gastroenterology Unit, those feasible for outpatient versus inpatient treatment; to analyze costs distribution in both management areas. METHODS: We retrospectively analysed 100 clinical records of admissions to A.O. San-Carlo-Borromeo, Milan between 2005-2008. Hospitalization criteria (at least 3): 1) occult/obscure gastrointestinal bleeding; 2) hemoglobin ≤ 8 gr/dL; 3) indication for blood transfusions; 4) urgent hospital admission. RESULTS: A total of 62 patients had urgent admission, 60 blood transfusions, 81 underwent EGD and colonoscopy, 8 enteroscopy and 5 surgery. Mean haemoglobin value was 8.67 g/dL. Capsule-endoscopy was positive in 70, uncertain in 8, negative in 22. Positive cases: 33 angiodyplasia, 18 ulcers/erosions, 13 polyps/masses, 5 overt bleeding, 1 celiac disease. 47/100 were appropriate as outpatient, saving 432 days of hospital stays. Admission coding was grouped into 7 DRGs (overall expense: 98,366 Euros). Considering EGD/colonoscopy outpatient costs and 1.100 euros as estimated value for capsule-endoscopy, the total expense could be 53.919. CONCLUSION: Outpatient capsule-endoscopy small bowel examination is feasible in half of the cases. It is cost saving, (about 45.000 Euros/100 patients), reducing inappropriate hospital stays.

Toward optimal classifier system performance in non-Markov environments.

Wilson's (1994) bit-register memory scheme was incorporated into the XCS classifier system and investigated in a series of non-Markov environments. Two extensions to the scheme were important in obtaining near-optimal performance in the harder environments. The first was an exploration strategy in which exploration of external actions was probabilistic as in Markov environments, but internal "actions" (register settings) were selected deterministically. The second was use of a register having more bit-positions than were strictly necessary to resolve environmental aliasing. The origins and effects of the two extensions are discussed.