National and Kapodistrian University of Athens
Χειρουργική χημειοπροφύλαξη σε προγραμματισμένες επεμβάσεις κοιλίας (ανώτερου - κατώτερου πεπτικού) σύγκριση μιας έναντι τριών δόσεων
Abstract
dc:descriptionIntroduction: Chemoprophylaxis with a single dose of an appropriate antibiotic(during anaesthesia induction) is an important issue. It ensures that sufficientlevels of antibiotic concentrations, will be present in the tissues during surgicalmanipulations, thus protecting patients against microbial spreading, which inhigh concentrations may be the cause of Surgical Site Infection (SSI).SSI is the third more common cause of infection at a rate 14%-16% of allNosocomial infection in the hospitalized patients. In the surgical patient is themost common Nosocomial infection at a rate up to 38% of all infections. All ofthe above increase postoperative morbidity and mortality, length of stay (LOS)and cost of hospitalization. Recommendations for prevention of SSI are based onwell-designed scientific studies. However, the number of studies validating riskfactors and preventive measures is large and often with conflicting results. Theterm ‘risk factor’ in pathologic-physiology and SSIs prevention refers to avariable, which has a significant independent association with the developmentof SSI after a particular surgical intervention. The international literature listspartially risk factors, identified by single-agent and multivariate analyzes.The aim of this study is:1. To assess and record SSI in elective procedures (potentially infected) of theupper and lower digestive after applying predetermined chemoprophylaxisschemes.2. To assess and record risk factors of SSI in relation to:• Patient characteristics• Surgical intervention characteristics3. To assess possible risk factors to be associated with SSI with no data in theliterature. Material – methods: The study was prospective and randomized. It wasconducted in the 1st Surgical Department of Sismanoglion General Hospital inAthens, from February 2002 to February 2005. The study protocol wasapproved by the Scientific Council of the Hospital and written consent wasobtained from all patients included in the study.Patients scheduled to undergo surgery in the upper or lower digestive tract wererandomized to receive one or three doses of antimicrobial prophylaxis.Chemoprophylaxis was administered to the patients during induction ofanaesthesia, 60 minutes before performing the surgical incision in the skin. Arepeat dose intra-operatively was done when operative procedure exceeded 3hours or/and in patients having blood loss greater than 300ml. In patientsrandomized to receive three doses, administration was done every 8 hours.Patients with preoperative hospitalization duration greater than or equal to 15days and those who had to be operated on urgently due to obstruction, bleedingor inflammation of the upper and lower digestive tract, as well as patients withactive infection or receiving systemic antibiotic treatment, were excluded fromthe study. Cefuroxime (1,5 gr) was administered for the upper gastrointestinaltract at one or three doses according to randomization. For lowergastrointestinal tract interventions in case of known and certified allergy to β-lactams alternative administration of aztreonam (2 gr) was chosen incombination with metronidazole (1 gr) at one or three doses again according torandomization. For each patient a special tracking form was completed in whichepidemiological data were recorded, those related to the surgery, infections ofthe surgical field and postoperative morbidity. The patients were monitored withdaily visits to the ward, until discharged and follow-up completed in 30 daysafter surgery by telephone contact and regular review at the surgical clinicoutpatients’ department.Statistical analysis: The numerical variables are described as a mean ±1Standard Deviation (SD) and the categorical as a rate (%). Normality ofdistribution was estimated with Kolmogorov – Smirnov control and graphicalmethods. Comparisons between qualitative parameters were performed by χ2and Fisher’s Exact test. Comparisons between independent continuous variableswere performed by Student’s unpaired t-test και Mann – Whitney’s U test, asappropriate. Logistic regression analysis was performed by model introduction to known basic parameters. Their selection was either hierarchically or by gradualretrograde insertion, based on maximum likelihood ratio. All checks carried outwere double steering. The chosen level of statistical significance was a=0.05.Results: Patients presented by postoperative SSI had statistically significantolder age (p<0.001), longer preoperative hospitalization (p<0.001), moreaggravating factors (p=0.003), greater number of transfusions received(p<0.001) and longer duration of operative procedure (p<0.001).From the multivariate models that were implemented to investigate thedifferences in rates of SSIs, that observed between subgroups of patients, whoreceived one or 3 doses of antibiotic formulation, revealed that the lowerdigestive surgery showed statistically significant more frequent SSI (p<0.001).Namely, the relative risk of SSI for surgery of the upper digestive tract is fourtimes lower (RR: 0.237, 95% CI: 0.115-0.485). By comparing the two groupswho received chemoprophylaxis in elective abdominal procedures for upper andlower digestive tract and they were randomized to receive one versus 3 doses,there was no statistically significant effect of the number of doses of antibioticson postoperative SSI, after normalization of aggravating factors, comorbiditiesand type of surgery.Conclusion:• In all multivariate models that constructed either by hierarchical choice ofparameters or by automatic selection in order to calculate the maximumlikelihood model, there was no statistically significant effect of the numberof doses of antibiotics on postoperative SSI.• The construction of logistic regression models to examine, whether SSIcan be predicted or not in the study participants and to identifystatistically significant predictors, revealed a Multivariate Forecast Modelof Logistic Regression. This model is able to provide statisticallysignificant prediction of SSI in participants with 85,7% sensitivity and88,2% specificity (Nagelkerke R2: 0.55, χ2: 91.3,05, p<10-12). This modeldemonstrated very good generalization ability in total control of unknowncases with sensitivity 85,5% and specificity 85,7%.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2013
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Alexiou, Konstantinos
- Αλεξίου, Κωνσταντίνος
Subjects
dc:subject × 10Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/38236