Page 163 - Prevention and Treatment of Incisional Hernia- New Techniques and Materials

Studies with one of the following characteristics were excluded: a mean follow-up of less than one year, less than 75% completion of follow-up of at least one year, or reporting on repair with a commercially not available mesh. Studies reporting series or cohorts of fewer than 10 patients operated over 3 years were excluded to eliminate small case series that were likely to be in uenced by learning curves, and to minimize selection and publication bias of ‘positive’ results. LIH in the iliac region or after lumbotomy were excluded. Studies were excluded if a full-text version was not available. Whenever multiple publications from institutions reporting the same cohort were encountered, only the most recent and complete article was included. Two reviewers independently assessed the titles and abstracts of all reports identi ed by electronic and manual searches. Any disagreement was resolved by discussion and consensus with the last author of this article.
In the results discrimination between ‘simple’ and ‘complex’ LIH was made. Simple LIH was de ned as a fascial defect over 10 cm (or surface over 100cm2) with intact soft tissue and skin and, if recurrent, with a not-infected mesh in situ from previous repair, comparable to the‘minor’complex abdominal hernias from the classi cation system of Slater (27). Complex LIH was de ned as a fascial defect over 10 cm (or surface over 100cm2) and associated problems of substantial loss of tissue, intra-abdominal infection, or if recurrent with infected mesh. LIH was also considered complex if during LIH repair a concomitant parastomal hernia was repaired. The category complex LIH includes most of the ‘moderate’ and ‘major’ complex abdominal hernias from the classi cation system of Slater(27).
Statistical analysis
To compare recurrence rates between repair techniques a generalized linear model (GLM) is used(28). Since the occurrence of individual recurrence is not reported, the risk of getting a recurrence is assumed equal during each month of follow-up of the study for not-a ected patients. Exponential survival curves are assumed, which are identical for all studies of a certain type of treatment, but which di er between treatments. Considering one study i, xi represents the sample size, ti the follow-up and yi the number of patients not experiencing the recurrence. The exponential survival curve is given by S(t)=exp(-at). a represents the angle of the slope of the curve and is estimated from the data of all inidividual studies reporting on one repair technique. The expected
8
Review treatment large IH
161