Q1 · MEDICINE
Article
Author: Laurent, S ; Bertrand, C ; Spaas, P ; Scalliet, P ; Jouret-Mourin, A ; Van Cutsem, E ; Buset, M ; Peeters, M ; Penninckx, F ; Duinslaeger, M ; Haustermans, K ; Vaneerdeweg, W ; Claeys, D ; Burnon, D ; Sempoux, C ; Van de Stadt, J ; Humblet, Y ; Van Eycken, E ; Vindevoghel, K ; Melange, M ; Cabooter, M ; Rahier, J ; Demey, W ; Van Laethem, J L ; Kartheuser, A ; Laenen, A ; Smeets, P ; Daubie, M ; De Coninck, D ; Mansvelt, B ; Pattyn, P ; Thijs, A ; Op de Beeck, B ; Leonard, D ; Demetter, P
AbstractBackgroundResearch on the relationship between hospital volume and quality of care in the treatment of rectal cancer is limited.MethodsProcess and outcome indicators were assessed in patients with rectal adenocarcinoma who underwent total mesorectal excision, registered on a voluntary basis in the PROCARE clinical database. Volume was derived from an administrative database and analysed as a continuous variable. Sphincter preservation, 30-day mortality and survival rates were cross-checked against population-based data.ResultsA total of 1469 patients registered in PROCARE between 2006 and 2011 were included in this study. A volume effect was observed regarding neoadjuvant therapy for stage II–III disease, reporting of the circumferential resection margin, R0 resection rate, sphincter preservation rate, and number of nodes examined after chemoradiotherapy. The global estimate of quality of care was highly variable, but surgery was the single domain in which quality correlated with volume. No volume effect was observed for recurrence and overall survival rates. In the population-based data set (5869 patients), volume was associated with 30-day mortality adjusted for age (odds ratio 0·99, 95 per cent confidence interval (c.i.) 0·98 to 1·00; P = 0·014) and adjusted overall survival (HR 0·99 (95 per cent c.i. 0·99 to 1·00) per additional procedure; P = 0·001), but not with the sphincter preservation rate. Because of incomplete and biased registration on a voluntary basis, results from a clinical database could not be extrapolated to the population.ConclusionSome volume effects were observed, but their effect size was limited.