206 An example of collaboration between a regional network and a cancer registry

Context One of the objectives of the ‘Plan-cancer 2003–2007’ was to offer to every patient with cancer a multidisciplinary care management of the disease. This multidisciplinary approach makes in particular reference to the implementation of Multidisciplinary Consulting Meetings (MCM) during which every new case must be in theory discussed. To estimate MCM exhaustiveness we should be able to identify all the new diagnosed cases of cancer. In France such a work is only possible in areas where cancer registries exist. Objective To estimate MCM exhaustiveness for patients with prostate cancer diagnosed in 2007 and who lived in Tarn9s department at diagnosis. To study factors influencing the probability for a patient to have his file discussed during MCM. Program All the prostatic cancers diagnosed in the population of Tarn in 2007 were identified by cancers registries. The constituted list was compared with all the cases discussed during MCM in Midi-Pyrenees. By considering various situations (a case had to be discussed during MCM within three months following diagnosis, a case had to be discussed during MCM before treatment), we calculated the exhaustiveness as the ratio between the number of discussed cases from Tarn (data from regional cancer network) and the number of cancers in Tarn (data from cancer registry). Various information collected by registry of Tarn (age, initial extension of the tumour, the score of Gleason, rate of initial PSA, realised treatment) were used to identify factors associated with the probability for a patient to be discussed during MCM. Univariate (χ 2 ) and multivariate analysis (logistic regression) were performed. Results Four hundred fifty nine cases of prostatic cancers were identified. Among them, 59% of the cases were discussed during MCM within 3 months following diagnosis and 58% were discussed above all treatment. The frequency of the discussion during MCM decreased with age, particularly after 80 years. Regarding gravity factors, only the fact to have a limited tumour was associated with a lower probability for patients to be discussed, whatever age. On the other hand this probability was strongly associated with the fact to be treated by prostatectomy. The prostatectomy was associated with a lower frequency of discussion during RCP and with a delayed discussion after treatment. These results were presented to the members of the regional cancer network (in particular urologists). It is too early to see an impact on the practice but a new comparison between registry data and data from the regional cancer network will be done for the cases diagnosed in 2009. Discussion There are in France cancer registries in 17 departments corresponding to 10 regions. The implementation of collaboration between these registries and the regional cancer network would allow having easily a good measure of the exhaustiveness of patients who have access to MCM on a large part of the French territory. In areas without registry, some methods based on the identification of cases using ‘PMSI’ could be developed. It would be interesting to test this method by comparing its results with those observed in areas with registry. Conclusion This type of study is easy to perform if regional cancer networks and the cancer registries work together. The interest of this collaboration exceeds widely the frame of the current study because it allows the network to take advantage of the experience of cancer registries about evaluation of medical practices in population. But this collaboration is also interesting for cancer registries by facilitating the access to information usually scattered. Contexte Un des objectifs du Plan cancer 2003–2007 etait d9offrir a chaque patient atteint de cancer une prise en charge pluridisciplinaire de sa maladie. Cette pluridisciplinarite fait notamment reference a la mise en place des Reunions de Concertation Pluridisciplinaire (RCP) au cours desquelles le dossier de tous les nouveaux patients doit etre presente et eventuellement discute. Evaluer l9exhaustivite du passage en RCP suppose d9etre capable d9identifier tous les nouveaux cas de cancer diagnostiques. En France un tel recensement n9est realise que dans les zones ou existe un registre de cancer. Objectif Evaluer l9exhaustivite de la presentation en RCP des cas de cancer de la prostate diagnostiques chez des residents du Tarn et etudier les facteurs modulant cette presentation. Programme (description, mise en œuvre, elements de suivi) Tous les cas de cancers prostatiques diagnostiques dans la population du Tarn en 2007 ont ete recenses par le registre des cancers. La liste ainsi constituee a ete confrontee a l9ensemble des cas passes en RCP dans la region Midi-Pyrenees. En considerant differentes situations (un cas devait passer en RCP dans les 3 mois apres son diagnostic, un cas devait passer en RCP avant tout traitement), nous avons calcule l9exhaustivite du taux de passage en RCP. Differentes informations collectees par le registre et permettant de connaitre les caracteristiques des cas (âge, extension initiale de la tumeur, score de Gleason, taux de PSA initial, traitement realise) ont permis d9etudier les facteurs associes a la presentation (ou non presentation) en RCP. Une analyse univariee (χ 2 ) puis multivariee (regression logistique) ont ete realisees. Resultats L9etude porte sur 459 cas de cancers prostatiques. Parmi eux 59% des cas ont ete presentes en RCP dans les 3 mois qui ont suivi leur diagnostic et 58% l9ont ete avant tout traitement. La frequence de la presentation chute avec l9âge, notablement apres 80 ans. Parmi les facteurs de gravite seul le fait de presenter une tumeur limitee est associee a une non presentation et cette non presentation n9est pas due a la mediation de l9âge. En revanche elle est fortement associee au fait d9etre traite par prostatectomie. La prostatectomie est associee a une moindre frequence de presentation en RCP et a une presentation tardive apres traitement. Les resultats de cette etude ont ete diffuses aux membres du reseau (notamment les urologues). Il est encore trop tot pour voir un impact sur les pratiques mais une nouvelle comparaison entre donnees du registre et donnees du DCC sera realisee pour les cas diagnostiques en 2009. Discussion (perspectives de developpement, limites) Il existe en France des registres de cancer dans 17 departements correspondant a 10 regions differentes. La mise en place d9une colaboration entre ces registres et les reseaux regionaux de cancerologie permettrait de disposer facilement d9une bonne mesure de l9exhaustivite du passage en RCP sur une fraction non negligeable du territoire francais. Dans les zones sans registre des methodes basees sur l9identification des cas par le PMSI peuvent etre developpees. Il serait interessant de les etalonner en les comparant avec les resultats obtenus dans les zones avec registres. Conclusion (lecons apprises et messages pour les autres) Ce type d9etude est facile a realiser si le reseau de cancerologie et un registre de cancer collaborent. L9interet de cette collaboration depasse largement le cadre de l9etude de l9exhaustivite car elle permet au reseau de profiter de l9experience du registre en matiere d9evaluation des pratiques en population. Mais elle est aussi benefique au registre en lui facilitant l9acces a des informations habituellement dispersees.