Date de publication
Nom du journal
European Journal of Cancer
Objectives: The study investigated the association between hospital volume and observed survival of patients with a head and neck squamous cell carcinoma (HNSCC).
Methods: Overall, 9245 patients diagnosed with HNSCC between 2009 and 2014, were identified in the population-based Belgian Cancer Registry. This database was coupled with other databases providing information on diagnostic and therapeutic procedures, vital status, and comorbidities. The overall and relative survival probabilities were estimated using the Kaplan-Meier and the Ederer II methods, respectively. The relation between hospital volume and observed survival since diagnosis was then assessed using Cox proportional hazard models adjusted for potential confounders.
Results: The care for patients with HNSCC in Belgium was dispersed over more than 99 centres with half of the centres treating four or less patients with HNSCC per year. Survival probabilities were significantly better for patients treated in higher volume centres (>20 patients/year): the median survival of patients treated in these centres was 1.1 year longer (5.1 versus 4.0 years) than in lower volume centres. This association was confirmed in analyses taking the case-mix between hospitals into account: the hazard to die of any cause decreased on average with 0.4% per increase of one additionally treated patient. Beyond 20 assigned patients per year, there was no further decrease in the hazard to die.
Conclusions: Statistically significant and clinically relevant improved survival probabilities were obtained in patients treated in higher volume centres (>20 patients/year) compared with their peers treated in lower volume centres. This supports the recommendation to concentrate the care for patients with HNSCC in reference centres.
Methods: Overall, 9245 patients diagnosed with HNSCC between 2009 and 2014, were identified in the population-based Belgian Cancer Registry. This database was coupled with other databases providing information on diagnostic and therapeutic procedures, vital status, and comorbidities. The overall and relative survival probabilities were estimated using the Kaplan-Meier and the Ederer II methods, respectively. The relation between hospital volume and observed survival since diagnosis was then assessed using Cox proportional hazard models adjusted for potential confounders.
Results: The care for patients with HNSCC in Belgium was dispersed over more than 99 centres with half of the centres treating four or less patients with HNSCC per year. Survival probabilities were significantly better for patients treated in higher volume centres (>20 patients/year): the median survival of patients treated in these centres was 1.1 year longer (5.1 versus 4.0 years) than in lower volume centres. This association was confirmed in analyses taking the case-mix between hospitals into account: the hazard to die of any cause decreased on average with 0.4% per increase of one additionally treated patient. Beyond 20 assigned patients per year, there was no further decrease in the hazard to die.
Conclusions: Statistically significant and clinically relevant improved survival probabilities were obtained in patients treated in higher volume centres (>20 patients/year) compared with their peers treated in lower volume centres. This supports the recommendation to concentrate the care for patients with HNSCC in reference centres.