Correcting bias due to missing stage data in the non-parametric estimation of stage-specific net survival for colorectal cancer using multiple imputation |
| |
| Institution: | 1. University College London, UK;2. London School of Hygiene and Tropical Medicine, UK;3. MRC Clinical Trials Unit at UCL, London, UK;1. Registre des cancers du Bas-Rhin, Laboratoire d’Épidémiologie et de Santé Publique, EA3430, FMTS, Université de Strasbourg, 4 rue Kirschleger, 67085 Strasbourg CEDEX, France;2. Service de Santé Publique, Hôpitaux Universitaires de Strasbourg, 1 place de l’hôpital, 67091 Strasbourg CEDEX, France;3. Service d’épidémiologie et de biostatistique, Centre Paul Strauss, 3 rue de la porte de l’hôpital, 67065 Strasbourg CEDEX, France;4. Registre des tumeurs de l''Hérault, Centre de Recherche, 208 rue des Apothicaires, 34298 Montpellier CEDEX 5, France;5. Registre des cancers de la Manche, Centre Hospitalier Public du Cotentin, 46 rue du Val de Saire, 50102 Cherbourg-Octeville, France;6. Registre des cancers du Tarn, 1, rue Lavazière BP 37, 81001 Albi CEDEX, France;7. Institut Claudius Regaud, IUCT-O, LEASP – UMR 1027 Inserm – Université Toulouse III, 1 avenue Irène Joliot-Curie, 31059 Toulouse CEDEX 9, France;8. Registre du cancer de la Somme, Service Épidémiologie, Hygiène et Santé Publique, Centre Hospitalier Universitaire Nord, 1 place Victor Pauchet, 80054 Amiens CEDEX 1, France;9. Registre des hémopathies malignes de Basse-Normandie, Unité Fonctionnelle Hospitalo- Universitaire n°0350, Centre Hospitalier Universitaire Nord, avenue de la Côte de Nacre, 14033 Caen CEDEX, France;10. Registre des cancers de Loire-Atlantique et Vendée, Centre Hospitalier Universitaire de Nantes, 50 route de Saint-Sébastien, 44093 Nantes CEDEX 1, France;11. Registre général des cancers de Lille et de sa région, GCS C2RC, Centre Hospitalier Régional Universitaire de Lille Hôpital Calmette, boulevard du Professeur Jules Leclercq, 59037 Lille CEDEX, France;12. Registre des tumeurs du Doubs et du Territoire de Belfort ? EA3181, Centre Hospitalier Régional Universitaire de Besançon Saint-Jacques, 2 place Saint-Jacques, 25030 Besançon CEDEX, France;13. Registre des tumeurs digestives du Calvados, Cancers & Préventions – U 1086 Inserm, Centre François Baclesse, 3 avenue du Général Harris, BP 5026 14076 Caen CEDEX 5, France;14. Registre des cancers de l’Isère, Centre Hospitalier Universitaire de Grenoble Pavillon E, boulevard de la Chantourne BP 217, 38043 Grenoble CEDEX 9, France;15. Registre général des tumeurs du Calvados, Cancers & Préventions – U 1086 Inserm, Centre François Baclesse, 3 avenue du Général Harris, BP 5026 14076 Caen CEDEX 5, France;p. Francim: Réseau français des registres des cancers, 31073 Toulouse, France;1. Division of Epidemiology and Biostatistics, School of Public Health, University of Illinois at Chicago, Chicago, United States;2. University of Illinois at Chicago Cancer Center, Cancer Control and Population Science Research Program, Chicago, United States;3. Institute for Health Research and Policy, Chicago, United States;4. Department of Otolaryngology-Head and Neck Surgery, University of Illinois at Chicago, Chicago, United States;5. School of Public Health, University of Alberta, Edmonton, Alberta, Canada;6. University of Illinois at Chicago, Department of Pediatrics, Chicago, United States;7. Survey Research Laboratory, Public Administration, University of Illinois at Chicago, 412 South Peoria Street, Chicago, 60607, United States;8. University of Illinois at Chicago, Department of Medicine, Chicago, United States;9. Ophthalmology and Visual Sciences, University of Illinois at Chicago, Chicago, United States;1. Division of Population Science, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA 02215, United States;2. Center for Clinical Epidemiology & Biostatistics, University of Pennsylvania, Philadelphia, PA 19104, United States;3. Division of Nephrology, Department of Medicine, Drexel University College of Medicine, Philadelphia, PA 19129, United States;1. Department of Public Health, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan;2. Department of Social and Behavioral Sciences, Harvard T.H. Chan School of Public Health, 677 Huntington Avenue, 7th floor, Boston, MA, 02115 USA;3. Department of Urology, Kanto Rosai Hospital, 1-1 Kizukisumiyoshi-cho, Nakahara-ku, Kawasaki, Kanagawa, 211-8510, Japan;1. Cancer Epidemiology and Health Services Research Group, Centre for Public Health, Queen’s University Belfast, Belfast, Northern Ireland, United Kingdom;2. School of Pharmacy, Queen’s University Belfast, Belfast, Northern Ireland, United Kingdom;3. Centre of Excellence for Public Health (NI), Centre for Public Health, Queen’s University Belfast, Belfast, Northern Ireland, United Kingdom;1. School of Public Health, Curtin University, Kent Street, Bentley, Western Australia, 6102, Australia;2. Department of Epidemiology and Biostatistics, St Mary’s Campus, Imperial College London, Norfolk Place, London, W2 1 PG, United Kingdom;3. School of Public Health, Edward Ford Building A27, University of Sydney, New South Wales, 2006, Australia;4. Department of Epidemiology and Preventive Medicine, Monash University, The Alfred Centre, 99 Commercial Road, Melbourne, Victoria, 3004, Australia;5. School of Population Health, University of Western Australia, 35 Stirling Highway, Crawley, Western Australia, 6009, Australia |
| |
| Abstract: | BackgroundPopulation-based net survival by tumour stage at diagnosis is a key measure in cancer surveillance. Unfortunately, data on tumour stage are often missing for a non-negligible proportion of patients and the mechanism giving rise to the missingness is usually anything but completely at random. In this setting, restricting analysis to the subset of complete records gives typically biased results. Multiple imputation is a promising practical approach to the issues raised by the missing data, but its use in conjunction with the Pohar-Perme method for estimating net survival has not been formally evaluated.MethodsWe performed a resampling study using colorectal cancer population-based registry data to evaluate the ability of multiple imputation, used along with the Pohar-Perme method, to deliver unbiased estimates of stage-specific net survival and recover missing stage information. We created 1000 independent data sets, each containing 5000 patients. Stage data were then made missing at random under two scenarios (30% and 50% missingness).ResultsComplete records analysis showed substantial bias and poor confidence interval coverage. Across both scenarios our multiple imputation strategy virtually eliminated the bias and greatly improved confidence interval coverage.ConclusionsIn the presence of missing stage data complete records analysis often gives severely biased results. We showed that combining multiple imputation with the Pohar-Perme estimator provides a valid practical approach for the estimation of stage-specific colorectal cancer net survival. As usual, when the percentage of missing data is high the results should be interpreted cautiously and sensitivity analyses are recommended. |
| |
| Keywords: | Cancer Informative censoring Multiple imputation Net survival Pohar-Perme estimator Uncongeniality |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
