An EM-based Semi-Parametric Mixture Model Approach to the Regression Analysis of Competing-Risks Data

We consider a mixture model approach to the regression analysis of competing-risks data. Attention is focused on inference concerning the effects of factors on both the probability of occurrence and the hazard rate conditional on each of the failure types. These two quantities are specified in the mixture model using the logistic model and the proportional hazards model, respectively. We propose a semi-parametric mixture method to estimate the logistic and regression coefficients jointly, whereby the component-baseline hazard functions are completely unspecified. Estimation is based on maximum likelihood on the basis of the full likelihood, implemented via an expectation-conditional maximization (ECM) algorithm. Simulation studies are performed to compare the performance of the proposed semi-parametric method with a fully parametric mixture approach. The results show that when the component-baseline hazard is monotonic increasing, the semi-parametric and fully parametric mixture approaches are comparable for mildly and moderately censored samples. When the component-baseline hazard is not monotonic increasing, the semi-parametric method consistently provides less biased estimates than a fully parametric approach and is comparable in efficiency in the estimation of the parameters for all levels of censoring. The methods are illustrated using a real data set of prostate cancer patients treated with different dosages of the drug diethylstilbestrol. Copyright (C) 2003 John Wiley Sons, Ltd.

Regional Advisory Committee on Cancer - Report on Skin Cancer (September 2004) (PDF 201 KB)

Skin cancer is the diagnosis for about a quarter of all patients with cancer and because most of the work is done on an outpatient basis the true extent of the disease has largely gone unrecognised. Skin cancers are related to ultraviolet radiation exposure. Geographic latitude as well as attitude affects the amount of ultraviolet exposure and the risk of skin cancer, with people from Northern Ireland exposing themselves to higher levels of ultraviolet radiation when on holiday abroad and artificially from sunbeds. Ozone depletion is known to increase the risk of ultraviolet exposure and skin cancer. The majority of people living in Northern Ireland have pale skin and are at increased risk of developing skin cancer, as are some patients with an increased genetic risk for cancer. Some pre-existing skin lesions are known to increase the risk of developing skin cancer. Data collection on the incidence of non melanoma skin cancer in Northern Ireland was not available before the establishment of the Cancer Registry in 1993. There is however good data on the incidence of melanoma before that period. In 1974 there were 39 cases in melanoma in Northern Ireland, by 2000 this had risen by almost 500% to 185 cases. åÊ

Chemotherapy for advanced gastric cancer.

BACKGROUND: Gastric cancer currently ranks second in global cancer mortality. Most patients are either diagnosed at an advanced stage, or develop a relapse after surgery with curative intent. Apart from supportive care and palliative radiation to localized (e.g. bone) metastasis, systemic chemotherapy is the only treatment option available in this situation. OBJECTIVES: To assess the efficacy of chemotherapy versus best supportive care, combination versus single agent chemotherapy and different combination chemotherapy regimens in advanced gastric cancer. SEARCH STRATEGY: We searched the Cochrane Central Register of Controlled Trials, MEDLINE and EMBASE up to March 2009, reference lists of studies, and contacted pharmaceutical companies and national and international experts. SELECTION CRITERIA: Randomised controlled trials on systemic intravenous chemotherapy versus best supportive care, combination versus single agent chemotherapy and different combination chemotherapies in advanced gastric cancer. DATA COLLECTION AND ANALYSIS: Two authors independently extracted data. A third investigator was consulted in case of disagreements. We contacted study authors to obtain missing information. MAIN RESULTS: Thirty five trials, with a total of 5726 patients, have been included in the meta-analysis of overall survival. The comparison of chemotherapy versus best supportive care consistently demonstrated a significant benefit in overall survival in favour of the group receiving chemotherapy (hazard ratios (HR) 0.37; 95% confidence intervals (CI) 0.24 to 0.55, 184 participants). The comparison of combination versus single-agent chemotherapy provides evidence for a survival benefit in favour of combination chemotherapy (HR 0.82; 95% CI 0.74 to 0.90, 1914 participants). The price of this benefit is increased toxicity as a result of combination chemotherapy. When comparing 5-FU/cisplatin-containing combination therapy regimens with versus without anthracyclines (HR 0.77; 95% CI 0.62 to 0.95, 501 participants) and 5-FU/anthracycline-containing combinations with versus without cisplatin (HR 0.82; 95% CI 0.73 to 0.92, 1147 participants) there was a significant survival benefit for regimens including 5-FU, anthracyclines and cisplatin. Both the comparison of irinotecan versus non-irinotecan (HR 0.86; 95% CI 0.73 to 1.02, 639 participants) and docetaxel versus non-docetaxel containing regimens (HR 0.93; 95% CI 0.75 to 1.15, 805 participants) show non-significant overall survival benefits in favour of the irinotecan and docetaxel-containing regimens. AUTHORS' CONCLUSIONS: Chemotherapy significantly improves survival in comparison to best supportive care. In addition, combination chemotherapy improves survival compared to single-agent 5-FU. All patients should be tested for their HER-2 status and trastuzumab should be added to a standard fluoropyrimidine/cisplatin regimen in patients with HER-2 positive tumours. Two and three-drug regimens including 5-FU, cisplatin, with or without an anthracycline, as well as irinotecan or docetaxel-containing regimens are reasonable treatment options for HER-2 negative patients.

Cancer in Iowa, 2003

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. In 2003 data will be collected on an estimated 14,700 new cancers among Iowa residents. A follow-up program tracks more than 97 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. Since 1973 the Iowa Registry has been funded by the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute (NCI). Iowa represents rural and midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2004 Cancer in Iowa Report, 2004

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. In 2004 data will be collected on an estimated 15,200 new cancers among Iowa residents. A follow-up program tracks more than 97 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. Since 1973 the Iowa Registry has been funded by the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute (NCI). Iowa represents rural and midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa is also providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2005 Cancer in Iowa Report: Prostate Cancer in Iowa, 2005

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 97 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2005 data will be collected on an estimated 15,800 new cancers among Iowa residents. Beginning with 2005 Cancer in Iowa, in situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2007 Cancer in Iowa Report, 2007

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2007 data will be collected on an estimated 15,700 new cancers among Iowa residents. In situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. In addition, the Registry receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2008 Cancer in Iowa Report, 2008

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for followup and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2008 data will be collected on an estimated 16,000 new cancers among Iowa residents. Noninvasive cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded primarily by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many National Cancer Institute publications. Beginning in 1990 a small percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2010 Cancer in Iowa Report, 2010

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2010 data will be collected on an estimated 16,400 new cancers among Iowa residents. In situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2011 Cancer in Iowa Report, 2011

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2011 data will be collected on an estimated 16,500 new cancers among Iowa residents. In situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many National Cancer Institute publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2012 Cancer in Iowa Report: A Focus on Lung Cancer, 2012

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2012 data will be collected on a projected 17,500 new cancers among Iowa residents. In situ cases of bladder cancer are included in the projections for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the NCI. Since 1973 the Iowa Registry has been funded by the SEER Program of the NCI. Iowa represents rural and Midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2013 Cancer in Iowa Report, 2013

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighbor-ing states to collect cancer data. Hospital cancer programs approved by the American College of Surgeons also report their data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2013 data will be collected on an estimated 17,300 new cancers among Iowa residents. In situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Starting in 2003, the University of Iowa has also been providing cost-sharing funds. In addition the Registry receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2014 Cancer in Iowa Report, 2014

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2014 data will be collected on an estimated 17,400 new cancers among Iowa residents. In situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

2015 Cancer in Iowa Report, 2015

Cancer is a reportable disease as stated in the Iowa Administrative Code. Cancer data are collected by the State Health Registry of Iowa, located at The University of Iowa in the College of Public Health’s Department of Epidemiology. The staff includes more than 50 people. Half of them, situated throughout the state, regularly visit hospitals, clinics, and medical laboratories in Iowa and neighboring states to collect cancer data. A follow-up program tracks more than 99 percent of the cancer survivors diagnosed since 1973. This program provides regular updates for follow-up and survival. The Registry maintains the confidentiality of the patients, physicians, and hospitals providing data. In 2014 data will be collected on an estimated 17,400 new cancers among Iowa residents. In situ cases of bladder cancer are included in the estimates for bladder cancer, to be in agreement with the definition of reportable cases of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute. Since 1973 the Iowa Registry has been funded by the SEER Program of the National Cancer Institute. Iowa represents rural and Midwestern populations and provides data included in many NCI publications. Beginning in 1990 about 5-10 percent of the Registry’s annual operating budget has been provided by the state of Iowa. Beginning in 2003, the University of Iowa has also been providing cost-sharing funds. The Registry also receives funding through grants and contracts with university, state, and national researchers investigating cancer-related topics.

Relation entre statut socio-économique et incidence du cancer du poumon dans le contexte d’une étude cas-témoins montréalaise

Des études ont observé que le statut socio-économique (SSE) est inversement associé avec le risque de cancer du poumon. Cette thèse approfondit différents aspects de la relation entre le SSE et le risque de cancer du poumon afin de contribuer à la compréhension des facteurs influençant le risque de cette maladie. Son objectif général consiste à étudier les associations entre le SSE et le risque de cancer du poumon en fonction à la fois : 1) de l’indicateur du SSE ; 2) des types d’ajustement pour le tabagisme, et ; 3) des facteurs de risque intermédiares potentiels de l’association entre le SSE et le cancer du poumon. Les données d’une vaste étude cas-témoins menée à Montréal entre 1996 et 2002 et portant sur l’identification de facteurs environnementaux impliqués dans l’étiologie du cancer du poumon ont été utilisées. L’échantillon analysé était constitué de 1203 cas (738 hommes et 465 femmes) diagnostiqués en 1996-1997 dans les 18 principaux hôpitaux du Grand Montréal et 1513 témoins (899 hommes et 614 femmes). Les témoins, recrutés au hasard à partir de la liste électorale, ont été appariés aux cas selon l’âge, le sexe et le district électoral. Les sujets ont été interviewés afin de colliger de l’information entre autres sur les caractéristiques sociodémographiques et économiques, les habitudes de tabagisme, l‘histoire professionnelle et l’alimentation. Une collecte de données additionnelle a été effectuée spécifiquement pour la présente thèse. En effet, l’évaluation foncière de la résidence correspondant à l’adresse de chacun des sujets à l’entrevue a été obtenue à partir des bases de données publiques de la ville de Montréal. Le premier article portait sur la comparaison de différents indices de disponibilité financière chacun en 5 catégories: un indice de valeur résidentielle, le revenu médian issu des données du recensement et le revenu familial auto-rapporté. En comparant au revenu familial auto-rapporté, une très bonne concordance a été obtenue pour 73% et 64% des individus respectivement pour l’indice de valeur résidentielle et le revenu du recensement. Le Kappa pondéré était plus élevé pour la comparaison du revenu familial auto-rapporté avec l’indice de valeur résidentielle (Kappap=0.36, IC 95%: 0.31-0.40) qu’avec le revenu médian issu du recensement (Kappap=0.22, IC 95%: 0.17-0.27). Le choix d’un indice dépend toutefois des objectifs et de la faisabilité de son utilisation dans la population étudiée. Le deuxième article présentait les associations entre les trois indices de disponibilité financière et le risque de cancer du poumon. Les fortes associations observées initialement sont disparues après un ajustement pour plusieurs dimensions du tabagisme. Ce dernier jouait un rôle de grande importance dans la relation entre le SSE et le risque de cancer du poumon. Finalement, les facteurs de risque intermédiares des associations entre trois différentes dimensions du SSE – éducation, revenu, classe professionnelle – ont été étudiés. Les résultats ont suggéré que les 3 facteurs de risque intermédiares potentiels étudiés: tabagisme, consommation de fruits et légumes riches en caroténoïdes, expositions professionnelles avaient un effet médiateur indépendant dans la relation entre le SSE et le cancer du poumon. Le plus important médiateur était le tabagisme, suivi de l’alimentation et à un degré moindre, les expositions professionnelles. Il nous paraît particulièrement pertinent d‘évaluer différents indicateurs du SSE lorsque possible, et d’effectuer un ajustement détaillé pour le tabagisme lorsque la relation entre le SSE et le risque de cancer du poumon est étudiée. Une bonne connaissance et la prise en compte des facteurs de risque intermédiares sont essentielles à la formulation des mesures de prévention du cancer du poumon.