Cancer Surveillance

CANCER SURVEILLANCE : CANCER SURVEILLANCE MPH 7023 EPIDEMIOLOGIC SURVEILLANCE Diego E Zavala, MSc, PhD

Cancer Surveillance : Cancer Surveillance Main objectives of this lesson To learn about the use epidemiologic surveillance methods con cancer prevention and control. To have an understanding of the issues associated with the design and implementation of a cancer registry. Core competencies that students should attain by the end of the session To gain an understanding of: the definition and applications of a cancer registry types of cancer registries the general principles of cancer prevention the steps for implementing a cancer registry the advantages and disadvantages of a cancer registry Minimum data elements for a cancer registry Data analysis and reporting

Cancer Surveillance : Cancer Surveillance Rationale Cancer is a leading cause of death worldwide. Rapidly rising cancer burden in low-income and medium-income countries. High-quality incidence data are needed from developing regions and countries to establish the cancer burden and to monitor its evolution.

Slide 4 : Need for data standardized classification of disease (injuries) comparisons between countries trends and changes over time disease (injury) related to other country/regional conditions (socioeconomic level) determine changes in characteristics of population over time and geographical area evaluate interventions Reasons for doing surveillance?

Distribution of age at death and numbers of deaths, World, 2004 : Distribution of age at death and numbers of deaths, World, 2004 Source: The global burden of disease: 2004 update. World Health Organization 2008

Distribution of deaths by leading cause groups, males and females, world, 2004 : Distribution of deaths by leading cause groups, males and females, world, 2004 Source: The global burden of disease: 2004 update. World Health Organization 2008

Leading causes of death by income group, 2004 : Leading causes of death by income group, 2004

Cancer Registry History* : Cancer Registry History* London, 1728- First (unsuccessful) cancer census attempted. Germany, 1900- Attempt to register all cancer patients under medical treatment questionnaires sent to MDs approximately 50% responded Netherlands, Spain, Portugal, Sweden, Hungary, Denmark, Iceland, 1902-1908. Similar efforts as in Germany. *O.M. Jensen, D.M. Parkin, R. MacLennan, C.S. Muir and R.G. Skeet. Cancer Registration: Principles and Methods

Cancer Registry History* : Cancer Registry History* Massachusetts, 1927- First cancer registry (pilot project). Mecklenburg, Germany, 1937- First continuous registry of cancer patients. US, late 1930’s. cancer incidence data collected from Ad Hoc morbidity surveys. All cases of cancer were recorded during one calendar year in 10 metropolitan areas in 1937-38; Second National Cancer Survey 1947-1948 Third (and last) Cancer Survey 1969-1971

Cancer Registry History* : Cancer Registry History* Population-based cancer registration Connecticut, 1935. The Connecticut Tumor Registry began operation on a statewide basis in 1941, registering cases retrospectively back to 1935. Denmark, 1942. Danish Cancer Registry founded. Voluntary reporting by physicians. Puerto Rico, 1950. Central Cancer Registry established,

Cancer Registry History* : Cancer Registry History* Worldwide establishment of cancer registries Recommendation from leading experts in the field of cancer control At a conference that took place in Copenhagen in 1946. Benefits: collection of data about cancer patients from as many different countries as possible; data recorded to be comparable; each nation should have a central registry; an international body should correlate the data and statistics obtained in each country.

Cancer Registry History* : Cancer Registry History* WHO 1965, established a research center: the International Agency for Research on Cancer (IARC) http://www.iarc.fr/ Tokyo, 1966. the International Association of Cancer Registries (IACR) was formed. http://www.iacr.com.fr/

Cancer Registries Worldwide* : Cancer Registries Worldwide*

Cancer Registries Worldwide* : Cancer Registries Worldwide*

Cancer Registries : Cancer Registries Purposes an essential part of any rational programme of cancer control.(Muir et al., 1985). cancer registry data can be used in a wide variety of areas of cancer control: etiological research primary prevention secondary prevention health-care planning (e.g. cancer care facilities) patient care (e.g. survival, screening)

Cancer Registries : Cancer Registries Main Objective of a Cancer Registry to collect and classify information on all cancer cases in order to produce statistics on the occurrence of cancer in a defined population and to provide a framework for assessing and controlling the impact of cancer on the community.

Cancer Registries : Cancer Registries Types of Cancer Registries Hospital Based Population Based Site specific (e.g. brain tumors, cancer in children)

Cancer Registry : Cancer Registry Definition* “Cancer registration may be defined as the process of continuing, systematic collection of data on the occurrence and characteristics of reportable neoplasms with the purpose of helping to assess and control the impact of malignancies on the community. The cancer registry is the office or institution which attempts to collect, store, analyze and interpret data on persons with cancer.” *O.M. Jensen, D.M. Parkin, R. MacLennan, C.S. Muir and R.G. Skeet. Cancer Registration: Principles and Methods

Cancer Registries : Cancer Registries Data collected Basic information (personal identification)- Avoids duplicates name sex date of birth or age usual residence address (permanent) Incidence date (diagnosis or hospitalization for first treatment the primary site and histological type of the cancer diagnosed (must be ICD-O-3 coded)

Cancer Registries : Cancer Registries Data collected Additional information ID number ethnic group/race stage of disease nature of the first treatment brain and nervous system follow–up for vital status death certificates used to update vital status active follow-up of the alive cases

Cancer Registries : Cancer Registries Classification and Coding 7th Revision (ICD–7; WHO, 1957) CI5C Vol I and II 8th Revision (ICD–8; WHO, 1967) CI5C Vol III and IV 9th Revision (ICD–9; WHO, 1977), CI5C Vol V, VI, VII 10th Revision (ICD–10 WHO, 1992) CI5C Vol VIII, IX

Cancer Registries : Cancer Registries Comparability and quality of data Numerator Definition of an incident case of cancer, Completeness of enumeration of cases in the population covered, Accuracy of abstraction and coding of information. Denominator Accurate estimate of person-years at risk

Cancer Registries : Cancer Registries Numerator

Cancer Registries : Cancer Registries Definition of incidence Three components Distinction between recurrence or extension of an existing cancer and the development of a new primary; Distinguish between multiple primaries and metastasis. The detection of cancers incidentally, in asymptomatic individuals; and Increase in incidence expected if these cases included. Examples include: breast and prostate cancer screening cases. The detection of cancers at autopsy. Percent of incident cases detected by autposy should be minimal.

Cancer Registries : Cancer Registries Completeness of enumeration of cancer cases. Completeness of registration is the proportion of all incident cases in the registry population that have been included in the registry database. Completeness should be as close to 100% as possible, so that comparison of incidence rates between registries reflects true differences in cancer risk.

Cancer Registries : Cancer Registries Completeness of enumeration of cancer cases. main concern is with the possibility of incompleteness in the data submitted is duplicate registration of cases indices of completeness include: Historic data methods: Stability of incidence rates over time Age-specific incidence curves Childhood cancer Proportion of cases microscopically verified Proportion of unknown basis of diagnosis Mortality:incidence (M:I) ratio Death certificate methods

CI5C Vol. IX : CI5C Vol. IX

CI5C Vol. IX : CI5C Vol. IX Age-specific incidence rates by site

Cancer Registries : Cancer Registries Coding practices Five cancer sites are of particular concern: skin, non-melanoma (ICD–10 C44) ovary (ICD–10 C56) urinary bladder (ICD–10 C67) brain and central nervous system (ICD–10 C70–72) Myeloproliferative disorders and myelodysplastic syndromes (MPD/MDS)

Cancer Registries : Cancer Registries Coding practices Skin, non-melanoma (ICD–10 C44) incidence of non-melanoma skin cancer (NMSC) is difficult to assess. these cancers are very common but rarely fatal, and completeness of registration varies widely depending on access to outpatient records and general practitioners. Most NMSCs are basal cell (BCC) or squamous cell (SCC) carcinomas; other skin cancers are rare. While some registries record the first occurrence of all NMSC, others register BCC only several registries collect information for lip and/or genital sites only many do not collect data on either SCC or BCC.

Cancer Registries : Cancer Registries Coding practices Ovary (ICD–10 C56) ovarian cystadenoma coded as borderline malignancy and borderline tumour of the ovary for these diagnoses, registration practice varies considerably. borderline ovarian diagnoses were considered as non-malignant tumours in ICD–O–1/ICD–9. they were considered malignant in ICD–O–2/ICD–10, and have been changed back to the /1 borderline category in ICD–O–3; As a result, they have been excluded from CI5C Vol IX

Cancer Registries : Cancer Registries Coding practices Urinary bladder (ICD–10 C67) problem of the coding of non-invasive tumors, taking into account recorded level of invasion and grade which to include as ‘cancer of the bladder’ has long been the subject of debate. the availability of data on histological type and behavior has made it include malignant cancer by excluding diagnoses with any behaviour code other than /3. many registries assign the behaviour code /3 to both non-invasive and unspecified diagnoses (making it impossible to distinguish such cases.) last two CI5C volumes include noninvasive diagnoses of bladder cancer (/1 and /2) and considered malignant the bladder cancer category includes the in situ and unspecified categories. some registries prefer not to include such cases in their dataset, for the sake of continuity over time.

Cancer Registries : Cancer Registries Coding practices Brain and central nervous system (ICD–10 C70–72) Some registries choose to include benign and unspecified tumors of the brain and central nervous system in their data because of the potentially serious clinical consequences of these tumors, and assign the behavior code /3 to both benign and unspecified diagnoses making it impossible to distinguish such cases benign and unspecified tumors of the brain and central nervous system are excluded from CI5C.

Cancer Registries : Cancer Registries Coding practices Myeloproliferative disorders and myelodysplastic syndromes (MPD/MDS) these diagnoses were considered as non-malignant disease (/1) in ICD–O–2 In ICD–10 (D45–D47) have changed behaviour code to malignant (/3) in ICD–O–3.

Cancer Registries : Cancer Registries Validity Defined as the proportion of cases with a given characteristic (e.g., site, age) that truly have the attribute. Five of the common indices of validity (Parkin et al., 1994): Internal consistency Histological verification Death certificate only Other and unspecified cases (ill-defined cases) Age unknown Use IARC-CHECK program

Cancer Registries : Cancer Registries Comparability and quality of data Numerator Definition of an incident case of cancer, Completeness of enumeration of cases in the population covered, Accuracy of abstraction and coding of information. Denominator Accurate estimate of person-years at risk

Cancer Registries : Cancer Registries Denominator

Cancer Registries : Cancer Registries A 10% error in the estimation of population at risk produces just as much inaccuracy in the calculated incidence rate as a 10% error in enumeration of cases. Population estimates? Source of the population-at-risk figures? methods used to produce estimates and projections? Shape of the population pyramids irregularities in the age-specific incidence curves may suggest errors in population estimates

Population pyramids : Population pyramids

Cancer Registries : Cancer Registries Cancer Registry Program in the US (and Puerto Rico) SEER NPCR 1997- Jan. 1 1997- report to CCRPR Jun. 1998,1999, 2000- hospital personnel training 1999 CCRPR software update free software to hospitals

Central Cancer Registry of Puerto Rico : Central Cancer Registry of Puerto Rico Notificación obligatoria de casos por toda entidad pública o privada que diagnostica o provea tratamiento a pacientes con cáncer. Ley Núm. 28 (20 de marzo de 1951) Ley Núm. 17 de 1953 Ley Núm. 74 de 1988 Ley Núm. 217 de 1999.

Cancer Cases - Puerto Rico, 2000* : Cancer Cases - Puerto Rico, 2000*

Cancer Deaths - Puerto Rico, 2000* : Cancer Deaths - Puerto Rico, 2000*

Age-Adjusted Cancer Incidence Rates- Puerto Rico Males, 2000* : Age-Adjusted Cancer Incidence Rates- Puerto Rico Males, 2000* *Tasas ajustadas por edad a la población Estándar Millón de PR año 2000. Fuente: Registro Central de Cáncer de PR, 2004 Prostate Stomach U. Bladder Lung Colon/rectum

Age-Adjusted Cancer Incidence Rates- Puerto Rico Females, 2000* : Age-Adjusted Cancer Incidence Rates- Puerto Rico Females, 2000*

Age-Adjusted Cancer Mortality Rates- Puerto Rico Males, 2000* : Age-Adjusted Cancer Mortality Rates- Puerto Rico Males, 2000*

Age-Adjusted Cancer Mortality Rates- Puerto Rico Females, 2000* : Age-Adjusted Cancer Mortality Rates- Puerto Rico Females, 2000*

Cancer in Puerto Rico : Cancer in Puerto Rico SUMMARY Approximately 10,300 new cases are reported annually. (1/3 prostate cancer and 1/3 breast cancer) There is a slight tendency of more male cases reported although the sex ratio is closer to 1:1 due to a decrease in prostate cancer incidence. Age is the main factor, specially for > 45 years of age. Fuente: Registro Central de Cáncer de PR, 2004

Cancer in Puerto Rico : Cancer in Puerto Rico SUMMARY Relative cancer survival to 5 years after diagnosis is 34% for all sites combined. Prognosis, breast and prostate cancer have best prognosis. Worstfor lung and stomach. An estimated 4,615 cancer deaths are reported annually. Fuente: Registro Central de Cáncer de PR, 2004

Cancer Registries : Cancer Registries Emerging Trends in the US National Program of Cancer Registries Modeling Electronic Reporting Project (NPCR-MERP) CDC NPCR-MERP Project a collaborative effort to position the cancer surveillance community to take advantage of the electronic health record (EHR) for cancer surveillance. To be accomplished by developing Unified Modeling Language (UML) and non-UML models to represent the flow of data through all levels of the cancer surveillance system. Including flow processes from the hospital’s EHR (which includes multiple database systems) and other cancer registry data sources (such as private pathology labs) to the hospital cancer registry; from the hospital cancer registry to the state central cancer registry; and from the state central cancer registry to the CDC.