DIAGNOSIS-SPECIFIC MORBIDITY STATISTICS’ DATA PROBLEMATIC ASPECTS Based on experiences from work executed within an action „Pilot projects on morbidity statistics” According to the methodology specified in the guidebook „Principles and guidelines for diagnosis-specific morbidity statistics” Agnieszka Broś Piotr Woch Focal points I. II. III. IV. V. General information on the diagnosis-specific morbidity statistics’ project in Poland • duration, aims, cooperation National data sources – enumeration, assessment, Methodology for producing best national estimates Examples & problematic aspects Future steps General information on the action Duration: 15 November 2009 - 14 May 2011 The reference year: 2006 Aims: 1. Inventory and description of all potential national sources for diagnosis- specific morbidity data which can be used to provide information about diseases listed in the Diagnosis-specific morbidity – European shortlist, SHORTLIST agreed by Eurostat: 60 diseases divided into 20 groups + 1 group covering „external causes of mortality and morbidity” (accidents, assault, poisoning, complications of medical procedures, etc.). 2. Elaboration of the methodology for producing best national estimates 3. Pilot data collection and testing of the proposed methodology, taking into account the results of former Eurostat projects. 4. Preparation of the final report on the action. Cooperation – a key issue Centre for Health Statistics in the Statistical Office in Krakow – leading role in the project Series of working meetings and consultations with external experts (from National Institute of Public Health, Oncology Centre, National Health Fund, Institute of Psychiatry and Neurology, Centre for Health Information Systems) …on all steps of the project discussion of templates for data sources description and assessment of data sources analysis of available figures for required measures (incidence, prevalence) Problematic issues discussed with EUROSTAT Templates in the project Inventory of national data sources for diagnosis-specific morbidity statistics – template for general overview of the potential data sources Broad description and evaluation of the data sources inventoried Relationships between the measures (items on Shortlist) and data sources (potential and finally kept) DATA SOURCES most important & commonly used Inventory of data sources for diagnosis-specific morbidity statistics (I) First step - identification of all potential data sources Register of tuberculosis Notifications of sexually transmitted diseases Notifications and registration of HIV/AIDS Notifications of infectious diseases, infections and poisoning Reports of influenza cases and suspicions of the influenza National Cancer Register General in-patient morbidity study General out-patient morbidity study Psychiatric morbidity study (out-patient & in-patient) Inventory of data sources (II) First step - identification of all potential data sources Database of provided health care benefits in the framework of the in-patient and out-patient specialist care – NHF (National Health Fund) Population Health Status Survey in Poland (2004) Statistical survey of mortality Police’ databases (on road traffic accidents, attempted suicides, crimes) Central Register of Occupational Diseases Databases on disabled people (The Social Insurance Institution & The Agricultural Insurance Fund) Inventory of national data sources for diagnosis-specific morbidity data (I) DIVISION & ASSESSMENT MAIN DATA SOURCES Name of the source ADDITIONAL SOURCES General assessment Name of the source Register of tuberculosis 4 Notification of sexually transmitted diseases 2 The Population Health Status Survey in Poland Notification and registration of HIV/AIDS 4 Statistical survey of the mortality Notifications of infectious diseases, infections and poisoning 4 Database on road traffic accidents National Cancer Register 4 General hospital morbidity study 4 General out-patient morbidity study 4 Psychiatric out-patient morbidity study 4 Psychiatric in-patient morbidity study 4 Central Register of Occupational Diseases Database of provided health care benefits in the framework of the in-patient and specialist care - NHF 4 Database on disabled people - SII Reporting of influenza cases and suspicion of the influenza 3 General assessment 2 3 2 Database on attempted suicides Database on crimes Database on disabled people - ASIF Assessment criteria: relevance, accuracy, timeliness & punctuality, accessibility & clarity comparability (geographical and over time), coherence Assessment scale: 1 - poor, 5 - very good 2 1 2 1 1 Inventory of national data sources (II) FURTHER DIVISION BASED ON PREVIOUS ASSESSMENT Highest rated (mark: 4) – used during project DATA SOURCES Name of the source General assessment Register of tuberculosis 4 Notification and registration of HIV/AIDS 4 Notifications of infectious diseases, infections and poisonings 4 National Cancer Register 4 General hospital morbidity study 4 General out-patient morbidity study 4 Psychiatric out-patient morbidity study 4 Psychiatric in-patient morbidity study 4 Database of provided health care benefits in the framework of the in-patient and specialist care - NHF 4 Main advantages: • confirmation of each case through medical diagnosis • continuity of data supply • whole population covered Inventory of national data sources (III) FURTHER DIVISION BASED ON PREVIOUS ASSESSMENT Lowest rated (mark: 1-2) - rejected ADDITIONAL SOURCES Name of the source General assessment Database on attempted suicides General Headquarter of Police 2 Database on crimes - General Headquarter of Police 1 Central Register of Occupational Diseases – Institute of Occupational Medicine 2 Database on disabled people – The Social Insurance Institution 1 Database on disabled people – The Agriculture Social Insurance Fund 1 Main disadvantages: • Lack of cases’ confirmation through medical diagnosis (police’s data) • Reference to population groups, not to general population (databases on disabled people, Register on Occupational Diseases) Examples & EXAMPLES problematic aspects & PROBLEMATIC ASPECTS Methodology for producing best national estimates Possible ways to approach the production of best estimates (proposed by Eurostat) and their usage during the realization of project: a one to one relation - with a direct connection between the source and the required measure (for a position of the shortlist of diseases), the most frequent one combination of data from various sources only for several diseases adjustment of data source in order to find the "perfect figure” period prevalence on the basis of „Data on out-patient and in-patient morbidity – NHF” incidence per episode on the basis of „General out-patient morbidity study” All calculated figures inserted in a table for data submission for Eurostat. Tuberculosis [A15-A18, B90] Data requirements: incidence by episode, period prevalence Potential data sources: Register of tuberculosis, General hospital morbidity study General out-patient morbidity study Database of provided health care benefits in the framework of the in-patient and out-patient specialist care – NHF Incidence by episode – calculated on the basis of TB cases reported to the Register of tuberculosis Period prevalence – calculated on the basis of data from National Health Fund Tuberculosis – a one to one relationship & adjustment Incidence by episode Register of tuberculosis Period prevalence APPLIED NHF database APPLIED All patients treated in hospitals and by specialists All TB cases subjected to the obligatory reporting Under-registration: • changeability in annual incidence occurred – lack of stability in the scope of detecting and registration, • insufficient knowledge about diagnostic procedures among physicians detected TB among children No information on GP’s patients General hospital morbidity study REJECTED General outpatient morbidity study Hospital study: only in-patients Unsatisfactory proportion of cases confirmed by bacteriological tests Out-patient study: no data on out-patients cured in specialist care All malignant neoplasms (cancer) [C00-C97] Data requirements: incidence by person, period prevalence (5 years) Potential data sources: National Cancer Register (NCR) General hospital morbidity study Database of provided health care benefits in the framework of the hospital and outpatient specialist care – NHF NCR as a basis for calculation: Cancer incidence - diagnosis of disease with histological or cytological symptoms or proved by imaging examination or clinic imaging. There can be a few primary cancer sites for a single person. 5-year prevalence – the number of people living with cancer disease, who have been diagnosed within the last 5 years. Total prevalence should be calculated on the basis of cancer registry data. The NCR does not possesses a long enough horizon of data (20-30 years) to determine the total prevalence, thus 5-year prevalence is applied. All malignant neoplasm – a one to one relationship Incidence by episode National Cancer Registry APPLIED Obligatory reporting: doctors → 16 regional registries (verification, completion) → NCR (next control and medical verification; Period prevalence National Cancer Registry APPLIED 5-year prevalence was estimated by NCR on the basis of incidence data and the 5-year survival rates calculated for the Polish population for patients diagnosed in 2000-2002 publishing annual report Estimated coverage of the NCR exceeds 85%: • before estimation: M – 63,9; W - 60,9 (in thous.) • after estimation: M – 75,2; W - 72,0 (in thous.) General hospital morbidity study REJECTED Hospital study: only in-patients Non-uniform under-registration across the country (high intervoivodship differences) Registration completeness depends on the cancer site (location) and age group considered NHF database No information on GP’s patients REJECTED Acute myocardial infarction (AMI) [I21, I22] Data requirements: incidence by person, period prevalence Potential data sources: General hospital morbidity study Database of provided health care benefits in the framework of the hospital and out-patient specialist care – NHF Acute myocardial infarction can be diagnosed based on clinical characteristics, electrocardiographic (ECG), biochemical and pathological. The guidelines apply to people with symptoms of ischemia and persistent ST segment elevation in the ECG (STEMI). In most of these patients stated a significant increase in levels of biochemical markers of myocardial necrosis and the formation of the typical heart attack pathological Q wave (according to the guidelines of the European Society of Cardiology - ESC). AMI – combination of data sources & adjustment Incidence by person Period prevalence General hospital morbidity study NHF database APPLIED APPLIED Statistical survey on mortality All patients treated in hospitals and by specialists in outpatient settings No information on GP’s patients Combination of data from 2 sources: General hospital morbidity study: number of discharged patients with AMI (including deaths in hospitals) Mortality data: number of deaths due to AMI outside the hospital (including persons not previously treated for the AMI in the hospital) Diabetes mellitus [E10-E14] Data requirements: incidence by person, period prevalence, point prevalence According to „Clinical recommendations for dealing with diabetes in 2010”, diagnostics, education and treatment of diabetes are conducted mailnly in primary care by GPs and in the specialised care by medical professionals. In case of complications, exacerbations and inability to achieve therapeutic effects in an outpatient care, there is a need for in-patient treatment As a part of the specialist care – there are made the specialist diagnostics of all types diabetes and treatment of monogenic diabetes and diabetes co-occurring with other diseases. Both - incidence by person and period prevalence - were estimated on the basis of the General out-patient morbidity study - the only one source of data on diabetes mellitus from primary out-patient care. Diabetes mellitus – a one to one relationship Incidence by person General outpatient morbidity study APPLIED Period prevalence General outpatient morbidity study APPLIED Data are provided by primary care physicians/ family doctors by whom DM is mainly diagnosed No data by sex and 5-year age groups available, only data for 0-18 and 19+ age groups NHF database REJECTED No information on GP’s patients Figure from this source 36,9% smaller than the number from General outpatient morbidity study Point prevalence NO DATA SOURCE Dementia (incl. Alzheimer disease) [F00-F03, G30] Data requirements: period prevalence Dementia case – recognized on the basis of clinical symptoms by a psychiatrist who orders proper pharmaceutical, psychological and psychoterapeutic treatment. Cases under consideration include: Dementia in Alzheimer’s disease [F00, G30], Vascular dementia (effect of brain infarction) [F01], Dementia in other diseases elsewhere classified (Pick’s, Creutzfeld-Jakob’s, Huntington’s diseases, HIV) [F02] Unspecified dementia [F03] These diagnoses can be derived from psychiatric in-patient morbidity study which is based on individual statistical cards of patients. Psychiatric out-patient morbidity study – wider range of codes [additionally: F04, F05, F06, F07, F09]; no identification of patients (only data on the aggregated level) Dementia – combination of data sources & adjustment Period prevalence NHF database (outpatients only) APPLIED Psychiatric out-patient morbidity study REJECTED Psychiatric in-patient morbidity study NHF – patients with a diagnosis corresponding to the required range of ICD-10 codes [F00-F03, G30] – counted only once (identified by PESEL number) Psychiatric in-patient study: • the required ICD-10 codes available, • individual records derived from statistical cards Psychiatric in-patient morbidity study Connection of these data sources is improper Reasons: Out-patient morbidity study – • wider range of codes than required • possibility of double-counting – a patient using in-patient and out-patient psychiatric care in the same calendar year • no possibility for identification an individual patient Human immunodeficiency virus disease (HIV/AIDS) [B20-B24, Z21] Data requirements: incidence by person, period prevalence, point prevalence Cases of HIV/AIDS are defined in the system of reporting communicable diseases. The basis of diagnosis are clinical symptoms and/or immunological confirmation. HIV infection – diagnosis based on laboratory criteria for HIV infection or AIDS diagnosis. There are detailed laboratory criteria for diagnosis, different for children under the age of 18 months and for the rest of people – adults, adolescents and children over 18 months. AIDS – includes persons infected with HIV who have any of 28 clinical conditions listed in the European case definition for AIDS applied for epidemiological surveillance (European AIDS surveillance case definition) HIV/AIDS register (notification and registration of HIV/AIDS) – was found as the best data source HIV/AIDS – a one to one relationship Incidence by episode Register of HIV/AIDS Period prevalence APPLIED Under-registration of seropositive cases: • unawareness of disease • confidentiality (sum of data in age groups ≠ total) General hospital morbidity study REJECTED Difficulties in estimation of the number of seropositive cases: • no proper indication of new (first time) cases overestimation • no all HIV/AIDS cases are hospitalized underestimation Register of HIV/AIDS APPLIED Adjustment = all registered – deceased (from the beginning of registration to the end of 2005) Point prevalence Register of HIV/AIDS APPLIED Adjustment = all registered as of 30 December 2006 – deceased (from the beginning of registration to the end of 2006) Missing data INCOMPLETE DATA IN AGE AND GENDER GROUPS Incidence by episode Sum of numbers in age groups ≠ total number NO RELAIBLE DATA SOURCE IDENTIFIED Incidence and prevalence HIV/AIDS Those listed on a register may retain their anonymity (age, gender) Land transport accidents’ victims For some cases no information on age and gender Pneumonia [J12-J18] Accidental falls [W00-W19] Accidental poisoning [X40-X49] Intentional self harm (incl. suicidal attempt) [X60-X84] Assault [X85-Y09] Medical and surgical complications [Y40-Y66, Y69-Y84] Period prevalence Rheumatoid arthritis [M05, M06] Arthrosis [M15-M19] Future steps Goal: regular morbidity data collection within the ESS Task Force on morbidity statistics (TF MORB) was established Fit existing methodological tools to that goal by Analysis of results of 16 pilot studies in MS (stress on quality, reliability and comparability across MS) 10 MS before 2009 (AT, CY, CZ, EE, HU, LT, LV, MT, SL, SI); for 6 MS (BE, DE, NL, RO, PL, FI) final report sent by autumn 2011 If needed, revise the existing methodology: guidelines, shortlist of diseases Deliverables discussed at a Technical Group MORB meeting Finalisation of the documents by November 2012 Thank you for the ATTENTION !!!