Coordinator research program Care Demand of the Chronically Ill and Disabled; honorary professor 'Pharmacy health services research', University of Groningen, the Netherlands
Publicatie
Publication date
Comparison of six electronic healthcare databases in Europe using standardized protocols: a descriptive study on the incidence of cancer.
Afonso, A.S., Groot, M.C.H. de, Ham, R. van den, Bruin, M.L. de, Huerta Alvarez, C., Gil, M., Hesse, U., Ronn, P.F., Souverein, P.C., Alvarez, Y., Slattery, J., Rottenkolber, M., Dijk, L. van, Schlienger, R., Reynolds, R., Klungel, O.H., Grimaldi-Bensouda, L. Comparison of six electronic healthcare databases in Europe using standardized protocols: a descriptive study on the incidence of cancer. Pharmacoepidemiology and Drug Safety: 2013, 22(suppl. 1), p. 285. Abstract: 29th International Conference on Pharmacoepidemiology & Therapeutic Risk Management. 25-28 augustus 2013, Montréal.
Read online
Background
There are several national cancer registries available across Europe, but information on cancer incidence from routine electronic healthcare record (EHR) databases (DBs), such as General Practitioners (GPs) and comparisons across different databases are rather scarce. It is important to compare this information, and also benchmark this against national registries in order to assess its usefulness for pharmacoepidemiological studies on cancer.
Objectives
To investigate sources of variation in the incidence of Cancer across routine EHR DBs in Europe, using a standardized methodology.
Methods
We used six EHR DBs from Spain-ES (BIFAP), the United Kingdom-UK (THIN and CPRD), the Netherlands-NL (Mondriaan: AHC and NPCRD), and Denmark-DK (National registrations of patients-NRP). Cancer incidences were calculated for the whole population between 2003 and 2008 and were stratified by sex, age and type of cancer (breast, prostate and colon). Overall incidence rates were age and sex standardized to the European 2008 reference population.
Results
The initially observed variation in cancer incidence decreased after standardization and ranged for any cancer from 25.2/10,000 in the NL (NPCRD) in 2004, to 71.5/10,000 in the DK (NRP) in 2008. The incidence of cancer increased in DK and doubled in NPCRD between 2003 and 2008, but decreased in BIFAP and THIN. Cancer incidence was higher for women in all DBs, except for BIFAP and AHC. In 2008, the incidence of breast cancer was the highest in the NPCRD (37.6/10,000) and the lowest in BIFAP (9.1/ 10,000), while the incidence of prostate cancer was the highest in DK (15.9/10,000). No major differences were observed between countries regarding colon cancer.
Conclusions
The incidence of cancer as measured in six routine EHR DBs differed between the four European countries using a standard methodology, despite the convergence seen after standardization for age and sex. Overall cancer incidence increased over time for most of the European countries, except for Spain. From our analysis we can infer that incidences are in line with the European cancer registries available. (aut. ref.)
There are several national cancer registries available across Europe, but information on cancer incidence from routine electronic healthcare record (EHR) databases (DBs), such as General Practitioners (GPs) and comparisons across different databases are rather scarce. It is important to compare this information, and also benchmark this against national registries in order to assess its usefulness for pharmacoepidemiological studies on cancer.
Objectives
To investigate sources of variation in the incidence of Cancer across routine EHR DBs in Europe, using a standardized methodology.
Methods
We used six EHR DBs from Spain-ES (BIFAP), the United Kingdom-UK (THIN and CPRD), the Netherlands-NL (Mondriaan: AHC and NPCRD), and Denmark-DK (National registrations of patients-NRP). Cancer incidences were calculated for the whole population between 2003 and 2008 and were stratified by sex, age and type of cancer (breast, prostate and colon). Overall incidence rates were age and sex standardized to the European 2008 reference population.
Results
The initially observed variation in cancer incidence decreased after standardization and ranged for any cancer from 25.2/10,000 in the NL (NPCRD) in 2004, to 71.5/10,000 in the DK (NRP) in 2008. The incidence of cancer increased in DK and doubled in NPCRD between 2003 and 2008, but decreased in BIFAP and THIN. Cancer incidence was higher for women in all DBs, except for BIFAP and AHC. In 2008, the incidence of breast cancer was the highest in the NPCRD (37.6/10,000) and the lowest in BIFAP (9.1/ 10,000), while the incidence of prostate cancer was the highest in DK (15.9/10,000). No major differences were observed between countries regarding colon cancer.
Conclusions
The incidence of cancer as measured in six routine EHR DBs differed between the four European countries using a standard methodology, despite the convergence seen after standardization for age and sex. Overall cancer incidence increased over time for most of the European countries, except for Spain. From our analysis we can infer that incidences are in line with the European cancer registries available. (aut. ref.)
Background
There are several national cancer registries available across Europe, but information on cancer incidence from routine electronic healthcare record (EHR) databases (DBs), such as General Practitioners (GPs) and comparisons across different databases are rather scarce. It is important to compare this information, and also benchmark this against national registries in order to assess its usefulness for pharmacoepidemiological studies on cancer.
Objectives
To investigate sources of variation in the incidence of Cancer across routine EHR DBs in Europe, using a standardized methodology.
Methods
We used six EHR DBs from Spain-ES (BIFAP), the United Kingdom-UK (THIN and CPRD), the Netherlands-NL (Mondriaan: AHC and NPCRD), and Denmark-DK (National registrations of patients-NRP). Cancer incidences were calculated for the whole population between 2003 and 2008 and were stratified by sex, age and type of cancer (breast, prostate and colon). Overall incidence rates were age and sex standardized to the European 2008 reference population.
Results
The initially observed variation in cancer incidence decreased after standardization and ranged for any cancer from 25.2/10,000 in the NL (NPCRD) in 2004, to 71.5/10,000 in the DK (NRP) in 2008. The incidence of cancer increased in DK and doubled in NPCRD between 2003 and 2008, but decreased in BIFAP and THIN. Cancer incidence was higher for women in all DBs, except for BIFAP and AHC. In 2008, the incidence of breast cancer was the highest in the NPCRD (37.6/10,000) and the lowest in BIFAP (9.1/ 10,000), while the incidence of prostate cancer was the highest in DK (15.9/10,000). No major differences were observed between countries regarding colon cancer.
Conclusions
The incidence of cancer as measured in six routine EHR DBs differed between the four European countries using a standard methodology, despite the convergence seen after standardization for age and sex. Overall cancer incidence increased over time for most of the European countries, except for Spain. From our analysis we can infer that incidences are in line with the European cancer registries available. (aut. ref.)
There are several national cancer registries available across Europe, but information on cancer incidence from routine electronic healthcare record (EHR) databases (DBs), such as General Practitioners (GPs) and comparisons across different databases are rather scarce. It is important to compare this information, and also benchmark this against national registries in order to assess its usefulness for pharmacoepidemiological studies on cancer.
Objectives
To investigate sources of variation in the incidence of Cancer across routine EHR DBs in Europe, using a standardized methodology.
Methods
We used six EHR DBs from Spain-ES (BIFAP), the United Kingdom-UK (THIN and CPRD), the Netherlands-NL (Mondriaan: AHC and NPCRD), and Denmark-DK (National registrations of patients-NRP). Cancer incidences were calculated for the whole population between 2003 and 2008 and were stratified by sex, age and type of cancer (breast, prostate and colon). Overall incidence rates were age and sex standardized to the European 2008 reference population.
Results
The initially observed variation in cancer incidence decreased after standardization and ranged for any cancer from 25.2/10,000 in the NL (NPCRD) in 2004, to 71.5/10,000 in the DK (NRP) in 2008. The incidence of cancer increased in DK and doubled in NPCRD between 2003 and 2008, but decreased in BIFAP and THIN. Cancer incidence was higher for women in all DBs, except for BIFAP and AHC. In 2008, the incidence of breast cancer was the highest in the NPCRD (37.6/10,000) and the lowest in BIFAP (9.1/ 10,000), while the incidence of prostate cancer was the highest in DK (15.9/10,000). No major differences were observed between countries regarding colon cancer.
Conclusions
The incidence of cancer as measured in six routine EHR DBs differed between the four European countries using a standard methodology, despite the convergence seen after standardization for age and sex. Overall cancer incidence increased over time for most of the European countries, except for Spain. From our analysis we can infer that incidences are in line with the European cancer registries available. (aut. ref.)
Gegevensverzameling