HealthDataNavigator Understand strengths and weaknesses of existing data bases and projects

France Data Source

Système National d’Information Inter-Régime de l’Assurance Maladie (SNIIR-AM)

This exhaustive nationwide database is at the heart of the financing system of diagnostics, pharmaceuticals, and physicians in the ambulatory care setting and of independent practitioners in private hospitals (essentially fee for service). It provides data on claims paid by the Social Security System is therefore the main source of information on ambulatory setting activity and associated expenditure.

Governance
Access to database
  • In French only
  • Some aggregate data and studies available online. A good presentation in French on its design and content can be found here.
  • Access to the raw database requires pre-authorization with the CNIL, public agency responsible for data confidentiality and is tightly controlled.
Coverage
  • 2 years + current year
  • Patient data : age, gender, town, long term and chronic diseases, date of birth, date of death, lower income indicator, healthcare care consumption and date
  • All consultations and visits to GPs and ambulatory care specialists but nothing about their content
  • All medical technical procedures
  • All dispensed drugs
  • All lab and diagnostics tests but not their results
  • All medical devices
  • Provider level data: their activity and sales turnover, geography, prescribing behaviors.
Linkage
  • Linkage possible at patient level with PMSI (hospital activity and expenditure data) and ESPS survey (social determinants, healthcare needs and access issues)
Data quality
  • Presentation of data is extremely complex and requires high expertise of the French Social Security payment system to providers and reimbursement rules (copayments, special waivers for some categories of doctors and patients etc.)
  • There are practically no manuals of utilization
  • Access is very difficult, requires often working physically in CNAMTS offices
Strengths and weaknesses
  • The SNIIR-AM is a source of very rich information on healthcare consumption and expenditure.
  • It can provide strategic information on patient care pathways and their cost implications, especially when linked to other databases.
  • However, the data are not for public and research use.

Programme de Médicalisation des Systèmes d’Information (PMSI)

This exhaustive nationwide database is at the heart of the financing system of hospitals (activity-based payment system, “tarification à l’activité”), especially for acute and surgical care. It provides data on almost all claims paid by the Social Security System to hospitals and is therefore the main source of information on hospital activity and associated expenditure.

Governance
  • Technical agency for hospital information (ATIH, Agence Technique de l’Information sur l’Hospitalisation) http://www.atih.sante.fr
  • Contact person: demande_base(at)atih.sante.fr
Access to database
  • In French only
  • Some aggregate data and studies available online
  • Access to the raw database requires pre-authorization with the CNIL, public agency responsible for data confidentiality. Fees apply. See further explanations here (in French)
Coverage
  • All public and private hospitals
  • Patient level data based on the DRG classification :

PMSI-MCO for medical, surgical and obstetrics wards, "médecine, chirurgie, obstétrique" : from 1996 onwards

PMSI-SSR for post-operative and rehabilitation wards, "soins de suite ou de réadaptation" : from 1999 onwards

PMSI-PSY for psychiatric wards : from 2006 onwards (IRDES report available)

PMSI-HAD for home-based hospitalization, "hospitalisation à domicile": from 2008 onwards

  • Medical, surgical and obstetrics wards (MCO) database provides best coverage: only MCO is used in payment system
  • Coverage of other databases likely to improve alongside introduction of domain specific payment systems
  • Themes covered : case-mix, activity and associated claims of hospitals based on the patient classification system (GHM)
Linkage
  • Linkage possible at patient level with SNIR-AM (ambulatory care claims data) and ESPS survey (social determinants, healthcare needs and access issues) since 2012
  • Linkage possible at hospital level with SAE (hospital characteristics)
Data quality
  • Raw data is complex and administrative-based and requires high expertise of the French hospital payment system
  • DRG classification and coding rules have been changed regularly since 2000
Strengths and weaknesses
  • The PMSI provides key information on hospital activity and expenditure
  • The acute care database has the most reliable data because it is used for payment to hospitals
  • But it has not been designed for analytical but for payment purposes and is therefore very complex to use for studies
  • However, when used correctly, it can provide strategic information on hospital case mix and activity with their cost implications.
  • When linked to other databases, it can provide very rich patient pathway information across the healthcare system.

 

 

Statistique Annuelle des Etablissements

This yearly complete administrative survey of all public and private hospitals nationwide aims at:1) Characterizing precisely hospitals’ infrastructure: their capacity, technical equipment, operating theatres and healthcare personnel 2) Describe their activity profile 3) Produce indicators to follow-up national policies, control regulated activities and quality assessments in the hospital setting.

Governance
  • Ministry of Health’s statistics and research administration, Direction de Recherche, Etudes, Evaluation et Statistiques (DREES)
  • Contact person: drees-sae(at)sante.gouv.fr
Access to database
Coverage
  • From 1974-current but major changes in 1994 and 2000.
  • Complete survey of all public (approximately 1000 hospitals) and private (2000 hospitals) of all type (acute care, rehabilitation, physiciatric)
  • Themes covered: type of wards (A&E, maternity…), capacity (number of beds, occupancy rates…), equipment (scanners, MRI, operating theatres...), personnel (salaried, independent practitioners, pharmacists…), activity (number of hospital days, number of radiotherapy sessions…)

Complete list available here.

Linkage
  • No linkage at individual level with mortality or socio-economic data.
  • Linkage possible at hospital level with PMSI database at hospital level (DRGs claims data).
  • Linkage with PMSI is used by the Ministry in combination with quality surveys to produce its hospital quality comparison tool PLATINES.
Data quality
  • In 1994 and 2000 the design of the survey underwent major changes: therefore data for these years are very fragile.
  • Before 2000, the data cannot be considered as representative
Strengths and weaknesses
  • SAE has the advantage of being exhaustive
  • Linkages with the DRG claims data (PMSI) enable fine analysis of the relationships between hospitals’ characteristics, their costs and their activity features
  • When linked to safety and quality indicators, the database is used by the health authorities to assess hospital performance

Enquête décennale de santé

This nationwide highly representative survey is rolled out approximately every 10 years with the following objectives: 1) Describe the population's health status: declared morbidity and perceived health 2) Provide quantitative information about the use of care facilities and prevention services (in volume and value) 3) Allow an analysis of health status, use of care facilities and prevention, with regard to the socio-demographic characteristics (age, gender, profession, etc.) of individuals and their households.

 

Governance
  • National Institute of Statistics and Economic Studies (INSEE)
  • Contact person: comite-secret(at)cnis.fr
Access to database
  • In French only
  • Some reports and studies derived from the survey are available online. Access to the database is highly restricted and requires to send a request to the Comité du Secret Statistique
Coverage
  • From 1960-, every ten-years
  • Nationwide non-panel interrogation of 40,900 people in 16,800 households in the last survey (2002-2003)
  • The 2002-2003 Health Survey had five regional extensions (Nord-Pas-de-Calais, Picardie, Champagne-Ardenne, Île-de-France, Provence-Alpes-Côte-D'azur)
  • The computer-assisted data collection method (CAPI) is used
  • Themes covered : living conditions, social protection, general state of health, disability, description of illnesses, use and contact with health care providers, surgical antecedents, interruptions of work due to illnesses and periods of bed rest, use of care facilities over the period (hospitalization, doctor, biology, paramedical interventions, etc.), incapacities, dietary habits and prevention [see further information]
Linkage
  • At aggregate postal level with other databases
Data quality
  • The last survey (2002-2003), while continuing to provide comparisons with previous surveys, was more consistent with other European surveys, especially with regard to its health module which was significantly more detailed than in previous surveys.

Baromètre Santé

They are set-up since 1992 by the national agency responsible for health promotion and hazards prevention (INPES). They therefore tend to focus on non-healthcare determinants of health such as individual behaviors (substance abuse, sexual behaviors etc.), occupational hazards and prevention measures taken in primary care (screening, health promotion).

Governance
  • Hosted by the national agency for health promotion and prevention (INPES)
  • Contact person: This email address is being protected from spambots. You need JavaScript enabled to view it.
    Téléphone : 01 49 33 22 22
    Télécopie : 01 49 33 23 9
Access to database
  • French only
  • Aggregate data and reports available online
  • Individual data can be obtained with the approval of the demand by Inpes
Coverage
  • Since 1992 there have been:

12 nationwide surveys targeted at adults

1 nationwide survey targeted at teenagers (12-19 years)

4 surveys targeted at general practitionners

Several regional surveys

  • Not regular, adhoc surveys (see presentation)
  • Randomized sampling by phone
  • Themes covered in the adult surveys : see list at the bottom of page
  • Themes covered in the GP surveys : see list
Linkage
  • No linkage possible
Data quality
  • Surveys method by phone is problematic
  • Questions change from one survey to other
Strengths and weaknesses
  • Baromètre Santé surveys are a reliable source for the study of behavioral determinants of health
  • However, the lack of consistency in time (not longitudinal) and the changes in the specific questions asked between each survey is a limit.

 

Enquête Santé et Protection Sociale (ESPS)

This survey is a key source of information on France’s health status, healthcare coverage and access, social determinants and healthcare needs and demand.

Governance
  • Conducted by the Institut de Recherche et Documentation en Economie de la Santé (IRDES) with the support of National Health Insurance Fund (CNAMTS)
  • Contact person: Stéphanie Guillaume +331 53 93 43 34 esps(at)irdes.fr
Access to database
Coverage
  • Every 2 years from 1988 to 2014
  • Nationwide panel interrogation of 22000 individuals in 8000 households
  • Half of the panel is surveyed every 2 years
  • 96% representativeness of the French population
  • Themes covered : health status, socio-economic situation, healthcare utilisation, insurance coverage (statutory and optional), social capital, life style [see complete list]
Linkage
  • At aggregate postal level with other databases
  • At individual level with the National Health Insurance Database (claims data)
Data quality
  • Questions on health status changed over time. Mini European module (self-perceived health) integrated in 2004.
  • The panel has been entirely renewed in 2010.
Strengths and weaknesses
  • ESPS is widely used for the evaluation of healthcare policy reforms, especially concerning equity of healthcare access issues.
  • Moreover, the linkage with the claims data of the French National Health Insurance Fund (SNIIR-AM) and with the Ministry’s PMSI enables refined analysis of healthcare resource consumption and expenditures.
  • But there is little information on the quality of the services and individuals’ experience with the health system.

 

Echelle Nationale des Coûts (ENC)

This survey of the cost weights for approximately 90 private and public hospitals is at the heart of the financing system of the hospital sector, as it is used by the Ministry to set the fee schedule of the DRG system.

Governance
Access to database
  • In French only
  • Some aggregate cost-weights data is available online.
  • Access to patient level cost data requires an authorization from CNIL (the public agency responsible for data confidentiality) and ATIH and very difficult to obtain.
Coverage
Data quality
  • Interpretation of data requires expertise of the French hospital funding system in France
  • Some manuals of utilization provide guidance
  • Access to patient level data is very difficult to obtain
Strengths and weaknesses
  • Detailed cost weights at the DRG level of a representative survey of public and private hospitals
  • However, interpretation and access remain very difficult