General Information

Background

Stroke is a general term used for acute vascular diseases of the brain. It includes brain infarctions (approx 85%), intracerebral haemorrhages (approx 10%) and subarchnoidal haemorrhages (approx 5%).

About 30,000 patients in Sweden suffer annually from a stroke, of which around 70 per cent are first-ever events. Patients with a stroke use more than a million bed days in hospital? per year within the healthcare system in Sweden and an even a higher number of bed days in nursing facilities and other municipal assisted living facilities. Stroke is the single somatic condition category requiring the most bed days in the country. The total estimated cost to the society has been estimated at SEK 18.3 billions (€ 1.9 billions) per year.

The national quality register Riksstroke was established in 1994. Since 1998 all hospitals admitting acute stroke patients participate. With a total of more than a 450 000 events in 2014 recorded, Riks-Stroke is one of the world’s largest stroke registers.

Approximately 10.000 Swedes face a transient ischemic attack (TIA) every year. The TIA registration started 2010 for patients treated in hospital and 2014 it was 59 of 72 hospitals are registrating TIA. From 2015, hopefully all hospitals in Sweden are recording TIA patients.

Aims

Riksstroke is a tool for continuous quality improvement of stroke care. The aim of the register is to support high and consistent quality of care for stroke patients throughout Sweden, ultimately to ensure patient benefit in the form of the best possible care.

Riksstroke is also an instrument for following up the National Board of Health and Welfare’s guidelines for stroke care.

In addition, Riksstroke provides a database for research on stroke above all management and outcomes in routine clinical settings.

Design, measurements and coverage

This section provides a brief description of the Riksstroke methodology. A more detailed description is available.

Information collected in Riksstroke

For each individual patient, a case record is used to collect data from before onset, during the hospital stay and at discharge. A separate record is used for the 3-months and 1 years follow-ups by questionnaire. English versions of the case records are available at http://www.riksstroke.org/forms/

Riksstroke includes information on several dimensions of stroke care:

  • The entire chain of stroke care, i.e. primary prevention, acute management, rehabilitation, secondary prevention and family and community support
  • Both process and outcome variables
  • Five dimensions of healthcare quality as defined by the US Institute of Medicine and the National Board of Health and Welfare are included in Riksstroke: Evidence-based health care, safe, provided in time, distributed fairly and patient-oriented.

To analyse processes and outcomes in relation to case-mix, a number of background variables are recorded.

 

 Examples of information collected

Background data

  • Living conditions (at home, in institution)
  • Marital/cohabitant status
  • Primary ADL functions before stroke
  • Previous stroke
  • Co-morbidity
  • Smoking
  • Level of consciousness on admission Time to arrival in hospital

 

Process indicators

  • Admission to stroke unit
  • Diagnostic procedures
  • Thrombolysis
  • Oral anticoagulants after cerebral embolism
  • Other antithrombotic therapy
  • Antihypertensive therapy
  • Statins
  • Access to rehabilitation after discharge
  • Length of hospital stay

 

Outcome indicators

  • Medical complications during hospital stay
  • Discharge status
  • Survival at 3 months
  • Living conditions at 3 and 12 months
  • ADL dependency at 3 and 12 months
  • Support by next-of-kin at 3 and 12 months
  • Patient satisfaction with hospital stay and rehabilitation after discharge
  • Low mood at 3 and 12 months
  • Perceived general health at 3 and 12 months

 

Validations

A detailed analysis of Riksstroke data quality showed that data entry errors were marginal. A 95% consistency rate was noted from the diagnosis stated in Riksstroke and the medical chart. Consistency of patient data was good, but it was lower (approx. 85%) for data related to the healthcare organisation, which reflects differences in how stroke care is organised in various hospitals.

The questions on functional capacity, low mood, general health and fatigue have been tested for validity. All questions showed good correlation with standardised measuring instruments for the respective area.

An extensive validation work took place under 2013 and 2014. More information on validations of the questions/instruments used in Riksstroke is available at Evaluations of variables in Riksstroke

Information to patients

Quality monitoring is, by law, required in the Swedish healthcare system. As an instrument for quality assessment, Riksstroke is a component of the regular healthcare system rather than a research project but patient information is compulsory and the patient can, after information, choose to refrain participation – an opt-out procedure. Riksstroke have chosen to provide the patient information. Patients are most often informed in writing

  • on the purpose of Riksstroke
  • on what type of information that is collected during the acute phase
  • on the 3-month follow-up questionnaire
  • that the register is confidential
  • that participation in Riksstroke is voluntary, that every participant has, by law, the right to know what data is being recorded and that it is possible to withdraw participation (opt-out procedure)
  • on permit by the Swedish Data Inspection Board
  • on contact addresses

If the patient is unable to communicate, a next-of-kin receives this information. A translation into English and other languages of the written information is available http://www.riksstroke.org/patient-information/

Data capture and handling

All acute stroke incidents in hospitals are to be registered. Data is usually collected through completion of printed forms during the hospital stay. At 3 months after stroke, patients are asked to respond to a questionnaire by letter, phone or at a return visit. If the patient is unable to respond, the questionnaire is filled in by a next-of-kin or by healthcare/social service staff.

The recorded data is transferred to the Riksstroke data base managed by Umeå University computer center. Control functions to ensure the accuracy of personal identity number, sex, dates, etc., are embedded in the data entry software and only valid response alternatives are accepted.

Patients who had a stroke in December should be follow-up during April and all incidents from the preceding full year and the associated 3 month follow-ups must be recorded by the participating hospital by 30th April. National data and between-hospital comparisons are processed by the Riksstroke secretariat in April/May.

In a similar way, TIA incidents treated in hospitals are registered through completion of printed forms during the hospital stay. This record shares many of the items from the acute stroke record but is more focused on diagnostics and treatment procedures and does not include a 3- and 12 months follow ups.

Annual number of registrations is around 25,000 records and has so been the last 10 years. In total the Riksstroke database contains more than 400,000 stroke episodes.

From 2010 most of the hospitals also are recording TIA and the TIA database contains around 30,000 events.

Coverage

Coverage of hospitals. From 1998, all hospitals in Sweden admitting patients with acute stroke report to Riksstroke.

 

Three-months and 12 months follow-ups. The participation rate for the 3-month follow-up is around 90 per cent, The follow-ups are performed by sending a questionnaire to the patient, or by telephone interviews or at a return visit. and has always been good. At the 12 month follow-up the follow-up is little lower around between 75-80 per cent. The 12-month follow-up is administrated by sending a questionnaire to the patients.

Follow-up of deaths. Information on deaths is performed by linkage to the Swedish cause-of-death register (using personal identification numbers) with no loss of individuals.

Patients not recorded in Riksstroke. Analyses of non-registered vs. registered patients in Riksstroke have generally shown only small differences in patient characteristics. However, patients admitted to stroke units are somewhat more likely than patients admitted to general wards to be covered by Riksstroke.

Expertise

Together, the Steering Committee of Riksstroke and the secretariat contribute with expertise in:

  • internal medicine
  • neurology
  • geriatrics
  • nursing
  • rehabilitation
  • physiotherapy
  • patient perspective
  • epidemiology
  • statistics
  • healthcare administration
  • programming
  • data handling procedures

 

Analysis and feedback

Analytical reports on process and outcome quality indicators are posted on the Riksstroke website. The variables are analysed at the national, regional and hospital levels. To facilitate data interpretation, background information, including patient characteristics and coverage, is presented at the hospital level.

Key quality indicators in Riksstroke are also presented in a report on the quality of Swedish healthcare produced annually by the National Board of Health and Welfare and the Swedish Association of Local Authorities and Regions.

Participating hospitals have access to a statistical and presentation package to analyse and present their own data more specifically and compare these with national data.

Scientific work

In addition to promoting the development of quality of stroke care in Sweden, Riksstroke provides a unique database for research on stroke services in routine clinical practice.

Articles based on Riksstroke data see here. In addition, Riksstroke data (with abstracts) have been presented at numerous national and international meetings.