External quality assurance in Norwegian primary health care

In Norway, with a population of 4.2 mill, there are about 1800 office laboratories run by general practitioners or private specialists. Most practices are group practices with 2-3 physicians, and most primary care are fee for service, including a fee for each laboratory procedure.

In 1992 the Norwegian health authorities and the Norwegian Medical Association made a joint project to secure the quality of laboratory analyses performed in primary health care. Two co-ordinated activities were established: in each of the 19 Norwegian counties a medical laboratory adviser linked to the central clinical chemistry laboratory in that county was engaged. These advisers assist general practitioners and primary care specialists in securing the quality of office laboratory procedures by visits, telephone contacts, and correspondence. In addition, a Norwegian centre for external quality assurance in primary care (Norsk senter for kvalitetssikring av laboratorieanalyser utenfor sykehus - NOKLUS ) was established. The cost of the project is NOK 10 mill per year.

NOKLUS

NOKLUS was established in the autumn of 1992. Participation in the EQAS is voluntary, there is no fee for participation, and the scheme has an educational purpose. In 1994 there were 1700 participants, comprising 95% of primary care practices. Several prerequisites must be taken into account when designing a primary care EQAS:

• The EQAS organiser has to communicate directly with the user of tests.
• The forms to be completed and the feed-back reports must be simple and understandable.
• The goals (and limits) for analytical quality must seem reasonable for the user.
• Advises concerning which actions to be taken must be clear and explicate.
• The workload and cost must seem reasonable and justifiable

Fresh materials are used for haemoglobin (EDTA blood ) and cholesterol (serum), and target values are therefore identical for all instruments and methods. The target values are determined with reference methods on two different sites. For glucose, CRP, prothrombin time, and urine commercial control material is used, and method specific target values are calculated using a truncated median.

Quality goals are based on target intervals and a percentage tolerated deviation

The goals for performance in primary health care should be similar to that on hospital laboratories. The tolerated limits are based on target intervals together with a percentage tolerated deviation.

Example

For haemoglobin the target interval = target value ± 0.1 g/dl. The percentage tolerated deviation from the target interval to be characterised as "good" is 3% (the percentage tolerated deviation for being acceptable is 8%), and is added to the limits of the target interval.

Using this method, the percentage tolerated deviation from the target value to be categorised as "good" will vary depending on the concentration of the constituent (unbroken lines) as opposed to a fixed percentage deviation (dotted lines) (see figure). These fences are comparable to the suggested "biological" fences (1) which is between 3 and 5 % for haemoglobin.

Evaluation: - good - acceptable - poor

It is important for the practitioner to receive understandable feedback reports, and we must be willing to express the results and evaluation in simple terms. The feed back reports generally state whether the measurements were "good", "acceptable" or "poor" together with suggestions for further action. To be categorised as good the performance has to be better than the quality goals suggested taking into account intra- and interindividual biological variation. This is the ultimate goal for the participants. Limits for acceptable are pragmatic looking to what can be a reasonable, taking care that the analytical performance is sufficient in most clinical circumstances.

An evaluation report is also made for the laboratory consultants with information of the performance of the participants in his/her county in addition to information pertinent to counselling. Participants with the evaluation "poor" are encouraged to take contact with their laboratory adviser.

Future

The goal for NOKLUS is to design quality assessment schemes for all constituents analysed in office laboratories. Since a lot of information about each practice is registered, it will be possible to relate analytical performance to different aspect of the practices e.g. use of internal controls, type of staff performing the analyses, number of analyses per week, types of instruments etc. The project will be evaluated with respect to consequences for analytical quality, as well as clinical practice, in order to elucidate the balance of cost and benefit.

Reference

1. Fraser CG, Hyltoft Petersen P. Quality goals in external quality assessment are best based on biology. Scand J Clin Lab Invest 1993; 53 suppl 212: 8-9.

Home page Department
Home page University

Department of Public Health and Primary Health Care, last updated 02.09.99