| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Department of Clinical Biochemistry, Pharmacology, and Genetics, Odense University Hospital, Odense, Denmark
2 Department of Clinical Biochemistry, Amager Hospital, Copenhagen, Denmark
3 Department of Biostatistics, Copenhagen University, Copenhagen, Denmark
4 Department of Quality and Research, Randers Regional Hospital, Randers, Denmark
5 Research Unit for General Practice, Centre of Health and Society, Copenhagen, Denmark
6 Management Division, Glostrup University Hospital, Copenhagen, Denmark
aAddress correspondence to this author at: Department of Clinical Biochemistry, Pharmacology, and Genetics, Odense University Hospital, DK-5000, Odense C, Denmark, Fax 45-65-41-19-11, E-mail Jonna.Skov.Madsen{at}ouh.regionsyddanmark.dk
To the Editor:
The results of biochemical tests often lead to diagnostic and therapeutic interventions, and the real value of a test can be assessed only by taking into account the subsequent health outcomes. The importance of outcomes studies, and the challenges in performing them, was reviewed by Bruns in 2001 (1), who argued that this type of study should be performed more frequently, and that such studies should be used to determine whether new tests should be implemented in clinical practice.
To investigate the extent to which this recommendation has been realized, a working group on evidence-based clinical biochemistry established by the Danish Society of Clinical Biochemistry undertook a pilot study to record the number and type of reports of diagnostic biochemical outcome studies published from January 2005 to January 2006 in 4 medical journals: Clinical Chemistry, Clinical Chemistry and Laboratory Medicine, Lancet and the New England Journal of Medicine. To be included as an outcome study, the reported study had to be designed to investigate outcomes in relation to a clinical or an economical variable of a well-defined clinical application of a biochemical test.
To identify reports of outcome studies, 2 authors manually went through reports published in each of the journals within a 12-month period. Detailed information on original full-length reports considered diagnostic biochemical outcome studies was registered together with the total number of original articles. Technical Briefs, Letters, Short Communications, Editorials, and Reviews were not included. When there were discrepancies in report selection by the 2 authors scrutinizing the same journal issues, a consensus decision was made in the entire author group. Selected outcomes reports were classified as investigating (A) direct clinical mortality or morbidity; (B) other clinical variables such as length of hospital stay, readmission rate, or satisfaction with care; or (C) economic outcomes.
A total of 829 original articles were registered, of which only 7 studies (0.8%) were classified as diagnostic biochemical outcome studies (Table 1
). Six (of 231) of these original articles were published in the New England Journal of Medicine and 1 of 179 original articles in Clinical Chemistry. Three articles were classified as A, 2 as A+B, 1 as A+C, and 1 as B.
|
An outcome study addresses the question of whether use of the studied intervention (in this case a laboratory test) leads to an anticipated outcome (1)(2). The randomized controlled trial is a powerful design strategy for such studies, avoiding many pitfalls that occur with other study designs (2). Three of the outcome studies identified in this pilot study used a randomized controlled trial design that involved some variant of a test-treat-counsel policy to be compared with a policy not involving the testing element.
Before-and-after diagnostic assessment of clinical impact is another appropriate design for evaluating clinical outcomes of the use of laboratory tests (3), as elegantly demonstrated in the study by Stramer et al. (Table 1
). This study investigated the effect of prospective screening of all blood donors for West Nile virus, and the results indicated that carriers of the virus were eliminated from the US Red Cross blood supplies, and no infections were detected among recipients after the introduction of this screening program.
We identified good examples of diagnostic biochemical outcome studies, but the absolute number of these studies was disappointingly low, indicating insufficient documentation of the health outcomes produced by diagnostic biochemical analyses. This insufficiency is probably attributable to multifaceted causes. An important aspect is undoubtedly the complexity and high costs of outcome studies (1)(4), because many steps lie between test findings and outcome. Research is needed that addresses methodological issues concerning design and conduct of test-outcome studies. Comprehensive discussions of this complex and important area are available (1)(2)(3)(4), and in the textbook Evidence-Based Laboratory Medicine the chapter on assessment of outcomes is especially relevant (5).
Our purpose is to highlight the importance of moving from diagnostic accuracy studies to evaluations of the effects of test results on clinical decision-making and subsequent health outcomes. In agreement with other investigators (1)(2)(3)(4)(5), we support efforts to increasing the use of outcome studies to enhance the effectiveness of healthcare policy and decision-making.
Acknowledgments
Grant/Funding Support: None declared.
Financial Disclosures: None declared.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
