Correspondence
Clinical Problem-Solving: Costly Errors
N Engl J Med 1995; 333:1080-1081October 19, 1995
- Article
To the Editor:
In the Clinical Problem-Solving case presented by Dr. Duffy (June 1 issue),1 concerning a patient with hemolytic anemia due to cold agglutinins who later received a diagnosis of large-cell lymphoma, there was a failure to probe for an important pathogenetic link that might have led to more rational and effective therapy. Serum electrophoretic and immunofixation studies in patients with cold-agglutinin anemia and lymphoma usually reveal a monoclonal immunoglobulin of the IgM type, and in rare cases of the IgG or IgA type.2
Coupled with phenotypic studies of lymphoma tissue (i.e., heavy and light chains and CD5 expression), such studies might have indicated the likelihood of monoclonal IgM production by the lymphoma, as in Waldenström's macroglobulinemia. Sometimes, the serum must be collected under warm conditions to prevent the false negative results produced by a type I cryoglobulin that may coexist with the cold agglutinin. Thus, the presumed “large-cell lymphoma” may have reflected the presence of plasmacytoid B lymphocytes, a finding consistent with the low-grade lymphoma of Waldenström's disease. Although there is no consensus on the definition of Waldenström's macroglobulinemia, the presence of monoclonal IgM and plasmacytoid B-cell lymphoma suggests this diagnosis and identifies patients who may benefit from treatments similar to those for macroglobulinemia.3
If monoclonal macroglobulinemia due to lymphoma were confirmed, the optimal treatment would differ from that usually given for large-cell lymphoma. Primary treatment of macroglobulinemia with a combination of an alkylating agent and a glucocorticoid has induced remissions in approximately 50 percent of patients, but we have reported a response rate of 80 percent among patients treated with two courses of the nucleoside analogue 2-chlorodeoxyadenosine.4 The median duration of remission after the completion of treatment has been 18 months, with an 80 percent frequency of recontrol with the same treatment among patients with relapses. The treatment has usually been well tolerated, and although a central venous catheter is currently required for a short period, recent studies suggest that subcutaneous administration is likely to be equally effective.5
5 ReferencesDonna M. Weber, M.D.
Danai Daliani, M.D.
Raymond Alexanian, M.D.
University of Texas M.D. Anderson Cancer Center, Houston, TX 770301
Duffy TP . Costly errors. N Engl J Med 1995;332:1503-1505
Full Text | Web of Science | Medline2
Crisp D ,Pruzanski W . B-cell neoplasms with homogeneous cold-reacting antibodies (cold agglutinins). Am J Med 1982;72:915-922
CrossRef | Web of Science | Medline3
Dimopoulos MA ,Alexanian R . Waldenstrom's macroglobulinemia. Blood 1994;83:1452-1459
Web of Science | Medline4
Dimopoulos MA ,Kantarjian H ,Weber D , et al. Primary therapy of Waldenstrom's macroglobulinemia with 2-chlorodeoxyadenosine. J Clin Oncol 1994;12:2694-2698
Web of Science | Medline5
Juliusson G ,Heldal D ,Hippe E , et al. Subcutaneous injections of 2-chlorodeoxyadenosine for symptomatic hairy cell leukemia. J Clin Oncol 1995;13:989-995
Web of Science | Medline
To the Editor:
We found the article by Dr. Duffy thought-provoking. It stimulated us to voice our concern about quality assurance in laboratory medicine and the interactions between laboratory professionals and clinicians in practice. We believe that the attribution of the “first mistake” to the blood bank was perhaps misplaced. Given what we know about staff training and standards of good laboratory practice in the United States, the failure of laboratory personnel to understand the implications of a positive Coombs' test seems unlikely. Most laboratorians would agree that a peripheral-blood smear should routinely accompany a bone marrow aspirate. Alarm bells may have rung in the ears of the laboratory technologists involved, but communication beyond that point clearly failed. It may not be so easy for concern at the laboratory to be passed on to a patient's physician. In 1978 Benson noted that it is often difficult for pathologists to communicate with clinicians.1 Shulman et al. have noted that when communication is successful, improvement occurs.2
When discussing diagnostic mishaps, clinicians and pathologists should focus more on communication and less on the errors of the individual persons involved. They must adopt a paradigm for quality that examines the interaction among all professionals who may be involved in a case (i.e., nurses, physicians, laboratory directors, medical technologists, and clerical staff). At the same time, all branches of the medical profession must collaborate with laboratorians to develop studies that examine the determinants of the quality of laboratory work as they affect clinical outcomes.3
Duffy has done us a great service by calling this case to our attention. We would be interested to know what steps were taken by the various professionals involved to improve communication after this experience.
3 ReferencesRichard A. Keenlyside, M.D.
Steven J. Steindel, Ph.D.
Centers for Disease Control and Prevention, Atlanta, GA 303331
Benson ES. Strategies for improved use of the clinical chemistry laboratory in patient care. In: Benson ES, Rubin M, eds. Logic and economics of clinical laboratory use. New York: Elsevier, 1978:245-58.
2
Shulman IA ,Lohr K ,Derdiarian AK ,Picukaric JM . Monitoring transfusionist practices: a strategy for improving transfusion safety. Transfusion 1994;34:11-15
CrossRef | Web of Science | Medline3
Boone DJ ,Steindel SJ . Conducting outcomes research: past experience and future directions. Clin Chem 1995;41:795-798
Web of Science | Medline
To the Editor:
I take strong exception to the statement that “In retracing the events of this complex case, it seems that the first mistake was made by the blood bank.” What did the blood bank report? Was the test performed as a pretransfusion test, or was it performed at the specific order of a physician?
Some blood banks still routinely perform a Coombs' test as part of a pretransfusion serologic evaluation, although this test is not required according to current standards.1,2 The Coombs' test is known to have a high false positive rate (i.e., relatively few patients with positive results have identifiable factors suggesting hemolysis). Easily 10 to 15 percent of hospitalized patients and a larger number of patients infected with the human immunodeficiency virus have positive Coombs' tests.2,3 Furthermore, as many as 1 in 1000 healthy blood donors has a positive test.2
Most blood banks routinely use a “nonspecific” reagent to perform an initial Coombs' test in order to detect molecules of either IgG or C3 on circulating red cells. Further testing to determine whether IgG or C3 (or both) is present is not recommended unless the patient has recently undergone transfusion or there is evidence of hemolysis.2
It is not the clinical laboratory's responsibility to do further testing unless the clinician responds to the initial test results by requesting further testing for the purposes of clarification and diagnosis. To suggest that, without a request by the clinician to perform further testing, the blood bank should have tested for cold agglutinins (a test optimally performed with blood collected and kept warm during the process of separating serum and cells) because hemolysis was suspected is unreasonable and not cost effective. It is also unreasonable to expect the blood bank to have collected the other data, such as the discrepancy between the hemoglobin and hematocrit values and the high sedimentation rate, in order to decide to do additional testing. That was the clinician's job, and the clinician failed. Perhaps the costly error here was the attempt to treat a severe (transfusion-requiring) hemolytic anemia by a clinician without enough experience in hemolytic syndromes and without the benefit of an initial evaluation by a more experienced hematologist. With the current emphasis on primary care and efforts to reduce costly, unnecessary laboratory testing, primary care physicians are going to need to learn when to call for help in time.
3 ReferencesMarilyn J. Telen, M.D.
Duke University Medical Center, Durham, NC 277101
Compatibility testing and selection of components. In: Standards for blood banks and transfusion services. 16th ed. Bethesda, Md.: American Association of Blood Banks, 1994:25-9.
2
Investigation of a positive DAT. In: Walker WH, ed. Technical manual. 11th ed. Bethesda, Md.: American Association of Blood Banks, 1993:355-87.
3
McGinniss MH ,Macher AM ,Rook AH ,Alter HJ . Red cell autoantibodies in patients with acquired immune deficiency syndrome. Transfusion 1986;26:405-409
CrossRef | Web of Science | Medline
Author/Editor Response
Dr. Duffy replies:
To the Editor: Telen's exemption of the blood bank from any role in the routine investigation of a positive Coombs' test is countered by Keenlyside and Steindel's suggestion that communication between pathologists and clinicians would improve the overall quality of patient care. It would not be cost effective for a blood bank to perform a full workup of every positive Coombs' test, but with the increasing reliance on primary care, there must be greater interaction among all professionals involved. Physicians trained in blood banking and laboratory medicine are ideally situated to communicate with and educate those who are the recipients of their laboratory data. Such communication would have been appropriate in the current case, when one considers that the long duration of the problem must have been evident to all the parties involved.
Weber et al. highlight the important role of 2-chlorodeoxyadenosine in the treatment of plasmacytoid lymphomas associated with monoclonal IgM production. The pathological features in the case under discussion, however, were not those of a plasmacytoid lymphoma but rather those of a diffuse, large-cell lymphoma. The “important pathogenetic link” in this case of cold-agglutinin hemolytic anemia and lymphoma is that successful treatment of the hemolytic problem resided in the recognition and treatment of the lymphoma.
Thomas P. Duffy, M.D.
Yale University School of Medicine, New Haven, CT 06520-8021
