Current issues of ACP Journal Club are published in Annals of Internal Medicine


Blood donors with persistently indeterminate HIV Western blot results were not infected with HIV

ACP J Club. 1991 July-Aug;115:22. doi:10.7326/ACPJC-1991-115-1-022

Source Citation

Dock NL, Kleinman SH, Rayfield MA, et al. Human immunodeficiency virus infection and indeterminate Western blot patterns. Prospective studies in a low prevalence population. Arch Intern Med. 1991 Mar;151:525-30.



To determine whether blood donors with persistent indeterminate Western blot (WB) patterns are infected with the human immunodeficiency virus (HIV).


3 cohorts of blood donors with different baseline HIV test results were followed for development of HIV infection.


3 urban U.S. Red Cross Blood Services regions from 1985 to 1988.


3 groups of donors (n = 387) were studied: 163 with confirmed positive HIV serologic results; 115 with indeterminate results; and 109 age-, sex-, and race-matched donors with negative results. Follow-up information was insufficient for 35 donors. Mean follow-up time was 12 months.

Description of test and diagnostic standard

Antibody tests consisted of enzyme immunoassays (EIA) supplemented by WB. A blot with bands of sufficient intensity to detect the major viral proteins (gag,pol, and env) was interpreted as positive; a blot with any bands, but not meeting the positivity criteria, was interpreted as indeterminate. Frozen peripheral blood mononuclear cells collected at all donor visits were assayed for proviral DNA by DNA amplification using the polymerase chain reaction (PCR).

At visits every 6 months, a general health examination and selected hematologic and immunologic assays were done; questionnaires on risk behavior for HIV infection were completed.

Main outcome measures

Laboratory and clinical indicators of HIV infection.

Main results

All seropositive donors' WB patterns remained stable during the study. 61 of 98 indeterminate donors had stable reactivity; 22 donors changed from initially negative to indeterminate, and 15 individuals changed from indeterminate to negative WB patterns. 1 participant in the indeterminate group, who had homosexual contact, progressed to positive reactivity to HIV-1. None of the donors with indeterminate or negative WB patterns showed abnormalities on physical examination or in laboratory measurements. Primary, secondary, or minimal risk factors were reported by 5, 1, and 8 donors, respectively, in the indeterminate group; and by 1, 1, and 4 donors, respectively, in the seronegative group (P > 0.2). 5 of 5 seropositive individuals' peripheral blood mononuclear cells were positive by PCR; all 98 indeterminate donors and 5 of 5 seronegative individuals tested by PCR were negative for gag and env sequences.


No evidence was found that blood donors with persistently indeterminate HIV Western blot patterns progress to HIV positivity.

Source of funding: American Red Cross.

Address for article reprint: Dr. N.L. Dock, Research and Development Laboratory, American Red Cross Blood Services, 636 South Warren Street, Syracuse, NY 13202, USA.


This paper addresses the important topic of the fate of persons with indeterminate results to HIV antibody analysis. Unfortunately, because the report provides insufficient information about the sample being studied, direct clinical application of the data may be limited. The subjects were taken from a sample of healthy blood donors described as "low prevalence," but the selection criteria were not clearly explained. 17% of the indeterminate group was lost to follow-up but no comparison with those followed up was made.

Risk behaviors for HIV infection were similar in the indeterminate and negative groups, but even if a difference was seen between the groups it may not have been detected because of the small sample size. A history of risk behaviors alone does not indicate that a patient with an indeterminate test is infected, but this information is important for counseling the patient. Although the clinical status of the donors was described, information about factors that might be important in the clinical applicability of these findings, such as sex and a history of autoimmune disease, were omitted. WB testing was well described.

This study concludes that a persistent indeterminate HIV result "is very unlikely to signify the presence of infection with HIV." If the characteristics of the sample group had been described in more detail, clinicians would be able to generalize the results to their patients more readily. Patients with indeterminate results require repeat testing; a person who remains indeterminate for 6 months without risk factors or symptoms is almost certainly not infected. Certain patterns have been identified (absence of p24 and env) that indicate freedom from HIV infection. Additional studies such as this will strengthen the confidence with which we tell our patients that a persistently indeterminate WB test result does not represent infection.

Neil Wenger, MD
University of CaliforniaLos Angeles, California, USA


The new test such a viral loading, the new generation of rapid enzyme assays, and screening with p24Ag plus the new subtypes of HIV infection may change our ability to identify patients who have early sero conversion. This may alter the clinical implication of the above study and how we counsel patients with indeterminant WB test results.

Fiona Smaill, MB BS
McMaster University, Faculty of Health SciencesHamilton, Ontario, Canada
Nancy L. Dock, Ph D
Greater Northeast RegionSyracuse, New York, USA