IngentaConnect Are our impressions of allergy test performances correct?

Authors: BrockWilliams, P¹; Ahlstedt, Staffan; Barnes, James H.²; Söderström, Lars³; Portnoy, Jay⁴

Source: Annals of Allergy, Asthma and Immunology, Volume 91, Number 1, July 2003 , pp. 26-33(8)

Publisher: American College of Allergy, Asthma, & Immunology

Abstract:

Background: The clinical diagnosis is often subjective and susceptible to bias, yet it is the primary standard by which diagnostic tests are judged. Consequently, our opinions regarding various diagnostic tests may not be entirely accurate.

Objective: To investigate the accuracy of the clinical history compared with concordant skin and quantitative specific IgE (s-IgE) measurements.

Methods: Consecutive, consenting patients (N = 152) at 2 different allergy centers were examined by history and physical examination (HPE) alone to determine their sensitivity to 7 common allergens. Results were classified as positive, negative, or indeterminate. The HPE results were then compared to concordant skin prick testing (SPT) and s-IgE measurements and to quantitative IgE antibody measurements with and without knowledge of the SPT results.

Results: Diagnosis by HPE deviated considerably from concordant SPT and s-IgE results. This deviation differed between allergists and allergens, reflecting a positive HPE bias that averaged 22%. Seventy-six percent of the HPE results judged indeterminate were resolved as negative. Using additional information from the quantification of s-IgE antibodies, considerable differences between the sites in the level of s-IgE associated with a positive HPE result with and without SPT results were observed.

Conclusions: Relative to the SPT and quantification of s-IgE antibodies, the diagnosis by HPE alone to common allergens is not consistent. Discrepancies were dependent on both allergen and allergist. The quantitative s-IgE data revealed that allergists use available information from the HPE and SPT differently. Since the HPE is the primary standard used in judging test efficacy (sensitivity and specificity), our current impressions of test performances are not likely to be accurate.

Document Type: Original article

Affiliations: 1: University of Missouri Medical School, Kansas City, Missouri. 2: School of Business Administration, University of Mississippi, Oxford, Mississippi (deceased). 3: Pharmacia Diagnostics AB, Uppsala, Sweden. 4: Children’s Mercy Hospitals and Clinics, Kansas City, Missouri.

The full text electronic article is available for purchase. You will be able to download the full text electronic article after payment.

$28.00 plus tax Refund Policy