Testing for allergic disease: Parameters considered and test value
© Szeinbach et al; licensee BioMed Central Ltd. 2008
Received: 18 February 2008
Accepted: 26 August 2008
Published: 26 August 2008
Test results for allergic disease are especially valuable to allergists and family physicians for clinical evaluation, decisions to treat, and to determine needs for referral.
This study used a repeated measures design (conjoint analysis) to examine trade offs among clinical parameters that influence the decision of family physicians to use specific IgE blood testing as a diagnostic aid for patients suspected of having allergic rhinitis. Data were extracted from a random sample of 50 family physicians in the Southeastern United States. Physicians evaluated 11 patient profiles containing four clinical parameters: symptom severity (low, medium, high), symptom length (5, 10, 20 years), family history (both parents, mother, neither), and medication use (prescribed antihistamines, nasal spray, over-the-counter medications). Decision to recommend specific IgE testing was elicited as a "yes" or "no" response. Perceived value of specific IgE blood testing was evaluated according to usefulness as a diagnostic tool compared to skin testing, and not testing.
The highest odds ratios (OR) associated with decisions to test for allergic rhinitis were obtained for symptom severity (OR, 12.11; 95%CI, 7.1–20.7) and length of symptoms (OR, 1.46; 95%CI, 0.96–2.2) with family history having significant influence in the decision. A moderately positive association between testing issues and testing value was revealed (β = 0.624, t = 5.296, p ≤ 0.001) with 39% of the variance explained by the regression model.
The most important parameters considered when testing for allergic rhinitis relate to symptom severity, length of symptoms, and family history. Family physicians recognize that specific IgE blood testing is valuable to their practice.
With the prevalence of allergic rhinitis estimated at 21% – 23% for the European population and 20% – 40% for the western population, appropriate diagnosis and treatment of allergic rhinitis is of global importance [1, 2]. Family physicians are usually first approached by patients experiencing symptoms; however, little information exists regarding the rationale to perform specific IgE blood testing, which parameters are most important, and the value of such testing. Given the need to determine if symptoms are truly attributed to allergic mechanisms, it is important that family physicians consider diagnostic testing in conjunction with a careful examination of patient history, clinical evidence, and environmental exposure factors to optimize patient care. The consequences of untreated symptoms can lead to multiple future complications while the consequences of misdiagnosis can lead to inappropriate treatments .
Chronic rhinitis has detrimental effects on quality of life and work productivity [4, 5]. Although medications may control symptoms in some patients, it is difficult to distinguish between allergic rhinitis and non-allergic rhinitis using clinical evaluation and medication trials. Two commonly applied methods are used to uncover an allergic etiology and identify possible causes. These include skin prick tests (SPT), and specific IgE tests that are thought to produce concordant measures on a dichotomous basis for specificity and sensitivity, as well as a propensity toward appropriate diagnoses in relation to the presence of specific IgE antibody levels [6, 7]. Decisions to utilize these tests are influenced by experience, patient history, diagnostic accuracy and efficacy of the test, and how well test results relate to symptoms [8, 9].
When presented with patient complaints and bothersome symptoms that may or may not be related to allergic rhinitis, physicians rely on numerous strategies to make an appropriate diagnosis. How family physicians weigh the importance of these patient-related parameters when recommending specific IgE testing is largely unknown, yet instrumental to determine appropriate treatment and follow-up therapy. To address this research question, we used a trade off approach (conjoint analysis) to evaluate family physicians' preference to recommend specific IgE blood testing with respect to patient symptoms, family history, and medication use. A second approach using visual analog scaling (VAS) was added to validate and compare findings obtained from the conjoint analysis. Visual analog scales have been used extensively in clinical assessment to quantify patient perceptions of disease severity and the impact of symptoms on health [10, 11]. Further evaluation was performed to determine if family physicians perceive that testing, as part of the care process was valuable to patient care. As healthcare gatekeepers, family physicians have the best opportunity to construct a baseline assessment of these patients to determine if current treatment strategies are effective, or if patients would benefit from a referral to an allergist or other specialist.
Primary care (family) physicians in a southeastern state in the United States were identified through already established medical societies and physician mailing lists that were complied at the Recruitment and Retention Shared Facility at the University of Alabama, Birmingham. Mailing addresses, telephone numbers, fax numbers, specialty area, and practice affiliation was verified for 424 physicians in Alabama. From the list of 424 physicians, a sample of 150 physicians was randomly selected to participate in the study. Three separate mailings containing 50 questionnaires were sent by priority mail, one week apart to these physicians at their respective practice sites. The questionnaire package contained a letter of invitation to participate in the study, along with a self-addressed stamped envelope for the returned questionnaire. Thirty-two questionnaires were completed and returned during the first month, follow-up reminder calls were conducted three weeks after the initial mailing (77 calls were answered), and 14 surveys were faxed per request by providers. A total of fifty completed surveys were returned within two months. The estimated sample size of 50 was determined for this study following examples for studies with repeated-measures designs . As an incentive, a gift certificate for a local department store was mailed to physicians who completed the questionnaire.
Techniques to evaluate preference include standard gamble, alternate rating (e.g., visual analog) scale, and time trade-off . Besides these techniques, conjoint analysis (a trade off approach among attributes) is another technique that is used to evaluate the importance of preference measurements . Choices are usually presented in the form of profiles that are ranked or rated (e.g., recommend specific IgE testing – yes or no). The part-worth values (coefficients) for each attribute are obtained from the random effects logistic regression model analysis with repeated measures (50 responses × 9 profiles = 450 observations), which follows stated choice experiments based on choice theory [15, 16]. Although developed in marketing research, the use of conjoint analysis in health care is becoming a valuable tool [17–20].
In an attempt to simplify the conjoint exercise for this study, each attribute was assigned three levels (see Additional file 1). For example, symptom severity was assigned "high," "medium," or "low," and symptom length appeared as symptoms for "less than 5 years," "symptoms for 10 years," and "symptoms for more than 20 years." Family history included "neither parent has allergic rhinitis," "mother has allergic rhinitis," and "both parents have allergic rhinitis." Medication use included the use of "over-the-counter medications to control allergy symptoms," "prescribed antihistamines," and the "prescribed nasal spray to control allergy symptoms." Effects coding was used to construct the numerical values of the profile attributes. The "low" level was the reference value and was denoted as -1. The "high" level was denoted as +1.
A one-third fractional factorial design using repeated measures was chosen to minimize the number of profiles to 9 thereby attempting to avoid respondent fatigue. Two additional profiles were produced manually as holdout profiles for use in validation . Each profile portrayed an individual with a pre-determined set of allergic symptoms and clinical indicators. For the dependent variable, family physicians were asked to provide a "yes" or "no" response to whether they would recommend specific IgE blood testing for this patient. For the purposes of this study, which specific IgE blood test was used by family physicians was not important, or what type of test (food or inhalant) was performed.
The hypothesis for this study was that the estimated part-worth values or coefficients, exponentiated to odds ratios in this study, for each of the four profile attributes were simultaneously equal to zero. In the next section, family physicians were asked to indicate if recommendations to test were influenced by managed care guidelines, value of testing, referral activities, familiarity with specific IgE testing, relation of test results to symptoms, and value of test to practice, the likelihood to use specific IgE testing using a ten-point scale "less likely to test" to "more likely to test." Specific IgE blood testing was rated using a scale ('1' = not valuable to '6' highly valuable) for overall value, value compared to skin testing, and value compared to not testing at all. Demographic characteristics of participating family physicians, such as age, gender, years in practice, and practice site information, were elicited in the last section of the questionnaire. To assess questionnaire validity, participants provided an estimate of a patient's overall health status given the impact of various symptoms and clinical indicators with 0 being the "Worst Possible State" and 100 being the "Best Possible State." The means for these items were compared to those rankings obtained from the conjoint exercise to offer additional information regarding testing.
Descriptive statistics were provided for demographic variables. The conjoint exercise data were analyzed using a random effects logistic regression model. This type of model was chosen since it produces standard errors that account for the intra-individual correlation. Assumptions of normality, linearity, and equal variances among the items were evaluated to ensure appropriate interpretation of statistical analyses. All statistical analyses were performed using Stata/SE version 9 (College Station, Texas, USA). An Institutional Review Board from the University of Alabama, Birmingham, granted approval for the study.
Demographic Characteristics of Family Physicians
Value (n = 50)
29 – 79
Years in Practice
2 – 52
Gender, no. (%)
Practice type, no. (%)
Managed care setting
Results from the random effects logistic regression model
Length of symptoms
5 years to 20 years
History of allergic rhinitis
OTC allergy medications*
Two methods were used to validate the results of the conjoint exercise – the use of a holdout profile and an alternate rating method. First, was to estimate predictive validity for the holdout profile using the regression model developed from the 9 orthogonal profiles. The relation between the observed response for the holdout profiles and the predicted responses was then examined. The predicted values for the holdout profiles were quite similar to the observed value. The predicted mean probabilities were 82.7% and 78.4% compared to the observed values 70% and 78%, respectively. The differences were not significantly different (t-test; p = 0.162, p = 0.996, respectively) suggesting that the conjoint model exhibits acceptable internal predictive validity. Second, was the use of a VAS where participants responded to each item from the conjoint study presented separately. Lower mean scores obtained for each domain indicated that the particular domain represented choices that were less desirable to the respondent. Symptom severity (mean = 36.7; SD = 16.4) and symptom length (mean – 36.0; SD = 16.6) were ranked the worst followed by medication use (52.6; SD = 21.5), and family history (mean = 61.0; SD = 24.0), revealing consistent response patterns between the conjoint study and the VAS.
Impact of testing issues on value
Descriptive Statistics and Scale Evaluation for Issues and Test Value
Testing issues *
Managed care practice guidelines
Patient's perceived value of the test
Reduced need to refer patients to allergists
Difficulty in interpreting test results
Familiarity with test use
Patient demand to have the test done
Type of allergic rhinitis (intermittent vs. persistent)
How well test results relate to symptoms
Value of testing to my practice
Testing value **
Overall value of specific IgE as a diagnostic tool
Compared to skin testing, usefulness of specific IgE blood testing
Compared to not testing at all, usefulness of specific IgE blood testing
Linear regression analysis was used to assess the relationship between testing issues and testing value. As hypothesized, results using composite scores for testing issues and testing value revealed a moderately positive association between these two dimensions (β = 0.624, t = 5.296, p ≤ 0.001) with (R2 = 0.39) 39% of the variance explained by the model.
According to our results, family physicians consider symptom severity to be the significant determinant, followed by symptom length and family history when recommending the use of specific IgE blood testing for patients suspected of having allergic rhinitis. Physicians in practice for 10 years or more placed greater value on specific IgE testing compared to those in practice for less than ten years. Moreover, results from VAS were consistent with findings from the conjoint study. Our findings were also corroborated in another recent study where VAS for symptom severity compared favorably with standard quality of life measures .
Professional organizations such as the American Academy of Allergy, Asthma, and Immunology and the European Academy of Allergology and Clinical Immunology recognize that allergic disease is a major health concern often requiring specific allergen avoidance and treatment strategies that are based on positive findings from history and diagnostic testing [26, 27]. Results from this study support the positions elicited from the Joint Task Force on Practice Parameters for Rhinitis and Allergic Rhinitis and its Impact on Asthma (ARIA) in that family physicians are capable of recommending specific IgE testing, using the test to confirm allergic disease and identifying possible allergens [28–30]. Also consistent with recommendations from the Joint Task Force, results from the VAS closely approximated the findings of the conjoint study, thus revealing the usefulness of VAS in clinical practice to assess symptom severity for patients suspected of having allergic rhinitis.
Values for each item relating to patient perceptions of the test, patient demand to have testing performed, other clinical indicators, and the type of allergic rhinitis were summated to create a composite score. This composite score for testing issues yielded a moderately positive correlation with testing value, thus providing initial evidence that issues associated with testing and the process of care were linked to outcomes such as testing value. Moreover, positively framing the information describing the benefits of testing and the value of testing to patients is also known to influence their expectations of benefits .
Limitations include a low response rate and a cross-sectional study representing one geographical region. In addition, family physicians may consider attributes that were not evaluated in this study when deciding to request specific IgE blood testing for patients suspected of having allergic rhinitis. Hypothetical profiles were developed for this study and may not include all aspects of information provided by patients to family physicians, reflect what happens in actual clinical practice, and represent the opinions of physicians in other geographical areas.
Given the economic burden of allergic rhinitis on society and the research evidence that supports an inverse relationship between health status and specific IgE antibody levels [32–34], current guidelines should be repositioned and possibly modified to allow family physicians to have a more active role in specific IgE blood testing. Although ARIA suggests the SPT as a first line choice when further evaluation of patients is needed, interpretation of test results requires extensive training and experience. Thus, specific IgE testing was examined in this study as a practical choice for primary care physicians. As suggested from this study and supported in the literature, with proper training family physicians would become more adept at quantifying the results from specific IgE blood testing and recognizing when to refer patients (e.g., continued treatment failure, complications, and beyond scope of expertise) to allergists or other specialists [35–38]. Another important aspect of training is the need to consider specific IgE blood test and SPT results in the context of patient history, especially when discrepancy exists between test results and symptoms. Diagnostic testing, per se, is no substitute for a thorough examination of patient symptoms, health status, and medical history. In summary, allergists and family physicians understand that test results coupled with the findings of a careful clinical examination serve as the foundation to establish a strategy for treatment, from which future health outcomes can be evaluated to determine the success of treatment.
Family physicians rely on symptom severity, and to some extent on length of time that symptoms are present and family history to determine whether patients should be tested to determine the presence of allergic disease. Physicians with more practice experience placed greater value on specific IgE testing. Findings also revealed a moderately positive association between the issues influencing the use of specific IgE blood testing and test value. Overall, family physicians valued specific IgE blood testing, especially compared to not testing.
From the study findings, family physicians can use symptom severity as a gauge in clinical practice to determine if patients should undergo detection and testing for allergic rhinitis or related conditions perhaps much earlier during the process of clinical evaluation, especially in the presence of severe symptoms and a positive family history. Baseline evaluation will also increase the likelihood of determining the correct diagnosis and appropriate treatment, and to ascertain the need for referral. Future research is needed to address the impact of patient expectations and treatment experience on value and other outcome measures.
- Malone DC, Lawson KA, Smith DH, Arrighi HM, Battista C: A cost of illness study of allergic rhinitis in the United States. J Allergy Clin Immunol. 1997, 99: 22-27.PubMedGoogle Scholar
- Bauchau V, Durham SR: Prevalence and rate of diagnosis of allergic rhinitis in Europe. Eur Respir J. 2004, 24: 758-764. 10.1183/09031936.04.00013904.View ArticlePubMedGoogle Scholar
- Zuberbier MM: Undertreatment of rhinitis symptoms in Europe: findings from a cross-sectional questionnaire survey. Allergy. 2007, 62: 1057-1063. 10.1111/j.1398-9995.2007.01367.x.View ArticlePubMedGoogle Scholar
- Leynaert B, Neukirch C, Laird R, Bousquet J, Neukirch F: Quality of life in allergic rhinitis and asthma. A population-based study of young adults. Am J Respir Crit Care Med. 2000, 162: 1391-1396.View ArticlePubMedGoogle Scholar
- Szeinbach SL, Seoane-Vazquez EC, Beyer A, Williams PB: The impact of allergic rhinitis on work productivity. Prim Care Respir J. 2007, 16 (2): 98-105.View ArticlePubMedGoogle Scholar
- Remaley AT, Sampson ML, DeLeo JM, Remaley NA, Farsi BD, Zweig MH: Prevalence-value-accuracy plots: a new method for comparing diagnostic test based on misclassification costs. Clin Chem. 1999, 45 (7): 934-941.PubMedGoogle Scholar
- Williams PB, Ahlstedt S, Barnes JH, Soderstrom L, Portnoy J: Are our impressions of allergy test performance correct?. Ann Allergy Asthma Immunol. 2003, 91: 26-33.View ArticlePubMedGoogle Scholar
- Gendo K, Larson EB: Evidence-based diagnostic strategies for evaluating suspected allergic rhinitis. Ann Intern Med. 2004, 140: 278-289.View ArticlePubMedGoogle Scholar
- Scadding GK, Richards DH, Price MJ: Patient and physician perspectives on the impact and management of perennial and seasonal allergic rhinitis. Clin Otolaryngol. 2000, 25: 551-557. 10.1046/j.1365-2273.2000.00417.x.View ArticlePubMedGoogle Scholar
- Senti G, Vavricka BMP, Graf N, Johansen P, Wüthrich B, Kündig M: Evaluation of visual analog scales for the assessment of symptom severity in allergic rhinoconjunctivitis. Annals of Allergy, Asthma & Immunology. 2007, 98: 134-138.View ArticleGoogle Scholar
- Nau DP, Steinke DT, Williams K, Austin R, Lafata JE, Divine G, Pladevall M: Adherence analysis using visual analog scale versus claims-based estimation. Annals of Pharmacotherapy. 2007, 41: 1792-1797. 10.1345/aph.1K264.View ArticlePubMedGoogle Scholar
- Lusk JL, Norwood FB: Effect of experimental design on choice-based conjoint valuation estimates. Amer J Agr Econ. 2005, 87 (3): 771-785. 10.1111/j.1467-8276.2005.00761.x.View ArticleGoogle Scholar
- Drummond MF, O'Brien BJ, Stoddart GL, Torrance GW: Methods for the Economic Evaluation of Health Care Programmes. 1997, Oxford, UK, Oxford University Press, 2Google Scholar
- Mark TL, Swait J: Using stated preference and revealed preference modeling to evaluate prescribing decisions. Health Econ. 2004, 13: 563-573. 10.1002/hec.845.View ArticlePubMedGoogle Scholar
- Green PE, Srinivasan V: Conjoint analysis in consumer research: issues and outlook. J of Consumer Res. 1978, 5: 103-123. 10.1086/208721.View ArticleGoogle Scholar
- Louviere JJ, Hensher DA, Swait JD: Stated Choice Methods: Analysis and Applications. 2000, Cambridge, UK, Cambridge University PressView ArticleGoogle Scholar
- Hall J, Kenny P, King M, Louviere J, Viney R, Yeoh A: Using stated preference discrete choice modeling to evaluate the introduction of varicella vaccination. Health Econ. 2002, 11 (5): 457-465. 10.1002/hec.694.View ArticlePubMedGoogle Scholar
- Johansson G, Stallberg B, Tornling G, Andersson S, Karlsson GS, Falt K, Berggren F: Asthma treatment preference study. Chest. 2004, 125 (3): 916-923. 10.1378/chest.125.3.916.View ArticlePubMedGoogle Scholar
- Harpe SE, Szeinbach SL, Caswell RJ, Corey R, McAuley JW: The relative importance of health related quality of life and prescription insurance coverage in the decision to pharmacologically manage symptoms of overactive bladder. J of Urology. 2007, 178: 2532-2536. 10.1016/j.juro.2007.08.006.View ArticleGoogle Scholar
- Flynn TN, Louviere JJ, Peters TJ, Coast J: Best-worst scaling: What it can do for health care research and how to do it. J Health Econ. 2007, 26: 171-189. 10.1016/j.jhealeco.2006.04.002.View ArticlePubMedGoogle Scholar
- Acito F, Jain AK: Evaluation of conjoint analysis results: a comparison of methods. J Mark Res. 1980, 17: 106-112. 10.2307/3151124.View ArticleGoogle Scholar
- Lancsar E, Louviere J, Flynn T: Several methods to investigate relative attribute impact in stated preference experiments. Soc Sci & Med. 2007, 64: 1738-1753. 10.1016/j.socscimed.2006.12.007.View ArticleGoogle Scholar
- Hair JF, Anderson RE, Tatham RL, Black WE: Multivariate Data Analysis. 1998, Upper Saddle River, NJ: Prentice Hall, 5Google Scholar
- Nunnally JC: Psychometric Theory. 1978, New York: McGraw-HillGoogle Scholar
- Bousquet PJ, Combescure C, Neukirch F, Klossek JM, Mechin H, Daures JP, Bousquet J: Visual analog scales can assess the severity of rhinitis graded according to ARIA guidelines. Allergy. 2007, 62: 367-372. 10.1111/j.1398-9995.2006.01276.x.View ArticlePubMedGoogle Scholar
- American Academy of Allergy, Asthma, and Immunology: The Allergy Report. 2001, Milwaukee, WI: American Academy of Allergy, Asthma, and ImmunologyGoogle Scholar
- EAACI Position: European Academy of Allergology and Clinical Immunology. (accessed August 6, 2007), [http://www.eaaci.net/media/PDF/E?820.pdf]
- Plaut M, Valentine MD: Allergic Rhinitis. NEJM. 2005, 353: 1934-1944. 10.1056/NEJMcp044141.View ArticlePubMedGoogle Scholar
- Bousquet J, Van Gauwenberge P, Khaltaev N: Allergic rhinitis and its impact on asthma. J Allergy Clin Immunol. 2001, 108 (Suppl 5): S147-S334. 10.1067/mai.2001.118891.View ArticlePubMedGoogle Scholar
- Dykewicz MS, Fineman S, Skoner DP, Nicklas R, Lee RE, Blessing-Moore J, Li JT, Burnstein IL, Berger W, Spector S, Schuller DE: Diagnosis and management of rhinitis: complete guideline of the Joint Task Force on Practice Parameters in Allergy, Asthma and Immunology. Ann Allergy Asthma Immunol. 1998, 81: 478-518.View ArticlePubMedGoogle Scholar
- O'Connor AM, Pennie RA, Dales RE: Framing effects on expectations, decisions, and side effects experienced: the case of influenza immunization. J Clin Epidemiol. 1996, 49 (11): 1271-1276. 10.1016/S0895-4356(96)00177-1.View ArticlePubMedGoogle Scholar
- Nathan RA: The burden of allergic rhinitis. Allergy and Asthma Proc. 2007, 28: 3-9. 10.2500/aap.2007.28.2934.View ArticlePubMedGoogle Scholar
- Custovic A, Murray C, Simpson A: Allergy and infection: understanding their relationship. Allergy. 2005, 60 (Suppl 79): 10-13. 10.1111/j.1398-9995.2005.00851.x.View ArticlePubMedGoogle Scholar
- Simpson A, Soderstrom L, Ahlstedt S, Murray CS, Woodcock A, Custovic A: IgE antibody quantification and the probability of wheeze in preschool children. J Allergy Clin Immunol. 2005, 116: 744-9. 10.1016/j.jaci.2005.06.032.View ArticlePubMedGoogle Scholar
- Spector SL, Nicklas RA, Chapman JA, Bernstein IL, Berger WE, Blessing-Moore J, Ddykewicz MS, Fineman SM, Lee RE, Li JT, Portnoy JM, Schuller DE, Lang D, Tilles SA: Symptom severity assessment of allergic rhinitis: part I. Ann Allergy Asthma Immunol. 2003, 91: 105-114.View ArticlePubMedGoogle Scholar
- Abraham CM, Ownby DR, Peterson EL, Wegienka G, Zoratti EM, Williams LK, Joseph CLM, Johnson CC: The relationship between seroatopy and symptoms of either allergic rhinitis or asthma. J Allergy Clin Immunol. 2007, 119: 1099-104. 10.1016/j.jaci.2007.01.024.View ArticlePubMedGoogle Scholar
- Dowdee A, Ossege J: Assessment of childhood allergy for the primary care practitioner. J Amer Acad of Nurse Pract. 2007, 19: 53-62. 10.1111/j.1745-7599.2006.00195.x.View ArticleGoogle Scholar
- Forrest CB, Nutting PA: Family physicians' referral decisions. J Fam Pract. 2002, 51 (3): 215-222.PubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2296/9/47/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.