Skin Prick Tests and Blood Tests for Allergy in Young Children

‘Disagreement between skin prick tests and specific IgE in young children’ by A.M Schoos, B.l.k. Chawes, N. Folsgaard, N. Samandari, K. Bonnelykke, and H. Bisgaard.

 This article appears as a pre-publication. It has been accepted by the journal ‘Allergy’. Of note, even as a pre-publication, it has been a very popular article. I presented this article at our section’s journal club today (November 4, 2014). What follows is somewhat lengthy synopsis and review of this article.

Purpose– to evaluate the agreement between skin prick tests (SPTs) and specific IgE (sIgE) results when diagnosing allergic sensitization. (Reviewers note- specific IgE is the blood test that is used for allergy- once upon a time it was referred to as RAST.)

Methods– The children were part of the Copenhagen Prospective Study on Asthma in Childhood (COPSAC). They were born to mothers who had asthma. The children were evalutated every six months for seven years. The evaluation for allergic sensitization was done by sIgE and SPT at 6 months, 18 months, 4 years, and 6 years. SPTs were done to milk, egg, wheat, soy, codfish, and peanut. The blood tests or sIgEs were performed for the same items that were used for the SPTs. At age 7 years the clinical history of food allergy and allergic rhinitis was obtained by a standardized interview (as opposed to a parental questionnaire).

The data was presented with prevalence curves, logistic regression, Kappa coefficients, and receiver operating characteristic curves. A Kappa value of 1.0 indicates perfect agreement between tests. The gradations below perfect were almost, substantial agreement, moderate, fair, slight, and poor strength of agreement between the two tests.


                Table 1-baseline characteristics of the population- there were 411 children enrolled of which 389 were tested and 22 were not. The table looked at gender, socioeconomic factors, early life exposure to cigarette smoke, and the allergic disposition of mother (who had to have asthma to qualify) and father. The only difference between those tested vs. not tested was income. Those children not tested were from families with lower incomes compared to those who were tested.

                Figure 1- Inhalant Allergens: Changes in prevalence over time.


  6 months 18 months 4 years 6 years
SPT 1.5% 3.8% 8.4% 15.4%
sIgE 0.6% 4.2% 18.1% 24.8%


This shows the age at which sensitization occurs and the impressive divergence between SPT and sIgE- the blood test results are almost twice the skin test results. The pollens (grass, birch, and mugwort) are seen first at age 4 years and mold sensitization was seen at age 6 years. The perennial allergens are seen starting at age 18 months. Note the few children who had allergic sensitization to inhalant allergens at ages 6 and 18 months.

                Table 2- Agreement of SPT and sIgE, these are the Kappa values. (Do the tests agree with each other?)

                6 months- -0.0011 Poor

                18 months- 0.48 Moderate

                4 years- 0.45 Moderate

                6 years- 0.55 Moderate

Note the scale- poor, slight, fair, moderate, substantial, and almost perfect. A 1.0 is perfect and the moderate range is 0.41 to 0.6.

                Figure 3- Food Allergens: Changes in prevalence over time

A note- cod was not included due to few positive results.


  6 months 18 months 4 years 6 years
SPT 5.3% 5.1% 3.7% 3.0%
sIgE 7.8% 12.1% 15.0% 18.9%


The number of SPT positive children declined from 6 months to 6 years for foods whereas the number of children positive by the blood test increased over the same interval.

                Table 2- Agreement of SPT and sIgE, these are the Kappa values.

                6 months- 0.46 Moderate

                18 months- 0.31 Fair

                4 years- 0.16 Slight

                6 years- 0.14 Slight

To repeat -Note the scale- poor, slight, fair, moderate, substantial, and almost perfect. A 1.0 is perfect and the moderate range is 0.41 to 0.6.

The next two tables do not appear in the paper. The reader is directed to an online repository for 4 more tables and 3 figures.

                Table E2- Clinical symptoms of inhalant allergy at 6 years of age and relation to sensitization.

Listed are the inhalants tested for; birch, grass, mugwort, horse, dog, cat, house dust mite, and molds. There were 266 of the children who had both SPT and sIgE performed at age 6 years.

Clinical Allergy (CA) 93
Total number with positive skin prick tests 78
Total number with positive blood tests 150
CA with no positive test 46
CA with both tests positive 26
CA with only a skin test positive 7
CA with only a blood test positive 14

Inhalant sensitization diagnosed by skin prick test compared to blood tests was better related to symptoms of nasal allergy: the positive predictive value (chance that the condition is present with a positive test) was 42.3% for the skin tests and 26.7% for the blood test. Note about half did not have any test positive.


Table E3- Clinical symptoms of food allergy at 6 years of age and relation to sensitization.

Listed are the foods tested for; wheat, egg, milk, soybean, cod, and peanut. There were 264 of the children who had both SPT and sIgE performed at age 6 years.


Clinical Allergy (CA) 12
Total number with positive skin prick tests 8
Total number with positive blood tests 75
CA with no positive test 6
CA with both tests positive 5
CA with only a skin test positive 0
CA with only a blood test positive 1

Skin prick test results for foods were better related to clinical symptoms than blood tests; the positive predictive value was 62.5% vs. 8.0% for the blood test, however half of the cases of clinical food allergy were not accompanied by a positive test.

Discussion– There is poor to moderate agreement between SPTs and sIgEs in diagnosing allergic sensitization in young children. This mismatching increases with age for foods. The choice of a test may have a major impact on results in research and guidance in clinical practice. Allergy testing should only be done in children with meaningful symptoms and not used as a screening tool.

Strengths of the Study- first evaluation of SPTs and sIgEs in a large birth cohort with the tests done simultaneously and frequently (4 times) over 6 years.

Weakness- no food challenges were performed. This was done in an at-risk population (mothers with asthma) not a general population.

The agreement of these two tests during the pre-school years was at best moderate, but a striking worsening of agreement was seen with foods as the child increased in age. The point prevalence for food sensitization diagnosed by SPT was in line with what we know about food allergy in children which will decrease with age. However the blood test for food antibodies increased with age. This suggests that this sIgE test does not reflect clinical food allergy.

Why does this happen? Extracts for skin testing may not contain all the components of foods- there may be more positive blood tests due to differences in what each contains. Cross-reactivity between foods and pollens may account for some of the difference, but that would not explain what was seen with milk where there were 35 sIgE positive children who did not have a positive SPT. The SPT and sIgE may measure different things- one is antibodies attached to mast cells in the skin while the other is found in the blood.

Warning- this may have an impact on research and clinical practice-

  1. Research- clarify the definition of sensitized vs. allergic children. The choice of test may select out different populations of sensitized children.
  2. Clinician- emphasize careful interpretation, base the diagnosis on the assessment method, the clinical history, the age of the child, and the type of allergy. Allergy testing should never be used indiscriminately for screening purposes.

Both tests had a low predictive value for clinical food allergy and symptoms of allergic rhinitis, with only 50% of the symptomatic children having a positive test.

There is substantial disagreement between SPT and sIgE for diagnosing allergic sensitization to common inhalant and food allergens in young children with increasing disagreement with age for foods. SPT and sIgE results cannot be used interchangeably.

The lack of agreement between tests and a poor correlation to clinical disease emphasizes the point that allergy testing should not be used as a screening tool in children.

Reviewers Comments

This is a very interesting, well done, and informative paper on a topic that has been the bane of pediatric allergy. Tests are done ‘just to see what the child is allergic to’ without a discriminating history of clinical conditions present in the child. Allergy tests are terrible for screening and this should not be done. The value of the test is dependent upon the story that supports what happens to the child with exposure. In allergy, the clinician uses the test to verify a clinical impression.

Note that the paper is very careful in stating that this is sensitization- meaning the making of an allergy antibody. Clearly, antibodies can be made and there may be absolutely no clinical reactivity. The allergic child has a history of an allergic reaction with an exposure and a positive test to that substance. Allergy is a story and a test whereas sensitization is only a test (which may never have a story to match it).

FEL 11-5-2014

November 6, 2014 · fleickly · No Comments
Tags: ,  · Posted in: Allergies, Allergy in Children, Allergy Testing, Interesting articles

Leave a Reply

You must be logged in to post a comment.