Are allergen batch differences and the use of double skin prick test important?

This study is part of the RAV study; a Western Denmark five centre study on risk factors for asthma in adults [ 12]. From the catchment area of each of the five centres 2,000 subjects aged 20–44 standardized by gender and age were randomly selected from the Danish Civil Registration System. In 2001–2003 an extended version of the ECRHS II screening questionnaire ( www.​ECRHS.​org) was sent to these 10,000 persons. 7271 responded. Among responders a random sample of 20 percent and a symptom group comprising all subjects who reported currently taking any medicine for asthma, asthma attack, or woken by an attack of shortness of breath at any time in the past 12 months were invited to a clinical investigation in 2003–2006.

The clinical investigation consisted of an interviewer-administered questionnaire based on the ECRHS II protocol [ 14], SPT, blood samples, spirometry and measurement of non-specific bronchial responsiveness.

Asthma, Rhinitis and smoking status was obtained from the interview. Current asthma was defined as an affirmative answer to the question “Are you currently taking any medication (including inhalators, aerosols or tablets) for asthma?” or the combination of an affirmative answer to the question “ Has your asthma ever been confirmed by a doctor?” and at least one of the following questions “Have you had wheezy breathing at any time during the last 12 months?”, “Have you woken up with a feeling of chest tightness at any time during the last 12 months?”, “Have you had breathlessness at rest in the daytime during the last 12 months” or “Have you woken up with an attack of breathlessness at any time during the last 12 months?”. “Current rhinitis” was defined as an affirmative answer to the question “Have you had problems of sneezing, running or stuffy nose without having a cold or flu within the last 12 months”. The subject was then asked in which months he/she had these problems. Seasons of the allergens were defined based on data from Danish Meteorological Institute 2003 [ 15]. Birch: April-May, Grass: June-July, Mugwort: July-August, Alternaria: July-September and Cladosporium: June-September. Seasonal rhinitis was defined when symptoms in all the months of the given allergen season was reported. Subjects with perennial symptoms were not included in the seasonal rhinitis group.

Dust mites, cat, dog and horse allergen were classified as “Perennial allergens”. Perennial rhinitis was diagnosed if the participant mentioned symptoms in at least 8 months of the year or could not specify “symptom months”.

The participants were asked to refrain from any antihistaminic medication 72 hours before the visit. SPT was carried out with commercial allergens from Alk-Abelló: Pollens: Birch, Grass ( Phleum pratense) and Mugwort ( Artemissa Vulgaris). Danders: Cat, Dog and Horse. Moulds Alternaria ( Alternaria Alternata) and Cladosporium ( Cladosporium Herbarum). House dust mites Dermatophagoides Pteronyssinus ( D pter) and Dermatophagoides Farinae ( D. Farinae). Positive control: Histamine 10 mg/ml. Negative control: 50% Glycerol in aqua solution. After disinfection with alcohol solution ALK number tape was placed on the volar side of the forearm and droplets of allergen extract were applied at one side of the tape in a predefined order. For each allergen a Phazet lancet was pricked perpendicular through the droplet. Double SPT: With the same lancet a second prick was made at the opposite side of the tape, using the allergen extract on the lancet from the first prick. This procedure was chosen above placing another drop of allergen to reduce the necessary time as well as allergen spending. After 15 minutes positive, negative and all allergen reactions were drawn up on the skin and transferred with transparent adhesive tape onto a record sheet. The SPTs were carried out by either physicians or especially trained nurses. The largest diameter and the perpendicular diameter were measured by two readers. If different, further measurements were made until agreement between the readers was obtained. (Largest diameter + perpendicular diameter)/2 > = 3 mm was considered as positive reaction and the actual weal size was recorded. If < 3 mm the result was recorded as negative as is usual in clinical settings.

In one centre a subsample of participants were at the same time enrolled in the GA2LEN Selenium project [ 13], namely 55 subjects from the random sample of the RAV study and 36 subjects from the RAV symptomatic sample who also fulfilled the GA2LEN asthma control and case criteria. Cases were subjects with a self-reported diagnosis of asthma and either wheezing, shortness of breath or waking at night with breathlessness in the previous 12 months. Controls were subjects with neither a diagnosis of asthma nor any of the three symptoms. In these subjects SPT’s were performed simultaneously with 10 allergens on the right forearm (The RAV Study) and 7 allergens on left forearm (The GA2LEN Study). All Allergens were delivered from ALK Abello. 5 allergens were in common (Birch, Grass, Cat, D Farinae and Alternaria), but from different production batches. The declared allergen activity (Birch, Grass, Cat, D Farinae) or concentration ( Alternaria) did not vary between batches. The GA2LEN allergens were only pricked once. In this subgroup, the actual weal size of all skin pricks were recorded, even when less than 3 mm. In the analysis of the correlation between SPT with different batches, the first of the two RAV study pricks was used.

The term disagreement between two SPT’s is used, when one is positive and the other negative.

Agreements in positive tests for each allergen were calculated using Kappa Statistics. To compare the ability of the tests to identify subjects with rhinitis the sensitivity and the specificity were calculated. When choosing the cut-off value for a test there will always be a trade-off between sensitivity and specificity. To encompass both, the Youden’s indexes [ 16] were calculated. The Pearson correlation coefficients were calculated to describe the correlation between the weal sizes of the two skin pricks in the “two batches” procedure. The difference between intrabatch variability and interbatch variability were tested as follow: For each participant the difference between the sizes of the two weals with same batch (Intrabatch) as well as the difference between Prick1 batch1 and prick batch 2 (Interbatch) were calculated. The equality of the interbatch and intrabatch differences were tested by Student’s t-test. The differences in number of disagreements were tested by Wilcoxon signed-rank test. As five comparisons are performed, p < = 0.01 was chosen for reporting significant results. All statistical analyses were conducted in Stata 10.0 IC (Stata Cooporation Texas, USA).

The study was approved by the Medical Ethics Committee of Ribe & Sonderjyllands County, Fyn & Vejle County, and Nordjyllands County. All subjects signed an informed consent form before participating in the study.

A high degree of concordance was seen between the two weals when SPT was carried out with the same allergen extract. The correlation coefficients varied from 0.88 in Alternaria and dog to 0.94 for horse and birch. The reactions to Dog allergen showed the highest rate of disagreements: 4%. A higher percentage disagreements was seen in the asthma group, whereas the Kappa values were the same, Table  3. Disagreements were seen in both directions except for Cladosporium. In 87% of the 1,135 participants, agreement was seen for all 10 allergens.

For Grass and Cat allergens the SPT mean weal size was significantly greater for the GA2LEN panel than the RAV panel: Grass 3.1 vs. 2.1 mm and cat 2.1 vs. 1.1 mm. For Alternaria, birch and D Farinae no significant differences in mean weal size was seen. The Pearson correlation coefficients between the average diameters in the two tests varied from 0.79 for grass to 0.94 for D Farinae. The percent disagreement in the 90 subjects varied from 2.2 for birch to 11.1 for cat, Table  4.

For five allergens it was possible to compare the intra- and inter batch variation in 69 subjects. In 54 participants (78%) agreement between batches was seen for all five allergens. The rate of disagreement was higher and the Kappa value lower for two different batches compared to double SPT with the same batch. This was true for all 5 allergens. For Alternaria, Grass and Cat the between batch differences in weal sizes were significantly larger than the within batch differences. When comparing number of disagreements significance was seen for Cat only Table  5.

Sensitivity, specificity and Youdens index for any positive SPT were calculated using rhinitis as the “true diagnosis” (Sensitivity = 0.57, specificity = 0,76 and Youdens index = 0.33). The same calculations were made for positive reaction to any perennial allergen with perennial rhinitis as the “true diagnosis” (0.37, 0.74 and 0.11) and finally for seasonal rhinitis with symptoms in the birch, grass and mugworth pollen seasons as the “true diagnosis”. For the single allergens the sensitivity were lower and the specificity higher (Table  6) Subgroup analysis on the random group and the asthma symptom group showed similar results (data not shown).

To analyse whether double SPT improved the diagnostic capability of the test, Prick 1 pos. was compared to (Prick 1 pos. OR Prick 2 pos.) and to (Prick 1 pos. AND Prick 2 pos.). As expected using (1 OR 2) showed a slightly higher sensitivity and lower specificity, while using (1 AND 2) lead to the reverse result. The Youden’s index’s hardly changed, Table  6. In the current asthma group the same patterns were seen, Table  7.

These calculations were not performed for moulds due to few cases.