None of the subjects who did not have a doctor’s diagnosis of asthma before the study responded to the bSICs or to any other SICs. As mentioned, only 27 subjects underwent the mannitol challenge. Of them, 10 responded to bSICs. There were no late responders (defined as a > 20% fall in PEF 3 to 8 h after the challenge). The bovine solution for skin testing induced slightly larger skin wheals than did the bovine solution for inhalation challenges (3.4 mm [95% CI, 2.6 to 4.2 mm] and 2.9 mm [95% CI, 2.1 to 3.7 mm], respectively; p = 0.005), but the correlation between the skin responses to the two solutions was close (r = 0.92; p < 0.0001).
The main results are expressed in Tables 1 and 2. The NPV for an SPT response to a bovine solution for skin testing of < 3 mm was 100%, and the PPV of a bIgE of > 5 IU/L was 100% (Fig 1). In the simple regression analysis, SPT response to bovine solution for inhalation challenges was shown to have the strongest association with RDR-bSIC (r = 0.65; p < 0.0001; Fig 2). The association between bIgE and RDR-bSIC (r = 0.59; p = 0.0001; Fig 3) was also clearly stronger than that between the RDR of the histamine challenge and RDR-bSIC (r = 0.30; p = 0.07; Fig 4). The RDR of the mannitol challenge, NO level, or prechallenge FEV1 percent of predicted did not show a statistically significant association with RDR-bSIC.
In the multiple regression analysis, the dependent variable was again RDR-bSIC. As mentioned, of the single variables, SPT response to bovine solution for inhalation challenges showed the strongest association with it (r = 0.65; r2 = 0.42). Adding bIgE to the model raised the r2 to 0.46. However, it did not increase further after adding the RDR of the histamine challenge to the model. The following equation could be created:
log RDR-bSIC = 0.126 X SPT + 0.223
X log bIgE + 0.427 (r2 = 0.46; p < 0.0001)
Discussion
This study confirmed that clinical history has a poor predictive value for establishing an association between workplace exposures and asthma. Only 13 of the 37 farmers (35%) with a clinical suspicion of occupational asthma actually responded to the SIC with occupational allergens. This figure is less than that reported by Malo. In that study, of the 162 subjects who were referred due to a suspicion of occupational asthma, 75 (46%) were objectively shown to have occupational asthma. 
Table 2—Sensitivity, Specificity, PPV, and NPV for Each Test, Based on the Response to bSIC*
| Test | Responders to bSIC, No. | Nonresponders to bSIC, No. | Sensitivity,% | Specificity,% | PPV,% | NPV,% |
| bIgE > 5 IU/L | 9/11 | 0/26 | 82 | 100 | 100 | 93 |
| Bovine skin reaction > 3 mm | 11/11 | 13/26 | 100 | 50 | 46 | 100 |
| Histamine PC20 ^ 8 mg/mL | 9/11 | 9/26 | 82 | 65 | 50 | 89 |
| FEVj fall > 15% in mannitol challenge | 2/10 | 1/17 | 20 | 94 | 67 | 67 |
| Exhaled NO > 46 ppb | 3/11 | 6/26 | 27 | 77 | 33 | 71 |
*See Table 1 for abbreviations not used in the text.
These subjects represented a wide variety of occupations. Our finding reflects the special difficulties the clinicians encounter when assessing the probability of occupational asthma among farmers and highlights a need for screening tests. We were able to demonstrate that simple tests, namely, SPTs and bIgE tests, can be used effectively for selecting patients for bSICs. These findings are clinically important, not only from the viewpoint of avoiding a challenge with potential for prolonged effects but also from an economic viewpoint, because challenge requires inpatient care and time lost from work.
Subscribe to instagram and twitter Canadian Health Care Mall
The major bovine allergen is Bos d 2, a protein of the lipocalin family. Subjects with occupational, cattle-associated asthma respond to an inhalation challenge with purified Bos d 2. Bos d 2 is produced in sweat glands and is transported to the skin surface as a carrier of the pheromone ligand. Therefore, cow dander preparations probably include Bos d 2. Indeed, the concentration of Bos d 2 recently has been measured from the commercial bovine inhalation challenge solution (ALK) that was used in our study, and it was 5.7 ^g/mL. In the present study, bovine skin test solutions, bovine inhalation challenge solutions, as well as bovine antigens in the IgE analysis were all of bovine dander origin. Therefore, we think that these tests were valid in the diagnosis of bovine asthma.
Simply by measuring bIgE, 35 of 37 patients (95%) could have been classified correctly without being subjected to bSIC. The PPV of bIgE > 5 IU/L was 100%. Accordingly, in the study of Taivainen all 19 patients with a bIgE level of > 3.5 IU/L responded to bSICs. Cow-allergic farmers with asthma have clearly higher bIgE levels than do cow-allergic farmers with rhinitis, which probably is explained by the larger size of the shock organ in asthma (ie, lungs vs the nose). Thus, a bIgE concentration of > 5 IU/L seems to be a reliable indicator for an occupational asthma induced by bovine allergens, bearing in mind that not all responders to bSICs have bIgE levels above that level.
Although the specificity of a positive SPT result was poor, the sensitivity was high, and the NPV was 100% (ie, none of the subjects with an SPT wheal of < 3 mm responded to bSICs). In the study of Taivainen the sensitivity of SPT was 94%. Thus, a negative result in SPT virtually excludes bovine asthma. This finding contrasts that of Aas, who reported that even 32% of symptomatic but skin test-negative subjects responded to SICs with various animal danders, including the cow. To our knowledge, there have been no international publications on this issue after that study, which is probably the reason why it has been a common practice in Finland to refer even skin test-negative subjects for bSICs.