Gehring U, Triche E, van Strien RT, Belanger K, Holford T, Gold DR, Jankun T, Ren P, McSharry J-e, Beckett WS, Platts-Mills TAE, Chapman MD, Bracken MB, Leaderer BP.
2004.
Prediction of Residential Pet and Cockroach Allergen Levels Using Questionnaire Information.
Environmental Health Perspectives, 112 (8): pp. 834-839.
As part of a larger study, this paper describes analyses conducted to assess the accuracy of questionnaire items on cat and dog ownership (presence at home) and presence of cockroaches in predicting measured allergen concentrations in surface dust collected in homes from the main living area and the infant child's bedding. The measured allergen levels in collected dust were described in three ways: below method detection limit (MDL), and above lower and higher cut points based on the distribution of study data. To report on accuracy, the researchers described their data in terms of sensitivity (e.g., if pets owned and present, then measured allergen levels in dust should be higher) and specificity (e.g., if pets not reported as owned, then measured levels of allergens in dust should be low and likely below MDL). As expected, for the upper cut point (for example, at or above reported sensitization thresholds for cat allergen), sensitivity and specificity were high. However, for the MDL and the lower cut point, specificity was high but sensitivity was low. Therefore, the authors stated for epidemiological studies, researchers should include measurements of allergens in settled dust on various surfaces in homes, schools, etc. Also, the data suggested levels of cat and dog allergen in dust were not high when the pets were reported not to be allowed into the infant's bedroom; low measured levels were likely due to the allergens being brought into the infant's bedroom on clothing, etc. |
Printer Friendly 
