METHODS: EMBASE (1947-) and Medline (1946-) were searched until December 8, 2015. A search strategy was used to identify studies on AR or NAR patients subjected to diagnostic local nasal provocation. All studies providing original NAPT data among the AR or NAR population were included. Meta-analysis of proportion data was presented as a weighted probability % (95% confidence interval [CI]).
RESULTS: The search yielded 4504 studies and 46 were included. The probability of nasal allergen reactivity for the AR population was 86.3% (95% CI, 84.4 to 88.1) and in NAR was 24.7% (95% CI, 22.3 to 27.2). Reactivity was high with pollen for both AR 97.1% (95% CI, 94.2 to 99.2) and NAR 47.5% (95% CI, 34.8 to 60.4), and lowest with dust for both AR 79.1% (95% CI, 76.4 to 81.6) and NAR 12.2% (95% CI, 9.9 to 14.7). NAPT yielded high positivity when defined by subjective end-points: AR 91.0% (95% CI, 86.6 to 94.8) and NAR 30.2% (95% CI, 22.9 to 37.9); and lower with objective end-points: AR 80.8% (95% CI, 76.8 to 84.5) and NAR 14.1% (95% CI, 11.2 to 17.2).
CONCLUSION: Local allergen reactivity is demonstrated in 26.5% of patients previously considered non-allergic. Similarly, AR, when defined by skin-prick test (SPT) or serum specific IgE (sIgE), may lead to 13.7% of patients with inaccurate allergen sensitization or non-allergic etiologies.