Applying GRADE for diagnosis 57 2 Haxel, 2016 [2] First author Haxel Year of publication 2016 Journal American Journal of Rhinology & Allergy Setting Tertiary care, Germany Study design Retrospective analysis Study population 161 patients, 60% female, mean age 35.95 yrs (sd 16.66). Clinically assumed house dust mite allergy Index test Specific IgE for D.pteronyssinus and D.farinae: ImmunoCAP, in venous blood; concentrations converted into 7 classes (0, 0.1–0.35 kU/L; I, 0.35– 0.70 kU/L; II, 0.70–3.50 kU/L; III, 3.50–17.5 kU/L; IV, 17.5–50.0 kU/L; V, 50.0–100 kU/L; and VI, >100 kU/L Note: for this systematic review dichotomised in positive or negative; threshold 0.35 kU/L. Reference test Nasal provocation: symptom scores (relevant acute nasal, ocular, cutaneous, bronchial and systemic symptoms), and nasal patency impairment (rhinomanometry) after allergen provocation. Allergen provocation: D.pteronyssinus and D.farinae 10.000 units/mL. Positive result: reduction of nasal airflow >40% at 150 Pa, symptom score > 3, combination of >20% reduction in airflow and symptom score >2. Performance of the index test D.pteronyssinus True positives 54/114=47% False positives 23/114=20% False negatives 10/114=9% True negatives 27/114=24% Sensitivity 84.38% [54/64] (95%CI: 73.14-92.24) Specificity 54.00% [27/50] (95%CI: 39.32-68.19) Pre-test probability 64/114=56% Performance of the index test D.farinae True positives 36/97=37% False positives 26/97=27% False negatives 7/97=7% True negatives 28/97=29% Sensitivity D.farinae: 83.72% [36/43] (95%CI: 69.30-93.19) Specificity D.farinae: 83.72% [28/54] (95%CI: 37.84-65.66) Pre-test probability 43/97=44% Risk of bias (QUADAS-2) Domain 1: Patient selection A. Risk of bias Describe methods of patient selection Retrospective analysis of patients presenting with clinically assumed house dust mite allergy Was a consecutive or random sample of patients enrolled? Yes/no/unclear Probably Was a case-control design avoided? Yes/no/unclear Did the study avoid inappropriate exclusions? Yes/no/unclear Retrospective analysis, no information about exclusions Could the selection of patients have introduced bias? RISK: LOW/HIGH/UNCLEAR B. Concerns regarding applicability
RkJQdWJsaXNoZXIy MTk4NDMw