Table 2.

Sensitivity and specificity of objective tests in asthma diagnosis

Strategy	Description*	Parameter*	Range of predictive values* (note that a single value indicates datum from a single study)				Comments**
			Sensi	Specii	PPViii	NPViv
Clinical assessment
Symptoms and signs	The commonest symptoms assessed were cough and wheeze, and, in adults, shortness of breath	Cough in adults Wheeze in adults Dyspnea in adults Cough in schoolchildren21 Wheeze in children21 Cough in preschool children Wheeze in preschool children Shortness of breath in preschool children	16%–66% 9%–76% 11%–73% 63% 59% 88% 54% 76%	26%–64% 34%–87% 38%–71% 75% 93% 7% 57% 52%	8%–44% 10%–81% 41%–59% 14% 34% 76% 80% 84%	18%–92% 28%–94% 26%–70% 97% 97% 15% 27% 40%	As isolated symptoms cough, wheeze and shortness of breath are neither sensitive, nor specific for asthma. Most children with asthma have intermittent cough, wheeze and exercise-induced symptoms, but only about a quarter of children with these symptoms have asthma. Note that the single study in preschool children compared current symptoms with a diagnosis of asthma two years later.
	Symptom variability	Episodic symptoms in adults Diurnal symptoms in adults Symptoms after exercise in adults Episodic symptoms in children22,23 Symptoms after exercise in children22,23 Nocturnal symptoms in children22.23	9%–40% 30%–56% 5%–40% 36%–93% 82%–94% 57%–84%	36%–91% 36%–83% 32%–93% 35%–93% 59%–73% 58%–78%	14%–86% 48%–76% 5%–81% 40%–94% 54%–86% 64%–85%	18%–93% 18%–67% 58%–84% 62%–90% 79%–91% 57%–82%	Asking about episodic symptoms improves the positive predictive values in children compared to current symptoms.
	Combinations of symptoms (typically cough, wheeze, chest tightness, dyspnea, exercise symptoms)	Symptom scores in adults Symptom scores in chiidren21–23 Symptoms of cough and wheeze in preschool children	60% 45%–83% 49%	66% 85%–97% 59%	44%–94% 80%	66%–97% 51%	Combinations of symptoms are clinically more helpful than isolated symptoms, especially in children. For example, two thirds of children with a cluster of cough, wheeze, chest tightness, dyspnea and exercise symptoms have asthma. Asthma is unlikely if a child does not have at least some of these symptoms.
History of atopy	Personal/family history of atopic/allergic diseases	Personal history of atopy in adults Personal history of rhinitis/eczema in preschool children Family history of atopy in adults Family history of atopy in children	54%–55% 47%–62% 26%–60% 43%–44%	68%–74% 20%–75% 56%–83% 57%–70%	46%–76% 72%–86% 44%–74% 51%–77%	45%–79% 14%–30% 38%–70% 24%–62%	History (personal or family) of atopic disease has poor sensitivity and specificity for asthma.
Strategies for demonstrating airway obstruction
Spirometry	Regard an FEV1:FVC ratio of less than 70% as a positive test for obstructive airway disease	Obstructive spirometry in adults Obstructive spirometry in children (5–18 years)	23%–47% 52%	31%–100% 73%	45%–100% 75%	18%–73% 49%	In the four larger studies (adults and children), the NPV was between 18% and 54%, which means that more than half of patients being investigated who have normal spirometry will have asthma (ie, false negatives).
Strategies for demonstrating variability in airway obstruction
Bronchodilator reversibility	In adults, regard an improvement in FEV1 of ≥12% and ≥200 mL as a positive test. In children regard an improvement in FEV1 of ≥l2% as a positive test.	Bronchodilator reversibility in adults Bronchodilator reversibility in schoolchildren (using a threshold of 9% change in FEV1)71	17%–69% 50%	55%–81% 86%	53%–82%	22%–68%	In these secondary care populations, about one in three people with a positive reversibility test will not have asthma (the cohorts all included people with COPD); and at least one in three people with a negative bronchodilator reversibility test will have asthma.
Challenge tests	Regard a PC20 value of 8 mg/mL or less as a positive test	Methacholine challenge in adults Methacholine challenge in children31,43,72	51%–100% 47%–86%	39%–100% 36%–97%	60%–100% 20%	46%–100% 94%	Challenge tests are a good indicator for those with a definitive diagnosis of asthma already (based upon clinical judgment, signs and symptoms and response to antiasthma therapy).
	Fall in FEV1 ≥15% at cumulative dose of ≤635 mg is positive Exercise challenge	Mannitol in adults Mannitol in children Exercise challenge in adults Exercise challenge in children	56% 63% 26%–80% 69%–72%	75% 81% 100% 69%–72%	80% 100% 90%–99%	49% 0% 5%–73%	These data are from a single study in adults and children with symptoms of asthma on questionnaire. The studies in adults had very small sample sizes. The larger study in children had a false positive rate of 1% (PPV 99%).
Peak flow charting	Monitor peak flows for 2–4 weeks, calculate mean variability. Regard ≥20% variability as a positive test	PEF charting in adults in a population study – using mean variability of >20% – using mean variability of >15% – using diurnal variation >15% on >3 days/week PEF charting in children – using variation >12.3% (95th centile)	46% 3%–5% 20% 50%	80% 98%–99% 97% 72%	97% 60%–67% 82% 48%	10% 60% 64% 74%	It is not clear whether the patients in these studies were symptomatic at the time of the charting, and results may not reflect clinical use in symptomatic populations. One study concluded that the number of days with diurnal variation was more accurate than calculating the mean variation.
Strategies for detecting eosinophilic inflammation or atopy
FeNO	Adults: Regard an FeNO level of 40 ppb or more as a positive test Children 5–16 years: regard an FeNO level of 35 ppb or more as a positive test.	FeNO in adults FeNO in schoolchildren	43%–88% 57%	60%–92% 87%	54%–95% 90%	65%–93% 49%	These studies are all in secondary care populations. Approximately one in five adults with a positive FeNO test will not have asthma (ie, false positives) and one in five adults with a negative FeNO test will have asthma (ie, false negatives).
Blood eosinophils	Suggested thresholds for blood eosinophils: Adults >4.15% Children ≥4%65	Blood eosinophils in adults Blood eosinophils in children	15%–36% 55%–62%	39%–100% 67%–84%	39%–100% 56%–69%	27%–65% 73%	Elevated blood eosinophil level is poorly predictive. The threshold varies in these studies from 4.0% to 6.3%.
IgE		Any allergen-specific IgE >0.35 kU/l in adults Total IgE in adults >100 kU/I	54%–93% 57%	67%–73% 78%	5%–14% 5%	95%–99% 99%	A normal IgE substantially reduces the probability of asthma in adults with a false negative rate of less than one in 10, although a positive result is poorly predictive.
Skin-prick testing		Any positive test (wheal ≥3 mm) in adults Any positive test (wheal ≥3 mm) in children	61%–62% 44%–79%	63%–69% 56%–92%	14%–81% 65%–92%	39%–96% 36%–79%

Notes:

^*Data derived from NICE evidence tables unless otherwise specified.64 Only studies reporting sensitivity, specificity, PPV, and NPV included here;

^**comments have been added by the guideline development group as an aid to interpretation of the data presented;

ⁱprobability of test being positive when asthma present;

ⁱⁱprobability of test being negative when asthma absent;

ⁱⁱⁱproportion of patients with positive test who actually have asthma (100 minus PPV is the proportion of patients with a false positive test);

^ivproportion of patients with negative test who do not have asthma (100 minus NPV is the proportion of patients with asthma, but in whom test was negative). In most studies, the reference test was spirometry plus either bronchodilator reversibility or a challenge test, although some studies also included a “typical history of attacks” or diurnal variation, or used physician diagnosis. Studies evaluating methacholine-challenge tests used physician diagnosis or bronchodilator reversibility and/or diurnal peak flow variability. In children, reference tests used were physician-diagnosed asthma plus spirometry or documented history of wheeze on at least two occasions and variability in FEV₁ over time or on exercise testing. This table is reproduced from BTS/SIGN British Guideline on the management of asthma by kind permission of the British Thoracic Society.22

Abbreviations: COPD, chronic obstructive pulmonary disease; FEV₁, forced expiratory volume in 1 second; FeNO, fractional exhaled nitric oxide; FVC, forced vital capacity; NICE, National Institute for Health and Care Excellence; NPV, negative predictive value; PC₂₀, provocative concentration causing a 20% fall in FEV₁; PEF, peak expiratory flow; PPV, positive predictive value; Sens, sensitivity; Spec, specificity.