Abstract
OBJECTIVES The objectives are: (1) To assess diagnostic test characteristics of six alternative index tests compared with the selected reference standard-a standardized exercise challenge test (ECT) in patients with suspected exercise-induced bronchoconstriction or asthma (EIB/EIA); (2) to determine the efficacy of a single prophylactic dose of four pharmacologic and one nonpharmacologic interventions vs. placebo to attenuate EIB/EIA in patients with diagnosed EIB/EIA; and (3) to determine if regular daily treatment with short-acting or long-acting beta-agonists (SABA or LABA) causes patients with EIA to develop tachyphylaxis when additional prophylactic doses are used pre-exercise. DATA SOURCES A systematic and comprehensive literature search was conducted in 14 electronic databases (Diagnosis) and the Cochrane Airways Register (Therapy). REVIEW METHODS Study selection, quality assessment, and data extraction were conducted independently by two reviewers. The primary outcome was the maximum percent fall in the post-exercise forced expiratory volume in 1 second (percent fall FEV1). The diagnostic threshold for a positive ECT was a percent fall FEV1 of 10% or more. Sensitivity (SN) and specificity (SP) were calculated. For therapy, mean differences (MD) in the percent fall FEV1 and 95% confidence intervals (CI) (random effects model) were calculated. A positive MD indicates the intervention works better than the control. RESULTS For the diagnostic reviews, 5,318 citations yielded 28 relevant studies; for the therapy reviews, 1,634 citations yielded 109 relevant RCTs. Diagnostic test results versus ECT: self-reported history (2 studies) SN=36-8 percent; SP=85-86 percent; sport specific challenges (5 studies) SN=0-100 percent, SP=0-100 percent; eucapnic voluntary hyperpnea (7 studies) SN=25-90 percent, SP=0-71 percent; free running asthma screening test (3 studies) SN=60-67 percent, SP=47-67 percent; mannitol (3 studies) SN=58-96 percent, SP=65-78 percent. All SN and SP calculations indicated substantial heterogeneity that could not be explained by sensitivity or subgroup analyses. Therapy results: SABA offered greater protection than mast cell stabilizers (MCS) (12 studies); MD=6.8 (95 percent CI: 4.5, 9.2) but combining them offered no additional benefit; SABA versus MCS plus SABA (5 studies) MD=1.3 (95 percent CI: -6.3, 8.9). Leukotriene receptor antagonists (LTRA), MCS, ipratropium bromide, and interval warmup routines provided statistically significant attenuation of EIA when compared with placebo; inhaled corticosteroids (ICS) and other warmup routines did not. Single-dose intervention versus placebo results are: LTRA (9 studies) MD=8.9 (95 percent CI: 6.9, 11.0); MCS (nedocromil sodium) (17 studies) MD=15.6 (95 percent CI: 13.2, 18.2); interval warmup versus no warmup (4 studies) MD=10.6 (95 percent CI: 6.5, 14.7); ICS (4 studies) MD=5.0 (95 percent CI: 0.0, 9.9); continuous low intensity warmup versus no warmup (3 studies) MD=12.6 (95 percent CI: -1.5, 26.7); continuous high intensity warmup versus no warmup (2 studies) MD=9.8 (95 percent CI: -6.4, 26.0). After daily LABA (salmeterol) use for 3 to 4 weeks (4 studies), the percent fall FEV1 following an ECT at 2 and 4 weeks was greater than at day 1 in the LABA arm indicating that tachyphylaxis to prophylactic LABA use occurred. Daily SABA use for 1 week (1 study) also indicated development of tachyphylaxis. However, both LABA and SABA continued to have an attenuating effect on EIA. CONCLUSIONS Given the small number of studies comparing EIB/EIA diagnostic tests, the heterogeneity of the study populations, and the varied study methodologies, there is no clear evidence that any of the index tests are a suitable replacement for a standardized ECT to diagnose EIB/EIA in the general population. All bronchodilator agents and most anti-inflammatory agents when used as pretreatment are somewhat effective in attenuating the percent fall FEV1 associated with EIA.
