Testing bronchial hyper-responsiveness: provocation or peak expiratory flow variability?

Assessing bronchial hyper-responsiveness (BHR) is a main diagnostic criterion of asthma. Provocation testing is not readily available in general practice, but peak expiratory flow (PEF) is. Several guidelines promote the use of PEF variability as a diagnostic tool for BHR. This study tested the agreement between histamine challenge testing and PEF variability, and the consequences for diagnosing asthma. To investigate the possibility of assessing BHR by PEF variability, using a histamine provocation test as a reference. Subjects with signs of symptoms indicating asthma (persistent or recurrent respiratory symptoms or signs of reversible bronchial obstruction) (n = 323) were studied. They had been identified in a population screening for asthma. A histamine provocation test and PEF variability were assessed over a three-week period. Asthma was defined as signs or symptoms together with a reversible airflow obstruction or BHR to the histamine challenge test. BHR was defined as a PC20 histamine of < or = 8 mg/ml or a PEF variability of > or = 15%. Overall correlation between PC20 and PEF variability was calculated using Spearman's rho. Furthermore, a decision tree was constructed to clarify the role of BHR in diagnosing asthma. Thirty-two patients had a reversibility in forced expiratory volume in 1 second (FEV1) of > or = 9% predicted, 131 patients showed a PC20 of < or = 8 and 11 patients had a PEF variability of > or = 15%. Overall correlation was poor at only -0.27 (P < 0.0001). One hundred and fourteen of the 131 patients diagnosed as having asthma when the histamine challenge test was used were not diagnosed by PEF variability. PEF variability cannot replace bronchial provocation testing in assessing BHR. This indicates that PEF variability and bronchial provocation do not measure the same aspects of BHR. If BHR testing is required in diagnosing asthma, a bronchial provocation test has to be used in general practice as well.