METHODS: Claims data from the MEDLINE, EMBASE, LILACS, CBM, CNKI, VIP, WanFang and the Cochrane Library databases were used to conduct retrospective cohort analyses. We review articles and reference lists of relevant articles prior to 4 September 2015 (Table 1) for human with restrictions as to language of English or Chinese. Two authors independently reviewed the titles and abstracts of the retrieved studies of GABA-agonist (baclofen or arbaclofen) which were matched using Review Manager 5.3 software. Then we extracted relevant data by outcome for therapy group of GABAB receptor agonists and for the control group of placebo or medicine. The outcome data were number of subjects whose cough threshold to inhaled capsaicin C5 increased after treatment or those symptom decline. Disagreements were resolved by consensus.