[Proposals for a rationale and for rational diagnosis of coughs]

Cough is probably the most frequent symptom in chest diseases. Hence, a rational and economical diagnostic procedure is essential to prevent unnecessary costs to the health services, i.e. acute bronchitis, a self-limiting disease, which is the most frequent cause for cough should not involve extensive per case costs. History, physical examination, chest X-ray and lung function testing--which constitute both the first and second, i.e. the basic level of a stepwise approach--allows to diagnose causes in most patients with cough. Without evidence of the cause after completing this basic diagnostic procedure patients with acute cough may require blood gases analysis, electrocardiography, echocardiography, lung perfusion study, spiral CT angiography, bronchoscopy or laboratory examinations for diagnosis of pulmonary embolism, aspiration or (seldom) pleuritis sicca. Chronic persistent cough (CPC) is diagnosed if the basic standard approach to chronic cough fails to lead to final diagnosis. Patients will then need further subtle diagnostic management, i.e. bronchial provocation testing, 24 hour pH probe, ENT- or neurological examination, high resolution CT of the thorax and bronchoscopy. We present two algorithms for the rational diagnostic approach to acute (figure 1) and chronic (figure 2) cough. Each algorithm considers spectrum and frequency of causes on the one hand, the positive predictive value, costs and patient discomfort due to the examination on the other. Nonetheless, despite extensive examination up to 20% of patients suffering from CPC the cause remains unclear [11]. Frequently, the capsaicin cough challenge test can reveal an idiopathic upregulation of the cough reflex as the hypothesised underlying condition. Psychogenic cough however, a rare condition in adults should not coincide with hypersensitivity of the cough reflex. Inconsistency and low reproducibility of results of the capsaicin challenge in patients with psychogenic cough preclude his routine clinical use. In conclusion, the very common acute bronchitis and the ACE inhibitor-induced cough do not require any other diagnostic procedure except patient history and physical examination. A simple basic diagnostic approach will usually allow to evaluate acute and chronic cough. In the remaining cases the proposed algorithm should be used for best results and to prevent excessive costs.