James Dubois

An Algorithmic Approach to Chronic Cough

Chronic cough is an important medical and economic problem. The prevalence of chronic cough in the United States among nonsmoking adults is reported to range from 14% to 23% [1, 2]. Not only is the symptom itself problematic, but it raises concerns about possible serious underlying disease [3]. Chronic cough is the fifth most common symptom seen by outpatient physicians [4] and is estimated to be the primary reason for 30 million physician visits annually [4]. In the United States alone, approximately

Cause and evaluation of chronic dyspnea in a pulmonary disease clinic.

600 million per year are spent on prescription and over-the-counter antitussives [5]. Major advances in the clinical approach to chronic cough have been made during the last 15 years. In a 1977 review [6], Irwin and colleagues proposed an approach to chronic cough based on the anatomic locations of the receptors and afferent pathways involved in the cough reflex. Using such an approach, Irwin and colleagues reported in 1981 [7] and again in 1990 [8] that the cause of chronic cough could be determined 100% of the time and that subsequent cause-specific treatment was almost always successful. The postnasal drip syndrome, mainly from chronic rhinitis, was the most common cause, followed by asthma [7, 8]. These two diagnoses, alone or in combination, accounted for cough in 75% of the patients [7, 8]. Gastroesophageal reflux was the next most common cause [7, 8]. Poe and colleagues [9] also reported that the postnasal drip syndrome or asthma caused chronic cough in most of their patients. We evaluated a sequential, stepped approach to chronic cough, emphasizing initial treatment of all patients with an antihistamine-decongestant for possible postnasal drip syndrome caused by rhinitis. We also determined the value of routine bronchoprovocation challenge for predicting whether asthma was a causative factor in cough. Methods Patient Selection All patients who came to our university-based pulmonary practice with a chief complaint of chronic cough were considered for inclusion in the study. Cough was considered to be chronic if it had been present for 3 weeks or longer [7, 8]. The following were exclusion criteria: 1) immunocompromise, including known lung cancer or other active malignancy, the acquired immunodeficiency syndrome, or current treatment with corticosteroids or other immunosuppressive agents; 2) cigarette use within 12 months; 3) use of an angiotensin-converting-enzyme inhibitor within 4 weeks; and 4) contraindication to the use of an antihistamine-decongestant or to bronchoprovocation challenge, including pregnancy and an FEV1 less than 70% of predicted value. Algorithm The approach involved a standardized initial evaluation, weekly follow-up, a series of sequential diagnostic and therapeutic steps, and a predefined end point. Initial Evaluation The initial evaluation included a history; physical examination; review of previous diagnostic studies; and a questionnaire on the duration, frequency, and severity of cough, postnasal drip symptoms, dyspnea, wheeze, and symptoms of gastroesophageal reflux. Spirometry was done next, followed by a methacholine bronchoprovocation challenge using a modification of the method of Hargreave and colleagues [10]. Bronchial hyper-responsiveness was defined as a 20% decrease in the FEV1 (Pc 20) at a methacholine concentration of 8 mg/mL or more [11]. Because of its low yield in previous studies [7-9], a chest roentgenogram was not included in the initial evaluation unless there was clinical suspicion of infection or neoplasm (such as fever, weight loss, or hemoptysis). Weekly Follow-up Patients were contacted by phone each week and seen in person whenever clinically necessary. We created a standardized scale for the patient to use in rating weekly cough severity and a second scale for rating side effects of therapy. The cough severity scale ranged from 0 (cough gone) to 7 (cough markedly worse). The descriptors for the scale are shown in Figure 1. Side effects were rated as none, mild, moderate, or severe (intolerable). Figure 1. Response to 1 week of antihistamine-decongestant therapy. End Point The end point for treatment was cough resolution, defined as a patient report of complete absence of cough for 2 consecutive weeks or of diminution to the point that the patient considered the cough insignificant. Step 1 An antihistamine-decongestant preparation containing 1 mg of azatadine maleate plus 120 mg of sustained-release pseudoephedrine sulfate (Trinalin, Key Pharmaceuticals Inc.; Kenilworth, New Jersey) was prescribed to be taken twice daily as empiric therapy for possible postnasal drip syndrome caused by rhinitis (postviral, allergic, or vasomotor). If the cough did not improve at the end of 1 week of antihistamine-decongestant therapy, step 2 was begun immediately. If the cough improved after 1 week of therapy, antihistamine-decongestant alone was continued either until it resolved or until no further improvement occurred. At that point, patients with persistent symptoms of the postnasal drip syndrome (sensation of postnasal drip, throat clearing, nasal congestion, or a tickle in the back of the throat) were given nasal corticosteroids in addition to the antihistamine-decongestant. Sinus imaging was obtained before nasal corticosteroids were prescribed if chronic sinusitis was suspected or subsequently if postnasal drip persisted despite the combination of antihistamine-decongestant and nasal corticosteroid therapy. Patients with sinus roentgenograms consistent with sinusitis (sinus opacification, air-fluid levels, or mucosal thickening) were treated with twice daily oxymetazoline hydrochloride nasal spray (Afrin, Schering-Plough HealthCare Products; Memphis, Tennessee) for 3 days, nasal corticosteroids twice daily, antihistamine-decongestant twice daily, and antibiotics for as long as 6 weeks. If cough still persisted, a computed tomographic study of the sinuses was obtained. If substantial abnormalities were present, the patient was referred for an otolaryngologic evaluation and possible sinus surgery. Substitutions for Trinalin could be made at any time if severe side effects developed. For severe drowsiness, astemizole (Hismanal, Janssen Pharmaceutica, Piscataway, New Jersey), 10 mg once daily, plus pseudoephedrine, 60 mg twice daily, were substituted. For severe insomnia, jitteriness, or urinary obstruction, astemizole once daily plus nasal corticosteroids twice daily were substituted. Step 2 Patients who were still coughing after step 1 were next evaluated for asthma. Because the treatment was the same, no attempt was made to distinguish between asthma and postviral bronchial hyper-responsiveness [12]. Patients who had bronchial hyper-responsiveness on bronchoprovocation challenge were treated for 1 week with an inhaled 2-agonist, albuterol, two puffs four times a day via metered-dose inhaler. If cough persisted, prednisone therapy (1 mg/kg body weight per day [maximum, 60 mg/d]) was added to the albuterol for a week. An oral 2-agonist or theophylline was substituted for the inhaled albuterol if a patient reported marked cough in response to inhaled albuterol. Step 3 Patients who continued to have persistent cough next had chest and sinus roentgenograms if they had not already been obtained. Any abnormalities considered to be clinically significant were evaluated and treated. If no significant abnormalities were found or the patient continued to cough despite appropriate therapy, the patient was advanced to step 4. Step 4 Patients were next evaluated for gastroesophageal reflux. Patients with symptoms consistent with gastroesophageal reflux first had a 2-week trial of therapy with ranitidine, 150 mg twice daily, along with antireflux measures (no eating or drinking for at least 2 hours before going to bed or lying down, 20-cm elevation of the head of the bed using blocks, and avoidance of caffeine, alcohol, chocolate, and other foods known to exacerbate gastroesophageal reflux). Lack of a response to the 2-week trial of ranitidine led to a 24-hour esophageal pH-probe study. Patients without symptoms of gastroesophageal reflux had 24-hour esophageal pH-probe monitoring before therapy for gastroesophageal reflux was begun. pH-probe monitoring was performed and interpreted according to the criteria of Demeester and colleagues [13]. All patients whose pH-probe studies showed pathologic gastroesophageal reflux were treated for at least 8 weeks with omeprazole, 20 mg daily, in addition to antireflux measures. An upper gastrointestinal barium swallow or a gastroenterology consultation or both were obtained if the patient had any persistent gastrointestinal complaints. Step 5 Patients who continued to cough next had bronchoscopy. If the bronchoscopy was nondiagnostic, the patient was treated (or retreated) with asthma therapy even if the bronchoprovocation challenge had been negative or if 2-agonist and prednisone therapy was previously ineffective. If cough persisted, uncommon causes were considered (Table 1) [6-9]. Psychogenic cough was considered a diagnosis of exclusion. Table 1. Uncommon Causes of Chronic Cough Three-Month Follow-up All patients were reevaluated using a standardized questionnaire 3 months after cough had resolved to determine whether cough had recurred and whether they were still taking cough-specific medications. Final Diagnostic Criteria No cause of cough was considered definitive until treatment for that cause had been effective (that is, was associated with marked improvement or resolution). When therapy for only one diagnosis resolved the cough, that diagnosis was considered to be the sole cause of cough. When therapy for a diagnosis effected marked improvement short of resolution, that diagnosis was considered to be one of the causes of cough and treatment was continued while additional diagnoses were investigated. Statistical Analysis Recurring data were analyzed by paired t-tests. The extent of association between two variables in a cross-tab