Distinguishing acute (less than 3 weeks), persistent (3–8 weeks), and chronic (more than 8 weeks) cough illness syndromes is a useful first step in evaluation. Postinfectious cough lasting 3–8 weeks has also been referred to as subacute cough to distinguish this common, distinct clinical entity from acute and chronic cough.
In healthy adults, most acute cough syndromes are due to viral respiratory tract infections. Additional features of infection such as fever, nasal congestion, and sore throat help confirm this diagnosis. Dyspnea (at rest or with exertion) may reflect a more serious condition, and further evaluation should include assessment of oxygenation (pulse oximetry or arterial blood gas measurement), airflow (peak flow or spirometry), and pulmonary parenchymal disease (chest radiography). The timing and character of the cough are not very useful in establishing the cause of acute cough syndromes, although cough-variant asthma should be considered in adults with prominent nocturnal cough, and persistent cough with phlegm increases the likelihood of chronic obstructive pulmonary disease (COPD). The presence of posttussive emesis or inspiratory whoop in adults modestly increases the likelihood of pertussis, and the absence of paroxysmal cough and the presence of fever decrease its likelihood. Uncommon causes of acute cough should be suspected in those with heart disease (heart failure [HF]) or hay fever (allergic rhinitis) and those with occupational risk factors (such as farmworkers).
2. Persistent and chronic cough
Cough due to acute respiratory tract infection resolves within 3 weeks in the vast majority (more than 90%) of patients. Pertussis should be considered in adolescents and adults who have persistent or severe cough lasting more than 3 weeks, who have not recently been boosted with Tdap, and who have been exposed to a person with confirmed pertussis. It should also be considered in selected geographic areas where its prevalence approaches 20% (although its exact prevalence is difficult to ascertain due to the limited sensitivity of diagnostic tests).
When angiotensin-converting enzyme (ACE) inhibitor therapy, acute respiratory tract infection, and chest radiograph abnormalities are absent, most cases of persistent and chronic cough are due to (or exacerbated by) postnasal drip (upper airway cough syndrome), asthma, or gastroesophageal reflux disease (GERD), or some combination of these three entities. Approximately 10% of cases are caused by nonasthmatic eosinophilic bronchitis. A history of nasal or sinus congestion, wheezing, or heartburn should direct subsequent evaluation and treatment, though these conditions frequently cause persistent cough in the absence of typical symptoms. Dyspnea at rest or with exertion is not commonly reported among patients with persistent cough; dyspnea requires assessment for chronic lung disease, HF, anemia, pulmonary embolism, or pulmonary hypertension.
Bronchogenic carcinoma is suspected when cough is accompanied by unexplained weight loss, hemoptysis, and fevers with night sweats, particularly in persons with significant tobacco or occupational exposures (asbestos, radon, diesel exhaust, and metals). Persistent and chronic cough accompanied by excessive mucus secretions increases the likelihood of COPD, particularly among smokers, or of bronchiectasis if accompanied by a history of recurrent or complicated pneumonia; chest radiographs are helpful in diagnosis. Chronic cough with dry eyes may represent Sjögren syndrome. A chronic dry cough may be the first symptom of idiopathic pulmonary fibrosis.
Examination can direct subsequent diagnostic testing for acute cough. Pneumonia is suspected when acute cough is accompanied by vital sign abnormalities (tachycardia, tachypnea, fever). Findings suggestive of airspace consolidation (crackles, decreased breath sounds, fremitus, egophony) are significant predictors of community-acquired pneumonia but are present in a minority of cases. Purulent sputum is associated with bacterial infections in patients with structural lung disease (eg, COPD, cystic fibrosis), but it is a poor predictor of pneumonia in the otherwise healthy adult. Wheezing and rhonchi are frequent findings in adults with acute bronchitis and do not indicate consolidation or adult-onset asthma in most cases.
Examination of patients with persistent cough should look for evidence of chronic sinusitis, contributing to postnasal drip syndrome or asthma. Chest and cardiac signs may help distinguish COPD from HF. In patients with cough and dyspnea, a normal match test (ability to blow out a match from 25 cm away) and maximum laryngeal height greater than 4 cm (measured from the sternal notch to the cricoid cartilage at end expiration) substantially decrease the likelihood of COPD. Similarly, normal jugular venous pressure and no hepatojugular reflux decrease the likelihood of biventricular HF.
Chest radiography should be considered for any adult with acute cough whose vital signs are abnormal or whose chest examination suggests pneumonia. The relationship between specific clinical findings and the probability of pneumonia is shown in Table 2–1. A large, multicenter randomized clinical trial found that elevated serum C-reactive protein (levels greater than 30 mg/dL) improves diagnostic accuracy of clinical prediction rules for pneumonia in adults with acute cough; procalcitonin added no clinically relevant information. A meta-analysis found that lung ultrasonography had better accuracy than chest radiography for the diagnosis of adult community-acquired pneumonia. Lung ultrasonography had a pooled sensitivity of 0.95 (95% confidence interval [CI], 0.93–0.97) and a specificity of 0.90 (95% CI, 0.86–0.94). Chest radiography had a pooled sensitivity of 0.77 (95% CI, 0.73–0.80) and a specificity of 0.91 (95% CI, 0.87–0.94). In patients with dyspnea, pulse oximetry and peak flow help exclude hypoxemia or obstructive airway disease. However, a normal pulse oximetry value (eg, greater than 93%) does not rule out a significant alveolar–arterial (A–a) gradient when patients have effective respiratory compensation. During documented outbreaks, clinical diagnosis of influenza has a positive predictive value of ~70%; this usually obviates the need for rapid diagnostic tests. No evidence exists to assess whether the initial evaluation of cough in immunocompromised patients should differ from immunocompetent patients, but expert recommendations suggest that tuberculosis be considered in HIV-infected patients in areas with a high prevalence of tuberculosis regardless of radiographic findings.
Table 2–1.Positive and negative likelihood ratios for history, physical examination, and laboratory findings in the diagnosis of pneumonia. ||Download (.pdf) Table 2–1. Positive and negative likelihood ratios for history, physical examination, and laboratory findings in the diagnosis of pneumonia.
|Finding ||Positive Likelihood Ratio ||Negative Likelihood Ratio |
|Medical history || || |
| Fever ||1.7–2.1 ||0.6–0.7 |
| Chills ||1.3–1.7 ||0.7–0.9 |
|Physical examination || || |
| Tachypnea (RR > 25 breaths/min) ||1.5–3.4 ||0.8 |
| Tachycardia (> 100 beats/min in two studies or > 120 beats/min in one study) ||1.6–2.3 ||0.5–0.7 |
| Hyperthermia (> 37.8°C) ||1.4–4.4 ||0.6–0.8 |
|Chest examination || || |
| Dullness to percussion ||2.2–4.3 ||0.8–0.9 |
| Decreased breath sounds ||2.3–2.5 ||0.6–0.8 |
| Crackles ||1.6–2.7 ||0.6–0.9 |
| Rhonchi ||1.4–1.5 ||0.8–0.9 |
| Egophony ||2.0–8.6 ||0.8–1.0 |
|Laboratory findings || || |
| Leukocytosis (> 11 × 109/L in one study or ≥ 10.4 × 109/L in another study) ||1.9–3.7 ||0.3–0.6 |
2. Persistent and chronic cough
Chest radiography is indicated when ACE inhibitor therapy–related and postinfectious cough are excluded. If pertussis is suspected, polymerase chain reaction testing should be performed on a nasopharyngeal swab or nasal wash specimen—although the ability to detect pertussis decreases as the duration of cough increases. When the chest film is normal, postnasal drip, asthma, or GERD are the most likely causes. The presence of typical symptoms of these conditions directs further evaluation or empiric therapy, though typical symptoms are often absent. Definitive tests for determining the presence of each are available (Table 2–2). However, empiric treatment with a maximum-strength regimen for postnasal drip, asthma, or GERD for 2–4 weeks is one recommended approach since documenting the presence of postnasal drip, asthma, or GERD does not mean they are the cause of the cough. Alternative approaches to identifying patients who have asthma with its corticosteroid-responsive cough include examining induced sputum for increased eosinophil counts (greater than 3%) or providing an empiric trial of prednisone, 30 mg daily orally for 2 weeks. Spirometry may help identify large airway obstruction in patients who have persistent cough and wheezing and who are not responding to asthma treatment. When empiric treatment trials are not successful, additional evaluation with pH manometry, endoscopy, barium swallow, sinus CT, or high-resolution chest CT may identify the cause.
Table 2–2.Empiric therapy or definitive testing for persistent cough. ||Download (.pdf) Table 2–2. Empiric therapy or definitive testing for persistent cough.
|Suspected Condition ||Step 1 (Empiric Therapy) ||Step 2 (Definitive Testing) |
|Postnasal drip ||Therapy for allergy or chronic sinusitis ||Sinus CT scan; ENT referral |
|Asthma ||Beta-2-agonist ||Spirometry; consider methacholine challenge if normal |
|GERD ||Lifestyle and diet modifications with or without proton pump inhibitors ||Esophageal pH monitoring |