Introduction

Cough is a widespread sign and symptom of diseases ranging from uncomplicated respiratory tract infections to serious illnesses affecting several organ systems. It is a source of discomfort for the young patient and anxiety for the parents. In the years 1995–1996, 24 million annual physician visits for cough took place in the USA, the largest number documented for a single symptom. Nearly half the patients were younger than 15 years old. In this young cohort, cough accounted for 8.5% of all medical appointments. Pediatric texts generally describe chronic cough as a condition that persists for more than 3 weeks. This observation suggests that chronic cough is likely to improve in time without treatment. A better informed classification by Irwin and colleagues divides cough into three categories: acute, lasting less than 3 weeks; subacute, lasting 3 to 8 weeks; and chronic, lasting more than 8 weeks. Irwin’s definition of chronic cough excludes most self-limiting cases. A chronic cough by these criteria often lasts much longer than 8 weeks and requires medical attention.

Viral infections of the upper respiratory tract are the most common causes of acute cough. Typically, the symptoms resolve within 10 to 14 days. Patients with subacute cough most often have a history of recent upper respiratory tract infection or seasonal allergic rhinitis (e.g. postinfectious cough, bacterial sinusitis and asthma). Children with chronic or recurrent episodes of dry, nonproductive cough over several months, require careful and systematic evaluation for the presence of specific diagnostic indicators. They pose a perplexing problem in pediatric practice and call for a careful evaluation. Cough may be a manifestation of an underlying disorder that must be identified and treated. Many children with chronic cough have experienced repeated treatment failures, and the families have come to regard the condition as permanent and untreatable. Fortunately, in most cases, this presupposition is inaccurate, but a systematic approach to the diagnosis is necessary, and therapy, to be effective, may have to be directed simultaneously at more than one involved cough mechanism. Evidence-based algorithms to manage the chronic cough of children based on validated outcome measures and a priori definitions to designate resolution should be put to use. Child-specific cough management protocols are advocated in Australia, the USA and the UK.

Differential Diagnosis ( Figure 27-1 , Box 27-1 )

The differential diagnosis of cough in childhood varies with the age of the patient, the duration, character and time of occurrence of the cough, associated signs and symptoms, and the patient’s exposure history. In the neonatal period, congenital abnormalities, especially pulmonary or cardiac, must be considered. Prematurity, especially in a patient who had required mechanical ventilation, may lead to bronchopulmonary dysplasia or the development of tracheal or bronchial stenosis. Vomiting and regurgitation may be the presenting signs and symptoms of gastroesophageal reflux (GER) or a tracheoesophageal fistula. Recurrent choking or cough associated with difficulty in sucking or swallowing suggests aspiration. Cough may occur in the course or following resolution of a respiratory infection. Attendance in daycare increases the risk of upper respiratory symptoms and infections in young children. In the toddler, foreign body aspiration and cystic fibrosis are added to the list of causes. A history of fever and/or presentation in winter suggests a viral etiology; seasonal occurrence suggests asthma or seasonal allergic rhinitis; year-round symptoms suggest perennial allergic rhinitis. Maternal smoking, in particular, appears to influence the development of respiratory symptoms in young children. In the older child, immune deficiency, tuberculosis and psychogenic cough enter into the differential diagnosis. Sinusitis, postnasal drip and GER may contribute to cough at any age. Cigarette smoking and psychogenic causes also require consideration among adolescents. A recent study showed that in otherwise healthy children with unexplained chronic cough, a significant proportion of the coughs was preceded by episodes of reflux. Most of these episodes were acidic in older children but not in infants ( Figure 27-3 ). Early evaluation and treatment of children with recurrent cough, sinusitis, foreign-body aspiration or GER are important to prevent bronchiectasis.

Algorithm for the evaluation and treatment of chronic cough in childhood. CT – computed tomography, Tx – therapy, GERD – gastroesophageal reflux disease, ELISA – enzyme-linked immunosorbent assay, PPD – purified protein derivative, UGI – upper gastrointestinal series, MRI – magnetic resonance imaging, ECG – electrocardiogram.

How often do normal children cough? Accurate answers come from studies that used cough recorders. Cough frequency over 24 hours was 11.3, with a range of one to 34 in 41 children free from respiratory infection for at least 1 month. Only two children coughed at night. In children with chronic cough, the frequency was 65/day and in normal controls 10/day. Unfortunately, most studies rely on parents to give an account of their child’s cough, a method that has been shown to provide inaccurate information. When questionnaires administered to parents about their child’s coughing were compared to overnight recordings performed in 145 homes, the agreement was low.

Pathophysiology ( Figures 27-2 and 27-3 )

Cough serves as a protective mechanism to clear the respiratory tract and to defend it against the aspiration of noxious materials. While mechanical barriers limit the exposure of the respiratory tract to inhaled pathogens, the mucociliary apparatus and cough act to expel any organisms that may have bypassed the primary defenses. Two associated processes, bronchoconstriction and mucus secretion, add to its effectiveness. Recurrent partial collapse or incomplete inflation of the lungs and pneumonia associated with ineffective cough attest to its importance.

Changes in flow rate, air volume, subglottic pressure and sound level generated during the act of coughing.

(From Bianco S, Robuschi M. Mechanics of cough. In: Braga PC, Allegra L, editors. Cough. New York: Raven; 1989.)

Proportion of children with cough preceded by reflux with pH 4–7 or <4, n = 8. The three small panels on the right present the data by age: <2 years (n = 15), >2–<6 years (n = 36), > 6 years (n = 32).

(From: Ghezzi M, Guida E, Ullmann N, Sacco O, Mattioli G, Jasonni V, Rossi GA, Silvestri M. Weakly acidic gastroesophageal refluxes are frequently triggers in young children with chronic cough. Pediatr Pulmonol 2013;48:295–302).

Cough is executed as a complex reflex, an automatic or involuntary response to a stimulus, completed by the afferent and efferent pathways and a putative cough center in the brain, but also, at least in part, intensified or restrained under voluntary control. The main afferent pathways of cough originate in nerve receptors immediately beneath the respiratory epithelium in the larynx and the tracheobronchial tree, and in extrapulmonary sites: the nose, the paranasal sinuses, the pharynx, ear canals and ear drums, the pleura, the stomach, the pericardium and the diaphragm. Nerve impulses from the tracheobronchial tree pass through the vagus, the principal afferent pathway. Cough may result from direct stimulation of this nerve. The trigeminal, glossopharyngeal and phrenic nerves conduct impulses from extrapulmonary sites. Axon reflexes traveling through branches of sensory end-organs may cause the release of neuropeptides and subsequent smooth muscle contraction, mucus secretion and epithelial injury. Thus, sensory signals taking part in cough may trigger or enhance bronchospasm. Reflexes regulate the parasympathetic nervous system, and chronic cough lowers the threshold for sensory signals. Efferent impulses of the cough reflex are transmitted to the respiratory musculature through the phrenic and other spinal motor nerves, and to the larynx through the recurrent laryngeal branches of the vagus. The vagus also provides efferent innervation to the tracheobronchial tree where its branches mediate bronchoconstriction.

Cough and Bronchospasm

Cough and bronchospasm are two closely related reflexes that enhance one another, but neither depends on the other for its action. Cough clears the airways effectively only at high lung volumes; sufficient air velocity to shear mucus from bronchial walls can be achieved only down to the sixth or seventh generation of airway branching. Co-existing bronchoconstriction adds to the effectiveness of cough by extending peripherally the region of rapid and turbulent airflow. Challenge with either methacholine or histamine provokes both cough and bronchoconstriction. However, the receptors for both reflexes are functionally distinct, and either response can arise independently. Challenge with hyperosmolar solutions causes both cough and bronchoconstriction, but hypo-osmolar solutions tend to bring about cough alone. Pretreatment divides induced cough from bronchoconstriction. When aerosolized water serves as the provoking agent, inhaled lidocaine blocks cough but not bronchoconstriction. When inhaled capsaicin is used to provoke cough, opiates administered systemically suppress cough whereas those administered by inhalation suppress bronchoconstriction. Bronchoconstriction, but not the urge to cough, can be blocked by pretreatment with intravenous atropine, consistent with the role of cholinergic pathways in the efferent limb of reflex bronchoconstriction. The mechanisms that trigger cough and bronchospasm following exercise or exposure to cold air appear to be different. Cough results mainly from excessive water loss, while bronchoconstriction follows airway rewarming. Cold, air-induced bronchoconstriction can be blocked by β-adrenergic agents, but cough cannot. Cough most often results from excitation of receptors concentrated in the larynx and proximal airways, while bronchoconstriction can be triggered from the lower airways as well. Finally, inflammatory changes in the airways may result in cough without simultaneously giving rise to bronchospasm.

Cough-Variant Asthma

Childhood asthma is a syndrome of inflammation in medium and small airways that gives rise to hyperresponsiveness and constriction of the bronchial smooth muscle, edema and disruption of the mucosa, and obstruction of the airway lumen. Inflammation may lead to airway remodeling with proliferation of smooth muscle and deposition of matrix proteins. Cough-variant asthma is associated with the same disordered physiological processes and presenting signs, but overt wheezing is absent, and cough is the most discernible clinical sign. However, substantial evidence shows that awareness of symptoms by children with asthma is poor, and both children and their parents may be more aware of cough than of other symptoms that may be present as well.

The diagnosis of asthma on the basis of cough alone accounts for the profusion of cases of cough-variant asthma that are open to doubt. In 1991, 10% of children with cough as the only symptom were diagnosed as having asthma; 2 years later, the figure had increased to 22.6%. Whereas in the past, cough may have been underrecognized as a sign of asthma, at present the opposite appears to be true. This is borne out by reports in which children with persistent nocturnal cough improved after 2 weeks of placebo therapy and received only modest additional benefit from a course of high-dose inhaled corticosteroids. Inhaled albuterol and beclomethasone in children with cough, but without wheezing, were no more effective than placebo in reducing cough frequency. Surprisingly, even the documentation of airway hyperreactivity did not predict a child’s response to these asthma medications. A study of nocturnal cough showed that in the absence of wheeze, shortness of breath or tightness of the chest, cough did not indicate hidden or atypical asthma in most children. Children under 4 years of age with frequent recurrent wheeze and a stringent index for the prediction of asthma at school age showed significantly higher median fractional exhaled nitric oxide (NO) levels (11.7 [11.85]) (median [interquartile range]) than children with recurrent cough but no history of wheeze (6.5 [5.5]; P < .001) and those with early recurrent wheeze and a loose index for the prediction of asthma at school age (6.4 [6.5]; P < .001). No difference in FeNO levels was found between children in the latter two groups ( P = .91). A prospective study of infants followed up to age 11 years, showed that recurrent cough present early in life resolved in the majority of children. Children with recurrent cough but without wheeze did not have airway hyperresponsiveness or atopy, and significantly differed from those with classical asthma, with or without cough. Brooke and colleagues reassessed, during the early school years, a cohort of children identified as having recurrent cough in the preschool period. Seventy of 125 (56.0%; 95% CI 47.3–64.5%) were symptom-free at follow-up, 46 (36.8%; 95% CI 28.7–45.5%) continued to have recurrent cough in the absence of colds, and only nine (7.2%; 95% CI 3.6–12.8%) reported recent wheezing. The authors concluded that long-term recurrent cough in some children is consistent with the diagnosis of cough-variant asthma, but that few progress to develop asthma characterized by wheeze.

Isolated cough is rarely due to asthma and often fails to respond to asthma medications. On the other hand, patients with a prolonged history of cough who respond to treatment with asthma medications or show evidence of bronchospasm or hyperresponsiveness without concurrent wheezing may be considered to have cough-variant asthma. Patients may be free of bronchoconstriction at the time of their evaluation. Their history of respiratory disease may be difficult to assess, while physical findings and routine pulmonary function tests may disclose no evidence of airway obstruction. In such cases, evaluation of airway function by bronchial provocation with methacholine, histamine or exercise is recommended. In children too young to perform pulmonary function testing, the diagnosis of cough-variant asthma may be confirmed by the patient developing unequivocal evidence of reversible airways obstruction later in the clinical course and by the patient’s response to asthma therapy.

Cough During and After Respiratory Infection

Children have an average of six to eight respiratory infections per year, a number that may be higher in those with siblings or in daycare. Repeated infections common in winter months may result in a chronic cough. Acute bronchitis usually follows the symptoms of upper respiratory illness. Cough associated with infection with respiratory syncytial virus (RSV), other respiratory viruses and cytomegalovirus, Mycoplasma pneumoniae, Chlamydia trachomatis, Ureoplasma urealyticum, Pneumocystis jiroveci (formerly carinii ), Corynebacterium diphtheriae and Bordetella pertussis often lasts beyond the acute stage. Measles causes a cough with coryza, conjunctivitis and fever. In the immunized patient, atypical measles is more likely to cause cough or pneumonia than the characteristic rash.

Persistent bacterial bronchitis (PBB) is an increasingly diagnosed form of chronic wet cough that occurs in children with a history of mild asthma or possibly misdiagnosed asthma. Some of the children have a history of invasive medical therapy (prolonged ventilation, cardiac surgery) and many have an underprivileged background. These children have a chronic wet productive cough with bacteria such as Streptococcus pneumoniae , Haemophilus influenzae and Moraxella catarrhalis persisting in the airways, and an associated neutrophilia. Spirometry and chest x-rays are typically normal. The cough responds to a course of antibiotic (e.g. amoxicillin-clavulanate for 2–4 weeks). Prolonged duration of cough and increased neutrophil counts are related to worse high-resolution computed tomography scan scores. A recent retrospective study identified a cohort of children with protracted bacterial bronchitis that for the most part had started in infancy. Almost three quarters of the children had an associated airway malacia. These children responded well to antibiotics, although a significant number relapsed and needed additional courses of treatment. A favorable response to a course of antibiotics confirms the diagnosis of PBB and further investigations may be unnecessary. Children who do not respond to treatment require investigation for specific causes of suppurative lung disease. This includes a sweat test and genotyping for cystic fibrosis, exhaled NO, evaluation of ciliary ultrastructure and beat frequency, white cell count, immunoglobulin levels and functional antibody studies, barium swallow, swallowing videofluorscopy and esophageal reflux studies.

The pathogenesis of postinfectious cough is not known. Children with persistent postinfectious cough do not have airway eosinophilia typical of untreated asthma, but some manifest increased reactivity of the airways. These observations suggest that postinfectious cough has different pathophysiological features from asthma. The infection causing the cough, in most cases, remains unidentified. The diagnosis is clinical and one of exclusion. It should be considered in patients with normal chest x-rays and pulmonary function tests who cough only after respiratory tract infections. Postinfectious cough generally regresses over time, but it often recurs. Its resolution may be accelerated by the administration of inhaled corticosteroids or ipratropium bromide.