Miriam Rabkin, M.D., M.P.H.



Upper respiratory infection (URI) is the most common acute illness in the United States, the most common reason for absence from school or work, and the most common acute complaint seen in the ambulatory setting. URI is also the syndrome most consistently mismanaged by primary care providers. Variably described as homely, prosaic and plebian in the medical literature, the topic is rarely perceived as exciting or controversial. Nonetheless, it is clear that physicians continue to overuse antibiotics in the treatment of URI and that this "prescriptive promiscuity" has directly contributed to the widespread emergence of antibiotic resistance.

This chapter will briefly address acute nasopharyngitis ("the common cold"), acute bronchitis, acute sinusitis, pharyngitis and influenza. Readers interested in further details – or in allergic rhinitis, community-acquired pneumonia, chronic cough, or the evaluation of fever and cough in immunocompromised patients – are encouraged to pursue the references at the end of the chapter.

The vast majority of URIs are mild, self-diagnosed and self-treated at home. Americans spend more than $1 billion each year on over-the-counter medications for URI symptoms. When patients do present for medical attention, however, you must efficiently and accurately be able to differentiate trivial from serious and potentially treatable infections. Is this an allergic or infectious problem? With acute infections, is there evidence for bacterial infection or superinfection? Have the symptoms lasted more than three weeks? Has a fever lasted longer than a week? Is there purulent nasal discharge with sinus pain, ear pain with discharge, severe odynophagia, chest pain, dyspnea, hemoptysis, stridor or difficulty breathing? These questions and others should give you clues with which to create a differential diagnosis. In addition, smokers with URI symptoms should be counseled that smoking cessation will reduce the incidence and severity of URIs.

The Common Cold

The term "common cold" refers to acute nasopharyngitis (sometimes called acute rhinosinusitis), a mild, self-limited syndrome caused by viral infection of the upper respiratory tract mucosa. Cardinal features include malaise, nasal discharge and obstruction, sneezing and sore or "scratchy" throat. Headache, mild conjunctivitis, hoarseness and cough may also be seen. High fever is uncommon. A prevalent myth among physicians is that the presence of purulent sputum – or the color of sputum – predicts bacterial infection and warrants antibiotics. This is clearly untrue. Discolored or purulent sputum or nasal discharge is consistent with the natural history of viral URI and is a normal and self-limited phase of the common cold. While two to three percent of colds may be complicated by secondary bacterial infection, including sinusitis and otitis media, these are usually evidenced by additional findings in the history and physical exam and are rarely subtle. Symptoms generally last one week, although 25 percent of patients (particularly smokers) may be ill for up to two weeks.

The majority of colds are caused by four different families of viruses - rhinovirus, coronavirus, adenovirus and respiratory syncytial virus. While there is seasonal variation in infection rates of the different viruses, they produce clinically indistinguishable symptoms. Influenza and parainfluenza cause upper tract symptoms but are usually associated with lower tract and systemic symptoms as well. The mechanisms of viral transmission are not well established, but most colds are thought to be spread by hand contamination with infectious secretions and subsequent auto-inoculation (nose-to-hand-to-hand-to-nose). Aerosol transmission and fomite transmission are also possible; prompt disposal of nasal secretions and handwashing are recommended interventions. Cold weather is not associated with more frequent or more severe upper respiratory infections.

Unfortunately for most patients, the old saying that "the treated cold lasts seven days and the untreated cold lasts a week" is still a truism. There are no safe and effective antiviral medications for this syndrome. Symptomatic treatment may include decongestants, antihistamines, warm saline gargles, lozenges and/or cough suppression with dextromethorpan or codeine (Table 1). Intranasal ipratropium bromide can be used to ameliorate sneezing if this is a particularly bothersome symptom, although it may cause nasal dryness and blood-tinged sputum. Inhaled cromolyn sodium was shown to improve symptoms of rhinorrhea, sore throat and cough when used within the first 24 hours of symptoms, but required q2 hour administration.

Herbal and complementary medicines are widely used for URI treatment and prevention, but not well studied. Chicken soup increases the clearance of nasal mucous. Careful analysis of the available evidence does not support the use of high dose vitamin C. Zinc lozenges may have a minor beneficial effect – the literature is conflicting – but are associated with noxious taste and gastrointestinal side effects. Several randomized trials of Echinacea show no benefit on either prevention or treatment of URIs. We do not recommend the use of cromolyn, ipratropium, Echinacea, zinc or vitamin C at this time.

There is no role for antibiotics in the treatment of the common cold – or, in fact, for most upper respiratory infections. This should be an uncontroversial statement as the common cold is caused by viruses, against which antibiotics are, obviously, useless. Despite this fact, more than 50 percent of patients presenting with URI symptoms are given antibiotics. This apparent paradox will be discussed at greater length below.

TABLE 1: Symptomatic treatment for the common cold


Generic name

Trade name


Side effects




Reduces sneezing, nasal mucus, symptom score




no difference from placebo



pseudoephedrine/ phenylephrine spray

many incl. Sudafed

Reduces congestion, sneezing

Tachycardia, dizziness, HTN, bladder outlet obstruction


oxymetazoline spray

Dristan, Afrin

Improved symptom score

Rebound nasal congestion




Does not reduce cough frequency, slight decrease in sputum production









minimally effective in acute cough due to common cold



Tessalon perles

effective antitussive

Numbs mouth, do not chew

Adapted and modified from reference .

Acute Bronchitis

While acute bronchitis is technically a lower respiratory tract infection, it is mentioned in this chapter because of its reputation as "any cough for which the doctor has decided to prescribe antibiotics." In other words, bronchitis is often invoked as a rationale for giving antibiotics to patients with URI symptoms. In fact, acute bronchitis is an inflammatory condition of the tracheobronchial tree caused almost exclusively by viral agents. A recent randomized controlled trial of Azithromycin showed that it was no more effective than vitamin C in the treatment of adults with acute bronchitis and no underlying lung disease. In the absence of specific alarm signs (dyspnea, hemoptysis, pleuritic chest pain, prolonged course, high fever) which may signify bacterial superinfection or pneumonia, there is no role for antibiotics in the management of bronchitis.

The diagnosis of acute bronchitis is a clinical one, and the syndrome does overlap with that of the common cold. While some authors argue that URI and acute bronchitis are variant presentations of the same illness, many agree than bronchitis is characterized by a more prominent cough (productive or nonproductive) and by fever; rhinorrhea is less prominent and duration is longer, often up to two weeks. There is no specific finding on history and physical that can rule in pneumonia, but the alarm signs above or physical exam consistent with pulmonary consolidation should prompt a chest Xray.

Acute Sinusitis

The term "sinusitis" means inflammation of one or more of the paranasal sinuses associated with ostial obstruction and impaired drainage. Symptoms of less than four weeks duration are considered acute. Historically, "acute sinusitis" has been used to signify a bacterial infection, but we now know that most common colds involve the sinuses (which is why that viral syndrome can be called acute rhinosinusitis). A small percentage (0.5-2.0%) of colds are complicated by acute bacterial infection of the sinuses; this overlap has made sinusitis difficult to study and explains our lack of information about its epidemiology and natural history. The fact that a clinical diagnosis of "acute sinusitis" is more likely to be associated with a viral infection than a bacterial one may also explain our failure to prove that antibiotics are indicated for this syndrome.

The diagnosis of acute sinusitis is initially a clinical one. No single symptom or sign is an accurate predictor. Prolonged rhinitis is usually the first clue, but there are five clinical indicators that have been shown to predict the syndrome: maxillary toothache, patient report of colored nasal discharge, mucopurulent discharge on examination, poor response to decongestants and inability to transilluminate the sinus. In a study of men with nasal symptoms, the presence of sinusitis when all five features were present was 92 percent. Headache and fever are also common. Radiologic confirmation is required only if the patient does not improve with empiric therapy or if s/he appears extremely ill. Sinus films (particularly the Waters occipitomeatal view) are most appropriate; CT scanning is sensitive but not specific. Sinus tap and culture are reserved for patients who appear toxic at presentation, those with severe facial pain, those with underlying immunocompromise, and those with recalcitrant infection.

Treatment of sinusitis is governed by how sick and/or uncomfortable the patient appears. Toxic patients or those with complicated sinusitis (periorbital cellulitis, dental abscess, intracranial extension) should be admitted to the hospital and treated with parenteral antibiotics under the supervision of an otolaryngologist. The vast majority of patients, however, have uncomplicated disease and can be effectively treated by primary care providers in the ambulatory setting. Decongestants are clearly indicated and should be offered to all patients; antibiotics are controversial.

There are many clinical trials of antibiotics for acute sinusitis, but most compare one drug to another. Although cultures obtained via sinus puncture demonstrate S. pneumonia, H. influenza, S. pyogenes and S. aureus are the most common bacterial pathogens, there have been trials of penicillins, cephalosporins, sulfonamides, quinolones and macrolides. These head-to-head drug studies do not help us to determine whether antibiotic therapy is indicated. A randomized, placebo-controlled trial of amoxicillin and decongestants vs. placebo and decongestants showed no improvement among those who received amoxicillin; these patients did, however, have a higher incidence of side effects. In older placebo-controlled trials, there was a 60 percent placebo response rate, suggesting that acute sinusitis has a high rate of spontaneous resolution;23 antibiotics were superior to placebo in some of these trials.

In spite of this underwhelming data, most guidelines recommend antibiotic treatment of acute sinusitis. Although the authors conclude that symptoms resolve in two-thirds of untreated patients within two weeks, the most recent (1999) AHCPR guideline recommends the use of amoxicillin or bactrim for acute uncomplicated sinusitis, stressing that patients are more likely to be cured and to be cured more quickly if they receive antibiotics. They agree that a 7 to 10 day period of watchful waiting before antibiotics are prescribed is a reasonable course of action, since most patients will recover without antibiotics and complications are rare.

Figure 1: Management of acute uncomplicated sinusitis

Acute uncomplicated sinusitis

(clinical diagnosis)


nasal decongestants every 4 hours x three days

(phenylephrine/NeoSynephrine or oxymetazoline/Afrin)

nasal saline ad lib

tylenol prn

if no improvement in 7-10 days:

Bactrim DS bid x 3-7 days or

Amoxicillin 500 mg tid x 7 days or

Clarithromycin 500 mg bid if allergic

If no improvement after 3 days of first-line antibiotic therapy:

Change to Cefuroxime 250 mg bid x 7 days or

Cefpodoxime 200 mg bid x 7 days

If no improvement after 7 days second-line antibiotic therapy:

Sinus xrays

Subspecialty referral



The flu syndrome is a severe illness, lasting from three days to two weeks with convalescence over one to four weeks. Symptoms include the abrupt onset of malaise, prominent myalgia, headache and high fever (up to 106F, lasting three to seven days), followed by cough, rhinorrhea, pharyngitis and weakness. Complications include prolonged airway hyperreactivity, particularly in patients with asthma and chronic bronchitis. Secondary bacterial bronchitis and pneumonia are seen in up to ten percent of patients, and usually present as purulent or bloody sputum and increased fever after a few days of apparent clinical improvement. The most common cause is S. pneumoniae, although H. influenza and S. aureus infections (associated with a 50 percent mortality) are also seen. Rare nonpulmonary complications include myositis (with or without myoglobinuria and acute renal failure), Guillain-Barre syndrome, encephalitis, transverse myelitis and Reye’s syndrome. Since the flu syndrome can be severe, patients who are elderly, debilitated by chronic disease, immunocompromised or who have severe dehydration or secondary complications may require hospitalization.

While may viruses can cause "flu," including parainfluenza, respiratory syncytial virus and adenovirus, 85 percent of cases are caused by influenza A or B. Epidemic spread of the influenza virus is due to the appearance of new antigenic variations of the virus in nonimmune populations. Viral transmission takes place through the aerosol route, by sneezing, coughing and talking; the virus has an extremely high attack rate. Major antigenic variations lead to pandemics, such as the influenza A pandemic of 1918, which killed 25 million people worldwide. Minor antigenic changes cause near-annual winter epidemics; 20,000 deaths a year in the United States are attributed to influenza. The CDC follow national and global influenza epidemiology extremely closely, and maintain an extensive website at Diagnosis is typically made clinically outside of research settings, although rapid diagnostic tests are available.

Treatment of the flu consists of symptomatic and supportive care, including bed rest, hydration, analgesics, antipyretic agents and antibiotics for secondary bacterial infection. In addition, four antiviral agents are approved for the treatment of influenza: amantadine, rimantadine, zanamivir and oseltamivir. Amantadine and rimantadine are approved for the treatment of influenza A infections; they inhibit viral uncoating after cell entry. Rimantadine has fewer CNS symptoms than amantadine but is more expensive. Zanamivir (inhaled) and oseltamivir (oral), both approved in 1999, are neuraminidase inhibitors which are effective against both influenza A and B. All four agents must be used within 24 to 48 hours of initial symptoms if they are to have any effect at all; they are not curative but shorten the duration of illness by one to two days. There are limited clinical data with which to assess the efficacy of zanamivir and oseltamivir and no trials comparing the four agents to each other. They are thought to be roughly comparable in terms of their palliative effects on flu symptoms; data suggest that zanamivir and oseltamivir may have fewer adverse effects.

Although heavily marketed, these antiviral agents have limited efficacy in the treatment of influenza and have no effect in patients with symptoms of more than 36-48 hours duration. It is reasonable to treat patients at high risk for morbidity and mortality who have the "flu syndrome" and who present within 48 hours of symptom onset – such patients might include unvaccinated persons with chronic pulmonary, cardiovascular, neuromuscular, metabolic and immunodeficiency diseases. These agents should not be prescribed for patients with simple URI symptoms – remember that severe fatigue, myalgia and high fever are characteristic of influenza.

Given the potential severity of illness, prophylactic measures are recommended for select patients. Vaccination guidelines are discussed in Chapter 2. Amantadine prophylaxis prevents clinical disease in 70 to 90 percent of patients; oseltamivir has also been shown to work as prophylactically. These agents may be considered for the same high risk groups outlined above, particularly those in closed settings such as nursing homes. Chemoprophylaxis is not a substitute for vaccination!


Acute pharyngitis is an inflammatory syndrome of the pharynx, usually caused by a virus but occasionally bacterial in origin. The main challenge for the primary care provider is to identify those patients with group A beta-hemolytic streptococcus (S. pyogenes) which can be associated with scarlet fever, acute rheumatic fever and acute glomerulonephritis. Streptococcal pharyngitis should be treated with antibiotics, but its low prevalence makes empiric antibiotic treatment of all adults with pharyngitis an extremely bad idea. In fact, Little et al. have demonstrated that this strategy – giving penicillin to all adults with sore throats – has the same probability of preventing one case of rheumatic fever or acute nephritis as it does of causing one death from penicillin-induced anaphylaxis.

Most cases of pharyngitis occur in the colder months, with the peak incidence of streptococcal pharyngitis in late winter to early spring. The severity of illness varies greatly, but cardinal features include sore throat, odynophagia, malaise, fever and headache. Signs of "strep throat" may include an exudative pharyngitis, tender tonsillar lymph nodes and a rare scarletiniform rash. Because these symptoms and signs are not specific, the formal diagnosis of strep throat requires a throat culture. Rapid strep antigen tests have good specificity but lack adequate sensitivity; negative antigen tests should be followed by a throat culture. Occasionally, bacterial pharyngitis is complicated by a retropharyngeal or peritonsillar abscess or epiglottitis, which are medical emergencies.

Treatment of viral pharyngitis is mainly symptomatic. Since acute rheumatic fever is very rare in adults, the principle goals of treatment are the amelioration of symptoms and the prevention of local suppurative complications and spread. Symptoms in the untreated patient may last up to five days, and early therapy (in the first 48 hours) is required for moderate symptomatic relief.

The question facing the primary care physician is to decide which patients to culture and which patients to treat with antibiotics. One strategy is to use the clinical prediction rule generated by a large multicenter prospective study which compared the results of throat cultures in patients using three clinical findings; tonsillar exudate, temperature greater than 100 F and tender anterior cervical lymphadenopathy. Of the patients with all three findings, 42 percent had positive throat cultures. Fourteen percent of patients with one finding and three percent of patients with none of these signs had positive throat cultures. Based on these data, it is reasonable to empirically treat patients with all three signs, to culture patients with only one sign and treat based on the results and to neither culture nor treat (with antibiotics) patients with no signs. All patients with a sore throat and a history of acute rheumatic fever (ARF), or young patients with a sore throat and a strong family history of ARF should be cultured and treated before the culture results are known. Treatment consists of parenteral benzathine penicillin 1.2 million units IM x 1 (preferably) or oral penicillin V 250 mg tid for 10 days. For patients allergic to penicillin, erythromycin 250 mg qid for 10 days is adequate. Other antibiotics, including cephalosporins, macrolides and clindamycin are effective, but penicillin is preferred because of its "proven efficacy, narrow spectrum and low cost."

Overuse of antibiotics

It is clear that there is no role for antibiotics in the management of simple upper respiratory infection and bronchitis. This statement is data-based, uncontroversial, and supported by every expert panel and management guideline. Why then do physicians continue to prescribe antibiotics – of wider and wider spectrum – for these syndromes? The scope of the problem is immense. A large 1997 survey indicated that antibiotics were prescribed for 52 percent of patients with URIs and 66 percent of patients with bronchitis;19 this practice did not vary by geographical area, physician specialty or patient sociodemographic or insurance status. These data are consistent with other U.S. surveys and with data collected in other industrialized countries. Antibiotics unnecessarily prescribed for URIs and bronchitis represent 31 percent of total antibiotic prescriptions in the U.S.

The cost of such overuse of antibiotics is profound. In addition to wasting money on the agents themselves, over-prescription promotes subsequent unnecessary office visits. More worrisome is the fact that widespread use of antibiotics has clearly contributed to the emergence of drug-resistant pathogens, what Neu has called "the crisis in antibiotic resistance." Twenty years ago, more than 99 percent of all pneumococcus isolates in the U.S. were sensitive to penicillin; resistance rates in some studies now approach 30 percent. Prior antibiotic use is the single most powerful predictor of antibiotic resistance. Data from other countries demonstrate that changing prescription patterns can decrease rates of antibiotic resistance, providing a powerful motivation to reform our criteria for antibiotic use.

Why do physicians prescribe antibiotics for patients with viral infections? Although the reason is probably a complex interaction of doctor and patient expectations, two answers are usually offered in the medical literature: "it might help the patient" and "the patient expects antibiotics." A third, less acknowledged reason, is that it is usually faster and easier for the physician to write a prescription than to explain to the patient why antibiotics are inappropriate.

Some physicians argue that the use of antibiotics may provide a benefit to patients. In essence, these doctors do not believe the outcomes research that demonstrates the converse. While one author has noted that "trying to determine the clinical predictors of antibiotic-responsive URIs may be like looking for the Holy Grail," these clinicians believe otherwise. I would emphasize that this is poor science and poor medicine. Side effects of antibiotics are not rare and drug-resistance is not merely a theoretical concern.

The second proffered reason is that patients "expect" antibiotics - when physicians perceive that patients desire antibiotics, they are 10 times more likely to be prescribed. An interview study of patients presenting with URI symptoms challenges this approach. While 50 percent of patients did expect antibiotics, all stated that they would have been satisfied with a "nonantibiotic medicine to relieve symptoms." The only variable associated with desire for antibiotics was past experience, including previous receipt of antibiotics for URI. Another survey showed that patient satisfaction was not associated with receipt of antibiotics but understanding the illness and with the perception of having adequate time with their care provider. Thus, a complex cycle of expectations is created as we teach patients to expect antibiotics by prescribing antibiotics. The responsible approach is to spend the time to educate patients about cold care, to emphasize that non-use of antibiotics is not the same thing as non-care, and to strictly adhere to practice guidelines.

Excluding patients with underlying illnesses such as chronic obstructive pulmonary disease, cystic fibrosis, HIV, chronic sinusitis, chronic bronchitis and bronchiectasis, we recommend the following strategy: