Sore throat
Click on any of the headings or subheadings to rapidly navigate to the relevant section of the guidemap

Introduction.

What causes a sore throat?

How does a "strep throat" infection present?

How does a physician clinically differentiate a "strep throat" pharyngitis from a viral pharyngitis?

Why is it important to diagnose and treat a "strep throat" infection?

Diagnostic tests to diagnose a "strep throat" infection.

Recommended approach for diagnosing, and treating, a "strep throat" infection.

What is the recommended antibiotic for "strep throat" infections?

What are the other common causes of a pharyngitis?

Commentary, criticism and controversy

Introduction:

I practiced emergency medicine for over 20 years and I frequently dealt with patients who presented to the ED with the complaint of a sore throat. I was constantly amazed by the attitude that many patients adopted with respect to this clinical problem, and I often discovered that their opinions were uninformed, illogical and not concordant with good quality medical practice. For instance, many patients expected antibiotic treatment for their sore throat, irrespective of whether they had an infection that warranted antibiotics (certain bacterial infections) or not (all viral infections). Their common refrain was that they always received antibiotics when they presented to their family doctor with a complaint of a sore throat, and they therefore expected me to prescribe antibiotics without any reservations. I usually refused to indiscriminately prescribe antibiotics on demand, because I believe that antibiotic therapy is only indicated for "strep throat" infections. In the USA, "strep throat" infections account for approximately 5-15% of adult sore throat infections. The remainder are mainly due to virus infections, which do not benefit from antibiotic therapy. However, in the USA, it is remarkable that 75% of adult patients who visit their family doctors with a complaint of a sore throat receive an antibiotic prescription. Why do US-physicians over-prescribe antibiotics? The reasons are multifactorial. One reason is that many physicians find it clinically difficult to determine which patients actually have a "strep throat" infection and they are reluctant to order additional diagnostic tests that are expensive, take too much time to perform and do not always offer an immediate answer. They also find it too time-consuming to explain to uninformed patients why antibiotics are only indicated under certain circumstances. It is much easier to simply write a prescription for antibiotics, and move on to the next patient. Another reason for the antibiotic over-prescribing behaviour of US-physicians is the medicolegal climate in the USA -- physicians are scared of the medical malpractice consequences of not treating a case of "strep throat" if there is a small chance of an adverse outcome (eg. rheumatic fever or suppurative complication), and they are therefore willing to treat many patients unnecessarily so as not to miss treating a single case of "strep throat".  Finally, it is an unfortunate social fact that many patients expect, and demand, antibiotics, and those patients are not readily satisfied if they leave a physician's office without an antibiotic prescription. In a consumer-oriented society such as the USA where "patient-satisfaction" (rather than good medical practice) is the mantra,  physicians are under constant pressure to satisfy the demands of the "patient-as-customer".

If you are an intelligent layperson, who only wants to take antibiotics for a sore throat if it clinically necessary, then you may find this guidemap invaluable. The chief purpose of this guidemap is to provide accurate, and useful, information on the clinical problem of a sore throat, so that an intelligent layperson can think through the clinical problem in a logical way. Hopefully, he should then be able to make rational choices in conjunction with his health care provider, and only use antibiotic therapy if it is clinically appropriate.

What causes a sore throat?

From a practical point of view, most patients who present to their family doctor with a complaint of a sore throat have pharyngitis (inflammation of the pharynx due to an infection), and/or tonsillitis (inflammation of the tonsils due to an infection). There are a number of other causes of a sore throat (eg. due to mechanical or chemical irritation), but the vast majority of sore throat complaints are due to a throat infection -- usually viral or bacterial. Therefore, a treating physician's main focus is usually targeted towards deciding whether the pharyngitis is likely to be due to an infecting virus, which only requires symptomatic therapy because there is no known effective antiviral therapy that can kill viruses and abort the clinical illness. Or, if the treating physician concludes that the infection is due to a bacterial agent, he then has to decide whether antibiotic therapy will benefit the patient and cure the problem. There are many bacteria that can cause a sore throat (and I will provide selective information on some of them later in this guidemap), but the vast majority of bacterial infections causing a sore throat are due to a specific bacteria called "Group A Beta-hemolytic Streptococcus" (henceforth abbreviated as GABHS). An acute GABHS infection is the specific cause of a "strep throat" infection and appropriate antibiotic therapy is usually warranted to treat the infection.

How does a "strep throat" infection present?

The "classic" presentation of a "strep throat" infection may involve many of the following symptoms and signs:-

Photograph of normal tonsils


Photograph of acute tonsillitis (inflammed tonsils)

Click here to see more photographs of acute tonsillitis.

Certain symptoms are very rarely found in GABHS pharngitis infections -- runny nose, cough, hoarseness, conjunctivitis, and diarrhea -- and their presence suggests a virus as the cause of the pharyngitis .

GABHS pharyngitis is more common in school-age children between the age of 5-15 years, especially during the late autumn, winter and early spring months in temperate climates (20-30% of sore throat cases), while adults with a sore throat have a much lower prevalence of GABHS infection (5-15% of sore throat cases).

Although many cases of "strep throat" present with some/all of the classical signs, it not uncommon for a GABHS infection to present with milder, and non-specific, signs -- no/mild fever, mild pharyngeal redness, no tonsillar exudates, no swollen neck glands -- that are much more suggestive of a viral pharyngitis. It then becomes very difficult for the treating physician to know whether the patient has a "strep throat" infection or a simple viral infection.

How does a physician clinically differentiate a "strep throat" pharyngitis from a viral pharyngitis?

Consider the situation of a primary care physician in clinical practice -- approximately 5-15% of his adult sore throat patients have pharyngitis due to a GABHS infection, while the remainder have a viral pharyngitis which doesn't require antibiotic treatment. If the primary care physician could accuratedly identify those "strep throat" patients, he would only have to prescribe antibiotics to 5-15% of all his adult pharyngitis patients. As previously mentioned, many physicians do not make a serious attempt to differentiate a GABHS pharyngitis from a viral pharyngitis infection, and those physicians indiscriminately prescribe antibiotics to 75% (or more) of their pharyngitis patients. From my perspective, that represents poor clinical judgement and bad medical practice.

Is it possible for an experienced physician to accuratedly identify "strep throat" patients if he used his clinical skills to the maximum possible degree?

When experienced physicians participated in a number of research studies to determine whether they could clinically diagnose a "strep throat" infection, those scientific studies demonstrated that they were only about 50-60% accurate in their clinical ability to diagnose a GABHS pharyngitis infection. Many primary care physicians are reluctant to acknowledge the fact that an experienced clinician has a limited ability to differentiate a GABHS pharyngitis from a viral pharyngitis . In fact, some over-confident physicians believe that they can always differentiate a "strep throat" pharyngitis from a viral pharyngitis -- despite the existence of considerable scientific evidence to the contrary. Other physicians, who accept the validity of the scientific evidence from multiple research studies, have attempted to develop clinical prediction rules that can more accuratedly predict whether a patient with a sore throat has a "strep throat" infection. There are many clinical prediction rules in current use, and the most favored clinical prediction rule is probably the Centor clinical prediction rule. Here is how it works.

Centor clinical prediction rule:

Assign 1 point for each of the following clinical characteristics:-

- history of fever
- anterior cervical lymphadenopathy (swollen glands alongside the front of the neck)
- exudates on the tonsils
- absence of a cough

Total Centor score varies between 0-4.

If one considers an adult population of sore throat patients that have a prevalence rate of "strep throat" infections of 5-15%, then a particular patient's likelihood of having a GABHS infection depends on his individual Centor score.

As a rough approximation, the following figures apply to adult patients with pharyngitis.

Centor score 0-1 => <5% likelihood of having a GABHS infection.
Centor score 2-3 => 5-30% likelihood of having a GABHS infection.
Centor score 4 => 30-60% likelihood of having a GABHS infection.

How does the Centor clinical prediction rule help a physician in clinical practice?

The Centor clinical prediction rule is most useful in ruling-out the possibility of a "strep throat" infection if the Centor score is very low (0-1). Patients with a Centor score of 0-1 have a <5% likelihood of having a GABHS infection and it is very reasonable not to use antibiotic therapy if the likelihood of a "strep throat" infection is that small. The same situation applies to patients who have prominent "common cold" symptoms (runny or congested nose, sinus pressure, mild cough) in addition to their sore throat symptoms. Those patients also have a very low likelihood of having a GABHS infection and they do not require antibiotic therapy.

Patients with a Centor score of 2-3 have an intermediate likelihood (5-30% likelihood) of having a GABHS pharyngitis infection. A 5-30% likelihood of GABHS pharyngitis infection is too low to warrant routine antibiotic treatment, because 60-95% of those patients would receive antibiotics unnecessarily. On the other hand, a recommendation to avoid antibiotics in all those patients is also not rational, because 5-30% of those patients theoretically have GABHS pharyngitis, and they would be deprived of effective treatment if not diagnosed, and treated. Therefore, the clinical prediction rule is not really useful in dealing with patients who have an intermediate likelihood of a GABHS pharyngitis infection.

A patient with a Centor score of 4 has a likelihood of a GABHS pharyngitis infection of approximately 30-60%. Should a patient with a Centor score of 4 receive empirical antibiotic therapy without undergoing further diagnostic testing?  If all sore throat patients with a Centor score of 4 are empirically treated with antibiotics, then more than 40% of those sore throat patients would receive antibiotics unnecessarily. Is it acceptable to over-treat sore throat patients to that degree just to ensure that all suspected "strep throat" patients are treated? The disadvantage of treating patients unnecessarily with antibiotics is that they are exposed to the expense and risks of therapy with no prospect of benefit. Risks include drug allergy reactions and drug side-effects. Another great disadvantage of unnecessary antibiotic use is that resident bacteria throughout the community become increasingly resistant to antibiotics.

I think that the Centor clinical prediction rule (like many other clinical prediction rules) in a sore throat patient population with a low prevalence of GABHS pharyngitis infection (5-15% prevalence) is really only useful if the patient has a low Centor score, because patients with a low score do not have a high enough likelihood of GABHS pharyngitis to warrant routine antibiotic therapy. By contrast, I think that patients with an intermediate or high Centor score may need further diagnostic testing to determine whether they have a sufficiently high likelihood of having a GABHS pharyngitis infection to warrant antibiotic therapy. I will discuss this issue at greater length in the "Recommended approach for diagnosing, and treating, a "strep throat" infection" section.

Why is it important to diagnose and treat a "strep throat" infection?

What would happen to "strep throat" patients if they did not receive antibiotics?

Most untreated patients would get significantly better in 3-4 days and considerably better in 8-10 days, and only a very small proportion of patients (~2%) would develop serious suppurative complications such as a peri-tonsillar abscess. During the active phase of the illness (and for 1 week after the acute illness) the patients would be contagious and they could easily spread the disease to another person via close contact (spread by oral secretions and respiratory droplets). A small number of "strep throat" patients would develop delayed complications such as rheumatic fever or post-streptococcal glomerulonephritis (a self-limited kidney disease). Rheumatic fever is the most serious consequence of an untreated GABHS pharyngitis infection, and a certain percentage of rheumatic fever patients develop carditis (inflammation of the heart) and serious damage to their heart valves. Therefore, it not surprising that physicians would consider prescribing antibiotics for all GABHS pharyngitis infections to prevent rheumatic fever, prevent acute glomerulonephritis, prevent suppurative complications, decrease contagion, and relieve the acute symptoms of the acute pharyngeal infection.

However, consider the matter more closely.

In the past, rheumatic fever was the most feared complication of untreated GABHS pharyngitis, and all physicians wanted to diagnose, and treat, every case of GABHS pharyngitis to avoid the possibility of rheumatic fever. Antibiotic treatment is very effective in preventing rheumatic fever if it is administered within 9 days of onset of the acute GABHS pharyngitis infection.  The imperative to diagnose every case of  GABHS pharyngitis has markedly diminished in recent decades in the USA because rheumatic fever has become very rare (occurring <60x as frequently today compared to its incidence in the 1970's). In fact, rheumatic fever is now so rare in the USA that many physicians think that it is not essential to diagnose and treat every case of GABHS pharyngitis to prevent that complication.

With respect to preventing acute post-streptococcal glomerulonephritis, there is no evidence that antibiotic therapy prevents this immune disease.

The risk of a serious suppurative complication, such as a peri-tonsillar abscess, is so small (2%) that many physicians think that it is not imperative to diagnose and treat every case of GABHS pharyngitis to attempt to avoid that complication. In fact, 50% of the cases of peri-tonsillar abscess present with this complication at the time of their first presentation, and those cases are not due to the fact that they were previously seen by a primary care physician who did not treat their acute phayngitis.

Therefore, the main advantage of treating a patient with GABHS pharyngitis is to relieve the symptoms. It has been shown that antibiotic therapy relieves the symptoms of acute pharyngitis 1-2 days sooner if it is given early in the disease course, and treated patients show significant improvement in 1-2 days instead of 3-4 days (as occurs in untreated GABHS pharyngitis). Treated patients are also cleared of the risk of spreading their infection after 24 hours of antibiotic therapy and they can therefore get back to work or school sooner.

In summary, these are the two major benefits of early antibiotic therapy for GABHS pharyngitis:- a 1-2 day earlier resolution of the patient's symptoms and a diminished likelihood of spreading the infection to other people.

Diagnostic tests to diagnose a "strep throat" infection

The most accurate means of diagnosing a GABHS pharyngitis is by means of a throat culture.

Throat culture

A throat culture is performed by first swabbing the patient's tonsils and posterior pharynx (back of the throat) with the cotton tip of a culture swab and then swabbing the stick onto a culture plate. The culture plate is incubated at the optimum temperature for 48 hours and the diagnosis of GABHS is made if the GABHS organism grows on the plate. It usually takes 18-48 hours before a definitive diagnosis can be made. The diagnostic test is 90-95% accurate in diagnosing the presence of GABHS bacteria. That means that the diagnostic test misses 5-10% of true-positive cases, and those missed cases are then labelled false-negative cases. False-negative cases occur for a variety of reasons eg. poor swabbing technique, poor plating technique, delay in transferring the swab to a laboratory for plating and incubation, inadequate incubation conditions, and technical laboratory errors in reading the plate.

Although a throat culture is regarded as the "criterion standard" diagnostic test in diagnosing a GABHS infection, it has the major disadvantage that a positive diagnosis of "strep throat" cannot be made at the initial point-of-care visit, and there is a 18-48 hour time period before a definitive diagnosis can be established. Many patients are unwilling to wait for that period of time before they receive antibiotics for a positive strep throat culture result. Also, a 18-48 hour delay in receiving antibiotic therapy means that the patient loses some of the benefit of early therapy (1-2 day earlier resolution of symptoms) and also remains contagious during the waiting period. Some physicians, who routinely use throat culture testing, therefore elect to treat the patient with antibiotic therapy if the clinical suspicion of a "strep throat" infection is reasonably high or if the patient is very symptomatic, and they subsequently instruct the patient to stop the antibiotics if the throat culture results come back negative. However, clinical studies have demonstrated that many patients continue to take the prescribed antibiotics despite a negative throat culture result. This injudicious pattern of behaviour obviates the benefit of performing a throat culture and results in substantial overuse of antibiotics with its attendant risks.

An alternative diagnostic approach is to use a rapid strep detection test (RSDT).

Rapid strep detection test (RSDT)

There are a number of RSDTs on the market and they allow the physician to obtain results within 10-30 minutes of obtaining a swab of the throat. That means that the patient can obtain a positive or negative test result at the time of the initial office visit -- which is a great advantage.

RSDTs are not as accurate as a throat culture, and the sensitivity of commercially available RSDT tests varies between 60-90% depending on the commercial product. A test sensitivity of 60-90% means that the test accuratedly diagnoses 60-90% of GABHS pharyngitis infections, and misses 10-30% of GABHS pharyngitis infections (10-30% false-negative rate). If the false-negative rate is really as high as 30%, then the RSDT test would be of little use to a clinician because it would miss too many cases of GABHS pharyngitis infection.

More recent RSDT tests using an optical immunoassay or chemiluminescent DNA probe technique have a better sensitivity (90%), and under optimal conditions they only miss 10% of GABHS pharyngitis cases (resulting in a false-negative rate of 10%). A RSDT test with a guaranteed sensitivity of 90% would be very useful in clinical practice, because its sensitivity would approximate the sensitivity of a throat culture in diagnosing a GABHS infection -- which means that very few GABHS pharyngitis infections would be missed.

Recommended approach for diagnosing, and treating, a "strep throat" infection.

Consider the different clinical approaches that physicians adopt with respect to their adult sore throat patients.

The most frequently used approach in the USA is the "clinical criteria and/or clinical prediction rule" approach. According to that approach, the treating physician usually prescribes antibiotics if he strongly suspects that the patient has GABHS pharyngitis based on his clinical evaluation, which may/may not include the use of a clinical prediction rule. As previously mentioned, this approach only has a 50-60% accuracy rate in predicting the presence of a "strep throat" if the treating physician strongly suspects that the patient has GABHS pharyngitis. By contrast, the "clinical criteria and/or clinical prediction rule" approach is very useful in excluding a GABHS pharyngitis if  the patient has a low clinical prediction score (Centor score of 0-1) or overt signs of a viral upper respiratory infection (runny or congested nose, cough, hoarseness, sinus pressure, post-nasal drip, conjunctivitis), because the likelihood of a GABHS infection is very low (<5%) under those circumstances. Patients with an intermediate likelihood of GABHS pharyngitis, as predicted by a clinical prediction rule and/or the treating physician's clinical "gestalt", could end up being treated with antibiotics if the treating physician is risk-averse, or not treated with antibiotics if the treating physician is risk-tolerant. Most physicians who use this empirical approach do not have a precise formula for dealing with patients who have an intermediate likelihood of GABHS pharyngitis, and they would probably have difficulty explaining the logical rationale underlying their clinical judgments.

Another approach that is less frequently used in the USA is the routine use of throat cultures for the majority of sore throat patients with clinical evidence of pharyngitis. That approach would miss very few GABHS pharyngitis infections, and it would avoid the unnecessary antibiotic treatment of non-GABHS pharyngitis. However, the major disadvantage of that approach is the fact that test results are not available at the time of the patient's visit. Most patients want immediate antibiotic treatment if it is clinically warranted, and they don't want to wait 18-48 hours for a throat culture result. Another major disadvantage is the expense -- many throat cultures would have to be performed if the community prevalence of GABHS pharyngitis infection is very low (prevalence of 5-15%), and the cost of having to perform many negative throat cultures would not represent cost-effective medical practice. Many professional organizations have therefore recommended a RSDT as an alternative diagnostic test that offers point-of-care results at the time of the initial patient visit. Although a RSDT diagnostic test performed in a physician's office is not as accurate as a throat culture, it will only miss a small percentage of GABHS pharyngitis infections. In fact, in an adult sore throat patient population with a prevalence of GABHS pharyngitis of 5-15%, the routine use of rapid strep testing would only miss <4% of GABHS pharyngitis infections if the RSDT diagnostic test's sensitivity is at least 80%. (See the mathematical calculations in the box below if you want to see how I calculated those figures).
 

If a RSDT diagnostic test is performed in a population of pharyngitis patients with an average prevalence of GABHS pharyngitis of 5-15%, then the ability of the RSDT diagnostic test to accuratedly exclude a GABHS pharyngitis infection depends on the test's sensitivity and specificity. If the RSDT diagnostic test has a test sensitivity of 80% and a test specificity of 95%, then the negative predictive value of a negative RSDT result will be between 96.4 - 98.9%. The negative predictive value (NPV) of a test is defined as the likelihood of a patient not having a disease if the test result is negative. By contrast, 100-NPV = Percentage of patients, who have a negative test result, but actually have the disease. Therefore, one can reasonably conclude that between 1.1-3.6% of patients, who have a negative RSDT result, will have a GABHS pharyngitis infection if the community prevalence of GABHS pharyngitis infection is between 5-15%. Those incorrectly diagnosed patients are labelled false-negative cases. 

Missing less than 4% of GABHS pharyngitis cases (false-negatives) in adults is not that serious, considering that the major effect of not treating a GABHS infection in adults is a slower recovery rate. Many physicians would therefore agree that when the community prevalence of GABHS pharyngitis is in the range of 5-15%, that the routine use of RSDT diagnostic testing for all adult pharyngitis patients (and selective antibiotic use only for RSDT-positive patients) is better than the alternative approach of empirically basing antibiotic treatment decisions on a "clinical  criteria and/or clinical prediction rule" approach.

Is there a cross-over point where empiric antibiotic treatment for clinically suspected GABHS pharyngitis infections -- rather than routine RSDT diagnostic testing -- would be the better approach?

This is a complex issue, that many laypeople will have difficulty understanding. However, even if you, as a layperson, cannot understand the following mathematical analysis, you will know much more than the average physician if you acquire an elementary understanding of the following analytical presentation.

The following table demonstrates how the positive predictive value (PPV) and negative predictive value (NPV) of a diagnostic test varies with changes in the prevalence of a disease (eg. GABHS pharyngitis).

The definition of the PPV of GABHS is the likelihood of a patient having a GABHS pharyngitis if a particular clinical prediction score is present, or if a particular diagnostic test is positive.

The definition of the NPV of GABHS is the likelihood of a patient not having a GABHS pharyngitis if a diagnostic test for GABHS pharyngitis is negative eg. RSDT diagnostic test for GABHS pharyngitis.

The prevalence of GABHS is the same as the pre-test probability of GABHS; they should be regarded as equivalent terms.
 

Prevalence of GABHS pharyngitis PPV of GABHS for a Centor score of 0 PPV of GABHS for a Centor score of 2 PPV of GABHS for a Centor score of 4 NPV of a negative RSDT test (80% sensitivity)
10% 2% 8% 41% 98%
20% 3% 16% 61% 95%
40% 10% 33% 81% 88%
50% 14% 43% 86% 83%

First of all, note that the PPV of GABHS (based on any particular clinical prediction score) increases as the prevalence of GABHS pharyngitis increases, while the NPV of a negative RSDT decreases as the prevalence of GABHS pharyngitis increases. It is too difficult to explain why this phenomenon occurs, but the clinical implications of this phenomenon are extremely important in medical practice.

Take the usual adult community practice situation where a physician is dealing with a sore throat patient population that has a 10% prevalence of GABHS pharyngitis. Note from the table, that if an individual patient has a Centor score of 0, that he only has a 2% likelihood of having a GABHS pharyngitis. The likelihood of a GABHS infection is so low that it is not rational sense to test, or treat, that particular patient. If one decided to order a RSDT diagnostic test on that patient, the NPV of the RSDT diagnostic test would be 98%, which means that there would be a 2% likelihood of a false-negative test result (negative RSDT result, but GABHS present). Therefore, after a negative RSDT test result, the patient still has a 2% likelihood of having a GABHS pharyngitis, which is no different from his estimated clinical likelihood of GABHS pharyngitis of 2% (based on the Centor clinical prediction rule). It should therefore be obvious that diagnostic testing offers no benefit in that situation. It should also be obvious that antibiotic treatment is not warranted if the patient only has a 2% likelihood of having a GABHS pharyngitis infection.

Consider another patient in a population of sore throat patients that has an average prevalence of GABHS pharyngitis of 10%. If that particular patient has overt signs of a GABHS pharyngitis and a Centor score of 4, then his estimated likelihood of a GABHS infection is 41% (from the table). If he is treated with antibiotics, then he has a 60% likelihood of being treated unnecessarily because he has a 60% chance of not having a GABHS pharyngitis infection. What should a physician do in that situation? If a physician performed a RSDT test on that particular patient (whose pre-test probability of a GABHS infection after clinical evaluation is 41%) and the RSDT result is positive, then the patient very likely has GABHS pharyngitis and he should definitely be treated with antibiotics. If the RSDT result is negative, one will note from the table that the estimated NPV of a negative test is 88% if the patient's pre-test probability (prevalence) of GABHS pharyngitis is 40%. That means that the patient's post-test probability of having GABHS pharyngitis after a negative test is 12% (100-NPV=Post-test probability of GABHS=100-88). What should his treating physician do if the RSDT result is negative? If he treats the patient with antibiotics, then he should anticipate a 88% likelihood of treating the patient unnecessarily because the patient has a 88% chance of not having a GABHS pharyngitis infection. However, if he chooses not to treat the patient with antibiotics, then he must accept a small chance of complications resulting from an untreated GABHS pharyngitis infection, because there is a 12% chance of the patient having GABHS pharyngitis, which was missed by the RSDT test (12% false-negative rate). The best choice in this scenario doesn't depend solely on the treating physician's clinical judgement, because an intelligent patient can discuss this complex situation with his doctor, and also participate in the clinical decision-making process.

Consider another scenario. Imagine that there is local epidemic of "strep throat" in a closed community, such as a boy's boarding school, and that the "average" prevalence of GAHBS pharyngitis during the local outbreak of "strep throat" is 50%. If a particular teenage boy develops pharyngitis, should he be treated empirically with antibiotics or should he first have a RSDT diagnostic test performed? If the physician's clinical suspicion of a GABHS pharyngitis is very high, because the patient has a Centor score of 4, then one can estimate from the table that the patient has a 86% likelihood of a GABHS pharyngitis infection. Would it not be rational for the treating physician to forgo diagnostic testing if the patient's likelihood of having a GABHS infection is that high -- especially if one considers that the NPV of a negative RSDT result is 83%, which means that a negative RSDT result  has a 17% chance of missing a GABHS pharyngitis infection (17% false-negative rate). By contrast, if a particular teenage patient had an intermediate Centor score of 2, then RSDT testing may be prudent, because the patient only has an estimated pre-test probability of GABHS pharyngitis of 43% after taking his clinical Centor score into account (see the table).

Consider a different scenario involving a pharyngitis patient who presents with scarlet fever. In this example, if a sore throat patient has clinical evidence of pharyngitis and a typical scarlatiniform rash -- the rash consists of fine red, raised spots beginning on the trunk and spreading to the extremities but sparing the palms and soles; the red rash blanches to pressure and has a sandpapery feel; the rash is more prominent in the skin creases and this is most notable in the antecubital area in front of the elbows --- the physician would presume that the patient is infected with a GABHS bacterial strain that is causing scarlet fever. The likelihood of a GABHS infection, based on the clinical assessment, would be >80%. Should the patient be empirically treated with antibiotics or should he first have a RSDT diagnostic test performed? The answer cannot be derived from that table, because I didn't supply numbers for a prevalence of GABHS pharyngitis of 80%. Would it be easier to answer that question, if I informed you that the NPV of a negative RSDT result is 54% if the prevalence (pre-test probability) of GABHS pharyngitis is 80%? That means that the patient still has a 46% chance of having a GABHS pharyngitis infection if the RSDT result is negative. So, why should the treating physician perform the test? It makes more sense for the patient's physician to treat the patient empirically with antibiotics, because there is only a small chance (<20%) that the patient doesn't have a GABHS pharyngitis infection. A positive RSDT result doesn't change the situation because the patient would still need to be treated with antibiotics, while a negative RSDT result doesn't really exclude the presence of a GABHS pharyngitis infection, because the test has too high a false-negative rate in that situation.

A similar situation would prevail if an individual pharyngitis patient has been in close contact with a known "strep throat" patient. That particular pharyngitis patient has a very high likelihood of having a GABHS pharyngitis infection, and the treating physician may appropriately choose to treat the patient on the basis of his clinical assessment (and/or clinical prediction rule) rather than depend on RSDT testing, which is associated with too high a rate of false-negative results.

If you find the previous analysis too difficult to understand, try and remember the following general rule: Routine RSDT testing is recommended for the majority of acute pharyngitis patients to establish whether they have a GABHS pharyngitis infection requiring antibiotic therapy. However, if a treating physician's clinical estimation of the likelihood of GABHS pharyngitis increases significantly (equivalent to an increase in the prevalence, or pre-test probability, of GABHS pharyngitis) because of the presence of additional epidemiological or clinical evidence, then empiric antibiotic therapy based on a clinical prediction rule may often become the better decision-choice, rather than routine RSDT diagnostic testing in all pharyngitis patients. This general rule is especially applicable if the RSDT diagnostic test has a relatively low sensitivity (<80-90% sensitivity), because the false-negative rate becomes proportionately greater as the RSDT's sensitivity decreases.

Are there special circumstances when a throat culture is indicated in adults -- if RSDT diagnostic testing is available? I think that a throat culture is indicated if the only available RSDT test has a very low test sensitivity  potentially resulting in too many "missed" GABHS cases; if there is an epidemic of rheumatic fever in the local community or if there is a high endemic incidence of rheumatic fever in the local area; or if the treating physician is participating in epidemiological studies of disease prevalence or studies of antibiotic resistance.

What about a pediatric population of sore throat patients?

The American Academy of Pediatrics (ACP) recommends routine RSDT diagnostic testing in children with pharyngitis. However, the ACP also recommends that all negative RSDT results be followed-up with a throat culture, because the throat culture has a higher test sensitivity and it will miss fewer GABHS pharyngitis infections. This is a very conservative recommendation that is based on a fear of rheumatic fever occurring in untreated GABHS pharyngitis patients. This fear is appropriate if rheumatic fever is endemic in the community or if there is any evidence of a local outbreak of rheumatic fever in the area. However, if the local incidence of rheumatic fever is very low, then the small gain in sensitivity may not justify the cost of having to order so many throat culture tests. For instance, if the chance of developing rheumatic fever from untreated GABHS pharyngitis is 1:3,000, and the prevalence of GABHS in a population of school-aged children is 20% (average prevalence is usually between 15-30%), and the sensitivity of the RSDT and throat culture tests are 80% and 95% respectively, then it would require 63,000 - 78,000 throat cultures to prevent a single case of rheumatic fever. (See the mathematical calculations in the box below if you want to see how I calculated those figures). By contrast, throat cultures are obviously much more cost-effective, and may be highly appropriate, if the risk of rheumatic fever is much higher (eg. 0.5-3% of untreated GABHS pharyngitis infections, which was the reported incidence of ARF during the 1940-1970's).
 

There are two methods of thinking through this problem.

Method number 1:

If 1,000 children present with pharyngitis and the prevalence of GABHS pharyngitis is 20%, then there will be 200 children with GABHS pharyngitis. If the RSDT has a test sensitivity of 80% and a test specificity of 95%, then the RSDT will detect 160 of those 200 GABHS cases and miss 40 cases (false-negative cases) -- presuming that the RSDT test is performed on all 1,000 patients. There will also be 760 true negative cases out of those 1,000 cases. That means that there will be a total of 800 negative RSDT tests. If 40 of those 800 cases are false-negatives (GABHS pharyngitis positive, but RSDT negative) then the prevalence of GABHS in that sample of 800 cases is 40/800 = 5%. If a throat culture (sensitivity 95% specificity 95%) is subsequently performed on those 800 cases, the throat culture will theoretically detect 38 out of those 40 false-negative cases. Therefore, it takes 800/38 = 21 throat cultures to detect one false-negative RSDT case. If only 1 out of every 3,000 untreated GABHS pharyngitis patients develops rheumatic fever, then it would require 21 x 3,000 = 63,000 throat cultures to prevent one case of rheumatic fever. The results could even be higher than 63,000 because the calculation is based on the assumption that the throat culture test will have a sensitivity of 95% (which normally only applies to an "average" study population of patients with an overall prevalence of GABHS of 5-30%). However, it is very likely that the throat culture will have a much lower test sensitivity in those RSDT false-negative cases because of the issue of spectrum bias -- a diagnostic test's sensitivity is much lower when the likelihood of significant disease is lower. It's quite possible that the "average" false-negative RSDT patient could have a much smaller number of bacterial organisms in their pharynx as a consequence of a spectrum bias issue.

Method number 2:

Let's presume that throat culture swabs were taken at the same time as the RSDT swabs and temporarily stored. If the RSDT test comes back negative, then the throat culture swabs of the RSDT negative patients would be sent for culture and processing. If the RSDT has a test sensitivity of 80% and a test specificity of 95%, then the RSDT will detect 160 of those 200 GABHS cases and miss 40 cases (false-negative cases) -- presuming that the RSDT test is performed on all 1,000 patients. There will also be 760 true negative cases out of those 1,000 cases. That means that there will be a total of 800 negative RSDT tests. That implies that 800 throat cultures would have to be sent off to the laboratory for culturing and laboratory analysis. How many of those throat culture results would be false-negative? Presuming that a throat culture has a sensitivity of 95% and a specificity of 95%, and the prevalence of GABHS pharyngitis is 20%, then there would be approximately 10 false-negative results if testing was done on all 1,000 patients. (* Because formal testing was actually only done on 800 patients, the number of false-negative throat culture results could even be lower than 10, and could be as low as 8). The false-negative result of 10 is 30 less than the number of RSDT false-negative results (total of 40). One can therefore presume that approximately 30 false-negative RSDT cases would be picked up by the throat culture test and that approximately 10 false-negative RSDT cases would still be missed. Therefore, it takes 800/30 = 26 throat cultures to detect one false-negative RSDT case. If only 1 out of every 3,000 untreated GABHS pharyngitis patients develops rheumatic fever, then it would require 26 x 3,000 = 78,000 throat cultures to prevent one case of rheumatic fever.

Note that GABHS infection is very rare in children < 3 years of age, and be aware that those infected infants often present with crusting of the nares and coryza rather than a red inflammed throat. Also, rheumatic fever is extremely rare in children <3 years, so testing for the presence of GABHS is not warranted in such young children.

Children with manifestations highly suggestive of viral infection (runny nose, conjunctivitis, hoarseness, cough), or anterior stomatitis (inflammation of the tongue and mouth and not throat), or discrete ulcerative lesions in the mouth and pharnyx, or diarrhea, are unlikely to have GABHS as the cause of their pharyngitis and should not be tested for GABHS.

A small percentage of children (3-5%) chronically harbor GABHS bacteria in their throat and they are regarded as chronic carriers of GABHS. If they develop an intercurrent viral pharyngitis infection, the throat culture test may be positive. This represents a false-positive result because the resident streptococcal bacteria are not really causing the pharyngitis. There is no easy  method of differentiating true-positive throat culture results from false-positive throat culture results during the acute illness. True-positive patients usually develop anti-streptococcal antibodies to the GABHS bacteria, and the presence of those anti-streptococcal antibodies can only be detected by anti-streptococcal antibody testing 1-2 weeks after the onset of the acute illness. Chronic carriers of GABHS (false-positive patients) do not develop an anti-streptococcal antibody response. Because there is no simple means of differentiating false-positive from true-positive throat culture results during the early stage of the illness, it is appropriate for a physician to treat all patients with a positive throat culture result, and simply accept a small degree of antibiotic overtreatment.

What is the recommended antibiotic for "strep throat" infections?

Because this guidemap is targeted toward a layperson audience, I have decided not to supply precise details about antibiotic therapy and I will merely delineate some general principles.

Considerable detail regarding antibiotic therapy can be found in the following emedicine.com chapters.

Pharyngitis, bacterial
Pharyngitis

The drug-of-choice for the treatment of "strep throat" is penicillin. It has been used for many decades and it remains the drug-of-choice because it is cheap, safe and highly effective. There is no evidence that GABHS bacteria have developed resistance to the drug.

Penicillin can be taken by mouth or administered by intramuscular injection as a single one-time dose. Intramuscular benzathine penicillin G is particularly useful in situations where the treating physician is unsure about the patient's ability to purchase the oral prescription, or if there is uncertainty that the patient will take a complete course of oral penicillin therapy. A complete course of oral penicillin V therapy lasts 10 days. Although a patient may feel much better in 3-4 days, it is imperative that the patient takes the drug for 10 days to prevent rheumatic fever and decrease the risk of a relapse. Penicillin V is usually taken 4x per day. However, it is well known that patient compliance is much better if a drug has to be taken less frequently, and penicillin V taken twice per day is also effective. Pediatricians often prescribe amoxicillin for children because the oral formulation tastes better than the oral formulation of penicillin V and it is also effective. Amoxicillin can be taken once per day, but it remains uncertain whether once-per-day dosing is as effective as standard therapy. Amoxicillin has the disadvantage that it nearly always produces a generalised itchy rash if inadvertently administered to a patient with infectious mononucleosis, which often presents in a similar fashion to "strep throat" pharyngitis.

The drug-of-choice for patients who are allergic to penicillin is erythromicin. Erythromicin is the original drug of a class of antibiotics called macrolides. There are a number of newer generation macrolides that have better gastro-intestinal tolerance than erythromicin, and they have the additional advantage that they can be taken once-or-twice per day. However, they are very expensive and they should not be used as the primary drug-of-choice in patients who are not allergic to penicillin. Over-prescribing of these macrolide antibiotics results in the emergence of bacterial resistance to the drug, and this became a major problem in a number of countries, especially Japan and Finland. The problem in Finland was partially reversed as a result of a vigorous public health campaign that resulted in a remarkable decrease in the over-prescribing of erythromicin antibiotics and a significant decrease in the number of erythromicin-resistant bacteria. At the present time, there is little evidence of erythromicin-resistance in the USA (<5% incidence).

Another class of antibiotics that is used to treat "strep throat" are the cephalosporins. There are many drugs in this class, usually starting with a "c" eg. cephalexin, cefadroxil, cefpodoxime, cefditoren. They are slightly more potent than penicillin and some of them are used very effectively in shorter 5-day courses. They are much more expensive than penicillin V and they should only be prescribed if patients develop a relapse after taking penicillin V, or if patients develop recurrent episodes of "strep throat". Cephalosporins should not be used in patients who have a history of an immediate anaphylactic-type allergic reaction to penicillin. If cephalosporins are prescribed, narrow spectrum (first generation) agents are preferred, rather than the more expensive newer-generation drugs.

In addition to antibiotics, home remedies are appropriate. Recommended home remedies include:-

As a general principle, there is no reason to have a post-treatment throat culture after completing a 10-day course of antibiotics to ensure that therapy was effective. Clinical improvement is the primary basis of deciding that drug therapy is successful. A patient should consult a physician if he fails to improve within 72 hours or if his condition worsens. Clinical symptoms that would be of major concern would be an inability to swallow liquids and drooling of saliva, any difficulty breathing especially if it is associated with hoarseness or noisy respiration, trismus (spasm of the jaw muscles and consequent inability to fully open the mouth), marked swelling of the neck (bull neck) or a relentless fever.

Post-treatment throat cultures 2 to 7 days after completion of therapy are only indicated in patients who remain symptomatic, who develop recurring symptoms, or who have had rheumatic fever and are at unusually high risk for recurrence.

Patients who have a recurrence of GABHS pharyngitis shortly after completing a 10-day course of a recommended oral antibiotic can be retreated with that same antibiotic; given an alternative oral antibiotic; or given an intramuscular dose of benzathine penicillin G, especially if inadequate compliance with oral therapy is likely.

Recurrent streptococcal pharyngitis (multiple episodes of GABHS pharyngitis within a 1-2 year time period) often requires special antibiotic therapy and referral to a throat specialist for an expert opinion regarding the potential benefit of elective tonsillectomy. It is extremely difficult to be certain that the recurrent episodes of pharyngitis are actually due to active infection with GABHS. The true situation could be that the patient is a carrier of GABHS and all-or-most of the recurrent pharyngitis infections are viral in origin.

Antibiotics are not indicated for the treatment of GABHS carriers (which is defined as the presence of a positive throat culture for GABHS in the absence of clinical pharyngitis or the development of an antibodies to GABHS antigens). However, according to the American Academy of Pediatrics Red Book "there are specific situations in which eradication of carriage may be indicated, and include the following: (1) during an outbreak of acute rheumatic fever or poststreptococcal glomerulonephritis, (2) during an outbreak of GABHS pharyngitis in a closed or semiclosed community, (3) when a family history of rheumatic fever exists, (4) when multiple episodes of documented symptomatic GABHS pharyngitis continue to occur within a family during a period of many weeks despite appropriate therapy, (5) when a family has excessive anxiety about GABHS infections, (6) when tonsillectomy is considered only because of chronic GABHS carriage, and (7) when a case of GABHS toxic shock syndrome or necrotizing fasciitis has occurred in a household contact.

Streptococcal carriage can be difficult to eradicate with conventional antibiotic therapy. A number of antimicrobial agents, including clindamycin, amoxicillin-clavulanate, and a combination of rifampin and either penicillin V or benzathine penicillin G, have been demonstrated to be more effective than penicillin in eliminating chronic streptococcal carriage."

What are the other common causes of a pharyngitis?

There are a large number of bacteria and viruses that can cause pharyngitis, and I have provided a limited amount of information on a select number of them. I have deliberatedly avoided describing all the bacterial and viral causes of pharyngitis because it would be too confusing for laypeople. I have focused on the more important causes of pharyngitis that should be considered by a diligent physician.

Diphtheria

The causal agent of diphtheria is the Corynebacterium diphtheriae bacteria and it is the most serious cause of pharyngitis because it often causes death if unrecognized and untreated.

Diphtheria is extremely rare in the USA because virtually all US children are immunized against the bacteria. It now mainly occurs in non-immunized or poorly immunized members of socioeconomically disadvantaged families.

Most US-based physicians have never seen a case of diphtheria pharyngitis and they may not immediately recognize its presence. Diphtheria pharyngitis is characterized by the presence of an extensive greyish-brown pseudomembrane that covers the tonsils and neighbouring soft palate. The pseudomembrane may also cover the entire pharnyx and extend down the throat to the larnyx and tracheobronchial airtubes. Because of the tight adherence of the pseudomembrane to the underlying soft tissue, bleeding will result if one attempts to scrape the pseudomembrane off the tonsils with a tongue blade. Patients often develop a "bull-neck" appearance due to extensive edema (swelling) of the soft tissues of the neck and due to marked enlargement of the neck lymph glands. The diphtheria bacteria produce powerful toxins that affect the heart and nervous system. Patients with diphtheria often appear critically ill.

If a physician suspected diphtheria in a patient with pharyngitis, he would immediately consult an infectious disease expert and arrange for immediate hospital admission to an isolation bed. All individuals who have been in close contact with a diphtheria patient need to be treated with antibiotics and diphtheria toxoid.

For detailed information on diphtheria, consult the following emedicine.com chapter

Diphtheria

Infectious mononucleosis

The Ebstein Barr virus is the cause of infectious mononucleosis (often called "mono").

Mono pharyngitis often appears similar in appearance to a severe case of "strep throat", and the diagnosis is often only recognized after a delay of 7-10 days when the patient, who was treated empirically for a "strep throat" infection, doesn't respond to the antibiotics. One week after empirically taking antibiotics for a presumed "strep throat", the patient will still have extensive evidence of a severe tonsillo-pharyngitis, because the Ebstein Barr virus doesn't respond to antibiotic therapy. It may only be at that point in time that the treating physician will seriously consider the diagnosis of Mono. Some patients (up to 30%) with Mono pharyngitis may even test positive for GABHS (positive RSDT and/or positive throat culture) due to the chronic presence of carrier GABHS bacteria in the throat. The presence of a positive GABHS test therefore doesn't exclude Mono, and the diagnosis of Mono should be actively pursued if the clinical picture is very suggestive of Mono.

Mono pharyngitis occurs most commonly in teenagers and young adults between the ages of 15-24 years. It is most commonly spread via saliva, and college students often refer to it as "kissing disease". Because the incubation period of Mono is 1-2 months, patients may not recall the potential source of the infection. The patient may develop a prodromal illness consisting of high fever, chills, sweats, and malaise before developing pharyngitis. Alternatively, the patient may complain of marked fatigue and malaise for weeks before developing overt signs of pharyngitis. The pharyngitis is often severe with marked redness of the tonsils and pharynx. Extensive yellow exudates may cover the tonsils, which also become markedly enlarged. The term "kissing tonsils" is used to describe tonsils that are so enlarged that they touch each other in the midline of the throat. The lymph glands in the neck are often markedly enlarged, and the enlargement is bilateral (both sides) and symmetrical. In fact, bilateral posterior cervical lymphadenopathy (enlarged glands on both sides of the back of the neck) is highly suggestive of Mono. Generalized lymphadenopathy (generalized swelling of the lymph glands) may occur, and the lymph glands in the axillae and groin may also become enlarged. 50% of patients develop enlargement of their spleen, usually later in the illness. The liver becomes enlarged in 15% of patients and jaundice may occur. A generalized itchy rash will occur in nearly all patients treated with amoxacillin antibiotic (an important clue to the presence of Mono) and the rash tends to be persistent. This is in sharp contrast to the transient non-itchy rash that can occasionally occur early in the disease course of Mono. Because Mono is a systemic disease it may involve many different body systems (eg. nervous system, liver and kidneys). Fatigue, which may be profound, is a common feature of Mono and it may last many months.

Click here to see a photograph of acute pharyngitis due to Mono.

The diagnosis of Mono can be confirmed by blood tests -- such as the Monospot blood test. Mono also produces a non-specific increase in the number of lymphocytes in the blood and the presence of >60% lymphocytes and/or >10% atypical lymphocytes in the blood is very suggestive of Mono. It is important to realize that the Monospot blood test is often negative during the first 1-2 weeks of the illness, and that a definite diagnosis of Mono cannot always be made during the first 14 days of the illness. A repeat Mono test may be required if an initial Mono test performed earlier in the illness was negative. The Monospot test may remain negative in children <4 years of age. For Monospot negative or atypical cases, EBV-specific antibodies to a number of viral antigens can be measured.

There is no curative therapy for Mono and the disease has to run its course, which may take many weeks. Steroids are occasionally prescribed for severe tonsillo-pharyngitis that threatens to obstruct the airway, but the drug is of no value in patients with mild or moderate pharyngitis. Steroids are also used to treat some of the rare complications of Mono -- hemolytic anemia (an autoimmune process in which red blood cells are destroyed), and severe thrombocytopenia (a decrease in platelets, which are clotting components in the blood).

Patients should be cautioned about strenuous physical activity and contact sports during the first month of the illness because of the risk of splenic rupture and death from intra-abdominal bleeding. The Ebstein Barr virus remains in the oropharyngeal secretions for months after the acute illness and the patient should be considered potentially contagious. Therefore, the patient should avoid exposing other people to his bodily secretions.

For detailed information on infectious mononucleosis, consult the following emedicine.com chapters.

Infectious mononucleosis
Infectious mononucleosis - a patient education guide

Group C and G streptococcal pharyngitis

Although Group A streptococcal infection is the most common cause of a "strep throat", there are other streptococcal bacteria that can cause pharyngitis -- Group C and G streptococci.

Group C and G streptococcal pharyngitis are usually associated with food-borne and waterborne outbreaks. The pharyngitis is usually less severe than GABHS pharyngitis and it is uncertain whether penicillin is effective therapy. Some experts recommend erythromicin therapy as the antibiotic-of-choice. Other experts think that antibiotic therapy is not effective, and that antibiotics may not even be indicated because untreated infection does not result in rheumatic fever. Because there is no risk of developing rheumatic fever, a shorter course of antibiotics (<10 days) is appropriate if the treating physician decides to prescribe antibiotics. Another important fact for a physician to consider is that these bacteria often reside as resident carrier commensals in the throat. Therefore, the presence of these bacteria in a throat culture does not necessarily imply that they are actively involved in the acute pharyngeal infection, which could be viral in origin.

Acute retroviral syndrome

It is increasingly recognized that that an acute monucleosis-type syndrome can develop within days to weeks after contracting a HIV infection, with a peak incidence after 3-5 weeks.

The syndrome usually lasts 14 days and is very non-specific and variable. The most common findings include fever, fatigue, headache, and generalized lymphadenopathy (swollen glands). Additional findings include pharyngitis, myalgia (muscle aches), nausea, vomiting and diarrhea. A truncal rash occurs in 40-80% of cases and it may have a variable presentation.

It may be differentiated from Mono because the clinical presentation is more acute in onset, and the tonsils are not massively enlarged or covered with exudate.

Standard HIV tests may be negative at this stage, and assays for HIV type 1 RNA or p24 antigen may be required to make the diagnosis.

Commentary, criticism and controversy

Insightful questions and comments from readers will be included in this section.

Disclaimer: My medical guidemaps reflect my personal approach to clinical problems, and although my personal approach is primarily based on a thorough evaluation of the evidence-based medical literature, my personal approach should not be regarded as the "standard of care". My medical guidemaps should only be used if the reader-user has substantial reason to believe that the clinical advice contained in the guidemaps is valid and accurate. My medical guidemaps do not represent a personal medical service and they should not be used as a substitute for consulting a physician or other health-care provider. My medical guidemaps should be perceived as being an altruistic educational resource that is only applicable to medical practice in the USA and certain other countries.