Emerging bacterial resistance patterns suggest
that clinicians should use restraint in prescribing antibiotics
for various infections. We used pharyngitis as the model for
studying prescribing habits, because many clinicians recommend
that culture results be analyzed before antibiotics are prescribed.
We compared prescribing habits of three groups: practitioners
in the Pediatrics, Adult Medicine, and Urgent Visit Departments.
Overall, 55% of patients were treated while culture results
were pending. Nurse practitioners and physician assistants were
more likely than physicians (including osteopaths) to treat
patients with pharyngitis who had negative culture results (57%
vs 38%). Our results thus show that practitioners should be
encouraged to avoid overprescribing for this common condition.
Introduction
Pharyngitis prompts a substantial proportion of outpatient visits
to primary care practitioners. Despite a plethora of guidelines,
continuing medical education (CME) activities, and assumed knowledge
about caring for this condition, its evaluation and management
continues to vary greatly. For example, in response to a national
questionnaire survey mailed to a random sample of primary care
practitioners,1 fully one quarter of 398 respondents
who performed microbiologic studies reported that they used
only a rapid antigen test. Of these respondents, 24% of pediatricians
and 58% of family practitioners failed to confirm negative test
results with cultures, despite recommendations to do so from
the Centers for Disease Control and Prevention (CDCP), American
Academy of Pediatrics (AAP), and the American Hospital Association
(AHA). Although the methods used for rapid antigen tests vary,
they are generally similar with regard to sensitivity and specificity.
Specificity of these tests approaches 95%; sensitivity averages
75% to 80%. Therefore, negative results of a rapid antigen test
should be confirmed by throat culture.
2
 |
| Figure 1. Viral symptoms of pharyngitis documented
in medical charts of 308 consecutive patients seen in
a one-month period (%). |
Other recommendations identify patients who should
be treated with antimicrobial agents. A recent practice guideline3
is consistent with previous recommendations concerning treating
individuals after an infectious organism's presence in the throat
is confirmed by microbiologic or immunologic means. In clinical
practice, this guideline is rarely followed; patients and parents
simply return to daycare, school, work, and other normal activities
when they perceive improvement in symptoms. For patients, the
idea of withholding antibiotic therapy is a barrier to returning
to these normal activities, even though most such patients are
told they probably have a viral infection.
The purpose of our study was to examine use of
antibiotic drugs for treatment of pharyngitis. We looked for
different patterns among clinical departments and among different
types of practitioners (ie, physician extenders and physicians).
We also investigated how culture results were used to guide
subsequent treatment.
Methods
We reviewed the medical charts of approximately 100 patients
from each of three clinical departments (Adult Medicine, Pediatrics,
Urgent Visit) in the Warren Paley Medical Office who were seen
for pharyngitis during a one-month period.
Charts were reviewed to determine presenting symptoms;
treatment selected; type of treating practitioner (nurse practitioner
(NP), physician assistant (PA), physician (MD), or osteopath
(DO)); whether treatment was started empirically; and whether
the patient was diagnosed with another condition that required
treatment. Documentation of cough, rhinorrhea, or viral symptoms
was also noted. In addition, a random sample of patients who
tested culture-negative were called after the visit to be informed
of the results and to determine their subsequent attitudes about
stopping antibiotic therapy.
Results
Charts of 308 patients were reviewed. Of these 308 patients,
119 were seen in the Pediatrics Department, 94 were seen in
the Urgent Visit Department, and 95 were seen in the Adult Medicine
Department. Overall, 30% of pediatric patients and 16% of adult
patients tested culture-positive. Charts lacked documentation
of cough for 33% of patients and lacked documentation of rhinorrhea
for 40% of patients. Of patients who tested culture-positive,
cough was documented in 17%, and rhinorrhea was documented in
17% (Figure 1).
 |
| Figure 2. Treatment of pharyngitis by health care practitioners
in three departments (%). |
Figure 2 lists treatment given by each of the
three clinical departments for patients who tested culture-positive
and were initially appropriately treated as well as for patients
who tested culture-negative and were initially not treated.
Overall accuracy of treatment (number of treated culture-positive
patients plus number of untreated culture-negative patients
divided by total number of patients) is also shown. We found
a significant relation between accuracy of treatment and treating
department (p < 0.001, 2 df, c2 = 34.01)) (Table
1).
Figure 3 shows the difference between treatment
selected by physicians (MD/DO) and by physician extenders (PA/NP).
Patients who were culture-negative and seen by a physician extender
were 66% more likely to receive antibiotic therapy than patients
seen by physicians (63% vs 38%). This practice did not result
in more culture-positive patients being treated (75% for PA/NP
vs 76% for MD/DO). Overall accuracy of treatment was thus higher
for physicians (79%) than for physician extenders (53%). We
found a significant relation between accuracy of treatment and
type of professional credential (p < 0.001, 2 df, c2
= 33.22) (Table 2).
Of 308 patients, 97 (31.5%) were instructed to
continue the prescribed antibiotic regimen regardless of culture
results. Practitioners' reasons for giving this instruction
are listed (Table 3).
The three most common reasons cited (sinusitis, otitis media,
bronchitis) accounted for 73.6% of reasons given.
Of the 308 patients, 96 (31.1%) were called by
our staff to be informed of culture results. Of these 96 patients,
31 (32.3%) were unwilling to stop antibiotic therapy. Some of
the more common comments from patients included "I don't
think my doctor wanted me to stop it"; "I feel better,
so I think I will continue"; "I think the doctor wanted
me to take it if I felt better"; and "My doctor told
me to take it just in case."
Discussion
Our study confirms the variation in evaluation and treatment
of pharyngitis. Indeed, although cough, hoarseness, and rhinorrhea
strongly suggest a viral cause, our study showed that almost
one third of medical charts contained no documentation of either
presence or absence of these symptoms. In addition, a substantial
number of patients had these symptoms and had cultures taken
anyway. Recovery of Group A streptococcus from the pharynx does
not distinguish patients with streptococcal infection (defined
by a serologic antibody response) from strep tococcal carriers
who have pharyngitis caused by a different organism (ie, a virus).
Given the viral symptoms, the pretest likelihood of streptococcal
pharyngitis is extremely low, making a positive test result
more likely to suggest a carrier state than a diagnosis of streptococcal
pharyngitis.
 |
| Figure 3. Treatment of pharyngitis by physicians and
physician extenders (%). |
During this era of antibiotic resistance, many
authors advocate restraint in presumptively starting therapy
until infection is confirmed.2-5 Our comparison of
treatment administered by the three clinical departments shows
that although Adult Medicine had the highest percentage of treated
culture-positive patients (92%), that department had the lowest
treatment accuracy (42%). This value was statistically significant,
but a potential bias may exist due to a higher likelihood that
pediatric practice strongly advocates conservative treatment
pending receipt of culture results. This position was formed
primarily as a result of the combination of lower incidence
of streptococcal pharyngitis in adults and higher number of
culture-negative patients who received antibiotic treatment
pending receipt of culture results. Presumptively starting antibiotics
is discouraged by most clinicians4 because treatment
is often continued despite a negative test result. Our study
showed that one third of patients are openly unwilling to stop
antibiotic therapy. We did not identify ourselves as physicians
when we called patients, and they may be less likely to stop
the antibiotic treatment unless a physician instructs them to
do so. However, in most offices, physicians are not the most
likely personnel to inform patients that their test results
are negative. Our finding therefore probably reflects patients'
experience in most cases.
When pooled together from all departments, physician
extenders were more likely than physicians to prescribe antibiotics
pending receipt of culture results. This result may reflect
bias, given the possible statistically significant impact of
each individual treating department. The Pediatric Department
employed only one nurse practitioner, so that department cannot
easily be compared with the others. The nurse practitioners
and physician assistants we surveyed said that they felt more
pressure to prescribe and that patients' comments added to this
pressure with comments such as "I had to come back last
time for an antibiotic," "I called the physician for
a prescription after I left here," or "I'd rather
see a physician if you're not going to give me an antibiotic."
Other studies6-8 have shown how practitioners differ
by credentials in responding to patient requests for antibiotics.
These studies and ours reflect an underlying need for clinicians
to have a repertoire of behaviors that positively respond to
patients' requests for antibiotics. Some examples of these behaviors
include: 1) having an open discussion among practitioners and
physician extenders within a practice about how they address
this issue; 2) during well visits, giving patients handouts
(available from the AAP, American Academy of Family Practice
(AAFP), and CDCP) about antibiotic resistance; 3) in examination
rooms, posting newspaper and magazine articles about antibiotic
resistance; 4) in waiting areas, posting posters (available
from the CDCP) about antibiotic resistance; 5) acknowledging
the difficulty of treating viral symptoms; and 6) discussing
with patients the therapeutic value of particular symptoms (eg,
fever boosts the immune system, cough protects the lungs from
pneumonia, runny nose washes out the virus).
We have preliminary data indicating that use of
a rapid antigen detection test substantially reduces the number
of prescriptions given for antibiotic drugs. Moreover, our analysis
of cost indicates that the resultant savings in antibiotic drugs
as well as reduced use of office resources may offset the cost
of the rapid antigen detection test. In addition, the substantial
reduction in antibiotic drug prescriptions is consistent with
the CDCP, AAP, and AAFP efforts to reduce unnecessary antibiotic
use.
Many patients are instructed to continue taking
their prescribed antibiotic drugs regardless of culture results
(Table 1). Tests should
logically not be done if they do not affect case management.
Exceptions to this proposition are possible; however, if our
study is representative of most practice, almost a third of
culture procedures could be eliminated. This reduction would
greatly increase the availability of resources (ie, receptionist
time, nurse time, medical recordkeeping time, laboratory time,
and practitioner time).
Conclusion
Pharyngitis is a common medical condition whose evaluation and
treatment varies greatly among primary care practitioners. Our
study confirms this variability. Clinicians should have an open
discussion about individual approaches to patients' requests
for antibiotic therapy in general and for antibiotic drugs to
treat pharyngitis in particular. Physician extenders need support
from physician colleagues to enhance appropriate patient care
and utilization of resources.
Viral symptoms should be inquired about in every
evaluation of pharyngitis. Nurses should be taught to refrain
from obtaining throat cultures in those situations, because
doing so promotes the patient's expectation that the culture
will be sent to the laboratory for analysis. Patients who have
viral symptoms (eg, rhinorrhea, cough, conjunctivitis, hoarseness,
diarrhea, anterior stomatitis, or discrete ulcerative lesions)
are unlikely to have Group A streptococcal infection and should
therefore not have specimens taken for culture.2
Presumptively starting therapy pending receipt
of culture results should be discouraged because treatment often
continues despite a negative test result. Early initiation of
antimicrobial therapy results in faster resolution of symptoms,
but two facts should be remembered. First, group A streptococcal
pharyngitis is usually a selflimited disease; fever and constitutional
symptoms disappear spontaneously within three or four days after
onset, even when antimicrobial therapy is not administered.9
Second, therapy can be safely postponed for as long as nine
days after onset of symptoms and still prevent the occurrence
of the major nonsuppurative sequela (acute rheumatic fever).10
Each intervention to reduce unnecessary antibiotic
drug use for common conditions such as pharyngitis can reduce
the number of serious infections that do not respond to conventional
antibiotic therapy. Clearly, education of health care practitioners
and patients is essential for changing behavior and for helping
to reduce the likelihood of antimicrobial resistance.11
References
1. Schwartz B, Fries S, Fitzgibbon AM, Lipman H. Pediatricians'
diagnostic approach to pharyngitis and impact of CLIA 1988 on
office diagnostic tests. JAMA 1994;271:234238.
2. Committee on Infectious Diseases. Group A streptococcal infections.
In: Academy of Pediatrics. Group A streptococcal infections.
Red book: report of the Committee on Infectious Diseases. 24th
ed. Elk Grove Village, IL: American Academy of Pediatrics; 1997.
p. 48394.
3. Bisno AL, Gerber MA, Gwaltney JM Jr, Kaplan EL, Schwartz
RH. Diagnosis and management of group A streptococcal pharyngitis:
a practice guideline. Infectious Diseases Society of America.
Clin Infect Dis 1997;25:57483.
4. Schwartz B, Marcy SM, Phillips WR, Gerber MA, Dowell SF.
PharyngitisPrinciples of judicious use of antimicrobial agents.
Pediatrics 1998;101S:1714.
5. Tanz RR, Shulman ST. Diagnosis and treatment of group A streptococcal
pharyngitis. Semin Pediatr Infect Dis 1995;6:6978.
6. Schwartz RH, Freij BJ, Ziai M, Sheridan MJ. Antimicrobial
prescribing for acute purulent rhinitis in children: a survey
of pediatricians and family practitioners. Pediatr Infect Dis
J 1997;16:185-90.
7. Mainous AG III, Hueston WJ, Love MM. Antibiotics for colds
in children: who are the high prescribers? Arch Pediatr Adolesc
Med 1998;152:349-52.
8. Pennie RA. Prospective study of antibiotics prescribing for
children. Can Fam Physician 1998;44:1850-6.
9. Brink WR, Rammelkamp CH Jr, Denny FW, Wannamaker LW. Effect
of penicillin and aureomycin on the natural course of streptococcal
tonsillitis and pharyngitis. Am J Med 1951;10:3008.
10. Catanzaro FJ, Stetson CA, Morris AJ, Chamovitz R, Rammelkamp
CH Jr, Stolzer BL, et al. The role of the streptococcus in the
pathogenesis of rheumatic fever. Am J Med 1954;17:74956.
11. Dajani AS. Adherence to physicians' instructions as a factor
in managing streptococcal pharyngitis. Pediatrics 1996;97(6
Pt 2):97680.
