ORIGINAL CONTRIBUTION

Antibiotics for Upper Respiratory Tract Infections

Follow-up Utilization and Antibiotic Use
William J. Hueston, MD; Arch G. Mainous III, PhD; Steven Ornstein, MD; Qin Pan, MS; Ruth Jenkins, MS

Objectives: To examine the effects of antibiotic prescribing during an initial visit for viral respiratory tract
infections on future care seeking and the cost of care.
Materials and Methods: Retrospective analysis of recorded visits for viral respiratory tract infections
(N = 49 862) between January 1, 1995, and December 31,
1997, to practices in a large network of affiliated practices that use the same electronic medical record.
Results: Patients receiving antibiotics at the initial visit
were less likely to return for a second visit, but this difference was small (15.4% vs 17.4%, P,.001). When returning for the second visit, those who received an an-

tibiotic on the initial visit were prescribed more expensive

antibiotics than those who had not received an antibiotic on the initial consultation. Overall, cost from initial
antibiotic use outweighed any benefit from reduced utilization in adults and children.
Conclusions: Antibiotic prescribing at an initial contact for a viral respiratory tract illness may reduce the likelihood that an individual will return for a subsequent visit,
but adds substantial costs to care for the initial antibiotic and for more expensive antibiotics used on subsequent visits.

Arch Fam Med. 1999;8:426-430

respiratory tract infections are

seen by many as trivial
conditions, they are a leading cause of acute morbidity and have substantial associated costs.
While viral respiratory tract infections are
generally mild, self-limited, and short term,
they are a primary reason for industrial and
school absenteeism.1-3 Each year in the
United States, upper respiratory tract infections (URIs) alone account for 170 million days of restricted activity, 23 million
days of school absence, and 18 million days
of work absence.3
LTHOUGH VIRAL

For editorial comment

see page 431

From the Department of Family

Medicine, Medical University of
South Carolina, Charleston.

According to estimates from the US

National Ambulatory Medical Care Survey,4 acute URIs accounted for 17 million visits in 1991 and were the fifth most
common reason for seeking care.In addition, microbiologic and laboratory diagnostic tests of dubious clinical value are
often performed and contribute additional unnecessary costs to the care of patients URIs.5 The cost of antibiotics prescribed unnecessarily further increases the
ARCH FAM MED/ VOL 8, SEP/OCT 1999
426

expense for these conditions.6 Finally, the

direct costs for URIs include nonprescription treatments. Americans spend between $1 and $2 billion annually on the
more than 800 over-the-counter cough and
cold preparations.7,8 Together, viral respiratory tract diseases constitute a substantial economic drain on health care systems.
While multiple factors contribute to
the use of resources in patients with URIs,
the factor under the most control of physicians is the prescribing of antibiotics in
cases of viral respiratory tract infections.
Because the overwhelming majority of
URIs are caused by viruses,1,9 antibiotics
are not indicated for their treatment.10-15
Nor do antibiotics prevent complications16,17 as evidenced by a double-blind,
placebo-controlled trial in which complications developed in 12% of those receiving placebo and 11% of those receiving antibiotics. 17 Yet, antibiotics are widely
prescribed for URIs.18,19 Recent studies using claims data18 and physician surveys20
indicate that more than 50% of episodes
for URIs are treated with antibiotics. Antibiotic overuse is not limited to URIs. Antibiotics have shown no benefit in the treatment of acute bronchitis,21,22 yet studies

Downloaded from www.archfammed.com at CLOCKSS, on November 7, 2009

1999 American Medical Association. All rights reserved.

MATERIALS AND METHODS

Data were compiled in primary care practices participating
in the Practice Partner Research Network (PPRNet). Practice Partner Research Network is a national network of practices that share a common electronic medical record system (Practice Partner; Physicians MicroSystems Inc, Seattle,
Wash) and combines data regarding diagnoses, preventive
services, selected laboratory and radiology services, and prescriptions for research purposes. Practice Partner Research
Network practices include 53 primary care practices and 4
specialty practice in 24 states in the United States. Each month
the practice collects data on a self-running program that is
then submitted to a central site for inclusion in a longitudinal database. The database includes more than 380 000 patients with data on 2.3 million outpatient contacts and 3.9
million diagnoses. For this study, the database containing
patient visit data between January 1, 1995, and December
31, 1997, was included for analysis.
Forty-five (79%) of the 57 practices in the PPRNet participated in this study. Of the 12 practices excluded, 5 had
few contacts for viral illnesses (4 were specialty practices
and 1 was primarily for patients with chronic illnesses), 5
practices had incomplete data for 1995, and 2 practices did
not record prescriptions for acute illnesses. All active patients were considered eligible for study. An active patient
was one whose medical record had been updated since January 1, 1994.
Information available in the PPRNet data set included the date of service, diagnosis, prescriptions written, duration of therapy, and demographic information about
the patient. Because some practices did not include all radiology and laboratory services in their medical records,
this information was not included in the study. Diagnoses
investigated in this study included those for viral

have shown antibiotic prescribing rates ranging from 66%

to 75% of the cases.20,23
One reason physicians may prescribe antibiotics for
viral conditions is to reduce subsequent office visits for
the same complaints. However, it is unclear what effect
prescribing antibiotics has on reconsultation rates. This
study investigated the effectof receipt of antibiotics on follow-up utilization for viral respiratory tract infections and
examined the treatment rendered at second visits.
RESULTS

The characteristics of the 45 participating practices are

given in Table 1. Most were single-specialty family physician practices and viral respiratory tract diseases constituted 38% of any type of encounter that involved a respiratory complaint.
Table 2 lists the number of contacts and episodes
for viral respiratory tract infections in adults and children. The vast majority of viral illnesses resulted in a single
contact and only 17% of the patients returned for a second visit. The rate of return was no different for children compared with adults. Table 2 also gives the percentage of episodes in which an antibiotic was prescribed

conditions coded under the following conditions: URI/

nasopharyngitis, acute bronchitis, influenza, and multiple
viral illnesses (which was used when more than one of the
index viral illnesses was diagnosed in the same visit, eg,
URI and acute bronchitis).
To evaluate how treatment with an antibiotic influenced subsequent care, rather than analyzing the data based
on individual visits, it was necessary to construct episodes of illness that served as the unit of analysis. An episode of illness was conceptualized to include all care related to one discrete viral illness. The definition was
constructed to assure that subsequent visits could be related to a previous index visit and that visits that signaled
a new episode were not likely to be related to a previous
viral illness. To meet these criteria, we defined an index
visit for a new episode as a visit with the diagnosis of URI,
acute bronchitis, or influenza with no other contact in the
preceding 14 days that included a diagnosis of URI, acute
bronchitis, influenza, pharyngitis, or viral disease not otherwise specified. All subsequent contact for a respiratory
condition or follow-up for potential complications such as
adverse drug reactions in the next 14 days was considered
part of the episode. An example of relevant diagnoses included any of the index diagnoses plus viral infection, pneumonia, rhinosinusitis, otitis media, nausea, diarrhea, tonsillitis, follow-up, and medical examination.
Bivariate analyses of categorical data were performed
using x2 statistic. Estimates of antibiotic costs for episodes
when antibiotics were prescribed were calculated using what
we considered the most conservative estimate of the potential price of an antibiotic, ie, the average wholesale price,24
assuming a 7-day course of therapy. Radiology and laboratory use were excluded from the model because we could
not determine how antibiotic prescribing influenced the use
of these ancillary services. An a level of .05 was considered statistically significant.

on the initial visit. Antibiotics were more likely to be used

on initial visits for adults than for children (P,.001)
The effect of prescribing antibiotics during the initial visit on the likelihood of seeking additional care at
the same practice is given in Table 3. Prescribing an antibiotic at the initial visit reduced the percentage of individuals who returned for a second appointment in adults
and children. For children, individuals who returned for
a second visit were no more likely to receive an antibiotic prescription on the second visit if they had received
an antibiotic initially. Just the opposite was found for
adults; those who did not receive an antibiotic on the first
encounter were more likely to receive an antibiotic if they
made a second visit (P,.001).
The types of antibiotics used during initial and second visits are listed in Table 4. Differences in drug selection for initial and return visits resulted in substantial variations in the average cost of drugs for each type
of visit. When prescribed on an initial visit, the average
price per prescription for initial visits was $18.85 for adults
and $11.57 for children. When patients who received an
antibiotic returned and were given another antibiotic, the
average price increased to $28.18 for adults and $29.28
for children. If a patient had not received an antibiotic

ARCH FAM MED/ VOL 8, SEP/OCT 1999

427

Downloaded from www.archfammed.com at CLOCKSS, on November 7, 2009

1999 American Medical Association. All rights reserved.

Table 1. Characteristics of Participating Practices

Family
Practice*

Internal
Medicine Multispecialty

Practices,
33 (73)
6 (13)
No. (%)
Physicians,
277 (87)
10 (3)
No. (%)
Active patients
195 794
9971
Patients per
736-24 004 1112-2444
practice, range

Total

6 (13)

45 (100)

31 (10)

318 (100)

50 799
1166-24 638

255 564
736-24 638

*Includes 8 academic family practices.

All multispecialty clinics were family practice and internal medicine.

Table 2. Episodes of Viral Illnesses in the Practice Partner

Research Network Population

Patients
Aged
,18 y
$18 y
Total Episodes

No. of
Episodes

Episodes With .1
Contact, No. (%)

Antibiotic Used
in Initial Visit of
Episode, No. (%)

20 484
38 358
58 842

3265 (16)
6474 (17)
9739 (17)

6701 (33)
18 484 (48)
25 185 (43)

Table 3. Frequency of Repeat Visits

and Treatment on Second Visit*
Initial Visit

Aged ,18 y
Percent with .1 visit
Received antibiotic on visit 2
Aged $18 y
Percent with .1 visit
Received antibiotic on visit 2
Total episodes
Percent with .1 visit
Received antibiotic on visit 2

Antibiotic
Received

No Antibiotic
Received

13.7
20.6

17.0
18.4

,.001
.15

16.0
17.1

17.7
21.4

,.001
,.001

15.4
17.9

17.4
19.4

,.001
.004

*All values expressed as a percentage.

on the initial visit, but was prescribed an antibiotic at the

second visit, the cost was $21.51 for adults and $16.38
for children.
The use of antibiotics on the initial encounter had
a notable effect on the overall cost of visits. Figure 1
and Figure 2 show a hypothetical population of 100 000
adults and children seeking care for viral respiratory tract
diseases. Based on the data from our population and an
estimated cost of $35 per patient visit, the average cost
of the episode increased 43% for adults and 26% for children when antibiotics were used on the initial visit. Antibiotics costs accounted for 33% of the episode cost in
adults and 24% in children when antibiotics were prescribed on the initial visit; when antibiotics were not used
on the initial visit, the overall cost of antibiotics contributed only 2% to the cost of the entire episode.
Because patients could seek care outside the PPRNet
of physicians in this study, we also performed a sensi-

tivity analysis to determine how patients who sought care

at other sites would alter our results. We found that 32%
of adults and 24% of children who had initial visits in
which they did not receive an antibiotic would have had
to make visits to other sources of care and all would have
to receive antibiotics for the cost of episodes in which
antibiotics were not used initially to exceed the cost when
antibiotics were prescribed on the initial visit.
Finally, we examined the diagnoses associated with
second visits in patients who did and did not receive antibiotics. Regardless of whether patients received an antibiotic on the initial visit, they were more likely to be
diagnosed with another viral infection or a bacterial infection if they received antibiotics on the second visit than
if no antibiotic was prescribed on the second visit (71%
vs 35%, P,.001). However, those who received an antibiotic on the initial visit were more likely to have a bacterial diagnosis on the second visit than those who did
not receive an antibiotic (31% vs 28%, P = .002).
COMMENT

Our data suggest that, first, individuals not given antibiotics at the initial visit for a presumably viral condition are more likely to return for a second visit, but overall second visit rates are low for both groups. Second, the
absolute differences between the groups are small. To reduce the revisit frequency by one individual, the number needed to treat with antibiotics would be 59 for adults
and 30 for children. Based on the cost of antibiotics, the
potential for side effects in the large number of patients
treated just to discourage reconsultation, and our observation that even more expensive antibiotics are used if
patients do return for another visit, antibiotics cannot be
considered a harmless placebo prescribed simply to discourage a second visit.
In addition to immediate cost and side effects from
antibiotic prescribing, antibiotic use for URIs may encourage patients to seek care for subsequent viral infections. Data suggest that providing courses of antibiotics
for primarily viral illnesses encourages patient beliefs in
the effectiveness of antibiotics for those illnesses.25,26 Patients associate the resolution of the self-limited illness
in time with the receipt of the antibiotics and, thus, draw
the conclusion that the antibiotics caused the resolution of the illness. Unfortunately, the results suggest that
the second antibiotic tends to be a more expensive broadspectrum agent, thereby increasing the cost substantially and conveying to the patient that they need the
strong antibiotic on future episodes.
Clinicians may believe that in practice they use antibiotics selectively and choose to prescribe antibiotics
only for patients who come into the visit with the expectation that they will receive an antimicrobial. However, evidence suggests that physicians are not accurate
at assessing patients expectations for antibiotics in URIs
and when unsure assume that the patient wanted an antibiotic.26 This error in perception will lead to overprescribing of antibiotics and will reduce the usefulness
of this strategy. In fact, most patients consult physicians
to clarify their symptoms and obtain a diagnosis.27,28 Patients are perfectly happy when they do not receive an-

ARCH FAM MED/ VOL 8, SEP/OCT 1999

428

Downloaded from www.archfammed.com at CLOCKSS, on November 7, 2009

1999 American Medical Association. All rights reserved.

Table 4. Antibiotics Used on Initial and Second Visits of Episode*

Initial Visit
Antibiotic Received
Aged ,18 y
Visit 1

Visit 2

Aged $18 y
Visit 1

Visit 2

No Antibiotic Received

Amoxicillin (54), cephalosporin (15), sulfamethoxazoletrimethoprim (12), erythromycin (8), other macrolide (7),
and all other (8)
Cephalosporin (35), other macrolide (17),
sulfamethoxazole-trimethoprim (17), amoxicillin (15),
and all others (16)

No antibiotic prescribed

Amoxicillin (24), other macrolide (22), erythromycin (17),

cephalosporin (16), sulfamethoxazole-trimethoprim (13),
and all others (16)
Other macrolide (30), cephalosporin (26), erythromycin (13),
amoxicillin (11), and all others (20)

No antibiotic prescribed

Amoxicillin (49), sulfamethoxazole-trimethoprim (15),

cephalosporin (14), other macrolide (8), and all others (14)

Other macrolides (24), amoxicillin (21), cephalosporin (17),

erythromycin (13), sulfamethoxazole-trimethoprim (13),
and all others (12)

*All numbers in parentheses are the percentage of patients.

100 000 Adult Patients

Initial Visit
Receive Antibiotics
n = 48 000

100 000 Pediatric Patients

Initial Visit
No Antibiotics
n = 52 000

Amoxicillin: 11 520
Macrolide: 10 560
Erythromycin: 8160
Cephalosporin: 7680
SulfamethoxazoleTrimethoprim: 6240
Other Antibiotics: 3800

No Second Visit
n = 40 320

Second Visit
n = 7680

Receive Antibiotics
n = 1313
Macrolide: 392
Cephalosporin: 340
Erythromycin: 170
Amoxicillin: 144
Other Antibiotics: 267

No Antibiotic
n = 6367

Cost per Case:

$60.22
Cost Attributed to Antibiotics: 33%

Initial Visit
Receive Antibiotics
n = 32 700

Initial Visit
No Antibiotics
n = 67 300

Amoxicillin: 17 280
Cephalosporin: 4672
SulfamethoxazoleTrimethoprim: 3840
Erythromycin: 2560
Macrolide: 2240
Other Antibiotics: 1480

Second Visit
n = 9204

Receive Antibiotics
n = 1969
Macrolide: 464
Amoxicillin: 404
Cephalosporin: 346
Erythromycin: 262
SulfamethoxazoleTrimethoprim: 260
Other Antibiotics: 233

No Second Visit
n = 42 796

No Antibiotic
n = 7234

Average Cost per Case:

$42.01
Cost Attributed to Antibiotics:
2%

No Second Visit
n = 28 220

Receive Antibiotics
n = 923
Cephalosporin: 319
Macrolide: 155
SulfamethoxazoleTrimethoprim: 155
Amoxicillin: 136
Other Antibiotics: 158

Second Visit
n = 4479

No Antibiotic
n = 3556

Cost per Case:

$52.96
Cost Attributed to Antibiotics: 23%

Second Visit
n = 11 441

Receive Antibiotics
n = 2105
Amoxicillin: 1246
SulfamethoxazoleTrimethoprim: 384
Cephalosporin: 353
Macrolide: 191
Erythromycin: 181
Other Antibiotics: 188

No Second Visit
n = 55 859

No Antibiotic
n = 9512

Average Cost per Case:

$41.91
Cost Attributed to Antibiotics:
2%

Figure 1. Treatment of viral upper respiratory tract diseases in adults.

Figure 2. Treatment of viral upper respiratory tract diseases in children.

tibiotics for their viral illness so long as they are provided with an explanation of the rationale behind that
decision.26,29 Because most patients simply want a diagnosis for their conditions, physicians may be overestimating the desire for antibiotics; rather than writing a
prescription to fulfull an unstated need, a focused discussion with the patient of the lack of benefits of antibiotic therapy and potential complications of antibotics
is a useful alternative stategy.
While the net costs of antibiotic prescribing add appreciable cost to the episode, another additional cost that
was not considered in this study is the indirect costs associated with the increased prevalence of antibioticresistant bacteria that has been linked to widespread use
of antibiotics for primarily viral conditions. Recent data

suggest that antimicrobial use is positively correlated with

the nasopharyngeal carriage of penicillin-resistant pneumococci in children.30 Similarly, nationwide reductions
of macrolide antibiotic use was shown to be associated
with a notable decrease in erythromycin-resistant group
A streptococci.31 The widespread development of antibiotic resistance means that use of antibiotics for URIs
and other primarily viral respiratory tract infections can
have serious repercussions and should not be viewed by
the physician as simply a harmless placebo to pacify the
presumed expectations of the patient.
There are several limitations of this study that should
be considered in interpreting the results. First, we could
not validate the diagnoses provided by the participating
physicians. Because viral respiratory tract infections are

ARCH FAM MED/ VOL 8, SEP/OCT 1999

429

Downloaded from www.archfammed.com at CLOCKSS, on November 7, 2009

1999 American Medical Association. All rights reserved.

part of a constellation of symptoms that is shared by other

respiratory infections (eg, rhinosinusitis or pneumonia), the coded diagnosis may be inaccurate and may be
an overestimate of ineffective antibiotic treatment. However, the coded diagnoses are what was determined by
the participating physician according to clinical and diagnostic criteria. Second, we only examined whether an
antibiotic was prescribed. Even though we used the most
conservative estimate of a cost for each drug (ie, the wholesale price), our cost models may have overestimated the
actual costs of antibiotic prescribing since other studies
suggest that a large proportion of patients in the United
States never get their medication prescription filled.
Finally, we could not account for utilization of
sources of care outside the PPRNet of primary care physicians in the study. If individuals who did not receive
antibiotics sought care at other sources more often than
those who received antibiotics, the differences between
these 2 groups would be an overestimate. However, as
noted in the sensitivity analysis, those who did not receive antibiotics would have had to use other sources of
care at a rate 150% to 200% higher to alter the overall
conclusions of the study.
In summary, this study suggests that initial antibiotic prescribing may reduce subsequent visits in episode
of viral respiratory tract illness. However, when patients
who previously received antibiotics do return, they are
likely to receive another antibiotic. This increases the cost
of care and may have implications in the development of
antibiotic resistance in common respiratory organisms.
Accepted for publication October 6, 1998.
This project was supported by a Robert Wood Johnson
Generalist Physician Faculty Scholar Award (Dr Hueston), Princeton, NJ, and support from IMS America, Philadelphia, Pa.
Corresponding author: William J. Hueston, MD, Department of Family Medicine, Medical University of South
Carolina, 295 Calhoun St, Charleston, SC 29401.
REFERENCES
1. Gwaltney J. The common cold. In: Mandell G, Douglas R, Bennett J (eds). Principles and Practice of Infectious Disease. New York, NY: Churchill Livingstone;
1990:489-493.
2. Koster F. Respiratory infections. In: Barker LR, Burton JR, Zieve PD (eds). Principles of Ambulatory Medicine. Baltimore, Md: Williams & Wilkins; 1987: 347-355.
3. Benson V, Marano MA. Current estimates from the National Health Interview Survey, 1993. Vital Health Stat 10. 1994, No. 189.
4. Schappert SM. National Ambulatory Medical Care Survey, 1991 summary. Hyattsville, Md: National Center for Health Statistics; 1994. Vital Health Stat 13. 1994,
No. 116.

5. Carroll K, Reimer L. Microbiology and laboratory diagnosis of upper respiratory

tract infections. Clin Infect Dis. 1996;23:442-448.
6. Mainous AG III, Hueston WJ. The cost of antibiotics in treating upper respiratory infections in a Medicaid population. Arch Fam Med. 1998;7:45-49.
7. Smith MB, Feldman W. Over-the-counter cold medications: a critical review of
clinical trials between 1950 and 1991. JAMA. 1993;269:2258-2263.
8. Kogan MD, Pappas G, Yu SM, Kotelchuck M. Over-the-counter medication use
among US prechool-age children. JAMA. 1994;272:1025-1030.
9. Hilding DA. Literature review: the common cold. Ear Nose Throat J. 1994;73:
639-643, 646-647.
10. Valman HB. ABC of 1 to 7: respiratory tract infection. BMJ. 1981;283:38-39.
11. Ackerman BD. Treatment of undifferentiated respiratory infections in infants. Clin
Pediatr. 1968;7:391-395.
12. Hardy LM, Traisman HS. Antibiotics and chemotherapeutic agents in the treatment of uncomplicated respiratory infections in children: a controlled study.
J Pediatr. 1956;48:146-156.
13. Todd JK, Todd N, Damato J, Todd WA. Bacteriology and treatment of purulent
nasopharyngitis: a double-blind placebo-controlled evaluation. Pediatr Infect Dis.
1984;3:226-232.
14. Steinwegg KK. Natural history and prognostic significance of purulent rhinitis.
J Fam Pract. 1983;17:61-64.
15. Wald ER, Guerra N, Byers C. Upper respiratory tract infections in young children: duration of and frequency of complications. Pediatrics. 1991;87:129-133.
16. Dingle JH, Badger GF, Jordan WS Jr. Illness in the Home: Study of 25,000 Illnesses in a Group of Cleveland Families. Cleveland, Ohio: The Press of Western
Reserve University; 1964.
17. Townsend EH, Radebaugh JF. Prevention of complications of respiratory illness
in pediatric practice: a double-blind study. N Engl J Med. 1962;266:683-689.
18. Mainous AG III, Hueston WJ, Clark JR. Antibiotics and upper respiratory infection: do some folks think there is a cure for the common cold? J Fam Pract. 1996;
42:357-361.
19. McCaig LF, Hughes JM. Trends in antimicrobial drug prescribing among officebased physicians in the United States [published correction appears in JAMA.
1998;279:434]. JAMA. 1995;273:214-219.
20. Gonzales R, Steiner JF, Sande MA. Antibiotic prescribing for adults with colds,
upper respiratory tract infections, and bronchitis by ambulatory care physicians. JAMA. 1997;278:901-904.
21. Orr PH, Scherer K, Macdonald A, Moffatt MEK. Randomized placebo-controlled
trials of antibiotics for acute bronchitis: a critical review of the literature. J Fam
Pract. 1993;36:507-512.
22. MacKay DN. Treatment of acute bronchitis in adults without underlying lung disease. J Gen Intern Med. 1996;11:557-562.
23. Mainous AG III, Zoorob RJ, Hueston WJ. Current management of acute bronchitis in ambulatory care: the use of antibiotics and bronchodilators. Arch Fam
Med. 1996;5:79-83.
24. 1996 Drug Topics Red Book. Montvale, NJ: Medical Economics Press, 1996.
25. Mainous AG III, Zoorob RJ, Oler MJ, Haynes DM. Patient knowledge of upper
respiratory infections: implications for antibiotic expectations and unnecessary
utilization. J Fam Pract. 1997;45:75-83.
26. Hamm RM, Hicks RJ, Bemben DA. Antibiotics and respiratory infections: are patients more satisfied when expectations are met? J Fam Pract. 1996;43:56-62.
27. Sanchez-Menegay C, Hudes ES, Cummings SR. Patient expectations and satisfaction with medical care for upper respiratory infections. J Gen Intern Med. 1992;
7:432-434.
28. Mayefsky JH, El-Shinaway Y, Kelleher P. Families who seek care for the common cold in a pediatric emergency department. J Pediatr. 1991;119:933-934.
29. Cowan PF. Patient satisfaction with an office visit for the common cold. J Fam
Pract. 1987;24:412-413.
30. Arason VA, Kristinsson KG, Sigurdsson JA, Stefansdottir G, Molstad S, Gudmundsson S. Do antimicrobials increase the carriage rate of penicillin-resistant pneumococci in children? cross-sectional prevalence study. BMJ. 1996;313:387-391.
31. Seppala H, Klaukka T, Vuopio-Varkila J, et al. The effect of changes in the consumption of macrolide antibiotics on erythromycin resistance in group A streptococci in Finland. N Engl J Med. 1997;337:441-446.

ARCH FAM MED/ VOL 8, SEP/OCT 1999

430

Downloaded from www.archfammed.com at CLOCKSS, on November 7, 2009

1999 American Medical Association. All rights reserved.