Correspondence

Multidrug-Resistant Streptococcus pneumoniae

N Engl J Med 2001; 344:1329-1331April 26, 2001

Article

To the Editor:

In their study of the prevalence of multidrug-resistant Streptococcus pneumoniae in the United States, Whitney et al. (Dec. 28 issue)1 report differences in the prevalence of resistance according to region, age, and race. We investigated whether these findings reflect patterns found in hospitalized patients and evaluated recent changes in the prevalence of drug-resistant streptococcus species among inpatients.

We used the Solucient projected inpatient data base for 1997 through 2000. This is an all-payer data base that contains information on more than 17 million inpatient discharges in the United States each year and has been used for comparisons of hospital-admissions data and the results of surveillance by the Centers for Disease Control and Prevention.2 Our cohort included 812,088 patients with one or more diagnosis codes from the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) indicating infection with streptococcus species, of whom 12,771 (1.57 percent) also had codes indicating antibiotic resistance.

We evaluated the increase in the prevalence of drug-resistant streptococcus in the inpatient setting from the first quarter of 1997 through the first quarter of 2000. We found a significant increase in the rate of resistant streptococcal infections (Figure 1Figure 1Prevalence of Drug-Resistant Streptococcal Infections among Inpatients in U.S. Hospitals from the First Quarter of 1997 through the First Quarter of 2000.). Patients who were infected with resistant organisms were 58 percent more likely to die than those with nonresistant streptococcal infections (P<0.001), male patients were 40 percent more likely to have a resistant organism than female patients (P<0.001), and whites were 18 percent more likely to have a resistant organism than blacks (P<0.001). Infection with a resistant strain was associated with a five-day increase in the length of stay. The prevalence rates ranged from a low of 1.49 percent in the Northeast to a high of 1.65 percent in the West (P<0.001).

David A. Foster, Ph.D., M.P.H.
Sivana T. Heller, M.D., M.P.H.
Janet K. Young, M.D., M.H.S.A.
Solucient, Ann Arbor, MI 48108

2 References

1
Whitney CG, Farley MM, Hadler J, et al. Increasing prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. N Engl J Med 2000;343:1917-1924
Full Text | Web of Science | Medline

2
Sullivan KM, Belay ED, Durbin RE, Foster DA, Nordenberg DF. Epidemiology of Reye's syndrome, United States, 1991-1994: comparison of CDC surveillance and hospital admission data. Neuroepidemiology 2000;19:338-344
CrossRef | Web of Science | Medline

To the Editor:

The cornerstone of the recommendations of Whitney et al. with regard to preventing pneumococcal disease is vaccination. There have been three double-blind, randomized, placebo-controlled trials that cast doubt on the ability of vaccination to prevent pneumococcal disease.1-3 It seems that the true value of pneumococcal vaccine is unclear. In the light of these studies, what recommendations should be made regarding the use of these vaccines?

Donald Tayloe, M.D.
Fresno Veterans Affairs Medical Center, Fresno, CA 93703

3 References

1
Simberkoff MS, Cross AP, Al-Ibrahim M, et al. Efficacy of pneumococcal vaccine in high-risk patients: results of a Veterans Administration Cooperative Study. N Engl J Med 1986;315:1318-1327
Full Text | Web of Science | Medline

2
Ortqvist A, Hedlund J, Burman LA, et al. Randomised trial of 23-valent pneumococcal capsular polysaccharide vaccine in prevention of pneumonia in middle-aged and elderly people. Lancet 1998;351:399-403
CrossRef | Web of Science | Medline

3
French N, Nakiyingi J, Carpenter LM, et al. 23-Valent pneumococcal polysaccharide vaccine in HIV-1-infected Ugandan adults: double-blind, randomised and placebo controlled trial. Lancet 2000;355:2106-2111
CrossRef | Web of Science | Medline

To the Editor:

The editorial on antibiotic resistance by Wenzel and Edmond1 covers a vitally important subject of global concern.2 However, their statement that the degree of antibiotic use in the United States “is equivalent to nearly 30 prescriptions per 100 persons per year and to 4.1 kg (9 lb) of antibiotics per 100 persons per year” is misleading. The implication is that each of these 30 prescriptions involves an average total dose of more than 136 g. Inappropriate prescribing is a serious problem, but it has yet to reach such colossal proportions. On the basis of the authors' figures, total U.S. consumption (both animal and human) would average 8.2 kg apiece. However, human usage should be quantified from actual data, not by dividing the total consumption of 22.7 million kg by 50 percent.

Wenzel and Edmond state that 160 million prescriptions for antibiotics are written annually in a country with a population of 275 million people, which they then calculate as being equivalent in terms of use to 30 prescriptions per 100 persons per year. The number would actually be slightly less than 60. This would cut the calculated average prescription size to less than 70 g, a high, but more reasonable figure.

Could the authors restate these data, which are so important to our future control of microbial infections, in both humans and animals?

Richard S. Wilbur, M.D., J.D.
Royal Society of Medicine Foundation, Lake Forest, IL 60045

2 References

1
Wenzel RP, Edmond MB. Managing antibiotic resistance. N Engl J Med 2000;343:1961-1963
Full Text | Web of Science | Medline

2
Soulsby L, Wilbur RS, eds. Antimicrobial resistance. London: Royal Society of Medicine Press, 2001.

Author/Editor Response

Dr. Whitney replies:

To the Editor: The data of Foster and colleagues are intriguing. The authors report significant associations between a diagnosis code indicating antibiotic resistance and male sex, the risk of death, and the length of the hospital stay. Unfortunately, insufficient details are provided regarding the search strategy they used, since the ICD-9-CM code for streptococcal sepsis (038.0) may reflect illnesses caused by a variety of pathogens (e.g., group A or group B streptococci) and it is not clear whether codes for S. pneumoniae (e.g., 038.2 and 481) were included. Furthermore, the sensitivity of the use of ICD-9-CM codes as a means of identifying antibiotic resistance is uncertain. A temporal increase in the frequency of codes indicating resistance might reflect increased testing, changes in coding practices, or increased resistance of particular pathogens.

The low prevalence of resistance identified makes us doubt that the strategy based on ICD-9-CM codes was a sensitive means of identifying pneumococcal infections and testing for resistance. In addition, patients with complicated hospital courses, and perhaps other infections with resistant organisms, may have been more likely to receive codes indicating antibiotic resistance than patients who had an uncomplicated infection with a resistant streptococcal species. Without information on the type of streptococci and the sensitivity and specificity of their search strategy, it is difficult to reach any conclusions about the findings of Foster and colleagues.

Pneumococcal polysaccharide vaccines are effective against invasive disease, such as bacteremia and bacteremic pneumonia, in many populations of patients.1,2 In immunocompromised persons, such as those with human immunodeficiency virus (HIV) infection or AIDS, such vaccines may be effective against invasive disease only in certain subgroups.3,4 Therefore, pneumococcal polysaccharide vaccine should be given as early as possible after the diagnosis of HIV infection; if the diagnosis is made later in the course, the vaccine should be given after the initiation of antiretroviral therapy. The efficacy of the vaccine against nonbacteremic pneumonia has not been clearly demonstrated; whether this is due to a true lack of efficacy or to problems in determining the cause of pneumonia is unclear. Although the study from Sweden5 that is cited by Dr. Tayloe did not find the vaccine to have a protective effect against pneumococcal pneumonia, the test used for the diagnosis was subsequently found to be nonspecific.6 Since invasive pneumococcal disease carries a high risk of death, especially in the elderly and persons with chronic illnesses, the recommendations for the use of pneumococcal polysaccharide vaccine for these patients are appropriate. Given the increasing difficulties in treating pneumococcal infections owing to the rise of multidrug-resistant organisms, the use of pneumococcal vaccines should be actively promoted.

Cynthia G. Whitney, M.D., M.P.H.
Centers for Disease Control and Prevention, Atlanta, GA 30333

6 References

1
Shapiro ED, Berg AT, Austrian R, et al. The protective efficacy of polyvalent pneumococcal polysaccharide vaccine. N Engl J Med 1991;325:1453-1460
Full Text | Web of Science | Medline

2
Butler JC, Breiman RF, Campbell JF, Lipman HB, Broome CV, Facklam RR. Pneumococcal polysaccharide vaccine efficacy: an evaluation of current recommendations. JAMA 1993;270:1826-1831
CrossRef | Web of Science | Medline

3
Dworkin MS, Ward JW, Carpenter LM, et al. Pneumococcal disease among HIV-infected persons: incidence, risk factors, and impact of vaccination. Clin Infect Dis (in press).

4
Breiman RF, Keller DW, Phelan MA, et al. Evaluation of effectiveness of the 23-valent pneumococcal capsular polysaccharide vaccine for HIV-infected patients. Arch Intern Med 2000;160:2633-2638
CrossRef | Web of Science | Medline

5
Ortqvist A, Hedlund J, Burman LA, et al. Randomised trial of 23-valent pneumococcal capsular polysaccharide vaccine in prevention of pneumonia in middle-aged and elderly people. Lancet 1998;351:399-403
CrossRef | Web of Science | Medline

6
Musher DM, Mediwala R, Phan HM, Chen G, Baughn RE. Nonspecificity of assaying for IgG antibody to pneumolysin in circulating immune complexes as a means to diagnose pneumococcal pneumonia. Clin Infect Dis 2001;32:534-538
CrossRef | Web of Science | Medline

Author/Editor Response

The editorialists reply:

To the Editor: In our editorial, we quoted Levy, who estimated that in 1996 160 million prescriptions were written for antibiotics in the United States and more than 50 million lb (22.7 million kg) of antibiotics was produced for use in people, animals, and agriculture.1 Approximately half those antibiotics (25 million lb [11.4 million kg]) were used by people. Because there are no specific data on prescribing profiles for the U.S. population, we calculated an average from the data: 25 million lb for 275 million people, or 9 lb (4.1 kg) per 100 persons per year. If 80 million prescriptions (half the total) were for use in people, the use of similar calculations would yield a value of 29.1 prescriptions per 100 persons per year.

The congressional Office of Technology Assessment estimated that in 1985, 17.6 million lb (8.0 million kg) of antibiotics was prescribed for use in animals alone — for treatment, disease prevention, and growth promotion.2 Assuming that equal quantities (17.6 million lb) were used for people in 1985 and that such use has increased over the past 15 years, we find that there is consistency in the pattern of the various gross estimates.

Nevertheless, Dr. Wilbur's point about our ability to estimate the average use per person or use per prescription without available data is well taken. Referring to data from the Office of Technology Assessment, the Institute of Medicine in 1998 reported annual antibiotic use in humans in different terms: approximately 190 million defined daily doses were used in hospitals, and approximately 145 million courses were used in the community.3 If one third to one half of the 35 million patients who are hospitalized in the United States each year receive antibiotics, the number of defined daily doses would be 16.3 to 10.9 per person. However, some patients in critical care units and others who have immunosuppression with fever and neutropenia receive multiple antibiotics at high doses for weeks at a time, far exceeding the average use. In a study of eight intensive care units, Gaynes and Monnet reported that the use of vancomycin ranged from 10 to 70 defined daily doses per 1000 patient-days and the use of third-generation cephalosporins ranged from 17 to 154 defined daily doses per 1000 patient-days.4 Wide variation in use among outpatients must also occur.

We conclude that a large tonnage of antibiotics is prescribed for people in the United States. Measures of use vary considerably from study to study. Most important, precise profiles of the distribution of antibiotics and the number of prescriptions written for people in the community, hospital, or extended care facilities are unknown and await accurate national surveillance.

Richard P. Wenzel, M.D.
Michael B. Edmond, M.D., M.P.H.
Virginia Commonwealth University, Richmond, VA 23298-0663

4 References

1
Levy SB. Antibiotic resistance: an ecological imbalance. In: Ciba Foundation. Antibiotic resistance: origins, evolution, selection and spread. Chichester, England: John Wiley, 1997:1-14.

2
Antibiotics in animal husbandry. In: Office of Technology Assessment, Congress of the United States. Impacts of antibiotic-resistant bacteria. Washington, D.C.: Government Printing Office, 1995:155-66.

3
Institute of Medicine, Forum on Emerging Infections. Antimicrobial resistance: issues and options — workshop report. Washington, D.C.: National Academy Press, 1998:40.

4
Gaynes R, Monnet D. The contribution of antibiotic use on the frequency of antibiotic resistance in hospitals. In: Ciba Foundation. Antibiotic resistance: origins, evolution, selection and spread. Chichester, England: John Wiley, 1997:47-60.

Citing Articles (1)