Correspondence

Smoking and Pneumococcal Disease

N Engl J Med 2000; 343:219-220July 20, 2000

Article

To the Editor:

On the basis of their population-based, case–control study of invasive pneumococcal disease, Nuorti and coworkers (March 9 issue)1 conclude that cigarette smoking is the strongest independent risk factor for invasive disease in immunocompetent persons between the ages of 18 and 64 years. Unfortunately, the authors did not adequately address the role of undiagnosed human immunodeficiency virus (HIV) infection as a risk factor for pneumococcal disease or as a potential confounder of the observed association between smoking and invasive pneumococcal infection.

As Nuorti and colleagues recently reported in another article, communities with a high prevalence of HIV infection, such as Atlanta and Baltimore, where 82 percent of the patients enrolled in this study lived, have a greater burden of pneumococcal disease than other communities.2 Although patients with known HIV infection were excluded from the study, pneumococcal disease is often the initial manifestation of asymptomatic HIV infection.3

Over a 14-month period in 1997 and 1998, we evaluated HIV seroprevalence among adults with pneumococcal bacteremia who were seen at Boston Medical Center in Boston. Forty patients were admitted and evaluated for HIV infection. Fifty percent of the patients were men; 19 were black, 14 were white, 5 were Hispanic, and 2 were members of other ethnic or racial groups. Their mean age was 47 years. Twenty-four patients (60 percent) were cigarette smokers, and 16 (40 percent) were known to have HIV infection. The HIV status of the other 24 patients was not known, but blinded testing revealed that 6 (25 percent) had HIV infection, as compared with 1 of 24 persons in a control group of hospitalized patients matched for age, sex, and race or ethnic group. Thus, 22 of the 40 patients (55 percent) who presented with pneumococcal bacteremia during this period had HIV infection.

In the study by Nuorti et al.,1 there were substantial differences between patients and control subjects with respect to age, sex, educational level, annual income, insurance status, presence or absence of a history of pneumonia, and presence or absence of an indication for the receipt of pneumococcal vaccine, suggesting a bias in the selection of controls by random-digit dialing. The demographic characteristics of the patients mirror those of persons affected by the urban HIV epidemic, further indicating that undiagnosed HIV infection may have been an important risk factor. Finally, patients with HIV infection may be more likely than the general population to smoke cigarettes. In evaluating risk factors for pneumococcal disease in patients with HIV infection, Gebo et al. reported that 69 percent of 85 patients with pneumococcal disease were current smokers, as were 67 percent of 85 HIV-infected control subjects without pneumococcal infection.4

All patients between the ages of 18 and 64 years who present with pneumococcal infection should be offered HIV testing. Since HIV-infected patients have a poor response to pneumococcal polysaccharide vaccine, their early identification and the appropriate institution of antiretroviral therapy and of trimethoprim–sulfamethoxazole for prophylaxis against pneumocystis infection are important additional strategies in the prevention of invasive pneumococcal disease.5

Catherine A. Fleming, M.D., M.P.H.
Donald E. Craven, M.D.
Boston University School of Medicine, Boston, MA 02118

5 References

1. 1

   Nuorti JP, Butler JC, Farley MM, et al. Cigarette smoking and invasive pneumococcal disease. N Engl J Med 2000;342:681-689
   Full Text | Web of Science | Medline

2. 2

   Nuorti JP, Butler JC, Gelling L, Kool JL, Reingold AL, Vugia DJ. Epidemiologic relation between HIV and invasive pneumococcal disease in San Francisco County, California. Ann Intern Med 2000;132:182-190
   Web of Science | Medline

3. 3

   Chirurgi VA, Edelstein H, McCabe R. Pneumococcal bacteremia as a marker for human immunodeficiency virus infection in patients without AIDS. South Med J 1990;83:895-899
   CrossRef | Web of Science | Medline

4. 4

   Gebo KA, Moore RD, Keruly JC, Chaisson RE. Risk factors for pneumococcal disease in human immunodeficiency virus-infected patients. J Infect Dis 1996;173:857-862
   CrossRef | Web of Science | Medline

5. 5

   1999 USPHS/IDSA guidelines for the prevention of opportunistic infections in persons infected with human immunodeficiency virusMMWR Morb Mortal Wkly Rep 1999;48:1-66
   Medline

Author/Editor Response

The authors reply:

To the Editor: Fleming and Craven make the important point that undiagnosed HIV infection is a risk factor for invasive pneumococcal disease. The incidence of invasive pneumococcal disease in persons with AIDS is extremely high. In Baltimore, one of our study areas, persons infected with HIV accounted for about one third of all nonelderly adults with invasive pneumococcal disease.1

Because HIV infection or AIDS and other immunocompromising conditions are overwhelming risk factors for invasive pneumococcal disease, we excluded immunocompromised persons from our case–control study. Restricting the study population to immunocompetent persons allowed us to control for potential confounding. The HIV status of the study patients was initially determined by a chart review and was subsequently verified during an interview. All the subjects were asked about the presence or absence of HIV infection.

It is possible that some proportion of patients and control subjects had undiagnosed HIV infection, which would be a potential confounder of the association between smoking and pneumococcal disease. Persons with known HIV infection are more likely to smoke than those without HIV infection,2 and one might assume that persons who are unaware that they are infected with HIV smoke as frequently as those who are aware of their infection. However, it is unlikely that undiagnosed HIV infection confounded the association between smoking and pneumococcal disease in our study, for several reasons.

First, the differences identified between patients and control subjects in the univariate analysis with respect to various characteristics (but not age, as Fleming and Craven suggest) are more likely to reflect true differences than a bias in the selection of control subjects. We did not match our study subjects according to these characteristics, many of which are associated with both pneumococcal disease3 and undiagnosed or diagnosed HIV infection. Our multivariate analysis was therefore controlled for a variety of known risk factors for HIV infection, including the study site, age, race, sex, and socioeconomic status. Second, the dose–response relations in our study constitute additional evidence against a confounding effect of undiagnosed HIV infection. It is improbable that undiagnosed HIV infection would be consistently associated with the intensity or duration of smoking or with the interval since the cessation of smoking. Third, it is unlikely that a substantial proportion of otherwise healthy, nonelderly adults at the study sites would not have been evaluated for HIV infection during hospitalization for pneumococcal bacteremia.4 Finally, further support for the association between smoking and pneumococcal disease comes from another recent case–control study, in which smoking was identified as an independent risk factor for invasive pneumococcal disease in a group of patients, all of whom were infected with HIV.5

J. Pekka Nuorti, M.D.
National Public Health Institute, 00300 Helsinki, Finland

Jay C. Butler, M.D.
Centers for Disease Control and Prevention, Anchorage, AK 99508-5902

Robert F. Breiman, M.D.
Centers for Disease Control and Prevention, Atlanta, GA 30333

5 References

1. 1

   Harrison LH, Dwyer DM, Billmann L, Kolczak MS, Schuchat A. Invasive pneumococcal infection in Baltimore, Md: implications for immunization policy. Arch Intern Med 2000;160:89-94
   CrossRef | Web of Science | Medline

2. 2

   Frankel RE, Virata M, Hardalo C, Altice FL, Friedland G. Invasive pneumococcal disease: clinical features, serotypes, and antimicrobial resistance patterns in cases involving patients with and without human immunodeficiency virus infection. Clin Infect Dis 1996;23:577-584
   CrossRef | Web of Science | Medline

3. 3

   Prevention of pneumococcal disease: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Morb Mortal Wkly Rep 1997;46:1-24
   Medline

4. 4

   Garcia-Leoni ME, Moreno S, Rodeno P, Cercenado E, Vicente T, Bouza E. Pneumococcal pneumonia in adult hospitalized patients infected with the human immunodeficiency virus. Arch Intern Med 1992;152:1808-1812
   CrossRef | Web of Science | Medline

5. 5

   Breiman RF, Keller DW, Phelan M, et al. Evaluation of effectiveness of the 23-valent pneumococcal capsular polysaccharide vaccine for HIV-infected patients. Arch Intern Med (in press).