496 • CID 2009:48 (15 February) • CORRESPONDENCE Table 1. Survival at 2 years, stratified by type of surgery. Variable No. of patients Monotherapy Combination therapy P Percentage of recipients Survival rate, % of patients (95% CI) Percentage of recipients Survival rate, % of patients (95% CI) 2-Stage exchange 17 53 100 (74–100) 47 87 (64–100) .3 Debridement and retention of the prosthesis 5 100 80 (45–100) … … Resection arthroplasty 22 64 77 (48–100) 36 74 (42–100) .6 We share Nsutebu and Hobson’s con- cerns about the issue of selection bias and the presence of potential confounders, given the observational nature of our study, and we stated those limitations in the discussion section of our work [2]. In Nsutebu and Hobson’s view, loosening of the prosthesis and 2-stage exchange sur- gery are factors associated with more-se- vere disease. Both of these variables were more common in the combination ther- apy group. To our knowledge, peripros- thetic loosening has not been identified as a marker of disease severity. In our study population, the presence of loosening was not a risk factor for failure ( and P p .5 in the monotherapy and combi- P p .3 nation therapy groups, respectively). A stratified analysis based on the presence of loosening did not reveal a difference in outcome between monotherapy and com- bination therapy groups ( ). Fur- P p .08 thermore, when therapy was stratified by type of surgery, no difference in outcome between the combination therapy group and the monotherapy group was observed (table 1). We do believe, on the basis of our study and clinical experience, that monotherapy may be sufficient, in combination with ag- gressive surgical treatment and the use of local antimicrobial therapy, for the man- agement of the majority of penicillin-sus- ceptible enterococcal prosthetic joint in- fections. Combination therapy may still be the best choice for the treatment of pa- tients with concomitant bacteremia or metastatic infection. A randomized con- trolled trial will further strengthen this conclusion, but given the rarity of this disease, such a trial will be difficult to conduct. Acknowledgments Potential conflicts of interest All authors: no conflicts. Odette C. El Helou, Elie F. Berbari, and Douglas R. Osmon Department of Internal Medicine, Division of Infectious Diseases, Mayo Clinic College of Medicine, Rochester, Minnesota References 1. Nsutebu EF, Hobson R. Is combination sys- temic therapy superior to monotherapy for en- terococcal prosthetic joint infection? Clin Infect Dis 2009; 48:495 (in this issue). 2. El Helou OC, Berbari EF, Marculescu CE, et al. Outcome of enterococcal prosthetic joint in- fection: is combination systemic therapy su- perior to monotherapy? Clin Infect Dis 2008; 47:903–9. 3. Baddour LM, Wilson WR, Bayer AS, et al. In- fective endocarditis: diagnosis, antimicrobial therapy, and management of complications: a statement for healthcare professionals from the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovas- cular Disease in the Young, and the Councils on Clinical Cardiology, Stroke, and Cardiovas- cular Surgery and Anesthesia, American Heart Association—endorsed by the Infectious Dis- eases Society of America. Circulation 2005; 111: e394–434. 4. Olaison L, Schadewitz K. Enterococcal endo- carditis in Sweden, 1995–1999: can shorter therapy with aminoglycosides be used? Clin In- fect Dis 2002; 34:159–66. Reprints or correspondence: Dr. Elie F. Berbari, 200 First St. SW, Rochester, MN 55905 (berbari.elie@mayo.edu). Clinical Infectious Diseases 2009; 48:495–6  2009 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2009/4804-0024$15.00 DOI: 10.1086/596547 Tuberculosis Transmission from Patients with Smear- Negative Pulmonary Tuberculosis in Sub-Saharan Africa To the Editor—We read with interest the report by Tostmann et al. [1] of a na- tional study of tuberculosis (TB) trans- mission in The Netherlands from 1996 through 2004. The relative TB transmis- sion rate among patients with smear-neg- ative, culture-positive pulmonary disease, compared with patients with smear-pos- itive disease, was found to be 0.24. Overall, 13% of TB transmission events were at- tributable to source patients with sputum smear-negative, culture-positive disease. These important findings are in close agreement with similar studies from San Francisco, California, and Vancouver, British Columbia [2, 3], collectively show- ing that in high-income countries, 10%– 20% of TB transmission at the population level is attributable to source cases with smear-negative pulmonary TB. Tostmann et al. [1] speculated on the relevance of their data for countries in which HIV in- fection is endemic and rates of smear-neg- ative TB disease are high. In sub-Saharan Africa, HIV infection has had a devastating impact on TB con- trol [4, 5]. In a study of a community in a township in Cape Town, South Africa, for example, the antenatal HIV seroprev- alence rate is ∼30%, and the annual TB notification rate has increased to 11500 cases per 100,000 population [5]—almost 200-fold higher than TB rates in The Netherlands [4]. This has been associated with a major and disproportionate in- crease in the rate of smear-negative disease among HIV-infected individuals [5]. Moreover, the prevalence of undiagnosed TB (a key determinant of transmission) among HIV-infected patients in this com- munity is very high, and disease duration before diagnosis is prolonged [6]. In view of these observations, transmission attrib- by guest on May 18, 2011 cid.oxfordjournals.org Downloaded from