Correspondence
Postexposure Treatment of HIV
N Engl J Med 1997; 337:499-501August 14, 1997
- Article
To the Editor:
Katz and Gerberding (April 10 issue)1 raise questions about public health policy related to the costs of antiretroviral therapy given immediately after an exposure to the human immunodeficiency virus (HIV) to prevent seroconversion in persons at high risk for HIV infection. Using their data and the techniques of decision analysis, we examined the cost effectiveness of prophylaxis against HIV.
We compared no prophylaxis with prophylaxis with zidovudine; zidovudine and lamivudine; and zidovudine, lamivudine, and indinavir — all given for four weeks at average wholesale costs of $260, $469, and $931, respectively. The cost of initial and follow-up HIV testing after a needle stick and the cost of an initial office visit would be $440. Patients given prophylaxis would each incur an additional cost of $152 for extra visits and tests. A case–control study2 has found that zidovudine has 79 percent efficacy in preventing seroconversion.
With needle-stick injuries (for which the associated seroconversion risk is 0.0032), zidovudine prophylaxis has a marginal cost-effectiveness ratio of $163,000 per seroconversion averted. For risks ranging from 0.0003 to 0.032, the corresponding marginal cost effectiveness of such therapy ranges from $1.7 million to $16,000 per seroconversion averted.
Seroconversion would shorten the life of a person 25 to 30 years old (with a life expectancy of about 50 more years) by 30 to 40 years. Assuming that society is willing to pay $50,000 to increase that person's life expectancy by 1 year, it should be willing to spend at least $2 million to prevent a seroconversion (that is, to increase the life expectancy by 40 years at $50,000 per year). Thus, zidovudine prophylaxis would appear to be reasonable even with low seroconversion rates.
For lamivudine and indinavir there are no data on prophylactic efficacy. Because of this uncertainty, we performed a two-way sensitivity analysis for a needle-stick injury, assuming that society is willing to pay $2 million to prevent one seroconversion (Figure 1Figure 1
Two-Way Sensitivity Analysis of Efficacy, Comparing Prophylaxis with Zidovudine Alone, with Zidovudine and Lamivudine, and with Zidovudine, Lamivudine, and Indinavir.). The estimated efficacies of lamivudine and indinavir determine the recommended strategy of prophylaxis. If these efficacies are low, zidovudine alone should be given. If the efficacy of lamivudine exceeds 65 percent, then only zidovudine and lamivudine should be given. If the efficacy of lamivudine is between 15 percent and 65 percent, the optimal strategy depends on the efficacy of indinavir.For needle-stick injuries, the cost of zidovudine prophylaxis per seroconversion prevented appears to be reasonable. Adding either lamivudine or lamivudine and indinavir raises the marginal cost-effectiveness ratio, and the optimal prophylaxis depends on the efficacy of the drugs, the risk of seroconversion, and society's willingness to pay. Our analysis does not consider the serious, difficult, and controversial social issues (increased high-risk behavior, antiretroviral-drug resistance, and restrictions on treatment) discussed by Katz and Gerberding. Cost-effectiveness analysis is only one of many tools used to make public health policy.
2 ReferencesRoger W. Li, M.D., M.P.H.
John B. Wong, M.D.
New England Medical Center, Boston, MA 021111
Katz MH ,Gerberding JL . Postexposure treatment of people exposed to the human immunodeficiency virus through sexual contact or injection-drug use. N Engl J Med 1997;336:1097-1100
Full Text | Web of Science | Medline2
Case-control study of HIV Seroconversion in health-care workers after percutaneous exposure to HIV-infected blood -- France, United Kingdom, and United States, January 1988-August 1994
Medline
To the Editor:
Katz and Gerberding review the limited evidence and recommend prophylaxis after high-risk exposures, including unprotected anal or vaginal intercourse, receptive oral intercourse with ejaculation, and the sharing of needles with an infected partner.1 Postexposure prophylaxis may also be indicated when the HIV status of the partner is unknown but there is a substantial risk of infection.
The authors note that a formal analysis of economic efficiency is needed. We propose a basic framework for assessing the cost effectiveness of postexposure prophylaxis. In our model, the probability that such prophylaxis prevents infection equals the product (A) of the probability that the partner is infected; the per-contact probability of transmission associated with a particular sexual act or drug injection; the effectiveness of chemoprophylaxis in preventing infection; and the percentage of people who complete the one-month regimen of zidovudine and lamivudine (plus a protease inhibitor, when appropriate).2 The cost–benefit ratio (the cost per quality-adjusted year of life saved) is (C - AT)/AQ, where C is the cost of postexposure prophylaxis, T is the lifetime cost of the medical care of a person with HIV disease; and Q is the number of quality-adjusted years of life saved by preventing one HIV infection. The following parameter values were used: E, 79 percent1; K, 67 percent1; C, $6001; T, $195,188 (in 1996 dollars)3; and Q, 11.23.3 The per-contact probabilities of transmission were those given by Katz and Gerberding.1
Table 1Table 1
Cost–Benefit Ratios for Postexposure Prophylaxis after a Sexual or Injection-Drug Exposure to HIV. shows the results of the analysis. Although there is no universal criterion, society is generally willing to spend up to $50,000, and in some cases $150,000, per quality-adjusted year of life saved.4 Thus, postexposure prophylaxis would not be especially cost effective for sexual exposures other than receptive anal intercourse, except perhaps when it is nearly certain that the partner is infected. In the case of receptive anal intercourse, postexposure prophylaxis is highly cost effective, even cost-saving in some cases (as indicated by a negative cost–benefit ratio). In the case of injection-associated exposures, postexposure prophylaxis is also likely to be cost effective, especially when there is a high probability that the partner is infected. Although the cost–benefit ratio is somewhat sensitive to the particular values assigned to the parameters, its implications for policy are generally robust.4 ReferencesSteven D. Pinkerton, Ph.D.
David R. Holtgrave, Ph.D.
Frederick R. Bloom, Ph.D.
Medical College of Wisconsin, Milwaukee, WI 532021
Katz MH ,Gerberding JL . Postexposure treatment of people exposed to the human immunodeficiency virus through sexual contact or injection-drug use. N Engl J Med 1997;336:1097-1100
Full Text | Web of Science | Medline2
Pinkerton SD, Holtgrave DR, Pinkerton HJ. Cost-effectiveness of chemoprophylaxis after occupational exposure to HIV. Arch Intern Med (in press).
3
Holtgrave DR, Pinkerton SD. Updates of cost of illness and quality of life estimates for use in economic evaluations of HIV prevention programs. J Acquir Immune Defic Syndr Hum Retrovirol (in press).
4
Fabian R. The Qualy approach. In: Tolley GL, Kenkel D, Fabian R, eds. Valuing health for policy: an economic approach. Chicago: University of Chicago Press, 1994:118-36.
Author/Editor Response
The authors reply:
To the Editor: The decision-analysis models of Li and Wong and Pinkerton et al. support the cost effectiveness of treating people after an exposure to HIV through sexual contact or injection-drug use. The analysis by Li and Wong indicates that zidovudine monotherapy would be cost effective for exposure to HIV through unprotected anal and vaginal intercourse and the use of contaminated injection-drug equipment. Their model also shows that depending on the degree of increased efficacy conferred by lamivudine and indinavir as compared with zidovudine alone, a three-drug regimen may also be cost effective.
In the model by Li and Wong, all the partners are assumed to be HIV-infected. In clinical settings, some people have partners who remain anonymous or refuse HIV testing. The analysis by Pinkerton et al. indicates that the cost effectiveness of treatment after some exposures depends on the probability that the partner is HIV-infected. For example, treatment would be cost effective in a case of receptive vaginal intercourse only if the odds that the source partner was infected exceeded 50 percent.
Both analyses assume that the efficacy of prophylaxis with zidovudine monotherapy after sexual or injection-drug exposures is the same as that after occupational exposures (a 79 percent reduction in the rate of seroconversion). This is a reasonable first assumption, but it must be remembered that there are no data on the efficacy of treatment after sexual or injection-drug exposure. Although studies in animals suggest that the immune response to mucosal exposures is similar to that observed in health care workers who have percutaneous exposures, the size of the viral inoculum and differences in antigen processing between the genital tract and skin may affect the efficacy of treatment. Extrapolation to injection-drug use, in which there may be a direct intravenous injection of a relatively large volume of blood, is also problematic.1
With both sexual and injection-drug exposures, it will be challenging to provide treatment as quickly as occurred in the occupational study of postexposure treatment (median, four hours).2 Delayed treatment may have less efficacy.3 On the other hand, two- or three-drug regimens are likely to be more efficacious than zidovudine monotherapy. Further research is needed to determine the efficacy of treatment after sexual and injection-drug exposures and to evaluate the other potential benefits and costs of treatment.
3 ReferencesMitchell H. Katz, M.D.
San Francisco Department of Public Health, San Francisco, CA 94102-6033Julie Louise Gerberding, M.D., M.P.H.
University of California, San Francisco, San Francisco, CA 941431
Gaughwin MD ,Gowans E ,Ali R ,Burrell C . Bloody needles: the volumes of blood transferred in simulations of needlestick injuries and shared use of syringes for injection of intravenous drugs. AIDS 1991;5:1025-1027
CrossRef | Web of Science | Medline2
Tokars JI ,Marcus R ,Culver DH , et al. Surveillance of HIV infection and zidovudine use among health care workers after occupational exposure to HIV-infected blood: the CDC Cooperative Needlestick Surveillance Group. Ann Intern Med 1993;118:913-919
Web of Science | Medline3
Martin LN ,Murphey-Corb M ,Soike KF ,Davison-Fairburn B ,Baskin GB . Effects of initiation of 3-azido, 3
CrossRef | Web of Science | Medline
- Citing Articles (2)
