RSS Feeds
Monitored Isoniazid Prophylaxis for Low-Risk Tuberculin Reactors Older Than 35 Years of Age: A Risk-Benefit and Cost-Effectiveness Analysis
- Shelley R. Salpeter, MD;
- Gillian D. Sanders, AB;
- Edwin E. Salpeter, PhD; and
- Douglas K. Owens, MD, MSc
- From Santa Clara Valley Medical Center, San Jose, California; Stanford University, Stanford, California; Cornell University, Ithaca, New York; and Veterans Affairs Palo Alto Health Care System, Palo Alto, California. Acknowledgments: The authors thank Robert F. Nease, Thomas G. Kelsey, and Lyn Dupre for comments on the manuscript and Andria Cardinalli for developing cost estimates. Grant Support: In part by the Santa Clara Valley Medical Center, San Jose, California, and by a Career Development Award from the Veterans Affairs Health Services Research and Development Service (Dr. Owens). Requests for Reprints: Shelley R. Salpeter, MD, Santa Clara Valley Medical Center, 2400 Moorpark Avenue, Suite 118, San Jose, CA 95128. Current Author Addresses: Dr. S.R. Salpeter: Santa Clara Valley Medical Center, 2400 Moorpark Avenue, Suite 118, San Jose, CA 95128.
Abstract
Background: Isoniazid chemoprophylaxis effectively prevents the development of active infectious tuberculosis. Current guidelines recommend withholding this prophylaxis for low-risk tuberculin reactors older than 35 years of age because of the risk for fatal isoniazid-induced hepatitis. However, recent studies have shown that monitoring for hepatotoxicity can significantly reduce the risk for isoniazid-related death.
Objective: To evaluate the effectiveness and cost-effectiveness of monitored isoniazid prophylaxis for low-risk tuberculin reactors older than 35 years of age.
Design: A Markov model was used to compare the health and economic outcomes of prescribing or withholding a course of prophylaxis for low-risk reactors 35, 50, or 70 years of age. Subsequent analyses evaluated costs and benefits when the effect of transmission of Mycobacterium tuberculosis to contacts was included.
Measurements: Probability of survival at 1 year, number needed to treat, life expectancy, and cost per year of life gained for individual persons and total population.
Results: Isoniazid prophylaxis increased the probability of survival at 1 year and for all subsequent years. For 35-year-old, 50-year-old, and 70-year-old tuberculin reactors, life expectancy increased by 4.9 days, 4.7 days, and 3.1 days, respectively, and costs per person decreased by $101, $69, and $11, respectively. When the effect of secondary transmission to contacts was included, the gains in life expectancy per person receiving prophylaxis were 10.0 days for 35-year-old reactors, 9.0 days for 50-year-old reactors, and 6.0 days for 70-year-old reactors. Costs per person for these cohorts decreased by $259, $203, and $100, respectively. The magnitude of the benefit of isoniazid prophylaxis is moderately sensitive to the effect of isoniazid on quality of life. The hypothetical provision of isoniazid prophylaxis for all low-risk reactors older than 35 years of age in the U.S. population could prevent 35 176 deaths and save $2.11 billion.
Conclusions: Monitored isoniazid prophylaxis reduces mortality rates and health care costs for low-risk tuberculin reactors older than 35 years of age, although reductions for individual patients are small. For the U.S. population, however, the potential health benefits and economic savings resulting from wider use of monitored isoniazid prophylaxis are substantial. We should consider expanding current recommendations to include prophylaxis for tuberculin reactors of all ages with no contraindications.
After declining in incidence for decades, tuberculosis has reemerged as a potent threat to the public health [1]. Isoniazid was introduced in 1952 for the treatment of active tuberculosis; in 1955, its use was expanded successfully to include chemoprophylaxis of inactive tuberculous disease [2] and a campaign for widespread prophylaxis was instituted [2-8]. In the early 1970s, however, enthusiasm for chemoprophylaxis was dampened by reports of several deaths due to isoniazid hepatotoxicity [9-13].
After this risk was identified, the American Thoracic Society recommended in 1974 that clinicians 1) restrict the use of prophylaxis to tuberculin reactors younger than 35 years of age or tuberculin reactors older than 35 years of age who were at increased risk for activation and 2) withhold isoniazid treatment from patients with acute liver disease [14]. A 1983 revision of the guidelines recommended that clinicians evaluate and periodically monitor liver function and discontinue treatment if aminotransferase levels exceed three to five times normal values [15]. Investigators believed that implementation of these guidelines would almost eliminate fatal hepatotoxicity related to isoniazid prophylaxis [16].
Over the past 15 years, studies have both supported and questioned the widespread use of isoniazid prophylaxis for tuberculin reactors of all ages [17-25]. All of these studies used data from the 1970s, data produced before the advent of routine monitoring for isoniazid-induced hepatotoxicity [9, 12, 13, 26-29]. Pooled data from large trials of unmonitored prophylaxis indicate that the average rate of fatal hepatitis was 0.01%, or 1 per 10 000 [9, 11, 13, 30-35].
Two recent studies [36, 37] reviewed data on patients who received prophylaxis according to the 1983 guidelines and found that the rate of fatal hepatitis was 1 to 1.7 per 100 000 for persons of all ages and approximately 2 per 100 000 for persons older than 35 years of age [36, 37]. Because this mortality rate is significantly lower than that found in the era before routine monitoring for liver toxicity, we reevaluated the use of monitored isoniazid prophylaxis in low-risk tuberculin reactors older than 35 years of age. Our objective was to quantitatively evaluate the risks and benefits of isoniazid prophylaxis in these reactors and to evaluate the cost-effectiveness of isoniazid prophylaxis, if it provided net health benefits. To accomplish these objectives, we developed a decision model that evaluated the risks and benefits to an individual patient and the public health benefit resulting from reduced transmission of Mycobacterium tuberculosis.
Methods
We used a decision model to evaluate the effectiveness and cost-effectiveness of a 6-month course of monitored isoniazid chemoprophylaxis for tuberculin reactors older than 35 years of age who have normal chest radiographs and are not at increased risk for tuberculosis activation. We evaluated monitored isoniazid prophylaxis in persons who were 35, 50, and 70 years of age.
Decision Model
We developed a Markov model (Figure 1, Table 1 and Table 2) with SMLTree software (version 2.9, 1990, J. Hollenberg, New York, New York) [51, 52] to track hypothetical cohorts of patients that either received or did not receive isoniazid prophylaxis. Patients who received isoniazid prophylaxis were at risk for death from isoniazid-induced hepatitis. We assumed that patients who did not die of isoniazid-induced hepatitis received either a full course of isoniazid with the full benefit of protection or an incomplete course (as a result of side effects or noncompliance) with only partial protection. All patients were at risk for developing active tuberculosis and for dying of tuberculosis or other causes. We assumed that all deaths from tuberculosis occurred within the first year after development of active disease. Life expectancies were obtained from the National Center for Health Statistics [43]. In our initial analyses, we assumed that the costs and benefits of isoniazid prophylaxis pertained only to the persons who received prophylaxis. In subsequent analyses, we included the reduction of transmission of M. tuberculosis to other members of the population in our evaluation of costs and benefits. We discounted health and economic outcomes by 3% annually [44].
Development of Active Tuberculosis
In large-population studies of tuberculin reactors whose duration of reactivity is unknown and who have more than 10-mm induration and normal chest radiographs, the annual activation rate, averaged over several years, was 0.07% to 0.1% [2, 4, 53]. One such study [53] found that the annual incidence of tuberculosis activation, recorded by years since the initial tuberculin reactivity was discovered, was 0.09% for 0 to 10 years, 0.06% for 10 to 15 years, and 0.04% for more than 15 years. The average activation rate for reactors older than 40 years of age was 0.095% per year [39]. To estimate the rate of activation over time in persons 40 years of age or older, we used the same time distribution for activation from patients of all ages to calculate the activation rates over time, assuming an average activation rate of 0.095% per year (Table 1).
Studies in some populations have found activation rates (0.04% to 0.06% per year) lower than those used in our base-case analysis [54-56]. These studies included either persons who had already received isoniazid prophylaxis or tuberculin reactors with 5- to 10-mm induration. We evaluated such rates in sensitivity analyses. Our estimates of the case-fatality rate in persons who have reactivation are based on the literature [40, 41, 57, 58] (Table 1).
Effectiveness of Isoniazid
No studies have evaluated the efficacy of a 6-month course of isoniazid for tuberculin reactors with more than 10-mm induration and normal chest radiographs. Many studies have evaluated courses of isoniazid prophylaxis lasting 6 months to 1 year and have included patients with abnormal chest radiographs or 5-mm induration on a tuberculin test; these studies indicate that isoniazid prophylaxis reduced the risk for activation by 69% to 93% [2, 4, 7, 30, 31, 38] and by 85% to 93% in patients who complete a full course [2, 31, 38]. Results of a large trial [31] indicated that if the chest radiograph showed no significant abnormalities, a 6-month course of isoniazid prophylaxis was as efficacious as a 12-month course. This study found that a 3-month course of isoniazid prophylaxis reduced the rate of activation of tuberculosis by 30%. For our analysis, we assumed that 80% of the patients received a full 6-month course and 20% received an incomplete (3-month) course of isoniazid prophylaxis [31, 36, 38]. We used conservative estimates of the risk reduction: 85% for a full course and 15% for a partial course. These assumptions resulted in an average reduction in the risk for activation of tuberculosis of 71% for all patients.
Deaths Due to Isoniazid-Related Hepatitis
A review of published and unpublished data on isoniazid prophylaxis for patients monitored according to the current guidelines showed the rate of fatal hepatitis to be 2 in 202 497, or 0.001% [36]. This study included data obtained from tuberculosis control officers in large cities throughout the United States and from 29 published trials that evaluated mortality rates among patients receiving monitored prophylaxis [8, 28, 38, 59-81]. The upper limit of the 95% CI for the true incidence of death, calculated by using the Poisson distribution, was 0.003%, or 6 per 200 000. From the available data on age, the estimated mortality rate for patients older than 35 years of age was 1 in 43 334, or 0.002% [36]. These deaths occurred during the first year after prophylaxis was initiated. Similarly, another recent study [37] estimated a mortality rate for monitored prophylaxis of 1.7 per 100 000 persons of all ages who receive isoniazid. We assumed that the mortality rate from isoniazid-induced hepatitis was 0.001% in persons 35 years of age and 0.002% in persons older than 35 years of age [36], and we examined higher rates [9, 11, 13, 30-35] in sensitivity analyses.
Quality of Life for Intermediate Health States
Our model included intermediate health states that may be associated with decrements in quality of life, including nonfatal and fatal hepatitis, nonfatal and fatal tuberculosis, and the use of a complete or an incomplete course of isoniazid prophylaxis. To adjust for such morbidity, our model included decrements in length of life for each of these health states (Table 1), an approach that is equivalent to multiplying the time spent in a health state by a quality adjustment. Because few empirical data about quality of life in these health states are available, we extensively examined the influence of quality of life in sensitivity analyses.
Costs
Our cost estimates for isoniazid prophylaxis, treatment of active tuberculosis, contact tracing [50, 82], and related expenses are shown in Table 2. We included the costs of isoniazid; visits with a registered nurse; liver function tests; and treatment of complications, including therapy for hepatitis and related gastrointestinal symptoms, associated physician visits, and hospitalization (in 0.05% of cases) [36, 45-50]. The treatment of active tuberculosis involves hospitalization (mean length of stay, 20 days) in 77% of cases [45]. Most patients with active tuberculosis receive a 6-month course of therapy, which involves medications; physician visits; and, in certain cases, daily observed treatment [45]. We also included, where appropriate, treatment costs for cases of active tuberculosis due to secondary transmission. We used published estimates of the costs (not charges) of hospitalization ($15 814) [45], outpatient treatment, and contact tracing for tuberculosis for our base-case analysis [36, 45, 50]. In sensitivity analyses, we examined a lower cost for tuberculosis-associated hospitalization based on an estimate from Medicare reimbursement [46-49]. We adjusted costs to 1996 U.S. dollars by using the Gross Domestic Product deflator [83].
Transmission Model
Isoniazid prophylaxis confers a public health benefit by reducing transmission of M. tuberculosis organisms. To assess the importance of this benefit, we used a transmission model that we had developed previously [42] to estimate how many cases of tuberculosis as a result of transmission would be prevented by isoniazid prophylaxis. The transmission model estimated that each case of active tuberculosis ultimately resulted in 1.2 additional cases of tuberculosis due to multiple successive transmissions over time [42]. We estimated that each death from tuberculosis in the patients who had the disease as a result of transmission resulted in a loss of life expectancy of 19.8 years [41, 43].
Population Study
We evaluated the effect of a policy that recommended isoniazid prophylaxis for all low-risk tuberculin reactors older than 35 years of age. The estimated prevalence of tuberculin reactivity ranges from 2.5% to 40% in small subgroups of the U.S. population [84-87] and from 4% to 6% for the entire U.S. population [42, 88]. For our analysis, we adopted 5% as the average prevalence in the U.S. population of 260 million people [89], for a total of 13 million tuberculin reactors. Of these reactors, it has been estimated that 90% have normal chest radiographs and thus are at low risk for activation of tuberculosis [2]. We estimated that there were 2.7 million low-risk reactors 35 to 50 years of age, 4.9 million 50 to 70 years of age, and 4.0 million older than 70 years of age, and we validated these estimates by comparing them with reported age-specific case rates in the United States [90]. Because little direct evidence is available about the number of tuberculin reactors in the United States, we examined a wide range of estimates in sensitivity analyses.
Role of the Funding Organizations
The funding agencies for our study had no role in the collection, analysis, or interpretation of data or in deciding whether the study results could be published.
Results
Isoniazid Prophylaxis
Use of isoniazid prophylaxis compared with no prophylaxis increased the probability of survival at 1 year and for all subsequent years (Table 3). Survival during the first year after initiation of prophylaxis increased because the number of deaths averted by the prevention of active tuberculosis was greater than the number of deaths caused by isoniazid-related hepatitis. Prevention of a tuberculosis-related death required administering isoniazid prophylaxis to 529 to 654 patients (Table 3). Prophylaxis increased life expectancy by 4.9 days for 35-year-old patients, 4.7 days for 50-year-old patients, and 3.1 days for 70-year-old patients. It reduced expected health care expenditures by $101 per person for 35-year-old patients, $69 per person for 50-year-old patients, and $11 for 70-year-old patients (Table 3, Table 5 and Table 4). Total costs were reduced because the savings from averted cases of active tuberculosis outweighed the costs of isoniazid prophylaxis (Table 4).
When we included the effect of secondary transmission of tuberculosis to potential contacts, the benefits of isoniazid prophylaxis increased, as did the cost savings (Table 3, Table 5 and Table 4). Prevention of one tuberculosis-related death required giving isoniazid prophylaxis to 268 to 435 patients (Table 3). The total gain in life expectancy for a person in the 35-year-old cohort who received isoniazid prophylaxis, and for his or her potential contacts, was 10.0 days; total expenditures were reduced by $259 per index case. Similarly, the use of isoniazid prophylaxis in the 50-year-old cohort increased life expectancy for the cohort and contacts by 9.0 days and saved $203 per index case (Table 4). In the 70-year-old cohort, patients and their contacts lived 6.0 days longer with isoniazid prophylaxis and expenditures decreased by $100 for each index case that received prophylaxis.
Giving isoniazid prophylaxis to all low-risk tuberculin reactors older than 35 years of age averted 35 176 deaths and reduced medical expenditures by $2.11 billion. If the prevalence of tuberculin reactivity was lower or higher than our base-case estimate, the number of deaths averted and the medical expenditures changed proportionally. For example, if the prevalence of tuberculin reactivity was only 3%, prophylaxis averted 21 105 deaths and reduced expenditures by $1.27 billion. Complete adherence to widespread screening and prophylaxis is unlikely. We estimated that administering isoniazid prophylaxis to only 20% of the low-risk tuberculin reactors (assuming a prevalence of 5%) would avert 7035 deaths in the United States and would save $422 million.
Sensitivity Analyses
Three variables were of primary importance in sensitivity analyses: the probability of fatal isoniazid-induced hepatitis, the efficacy of isoniazid in reducing the probability of tuberculosis activation, and the average annual probability of tuberculosis activation. Compared with our base-case estimates of the probability of isoniazid-induced death, 0.001% for 35-year-old patients and 0.002% for 50-year-old and 70-year-old patients (Table 1), the probability of isoniazid-induced death must exceed 0.12% for 35-year-old patients, 0.14% for 50-year-old patients, and 0.16% for 70-year-old patients (including the benefit of reduced transmission) for no prophylaxis to be preferred. Even when the benefit of secondary transmission of tuberculosis was excluded, the probability of isoniazid-induced death had to be at least 35 times higher than our base-case estimate for the no-prophylaxis strategy to be preferred (Figure 2, top). Because the isoniazid-associated mortality rate is 0.001% to 0.002%, monitored prophylaxis increased life expectancy even when the efficacy of isoniazid was reduced substantially from our base-case estimate (Figure 2, upper middle).
The benefit of isoniazid prophylaxis was derived from a reduction in the rate of activation of tuberculosis without isoniazid prophylaxis. The lower the rate of activation without prophylaxis, the less benefit prophylaxis could provide. We examined rates of activation that were substantially lower than our base-case estimate (Figure 2, lower middle). In 70-year-old patients (excluding the benefit from reduced transmission), isoniazid prophylaxis increased life expectancy if the annual activation rate without prophylaxis was at least 0.0142%-a value that is substantially lower than both the observed rates and our base-case estimate (0.095% for patients older than 40 years of age). In all other cohorts analyzed, prophylaxis conferred even greater benefits.
If the cost of isoniazid prophylaxis is increased by 50%, prophylaxis still reduces expenditures (except in the 70-year-old cohort when transmission is excluded; for this group, each additional quality-adjusted life-year costs $4984). When we used estimated hospitalization costs for tuberculosis that were substantially lower than our base-case estimate, isoniazid prophylaxis still reduced expenditures for all age groups when we included the effects of secondary transmission of tuberculosis organisms. When we excluded transmission, prophylaxis reduced expenditures for 35-year-old and 50-year-old patients but cost $1561 per life-year saved in 70-year-old patients (Figure 2, bottom).
Our analysis included a small disutility for taking medication (a reduction in life expectancy of half a day) and decrements in quality of life for the development of hepatitis or tuberculosis. Isoniazid was preferred unless the disutility for isoniazid prophylaxis became approximately 5.9 days (secondary transmission excluded) or 11.6 days (secondary transmission included) for 35-year-old patients. Stated differently, the quality of life that results from receiving isoniazid (without hepatitis) would have to be less than 0.97 (transmission excluded) or 0.935 (transmission included) for prophylaxis not to be preferred. By comparison, patients rate the utility of warfarin therapy, including the inconvenience of periodic laboratory tests, as 0.997 (83 patients), which is well above these thresholds [91]. Our finding that isoniazid prophylaxis was preferred was not sensitive to other quality adjustments.
Discussion
Using recent information about the safety of monitored isoniazid prophylaxis, we reevaluated the usefulness of prophylaxis in low-risk tuberculin reactors older than 35 years of age. Our findings indicate that relative to no prophylaxis, monitored isoniazid prophylaxis increased the probability of survival during the first year (the period during which all cases of fatal isoniazid-induced hepatitis would occur) and during all subsequent years, decreased the lifetime probability of death from active tuberculosis, and increased life expectancy for each age group evaluated. Furthermore, the use of isoniazid reduced medical expenditures in all age groups. Although the increase in life expectancy from prophylaxis is modest for an individual patient (3 to 5 days, secondary transmission excluded), the importance of tuberculosis prevention for the population is underscored by our finding that applying broader indications for prophylaxis could avert more than 7000 deaths and reduce health care expenditures by more than $400 million, even if only 20% of low-risk tuberculin reactors older than 35 years of age receive prophylaxis. Thus, our analysis suggests that we should reevaluate the current recommendation that clinicians restrict isoniazid prophylaxis to patients 35 years of age or younger and high-risk reactors older than 35 years of age [15]. Such a recommendation may not improve patients' health and may increase total health care expenditures.
Our analysis differs from previous studies [18, 20] in three important respects. First, we used rates of isoniazid-related death that reflect the safety of monitored isoniazid prophylaxis. Salpeter [36] reviewed reports on 202 497 patients who received isoniazid prophylaxis. Only 2 of these patients died of isoniazid-induced hepatitis, for an overall mortality rate of 0.001% and an estimated mortality rate of 0.002% in patients older than 35 years of age [36]. Another study [37] found similarly low rates. The rate of fatal hepatitis in these studies is equivalent to the rate of death from acute hepatitis and acute liver necrosis from any cause in the general population, which is 0.001% for persons of all ages and 0.002% for persons older than 35 years of age [90]. Second, we formally evaluated the effect of secondary transmission of M. tuberculosis on the benefits and costs of prophylaxis. Including this public health consequence approximately doubled the estimated benefit conferred by isoniazid prophylaxis. Finally, we evaluated the economic implications of isoniazid prophylaxis, including the additional costs of monitored prophylaxis and the costs and benefits from reduced secondary transmission of M. tuberculosis. Together, these factors enabled us to comprehensively assess the costs and benefits of monitored isoniazid prophylaxis.
Current guidelines recommend isoniazid prophylaxis for tuberculin reactors older than 35 years of age who are at high risk for activation of their latent tuberculous infection [15]. Such reactors include persons who are close contacts of patients with active pulmonary tuberculosis, persons whose skin tests converted recently, persons who have abnormal chest radiographs consistent with old tuberculosis, alcoholic persons, and intravenous drug users [92]. In addition, medical conditions that increase the risk for activation of tuberculosis are immunosuppression (including that caused by corticosteroid use); silicosis; end-stage renal disease; diabetes mellitus; malnutrition or notable weight loss; and some malignant neoplasms, including leukemia, lymphoma, and carcinoma of the upper gastrointestinal tract [92]. Studies indicate that persons in these high-risk groups develop active tuberculosis significantly more often than do low-risk reactors, with activation rates as high as 2% to 5% annually [31, 93, 94]. Our analysis evaluated the benefit of isoniazid prophylaxis to low-risk reactors only. The benefit to high-risk reactors older than 35 years of age are significantly higher than the benefits for cohorts evaluated in this study [94, 95].
In some areas of the United States, tuberculosis has reached epidemic proportions [95]. One survey done to evaluate the apparent failure to control tuberculosis found that of patients who were known to be tuberculin reactors before they developed tuberculosis, two thirds had not been offered isoniazid prophylaxis [96]. Another retrospective study performed through the Indian Health Services, whose current recommendation is to offer prophylaxis to reactors of all ages [97], found that approximately 75% of cases of tuberculosis could have been prevented if the standard U.S. guidelines for screening and prophylaxis had been followed. An additional 7% of these patients with tuberculosis were older than 35 years of age and had no known risk factors for tuberculosis activation and thus would have been eligible for prophylaxis under the broader Indian Health Services guidelines [97].
The failure of practitioners to offer isoniazid prophylaxis, even according to current guidelines, may be due to the belief that the toxicity of isoniazid significantly outweighs its benefits. It is also possible that prophylaxis is withheld in many cases because the current guidelines are complex and practitioners may be confused about who should receive prophylaxis.
Our analysis has several limitations. Our estimates of the risk for isoniazid-related death are based on comprehensive but retrospective reviews of the safety of monitored isoniazid prophylaxis [36, 37]. To the extent that isoniazid-related deaths are underreported, these studies may have underestimated the rate of death from isoniazid prophylaxis. However, our sensitivity analyses indicate that for prophylaxis not to be preferred, the rate of fatal hepatitis must be at least 35 times greater than our base-case estimate. Costs of prophylaxis for and treatment of tuberculosis vary depending on the delivery setting; our estimates, although reasonable, are not applicable to all settings. Although we did not explicitly consider the role of false-positive results on tuberculin tests, our estimated activation rates are based on studies of all positive tuberculin tests (with indurations ≥ 10 mm), not just tests with true-positive results. Thus, our estimates of the average benefit for prophylaxis account for the false-positive rate.
Finally, the average benefit of isoniazid prophylaxis for an individual patient is small and modestly sensitive to decrements in quality of life. The decrements in quality of life that we included in the analysis, although reasonable, were arbitrary: Little empirical evidence is available about quality of life for patients while they are receiving isoniazid or for patients in other relevant health states. Therefore, the decision to initiate prophylaxis should be tailored to each individual patient after discussion of the patient's preferences and appraisal of patient-specific factors that may affect the risks and benefits of prophylaxis [24]. For example, clinicians should carefully consider factors that would increase the probability of a false-positive tuberculin test, such as borderline reactivity.
Our analysis indicates that these modest benefits for individual patients, along with reduction in transmission as a result of prophylaxis, provide a compelling public health benefit. Further prospective studies could provide confirmatory evidence of the benefits and risks of isoniazid prophylaxis. However, such information will not be available in the near future, if ever. Although our top priority still should be to emphasize prevention for those high-risk persons who will receive the greatest benefit from isoniazid prophylaxis, our analysis suggests that we should consider expanding and simplifying current recommendations to include monitored isoniazid prophylaxis for tuberculin reactors of all ages who have no contraindications.
Ms. Sanders: Section on Medical Informatics, Medical School Office Building 215, Stanford University School of Medicine, 251 Campus Drive, Stanford, CA 94305.
Dr. E.E. Salpeter: Center for Radiophysics and Space Research, 612 Space Sciences Building, Cornell University, Ithaca, NY 14853.
Dr. Owens: Veterans Affairs Palo Alto Healthcare System, Mail-code 111A, 3801 Miranda Avenue, Palo Alto, CA 94304.
- Copyright ©2004 by the American College of Physicians
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
Related Clinical Content
