Adherence to combination antiretroviral therapy in HIV disease has been shown to be an important predictor of virologic suppression and of clinical outcomes.1-3 The degree of adherence to medications required for optimal virologic suppression seems to be >95%, a level that is challenging to achieve and to sustain over long periods.1 Studies in HIV disease and in other chronic diseases document substantial decreases in adherence to medications over time.4-7
Various interventions to improve adherence to antiretroviral therapy in HIV have been studied. These include the use of electronic reminders, behavioral training, nursing interventions, and directly observed therapy.8-12 The efficacy of these interventions has ranged from 0% to 35% improvement in virologic suppression over a wide variety of different study periods and time lines.13-19
One of the concerns with instituting new interventions in HIV management in the United States and in other countries is their impact on the overall cost of HIV care.20,21 The current standard of HIV care in the United States includes combination antiretroviral therapy, opportunistic infection prophylaxis, technologically sophisticated laboratory monitoring, and regular preventive primary care as well as the management of opportunistic infections, concordant diseases, and toxicity from medications.22 Costs generally exceed $20,000 per person per year.20,22-24 Thus, any new proposed interventions for care need to be considered in terms of their clinical efficacy and their cost-effectiveness.20,25-28 We hypothesized that improved adherence to first- and second-line regimens might lead to the postponement of expensive subsequent and salvage regimens, including boosted protease inhibitor regimens, leading to a partial or complete offset of intervention costs.
Previously, we have used a simulation model of chronic HIV disease to examine the potential impact and cost-effectiveness of various strategies for HIV care and of potential interventions to improve adherence to antiretroviral therapy.29-31 Our prior cost-effectiveness analysis on adherence interventions was exploratory because of the lack of empiric data at the time. Our objective in the current study was to incorporate data from a recently completed randomized controlled trial of an intensive nursing intervention in HIV-infected patients into a simulation model to estimate the long-term clinical impact and cost-effectiveness of this intervention.
We used data from a randomized controlled clinical trial (HAART to HEART) conducted at 2 urban community health centers in Boston from September 2000 to April 2003. Eligible patients were HIV infected and initiating or changing antiretroviral therapy regimens or had a known history of adherence problems. Specifically, 27% were changing antiretroviral regimens, 12% were resuming antiretroviral regimens, and 4% were starting their first-line antiretroviral regimen. Fifty-seven percent of the subjects self-reported nonadherence. More than 75% of the subjects had a history of substance abuse, a psychiatric diagnosis, or both. Other details of the trial have been reported elsewhere.32 In brief, 54 patients were randomized to receive standard HIV care or standard care and a home nurse intervention (Table 1). The home nurse intervention included 2 1-hour home visits per week for the first 6 weeks of therapy. Data were collected at study baseline and at 3, 6, and 12 months of follow-up. Adherence data were self-reported using 3-day recall, last dose missed, and number of missed doses per week. The primary outcome of interest in the trial was HIV RNA suppression to <400 copies/mL at the study endpoint. All trial data were analyzed using an intent-to-treat approach. The intervention increased viral suppression at 48 weeks by 63%, from 27% to 44%.
Data from the trial were incorporated into the “Cost-Effectiveness of Preventing AIDS Complications” (CEPAC) model. The CEPAC model is a computer-based state transition model of HIV disease. It has previously been used to project the long-term cost and cost-effectiveness of various HIV interventions, including prophylaxis for opportunistic infections, use of combination antiretroviral therapy, use of genotype testing to guide subsequent treatment regimen choice, and the role of routine voluntary HIV testing.29-31,33-36
The model has been described in detail elsewhere.29-31 Briefly, disease progression is characterized as a sequence of monthly transitions from one health “state” to another. Unique states are defined by CD4 cell counts and HIV RNA levels. In addition, the model accounts for current antiretroviral therapy regimen and prophylaxis for opportunistic infections. The main determinant of clinical outcomes in the model is CD4 cell count; baseline CD4 cell count and HIV RNA level determine the trajectory of CD4 cell count decline over time.37 If antiretroviral therapy is successful, HIV RNA is suppressed and CD4 cell count rises. When virologic failure occurs, HIV RNA rises, CD4 cell count then declines, and the risks of opportunistic infections and death increase.37
The efficacy of each antiretroviral regimen, defined as the proportion of patients with suppressed HIV RNA at a given time, is determined by the trial data and is used in the model based on current and setpoint HIV RNA and other published data.31,38 Trials incorporated into the model include the Dupont 006 trial, the T-20 versus Optimized background Regimen Only (TORO) trials, and several trials on boosted protease inhibitors in combination with nonnucleoside reverse transcriptase inhibitors, among others.39-41 The model can thus project long-term virologic suppression and failure over time based on data from a short-term study, such as the trial described here.
For this analysis, cohort characteristics, including age, gender, mean CD4 cell count, and distribution of HIV RNA, were derived from the clinical trial (Table 2). Risks of opportunistic infections and death in each month were derived from the Multicenter AIDS Cohort Study, as previously reported.33 Quality-of-life estimates were derived from the HIV Cost and Services Utilization Study using utility weights from the short form (SF)-6D health state classification and were stratified by CD4 cell count and opportunistic infection history.42
Because only 4% of patients in the trial were antiretroviral-naive, all patients were assumed to be starting their second antiretroviral regimen.32 To establish the prevalence of a history of an opportunistic infection in the cohort, we ran a preliminary version of the model to simulate the natural history of HIV disease up until the mean CD4 cell count at the trial baseline. After virologic failure occurred in the model, all patients were assumed to be given a third line of therapy, consisting of a boosted protease inhibitor, a nonnucleoside reverse transcriptase inhibitor, and a nucleoside reverse transcriptase inhibitor (virologic suppression of 31% at 48 weeks).41 Third-line failure was followed by salvage therapy with an optimized background regimen (virologic suppression of 12% at 48 weeks).40 After virologic failure of the final line of therapy, the optimized background regimen was continued, simulating the common practice of continuing antiretroviral therapy for any beneficial effects it may have, even in the presence of virologic failure.43
To derive the incremental direct cost of the nursing intervention, provider labor costs for adherence services, participant incentives, adherence tools provided, and provider training time costs were measured for both study arms on an intent-to-treat basis. Results were summarized as the average direct cost per participant per month and were then annualized for the 48-week duration of the trial. The cost of the intervention was defined as the incremental adherence services cost per patient in the nursing intervention arm compared with HIV standard care. This included not only the cost of home nursing visits but any additional adherence services used at the clinic site by patients who received the home visits compared with those receiving HIV standard care.
Details on the costing methodology have previously been reported.28 For every participant encounter, the following data were collected on a form completed by the provider who saw the patient: (1) the type of adherence encounter, including adherence support counseling, addictions treatment, case management, clinical care, HIV education, medication delivery, mental health treatment, or regimen review; (2) service delivery setting (program facility or participant's home); (3) name of provider, which was cross-referenced with a list of providers' licensure and training; (4) intervention delivery mode (individual or group); and (5) adherence tools (eg, pill organizers) newly provided. The following data were estimated from interviews with site personnel and the adherence nurses: (1) average duration of time spent conducting the adherence intervention at each type of encounter, (2) structure of incentive payments, (3) cost of adherence tools used by the site, and (4) trainer and trainee time for staff adherence training sessions. The costing methods conform to the recommendations for conducting a cost-effectiveness analysis established by the US Panel on Cost-Effectiveness in Health and Medicine.44
National labor rates were used to allow the measurement of cost-effectiveness from a US perspective. National hourly wage rates for each provider category were multiplied by 1.39 to account for non-wage-related benefits (including vacations, holidays, sick leave, and other leave).45 For home nursing visits, we assumed that the average 2002 Medicare national reimbursement rate for a home nursing visit of $98 represented the combined labor and transportation cost of this visit. To determine direct costs from a societal perspective, costs were also assigned to patient time spent receiving the intervention and to unpaid peer provider time. The hourly labor rate for patient and peer time was $13.71, which was the median hourly wage rate reported by participants in the baseline HIV Cost and Services Utilization Survey. All costs were updated to constant 2004 dollars using the medical care component of the Consumer Price Index.46,47
We modeled long-term clinical outcomes and cost for all patients in the trial, including treatment-naive and treatment-experienced patients. We conducted the analysis from the societal perspective, discounting life expectancy and costs at 3% per year as recommended by the US Panel on Cost-Effectiveness in Health and Medicine.44 The comparative performance of the nursing intervention was expressed as the incremental cost per quality-adjusted life year (QALY) gained compared with standard care. Sensitivity analyses focused on exploring the impact of alternative assumptions related to the impact of the intervention on antiretroviral efficacy, the duration of effect of the intervention, the availability of subsequent therapeutic regimens, and the initial cost of the intervention as well as its cost over time.
Base Case Analysis
Using data for all patients in the trial, beginning with a mean CD4 count of 319 cells/mm3 and starting second-line antiretroviral therapy, mean projected per person survival increased by 6.4 quality-adjusted life months (QALMs), from 94.5 QALMs in the standard care arm to 100.9 QALMs with the nursing-based adherence intervention (Table 3). Projected lifetime direct medical costs increased from $253,800 to $261,300, resulting in a cost-effectiveness ratio of $14,100 per QALY gained for the nursing intervention compared with standard care.
Sensitivity Analysis on Efficacy
Results of the analyses were most sensitive to the impact of the adherence intervention on the efficacy of antiretroviral therapy, as measured by viral suppression over time (Fig. 1). In the base case analysis, the intervention increased viral suppression at 48 weeks by 63%, from 27% to 44%. If the intervention resulted in only a 30% increase in suppression over the standard care arm (increasing absolute suppression from 27% to 35%), the cost-effectiveness ratio was $16,700 per QALY gained. Even a 6% increase in suppression over the standard care arm (from 27% to 28.6%), an increase too small to be measurable in a reasonably sized clinical trial, would likely translate into a survival benefit of 1.4 QALMs, with a cost-effectiveness ratio of only $24,200 per QALY gained.
We considered the impact of changes in the duration of effect of the intervention by examining outcomes associated with decreased viral suppression beyond 48 weeks. When projected suppression at 72 weeks fell to 25% from the base case value of 30%, the survival benefit of the intervention was 3.9 QALMs. Further decreasing the projected suppression at 72 weeks to 16% yielded a s urvival benefit of 2.7 QALMs. Cost-effectiveness ratios remained in the range of $13,800 to $14,300 per QALY gained.
If we did not include the availability of third- and fourth-line regimens, with patients remaining on second-line therapy after failure, projected lifetime costs were $182,300 and $194,800 for the standard and intervention arms, respectively, with an incremental cost-effectiveness ratio of $19,800 per QALY gained for the intervention. In this scenario, patients in the intervention arm had a mean duration of 53.6 QALMs on second-line therapy before failure (more efficacious and less expensive than subsequent lines), whereas patients in the standard arm had a mean duration of only 44.1 QALMs on second-line therapy.
Sensitivity Analysis on Cost
Results of the analysis were relatively insensitive to changes in the cost of the intervention. When we varied the intervention cost from the base case of $340 down to $170 and up to $680, the cost-effectiveness ratios ranged from $13,800 per QALY at the lower cost to $14,800 per QALY at the higher cost. In a sensitivity analysis using the highest cost adherence intervention from the national Special Projects of National Significance (SPNS) initiative, $725, the cost-effectiveness ratio was $14,900 per QALY gained.28 Even increasing the intervention cost 3-fold, to $1020, only increased the cost-effectiveness ratio to $15,400 per QALY gained. Finally, if we assumed that the baseline intervention cost needed to be maintained at $28 per month (one twelfth of the 1-time intervention cost) beyond the initial trial period to sustain the virologic and immunologic benefit observed in the trial, total costs increased from $261,300 to $263,900, with a cost-effectiveness ratio of $18,900 per QALY gained.
Despite the potency of current therapeutic options for HIV, which have significantly decreased morbidity and increased survival, imperfect adherence to combination antiretroviral therapy remains a major cause of treatment failure among HIV-infected patients.1,3,48 It is likely that individuals whose prior antiretroviral adherence was poor might experience better outcomes if they were to improve their adherence with second-line and subsequent effective regimens.
The results of the Health Resource and Services Administration (HRSA)-sponsored SPNS randomized clinical trial conducted in Boston showed that an intensive nursing intervention in patients with a history of adherence problems can lead to significant improvements in virologic suppression and CD4 cell counts at 48 weeks. We linked these results with an HIV simulation model using the clinical and cost data from the trial and found that an adherence intervention with this efficacy and cost is likely to result in substantial improvement in survival over the long term, with a cost-effectiveness ratio of approximately $14,000 per QALY gained. Furthermore, we found that even if improvements in antiretroviral efficacy were small with this intervention, the likely long-term improvement in clinical outcomes and life expectancy would still provide good value for this type of adherence intervention.
Sensitivity analyses showed that the cost of the intervention was not a critical variable in this analysis, because the intervention costs of several hundred dollars were minor compared with the overall costs of HIV care.20,24,28 Additionally, we found that an effective adherence intervention allows patients to spend relatively more time on earlier regimens, which are more effective and less expensive. Because continued efficacy of the intervention led to continued costs of care, the duration of efficacy did not have a major impact on the cost-effectiveness results.
There are several limitations to this study. As with many clinical trials, this study is limited by its relatively short follow-up time and by the fact that it was designed with virologic, immunologic, and adherence outcomes rather than to demonstrate the impact of such an intervention on long-term clinical outcomes. Most efficacy studies in HIV disease use surrogate markers at 48 weeks as their endpoint, however. The study was also small, with a total of 54 subjects randomized; thus, it is important to reproduce the efficacy findings from this type of focused intervention in other clinical settings and with larger sample sizes.
Some adherence intervention studies using less intensive nursing interventions in selected patients have not found an improvement in virologic suppression.16 Additionally, the mechanism of action of this intervention's effect on virologic suppression is unclear, because the patients in the intervention arm did not self-report better adherence than the patients in the control arm. This may be attributable to limitations in the instrument used in the study to assess adherence.32 The study was also conducted in predominantly treatment-experienced patients; thus, the results may not be generalizable to treatment-naive patients. Finally, overhead costs were not assigned and included in the analysis. There is wide variability in the methods for determining these costs among different sites in different institutional settings.49 Although the cost of the intervention itself is unlikely to alter the cost-effectiveness finding, readers may want to adjust the total direct costs reported here by the overhead rates typically applied by payers in their own institutional settings before drawing conclusions about costs from a payer perspective. From a societal perspective, however, overhead costs that would remain unchanged in the long term with or without the adherence intervention should be excluded.44
There are, however, several aspects of the study that strengthen the conclusions. First, subjects were chosen to be those who reported prior nonadherence or were concerned about adherence. The trial took place in 2 community health centers serving predominantly ethnic minorities, and there were high rates of substance abuse and psychiatric diagnoses in the subjects. Thus, there were definite challenges to improving adherence. Further, the sites had comprehensive HIV care teams in place; thus, the standard-of-care arm may have received more adherence counseling than is typical in many settings.
Based on this randomized controlled trial showing improved virologic suppression and higher CD4 cell counts at 48 weeks, we found that an intensive nursing adherence intervention is likely to have a long-term survival benefit for patients and to be highly cost-effective compared with other uses of HIV care funds.27,50,51 These include prophylaxis for Mycobacterium avium complex ($49,200 per QALY gained) and genotypic resistance testing at time of antiretroviral therapy failure ($22,900 per QALY gained).31,33 Although there is no formally accepted threshold for “cost-effectiveness” of a medical care intervention in the United States, the intervention in this trial was far more cost-effective than commonly cited thresholds, which range from $50,000 per QALY gained to more than $200,000 per QALY gained.29,52
Although the trial evaluated has limitations, this analysis illustrates the value of applying a cost-effectiveness approach to trials of adherence interventions. As more and larger trials are completed and more trial data become available, cost-effectiveness analyses such as the one performed here can yield valuable information on the role of adherence interventions in the overall package of HIV care. Payers such as Medicaid, Medicare, private insurers, and AIDS Drug Assistance Programs should support adherence interventions with this degree of efficacy so as to maximize the benefit of combination antiretroviral therapy in patients with a history of adherence problems.
The authors thank participants in the HAART to HEART trial and other members of the CEPAC team, including Wendy Aaronson, Zhigang Lu, Lauren Mercincavage, Sara Sadownik, Paul Sax, Callie Scott, and Rochelle Walensky, for helpful comments on the analysis.
1. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med
2. Gross R, Bilker WB, Friedman HM, et al. Effect of adherence to newly initiated antiretroviral therapy on plasma viral load. AIDS
3. McNabb J, Ross JW, Abriola K, et al. Adherence to highly active antiretroviral therapy predicts virologic outcome at an inner-city human immunodeficiency virus clinic. Clin Infect Dis
4. Ickovics JR, Cameron A, Zackin R, et al. Consequences and determinants of adherence to antiretroviral medication: results from Adult AIDS Clinical Trials Group protocol 370. Antivir Ther
5. Liu H, Golin CE, Miller LG, et al. A comparison study of multiple measures of adherence to HIV protease inhibitors. Ann Intern Med
6. de Klerk E, van der Heijde D, Landewe R, et al. Patient compliance in rheumatoid arthritis, polymyalgia rheumatica, and gout. J Rheumatol
7. Wetzels GE, Nelemans P, Schouten JS, et al. Facts and fiction of poor compliance as a cause of inadequate blood pressure control: a systematic review. J Hypertens
8. Safren SA, Otto MW, Worth JL, et al. Two strategies to increase adherence to HIV antiretroviral medication: life-steps and medication monitoring. Behav Res Ther
9. Kirkland LR, Fischl MA, Tashima KT, et al. Response to lamivudine-zidovudine plus abacavir twice daily in antiretroviral-naive, incarcerated patients with HIV infection taking directly observed treatment. Clin Infect Dis
10. Altice FL, Mezger JA, Hodges J, et al. Developing a directly administered antiretroviral therapy intervention for HIV-infected drug users: implications for program replication. Clin Infect Dis
. 2004;38(Suppl 5):S376-S387.
11. Macalino GE, Mitty JA, Bazerman LB, et al. Modified directly observed therapy for the treatment of HIV-seropositive substance users: lessons learned from a pilot study. Clin Infect Dis
. 2004;38(Suppl 5):S393-S397.
12. Grossberg R, Zhang Y, Gross R. A time-to-prescription-refill measure of antiretroviral adherence predicted changes in viral load in HIV. J Clin Epidemiol
13. Weber R, Christen L, Christen S, et al. Effect of individual cognitive behaviour intervention on adherence to antiretroviral therapy: prospective randomized trial. Antivir Ther
14. Rawlings MK, Thompson MA, Farthing CF, et al. Impact of an educational program on efficacy and adherence with a twice-daily lamivudine/zidovudine/abacavir regimen in underrepresented HIV-infected patients. J Acquir Immune Defic Syndr
15. Pradier C, Bentz L, Spire B, et al. Efficacy of an educational and counseling intervention on adherence to highly active antiretroviral therapy: French prospective controlled study. HIV Clin Trials
16. Samet JH, Horton NJ, Meli S, et al. A randomized controlled trial to enhance antiretroviral therapy adherence in patients with a history of alcohol problems. Antivir Ther
17. Goujard C, Bernard N, Sohier N, et al. Impact of a patient education program on adherence to HIV medication: a randomized clinical trial. J Acquir Immune Defic Syndr
18. Conway B, Prasad J, Reynolds R, et al. Directly observed therapy for the management of HIV-infected patients in a methadone program. Clin Infect Dis
. 2004;38(Suppl 5):S402-S408.
19. Tuldra A, Fumaz CR, Ferrer MJ, et al. Prospective randomized two-arm controlled study to determine the efficacy of a specific intervention to improve long-term adherence to highly active antiretroviral therapy. J Acquir Immune Defic Syndr
20. Bozzette SA, Joyce G, McCaffrey DF, et al. Expenditures for the care of HIV-infected patients in the era of highly active antiretroviral therapy. N Engl J Med
21. Moore RD. Cost effectiveness of combination HIV therapy: 3 years later. Pharmacoeconomics
22. Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Panel on Clinical Practices for Treatment of HIV Infection convened by the Department of Health and Human Services (DHHS), October 6, 2005. Available at: http://aidsinfo.nih.gov/ContentFiles/AdultandAdolescentGL.pdf
. Accessed March 9, 2006.
24. Gebo KA, Chaisson RE, Folkemer JG, et al. Costs of HIV medical care in the era of highly active antiretroviral therapy. AIDS
25. Yeni PG, Hammer SM, Hirsch MS, et al. Treatment for adult HIV infection: 2004 recommendations of the International AIDS Society-USA Panel. JAMA
26. Flexner CW. Advances in HIV pharmacology: protein binding, pharmacogenomics, and therapeutic drug monitoring. Top HIV Med
27. Sax PE, Losina E, Weinstein MC, et al. Cost-effectiveness of enfuvirtide in treatment-experienced patients with advanced HIV disease. J Acquir Immune Defic Syndr
28. Schackman BR, Finkelstein R, Neukermans CP, et al. The cost of HIV medication adherence support interventions: results of a cross-site evaluation. AIDS Care
29. Goldie SJ, Paltiel AD, Weinstein MC, et al. Projecting the cost-effectiveness of adherence interventions in persons with human immunodeficiency virus infection. Am J Med
30. Freedberg KA, Losina E, Weinstein MC, et al. The cost effectiveness of combination antiretroviral therapy for HIV disease. N Engl J Med
31. Weinstein MC, Goldie SJ, Losina E, et al. Use of genotypic resistance testing to guide HIV therapy: clinical impact and cost-effectiveness. Ann Intern Med
32. Hirschhorn LR, Ruhlmann L, Katz C, et al. Randomized study comparing addition of home nurse to a multidisciplinary adherence program to improve success with HAART in 2 community health centers (CHC) [abstract]. Presented at: International Conference on Antimicrobial Agents and Chemotherapy; 2003; Washington, DC.
33. Freedberg KA, Scharfstein JA, Seage GR III, et al. The cost-effectiveness of preventing AIDS-related opportunistic infections. JAMA
34. Paltiel AD, Goldie SJ, Losina E, et al. Preevaluation of clinical trial data: the case of preemptive cytomegalovirus therapy in patients with human immunodeficiency virus. Clin Infect Dis
35. Paltiel AD, Weinstein MC, Kimmel AD, et al. Expanded screening for HIV in the United States-an analysis of cost-effectiveness. N Engl J Med
36. Walensky RP, Weinstein MC, Kimmel AD, et al. Routine human immunodeficiency virus testing: an economic evaluation of current guidelines. Am J Med
37. Mellors JW, Munoz A, Giorgi JV, et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med
38. Hammer SM, Squires KE, Hughes MD, et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team. N Engl J Med
39. Staszewski S, Morales-Ramirez J, Tashima KT, et al. Efavirenz plus zidovudine and lamivudine, efavirenz plus indinavir, and indinavir plus zidovudine and lamivudine in the treatment of HIV-1 infection in adults. Study 006 Team. N Engl J Med
40. Nelson M, Arasteh K, Clotet B, et al. Durable efficacy of enfuvirtide over 48 weeks in heavily treatment-experienced HIV-1-infected patients in the T-20 versus optimized background regimen only 1 and 2 clinical trials. J Acquir Immune Defic Syndr
41. Losina E, Islam R, Pollock AC, et al. Effectiveness of antiretroviral therapy after protease inhibitor failure: an analytic overview. Clin Infect Dis
42. Schackman BR, Goldie SJ, Freedberg KA, et al. Comparison of health state utilities using community and patient preference weights derived from a survey of patients with HIV/AIDS. Med Decis Making
43. Cole SR, Hernan MA, Robins JM, et al. Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. Am J Epidemiol
44. Gold MR, Siegel JE, Russel LB, et al, eds. Cost-Effectiveness in Health and Medicine
. New York: Oxford University Press; 1996.
45. US Department of Labor Bureau of Labor Statistics. National compensation survey. Available at: http://data.bls.gov
. Accessed January 3, 2004.
46. HIV Cost and Services Utilization Study. Public Use Interview Data Set
. Rockville, MD: Agency for Healthcare Research and Quality; 2000.
47. Bureau of the Census. Statistical Abstract of the United States: 2003
. Washington, DC: Bureau of the Census; 2003.
48. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med
49. Drummond M, O'Brien B, Stoddart G, et al. Methods for the Economic Evaluation of Health Care Programmes
. 2nd ed. Oxford, UK: Oxford University Press; 1997.
50. Sanders GD, Bayoumi AM, Sundaram V, et al. Cost-effectiveness of screening for HIV in the era of highly active antiretroviral therapy. N Engl J Med
51. Goldie SJ, Kaplan JE, Losina E, et al. Prophylaxis for human immunodeficiency virus-related Pneumocystis carinii
pneumonia: using simulation modeling to inform clinical guidelines. Arch Intern Med
52. Ubel PA, Hirth RA, Chernew ME, et al. What is the price of life and why doesn't it increase at the rate of inflation? Arch Intern Med