The economic costs of caring for people with HIV infection and AIDS in England and Wales

Cost-Effectiveness Uncertainty Analysis Methods: A Comparison of One-Way Sensitivity, Analysis of Covariance, and Expected Value of Partial Perfect Information

OBJECTIVES

To compare model input influence on incremental net monetary benefit (INMB) across 3 uncertainty methods: 1) 1-way sensitivity analysis; 2) probabilistic analysis of covariance (ANCOVA); and 3) expected value of partial perfect information (EVPPI).

METHODS

In a preliminary model, we used a published cost-effectiveness model and assumed £20,000 per quality-adjusted life-year (QALY) willingness-to-pay (Case 1: lower decision uncertainty) and £8000/QALY willingness-to-pay (Case 2: higher decision uncertainty). We conducted 1-way sensitivity, ANCOVA (10,000 Monte Carlo draws), and EVPPI for each model input (1000 inner and 1000 outer draws). We ranked inputs based on influence of INMB and compared input ranks across methods within case using Spearman's rank correlation. We replicated this approach in 3 follow-up models: an additional linear model, a less linear model with uncorrelated inputs, and a less linear model with correlated inputs.

RESULTS

In the preliminary model, lower and higher decision uncertainty cases had the same top 3 influential parameters across uncertainty methods. The 2 most influential inputs contributed 78% and 49% of variation in outcome based on ANCOVA for lower decision uncertainty and higher decision uncertainty cases, respectively. In the follow-up models, input rank order correlations were higher across uncertainty methods in the linear model compared with both of the less linear models.

CONCLUSIONS

Evidence across models suggests influential input rank agreement between 1-way and more advanced uncertainty analyses for relatively linear models with uncorrelated parameters but less agreement for less linear models. Although each method provides unique information, the additional resources needed to generate and communicate advanced analyses should be weighed, especially when outcome decision uncertainty is low. For less linear models or those with correlated inputs, performing and reporting deterministic and probabilistic uncertainty analyses appear prudent and conservative.

Cost effectiveness of darunavir/ritonavir 600/100 mg bid in protease inhibitor-experienced, HIV-1-infected adults in Belgium, Italy, Sweden and the UK

BACKGROUND

Two phase II trials (POWER 1 and 2) have demonstrated that darunavir co-administered with low-dose ritonavir (DRV/r) provides significant clinical benefit compared with control protease inhibitors (PIs) in highly treatment-experienced, HIV-1-infected adults, when co-administered with optimized background therapy (OBR).

OBJECTIVE

To determine whether DRV/r is cost effective compared with control PIs, from the perspective of Belgian, Italian, Swedish and UK reimbursement authorities, when used in treatment-experienced patients similar to those included in the POWER 1 and 2 trials.

METHODS

An existing Markov model containing health states defined by CD4 cell count ranges (> 500, 351-500, 201-350, 101-200, 51-100 and 0-50 cells/mm³) and death was adapted for use in four European healthcare settings. Baseline demographics, CD4 cell count distribution and antiretroviral drug usage reflected those reported in the POWER 1 and 2 trials. Virological/immunological response rates and matching transition probabilities over the patient's lifetime were based on results from the POWER trials and published data. After treatment failure, patients were assumed to switch to a tipranavir-containing regimen plus OBR. For each CD4 cell count range, utility values and HIV-related mortality rates were obtained from the published literature. National all-cause mortality data and published data on the increased risk of non HIV-related mortality in HIV-infected individuals were taken into account in the model. Data from observational studies conducted in each healthcare setting were used to determine resource-use patterns and costs associated with each CD4 cell count range. Unit costs were derived from official local sources; a lifetime horizon was taken and discount rates were selected based on local guidelines.

RESULTS

In the base-case analysis, quality-adjusted life-year (QALY) gains of up to 1.397 in Belgium, over 1.171 in Italy, 1.142 in Sweden and 1.091 in the UK were predicted when DRV/r-based therapy was used instead of control PI-based treatment. The base-case analyses predicted an incremental cost-effectiveness ratio (ICER) of €11,438/QALY in Belgium, €12,122/QALY in Italy,€10,942/QALY in Sweden and €16,438/QALY in the UK. Assuming an acceptability threshold of €30,000/QALY, DRV/r-based therapy remained cost effective over all parameter ranges tested in extensive one-way sensitivity analyses. Probabilistic sensitivity analysis revealed a 95% (Belgium), 97% (Italy), 92% (Sweden) or 78% (UK) probability of attaining an ICER below this threshold.

CONCLUSION

From four European payer perspectives, DRV/r-based antiretroviral therapy is predicted to be cost effective compared with currently available control PIs, when both are used with an OBR in treatment-experienced, HIV-1-infected adults who failed to respond to more than one PI-containing regimen.

Predicting direct costs of HIV care during the first year of darunavir-based highly active antiretroviral therapy using CD4 cell counts: evidence from POWER

BACKGROUND

Given the association between CD4 cell counts and HIV-related morbidity/mortality, new antiretroviral therapies could potentially lower the direct costs of HIV care by raising CD4 cell counts.

OBJECTIVES

To predict the effects of the ritonavir-boosted, HIV protease inhibitor (PI) darunavir on the direct costs of care, while accounting for CD4 cell counts, during the first year of therapy in highly treatment-experienced, HIV-infected adults in different healthcare settings.

METHODS

The mean annual per-patient cost of darunavir/ritonavir (DRV/r) and control PI-based highly active antiretroviral therapy (HAART) was calculated from the proportional use of antiretroviral agents in the DRV/r and control PI arms of the pooled POWER 1 and 2 trials, applying drug-acquisition costs for five healthcare settings. Non-antiretroviral-related costs by CD4 cell count, derived from non-interventional studies in the same settings, were applied to the POWER data (proportion of patients with CD4 cell counts in different strata at week 48) to estimate mean annual non-antiretroviral-related costs per patient in patients receiving DRV/r or control PI-based HAART during year 1.

RESULTS

Across all settings, the mean annual per-patient cost of DRV/r-based treatment was 2-19% higher than that of control PI-based therapy during the first year of therapy. By raising CD4 cell counts, however, DRV/r-based regimens were predicted to lower mean annual non-antiretroviral-related costs by 16-38% compared with control PI-based therapy. When combined, the total annual per-patient cost of HIV care during the first year of therapy was estimated to be 7% lower in the DRV/r compared with the control PI arm using US data, 8% lower using Swedish data, budget neutral using UK and Belgian data and 5% higher using Italian data.

CONCLUSIONS

Darunavir-based HAART may lower non-antiretroviral-related costs compared with control PI-based therapy in highly treatment-experienced, HIV-infected patients during the first year of therapy by improving patients' CD4 cell counts. These costs could partly/fully offset the increased acquisition cost of DRV/r in this patient population over the same period.

Cost effectiveness of darunavir/ritonavir 600/100 mg bid in treatment-experienced, lopinavir-naive, protease inhibitor-resistant, HIV-infected adults in Belgium, Italy, Sweden and the UK

BACKGROUND

Using data from the phase IIb POWER trials, darunavir boosted with low-dose ritonavir (DRV/r; 600/100 mg twice daily; bid)-based highly active antiretroviral therapy (HAART) was shown to be significantly more efficacious and cost effective than other protease inhibitor (PI)-based therapy in highly treatment-experienced, HIV-1-infected adults. Furthermore, in the phase III TITAN trial (TMC114-C214), DRV/r 600/100 mg bid-based HAART generated a superior 48-week virological response rate compared with standard-of-care lopinavir/ritonavir (LPV/r; 400/100 mg bid)-based therapy in treatment-experienced, lopinavir-naive patients, and in particular those with one or more International AIDS Society - USA (IAS-USA) primary PI resistance-associated mutations at baseline. These patients had a broader degree of previous PI use/failure (0 - ≥ 2) than the POWER patients.

OBJECTIVES

To determine whether DRV/r 600/100 mg bid-based HAART is cost effective compared with LPV/r-based therapy, from the perspective of Belgian, Italian, Swedish and UK reimbursement authorities, when used in treatment-experienced patients similar to TITAN patients with one or more IAS-USA primary PI mutations at baseline.

METHODS

An existing Markov model containing health states defined by CD4 cell count ranges (>500, 351-500, 201-350, 101-200, 51-100 and 0-50 cells/mm³) and an absorbing state of death was adapted for use in the above-mentioned healthcare settings. Baseline demographics, CD4 cell count distribution, antiretroviral drug usage, virological/immunological response rates and matching transition probabilities were based on data collected during the first 48 weeks of therapy in the modelled subgroup of TITAN patients and the published literature. After treatment failure, patients were assumed to switch to a follow-on combination regimen. For each health state, utility values and mortality rates were obtained from the published literature. Data from local observational studies (Belgium, Sweden and Italy) or the published literature (UK) were used to determine resource-use patterns and costs associated with each CD4 cell count range. Unit costs were derived from official local sources; a lifetime horizon was taken and discount rates were chosen based on local guidelines.

RESULTS

The base-case analysis predicted quality-adjusted life year (QALY) gains of 0.785 in Belgium, 0.608 in Italy, 0.584 in Sweden and 0.550 in the UK when DRV/r-based therapy was used instead of LPV/r-based treatment. The estimated base-case incremental cost-effectiveness ratios (ICERs) were €6964/QALY gained in Belgium, €9277/QALY gained in Italy, €6868 (SEK69,687)/QALY gained in Sweden and €14,778 (£12 612)/QALY gained in the UK. Assuming a threshold of €30,000/QALY gained, DRV/r-based therapy remained cost effective over most parameter ranges tested in extensive one-way sensitivity analyses. The variation of immunological response rates and the time horizon were identified as important drivers of cost effectiveness. Probabilistic sensitivity analysis revealed a greater than 70% probability of achieving an ICER below this threshold in all four healthcare settings.

CONCLUSION

From the perspective of Belgian, Italian, Swedish and UK payers, DRV/r 600/100 mg bid-based HAART is predicted to be cost effective compared with LPV/r 400/100 mg bid-based therapy, when used to manage treatment experienced, lopinavir-naive, PI-resistant, HIV-infected adults with a broad range of previous PI use/failure.

Analysis of costs by CD4 count category for the darunavir/r 600/100 mg bid and control protease inhibitor arms of the POWER 1 and 2 trials

BACKGROUND

HIV-infected people with low CD4 counts are at higher risk of AIDS and incur increased health care costs from in-patient stays and medications.

METHOD

In the POWER 1 and 2 trials, patients were treated with optimized nucleoside reverse transcriptase inhibitors and optional enfuvirtide (T-20), plus darunavir/ritonavir (DRV/r) or control protease inhibitor (PI). UK data on costs of care by CD4 count were combined with the data on antiretroviral treatment use and CD4 counts from the POWER trials to calculate expected health care costs.

RESULTS

The mean annual UK cost of care (excluding antiretrovirals [ARVs]) was 23,780 pounds, 13,762 pounds, 7,032 pounds, and 7,032 pounds for patients with CD4 <50, 50-200, 200-350, and >350 cells/muL respectively. In the POWER trials, at week 48, the proportions of patients with CD4 counts in these categories were 7%, 36%, 29%, and 27% for the DRV/r arm versus 23%, 28%, 27%, and 22% for the control PI arm. The mean predicted annual per-patient cost of care was 11,170 pounds for DRV/r versus 12,873 pounds for control PI.

CONCLUSION

By raising CD4 counts to levels where the risk of AIDS events is reduced, DRV/r treatment is predicted to lower patient care costs for ARV-experienced, HIV-infected individuals in the first year of treatment.

Cost-of-illness studies : a review of current methods

The number of cost-of-illness (COI) studies has expanded considerably over time. One outcome of this growth is that the reported COI estimates are inconsistent across studies, thereby raising concerns over the validity of the estimates and methods. Several factors have been identified in the literature as reasons for the observed variation in COI estimates. To date, the variation in the methods used to calculate costs has not been examined in great detail even though the variations in methods are a major driver of variation in COI estimates. The objective of this review was to document the variation in the methodologies employed in COI studies and to highlight the benefits and limitations of these methods. The review of COI studies was implemented following a four-step procedure: (i) a structured literature search of MEDLINE, JSTOR and EconLit; (ii) a review of abstracts using pre-defined inclusion and exclusion criteria; (iii) a full-text review using pre-defined inclusion and exclusion criteria; and (iv) classification of articles according to the methods used to calculate costs. This review identified four COI estimation methods (Sum_All Medical, Sum_Diagnosis Specific, Matched Control and Regression) that were used in categorising articles. Also, six components of direct medical costs and five components of indirect/non-medical costs were identified and used in categorising articles.365 full-length articles were reflected in the current review following the structured literature search. The top five cost components were emergency room/inpatient hospital costs, outpatient physician costs, drug costs, productivity losses and laboratory costs. The dominant method, Sum_Diagnosis Specific, was a total costing approach that restricted the summation of medical expenditures to those related to a diagnosis of the disease of interest. There was considerable variation in the methods used within disease subcategories. In several disease subcategories (e.g. asthma, dementia, diabetes mellitus), all four estimation methods were represented, and in other cases (e.g. HIV/AIDS, obesity, stroke, urinary incontinence, schizophrenia), three of the four estimation methods were represented. There was also evidence to suggest that the strengths and weaknesses of each method were considered when applying a method to a specific illness. Comparisons and assessments of COI estimates should consider the method used to estimate costs both as an important source of variation in the reported COI estimates and as a marker of the reliability of the COI estimate.

Cost of AIDS care in Mexico: what are its main individual predictors?

BACKGROUND

The Mexican government is offering universal access to antiretroviral (ARV) drugs. The cost of doing so, despite aggressive price negotiation with the pharmaceutical industry, remains high. Even with a low prevalence, about 150,000 Mexicans are estimated to be living with HIV and will require ARV treatment. Estimating the resources needed to fund this gap should consider how patient and provider characteristics affect health care costs.

METHODS

Using a sample of patients from 11 facilities in three Mexican cities representing a large proportion of AIDS patients in the country, we developed a fixed-effect model, which by controlling the facilities and individual heterogeneity estimates predicted costs using patient demographic and socioeconomic characteristics as well as physician training.

RESULTS

The estimated model explains about 45% of the variation in costs. Additional education is significantly and positively associated with cost. Increasing age is also associated with higher costs.

CONCLUSIONS

Socioeconomic status and demographic characteristics explain an important proportion of variation in care costs for AIDS, despite AIDS being such a heterogeneous disease. Such characteristics will need to be taken into account when resource needs are estimated. A priority-setting process considering the principles of equity in the fair distribution of resources is needed to help reduce the social burden of HIV/AIDS in Mexico.

The cost of HIV treatment and care. A global review

This review of published studies on the costs of HIV treatment and care describes some of the recent developments that have influenced these costs in industrialised and industrialising countries, especially within the context of changing drug treatments. Some of the different approaches to estimating the economic impact of HIV infection are briefly presented. The methods used to review the literature are described, particularly the criteria of a scoring system that was specifically developed to systematically screen some of the studies identified. The mean review score for studies dealing with direct hospital costs increased significantly (p = 0.003) over the 3 periods analysed (before 1987, 1987 to 1995, and 1996 and beyond), indicating that the overall 'quality' of studies increased over time. All cost estimates, other than those from non-industrialised regions, were converted to 1996 US dollars using country-specific total health expenditure inflaters and country-specific Gross Domestic Product Purchasing Power Parity converters. A summary of hospital cost estimates over time and by region demonstrated that the costs of treating asymptomatic individuals and people with symptomatic non-AIDS increased over the period, but that the costs of treating individuals with AIDS appears to have stabilised since the late 1980s. As fewer studies could be identified on the costs of community and informal care, indirect productivity costs and population cost estimates, and costs of care for children with HIV infection, all of these studies were reviewed without the use of the scoring system. Finally, the discussion explores the evidence on the global costs of HIV in non-industrialised economies and the affordability of HIV treatment and care. Some suggestions for the direction of future HIV costing studies are also presented. A need remains for good quality cost data. Adequate research effort should be directed to improving the scope and quality of information on costs of HIV service provision around the world.

Cost effectiveness of epoetin-alpha to augment preoperative autologous blood donation in elective cardiac surgery

OBJECTIVE

The objective of this study was to assess the cost effectiveness of using epoetin-alpha (erythropoietin) to augment preoperative autologous donation (PAD) of blood prior to elective cardiac surgery.

DESIGN AND SETTING

We designed a decision-analytic model incorporating the risk of receiving allogeneic blood, the costs of blood products, the likelihood of developing transfusion-related diseases, the costs of transfusion-related diseases and their impact on life expectancy, and the effect of epoetin-alpha on the probability of transfusion.

INTERVENTIONS

The efficacy of epoetin-alpha was derived from data from a meta-analysis of published randomised trials comparing the use of epoetin-alpha to augment PAD with the use of PAD alone. Estimates for the other parameters were obtained by a systematic review of the literature.

MAIN OUTCOME MEASURES AND RESULTS

The use of epoetin-alpha reduced the proportion of patients receiving allogeneic transfusions by 60% (from 31.6 to 12.7%). However, this led to only a modest benefit of 0.000035 life years gained per patient and an incremental cost per life year gained of $Can44.6 million (1998 Canadian dollars). A detailed sensitivity analysis confirmed that the cost-effectiveness ratio was larger than that which is generally considered acceptable.

CONCLUSIONS

Our study indicates that the use of epoetin-alpha to reduce perioperative allogeneic transfusions in cardiac surgery is not cost effective.

Healthcare Economics in HIV

In an era of cost-consciousness in the delivery of medical care, the economics of healthcare delivery for HIV-infected persons has been an area of active interest. Interested parties include the payors of HIV care, particularly public insurers, who are paying for an increasing amount of the overall cost of HIV care in the US; providers of care, many of whom are finding it increasingly difficult to provide HIV care in a capitated market; and those persons who are HIV-infected and increasingly caught in the economic turmoil of the HIV healthcare marketplace. This paper will review the literature published over the past year regarding the economics of healthcare for HIV in the US.

Cost effectiveness of combination HIV therapy: 3 years later

Since 1997, expert panel guidelines for HIV care have recommended the use of combination antiretroviral therapy with at least 3 antiretroviral drugs. Several studies have examined the cost effectiveness of 3-drug combination antiretroviral regimens for the treatment of HIV infection. Analyses comparing a 3-drug protease inhibitor-containing regimen with a 1- or 2-drug non-nucleoside reverse transcriptase inhibitor regimen have consistently yielded incremental direct cost estimates ranging from $US10,000 to just over $US13,000 per year of life saved. In Western societies, such an incremental cost per year of life saved compares favourably with chronic therapy for other diseases and argues for the adoption of these drugs by payors and policy makers. The reason for this favourable cost-effectiveness ratio appears to be the decrease in opportunistic complications and hospitalisation associated with the effective use of combination antiretroviral therapy. Whether this initial benefit will be maintained is not yet known. Other comorbid illnesses such as hepatitis C or renal failure may subsequently increase the cost of HIV care, and some analyses suggest that resistance may develop to these drugs over the long term. In addition, studies are needed to assess the cost effectiveness of these therapies in developing countries where the expense of these drugs appears to put them out of reach. The collection and analysis of economic data will continue to be needed as newer HIV therapies become available and the HIV healthcare environment evolves. Quantifying medical care costs and calculating cost effectiveness involve assessing a moving target. Economic analyses of HIV infection must evolve in tandem with therapeutic changes to continue to be relevant to policy makers, payors of care, and those who provide and receive HIV care.

Demograghic and socio-economic determinants of community and hospital services costs for people with HIV/AIDS in London

We examined the influence of demographic, social and economic background of people with HIV/AIDS in London on total community and hospital services costs. This was a retrospective study of community and hospital service use, needs and costs based on structured questionnaires administered by trained interviewers and costing information obtained from the service purchasers and providers, based on two Genito-urinary Medicine clinics in London: the Jefferiss Wing at St. Mary's Hospital and Patric Clements at the Central Middlesex Hospital, London, England. The subjects were 225 HIV infected patients (105 asymptomatic, 59 symptomatic non-AIDS and 61 AIDS). We found that over and above well established determinants of health care costs for HIV infected people such as disease stage and transmission category, social and economic factors such as employment and support of a living-in partner significantly reduced community services costs. Private health insurance had a similar effect, though only a small proportion of HIV people had such cover. The cost of community services for HIV infected non-European Union nationals, mainly of African origin, was one quarter that for the European Union nationals. Community services costs were highest for heterosexually infected women and lowest for heterosexually infected men after adjusting for other factors. Hospital services costs were significantly higher for HIV infected people lacking educational qualifications and employment. We conclude that access to community care for HIV infected non-EU nationals appears to be very poor as the cost of their community services was one quarter that for the EU nationals after adjusting for the effects of transmission category, disease stage, living with a partner, employment and having a private health insurance. Additional incentives for informal care for HIV infected people could be a cost-effective way to improve their community health service provisions.