Modeling the economic outcomes of immuno-oncology drugs: alternative model frameworks to capture clinical outcomes.

BACKGROUND: Economic models in oncology are commonly based on the three-state partitioned survival model (PSM) distinguishing between progression-free and progressive states. However, the heterogeneity of responses observed in immuno-oncology (I-O) suggests that new approaches may be appropriate to reflect disease dynamics meaningfully. MATERIALS AND METHODS: This study explored the impact of incorporating immune-specific health states into economic models of I-O therapy. Two variants of the PSM and a Markov model were populated with data from one clinical trial in metastatic melanoma patients. Short-term modeled outcomes were benchmarked to the clinical trial data and a lifetime model horizon provided estimates of life years and quality adjusted life years (QALYs). RESULTS: The PSM-based models produced short-term outcomes closely matching the trial outcomes. Adding health states generated increased QALYs while providing a more granular representation of outcomes for decision making. The Markov model gave the greatest level of detail on outcomes but gave short-term results which diverged from those of the trial (overstating year 1 progression-free survival by around 60%). CONCLUSION: Increased sophistication in the representation of disease dynamics in economic models is desirable when attempting to model treatment response in I-O. However, the assumptions underlying different model structures and the availability of data for health state mapping may be important limiting factors.

Cost-effectiveness of cytochrome P450 2C19 *2 genotype-guided selection of clopidogrel or ticagrelor in Chinese patients with acute coronary syndrome.

The choice of antiplatelet therapy among Asian populations for the treatment of acute coronary syndrome (ACS) is complicated owing to the high prevalence of cytochrome P450 2C19 (CYP2C19) genetic polymorphism that has been associated with reduced efficacy of clopidogrel. Ticagrelor is a potent but more expensive alternative antiplatelet agent that is not affected by CYP2C19 polymorphism. This study aimed to evaluate the cost-effectiveness, from the Hong Kong health-care provider's perspective, of CYP2C19*2 genotype-guided selection of antiplatelet therapy compared with the universal use of clopidogrel or ticagrelor among ACS patients who undergo percutaneous coronary intervention (PCI). In the present study, a two-part model consisting of a 1-year decision tree and a lifetime Markov model was built to simulate the progress of a typical cohort of 60-year-old Chinese patients until age 85 years and compare three treatment strategies: (i) generic clopidogrel or ticagrelor based on CYP2C19*2 genotype, (ii) universal use of generic clopidogrel or (iii) universal use of ticagrelor for all patients. Incremental cost-effectiveness ratios (ICERs) of <1 gross domestic product per capita locally (US dollar (USD)42 423/quality-adjusted life year (QALY)) were considered cost-effective. Base-case results showed universal ticagrelor use was cost-effective compared with universal clopidogrel, but was dominated by genotype-guided treatment. Genotype-guided treatment was cost-effective compared with universal clopidogrel use (ICER of USD2560/QALY). Sensitivity analysis demonstrated that with the cost of genotype testing up to USD400, CYP2C19*2 genotype-guided antiplatelet treatment remained a cost-effective strategy compared with either universal use of generic clopidogrel or ticagrelor in post-PCI ACS patients in Hong Kong.

Forecasting the future burden of opioids for osteoarthritis.

OBJECTIVE: To quantify the current national burden of opioids for osteoarthritis (OA) pain in Australia in terms of number of dispensed opioid prescriptions and associated costs, and to forecast the likely burden to the year 2030/31. DESIGN: Epidemiological modelling. METHODS: Published data were obtained on rates of opioid prescribing for people with OA and national OA prevalence projections. Trends in opioid dispensing from 2006 to 2016, and average costs for common opioid subtypes were obtained from the Pharmaceutical Benefits Scheme and Medicare Australia Statistics. Using these inputs, a model was developed to estimate the likely number of dispensed opioid prescriptions and costs to the public healthcare system by 2030/31. RESULTS: In 2015/16, an estimated 1.1 million opioid prescriptions were dispensed in Australia for 403,954 people with OA (of a total 2.2 million Australians with OA). Based on recent dispensing trends and OA prevalence projections, the number of dispensed opioid prescriptions is expected to nearly triple to 3,032,332 by 2030/31, for an estimated 562,610 people with OA. The estimated cost to the Australian healthcare system was $AUD25.2 million in 2015/16, rising to $AUD72.4 million by 2030/31. CONCLUSION: OA-related opioid dispensing and associated costs are set to increase substantially in Australia from 2015/16 to 2030/31. Use of opioids for OA pain is concerning given joint disease chronicity and the risk of adverse events, particularly among older people. These projections represent a conservative estimate of the full financial burden given additional costs associated with opioid-related harms and out-of-pocket costs borne by patients.

RESPOND: a programme to prevent secondary falls in older people presenting to the emergency department with a fall: protocol for an economic evaluation.

BACKGROUND: Falls remain common for community-dwelling older people and impose a substantial economic burden to the healthcare system. RESPOND is a novel falls prevention programme that aims to reduce secondary falls and fall injuries among older people who present to a hospital emergency department (ED) with a fall. The present protocol describes a prospective economic evaluation examining the incremental cost-effectiveness of the RESPOND programme, compared with usual care practice, from the Australian health system perspective. METHODS AND DESIGN: This economic evaluation will recruit 528 participants from two major tertiary hospital EDs in Australia and will be undertaken alongside a multisite randomised controlled trial. Outcome and costing data will be collected for all participants over the 12-month trial. It will compare the RESPOND falls prevention programme with usual care practice (current community-based falls prevention practices) to determine its incremental cost-effectiveness according to three intermediate clinical outcomes: (1) falls prevented, (2) fall injuries prevented and (3) injurious falls prevented. In addition, utilities will be derived from a generic quality-of-life measure (EQ-5D-5L) and used to calculate the 'incremental cost per quality-adjusted life years gained'. DISCUSSION: The results of this study will provide healthcare decision makers with evidence to assist with setting spending thresholds for preventive health programmes and inform selection of emergency and community service models of care. TRIAL REGISTRATION NUMBER: The protocol for this study is registered with the Australian New Zealand Clinical Trials Registry (ACTRN12614000336684); Pre-results.

Is a biomarker-based diagnostic strategy for invasive aspergillosis cost effective in high-risk haematology patients?

Empirical antifungal therapy is frequently used in hematology patients at high risk of invasive aspergillosis (IA), with substantial cost and toxicity. Biomarkers for IA aim for earlier and more accurate diagnosis and targeted treatment. However, data on the cost-effectiveness of a biomarker-based diagnostic strategy (BDS) are limited. We evaluated the cost effectiveness of BDS using results from a randomized controlled trial (RCT) and individual patient costing data. Data inputs derived from a published RCT were used to construct a decision-analytic model to compare BDS (Aspergillus galactomannan and PCR on blood) with standard diagnostic strategy (SDS) of culture and histology in terms of total costs, length of stay, IA incidence, mortality, and years of life saved. Costs were estimated for each patient using hospital costing data to day 180 and follow-up for survival was modeled to five years using a Gompertz survival model. Treatment costs were determined for 137 adults undergoing allogeneic hematopoietic stem cell transplant or receiving chemotherapy for acute leukemia in four Australian centers (2005-2009). Median total costs at 180 days were similar between groups (US$78,774 for SDS [IQR US$50,808-123,476] and US$81,279 for BDS [IQR US$59,221-123,242], P = .49). All-cause mortality was 14.7% (10/68) for SDS and 10.1% (7/69) for BDS, (P = .573). The costs per life-year saved were US$325,448, US$81,966, and US$3,670 at 180 days, one year and five years, respectively. BDS is not cost-sparing but is cost-effective if a survival benefit is maintained over several years. An individualized institutional approach to diagnostic strategies may maximize utility and cost-effectiveness.

The cost-effectiveness of hospital-based telephone coaching for people with type 2 diabetes: a 10 year modelling analysis.

BACKGROUND: Type 2 diabetes (T2DM) is a burdensome condition for individuals to live with and an increasingly costly condition for health services to treat. Cost-effective treatment strategies are required to delay the onset and slow the progression of diabetes related complications. The Diabetes Telephone Coaching Study (DTCS) demonstrated that telephone coaching is an intervention that may improve the risk factor status and diabetes management practices of people with T2DM. Measuring the cost effectiveness of this intervention is important to inform funding decisions that may facilitate the translation of this research into clinical practice. The purpose of this study is to assess the cost-effectiveness of telephone coaching, compared to usual diabetes care, in participants with poorly controlled T2DM. METHODS: A cost utility analysis was undertaken using the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model to extrapolate outcomes collected at 6 months in the DTCS over a 10 year time horizon. The intervention's impact on life expectancy, quality-adjusted life expectancy (QALE) and costs was estimated. Costs were reported from a health system perspective. A 5 % discount rate was applied to all future costs and effects. One-way sensitivity analyses were conducted to reflect uncertainty surrounding key input parameters. RESULTS: The intervention dominated the control condition in the base-case analysis, contributing to cost savings of $3327 per participant, along with non-significant improvements in QALE (0.2 QALE) and life expectancy (0.3 years). CONCLUSIONS: The cost of delivering the telephone coaching intervention continuously, for 10 years, was fully recovered through cost savings and a trend towards net health benefits. Findings of cost savings and net health benefits are rare and should prove attractive to decision makers who will determine whether this intervention is implemented into clinical practice. TRIAL REGISTRATION: ACTRN12609000075280.

Direct hospital cost determinants following hip and knee arthroplasty.

OBJECTIVE: Total joint arthroplasty (TJA) places a significant economic burden on health care resources. This cohort study examines the costs associated with arthroplasty in 827 patients undergoing hip and knee TJA from January 2011 to June 2012 at a single center in Melbourne, Australia. METHODS: Data included total inpatient, outpatient, and readmissions costs in the 30 days following TJA. Factors associated with cost were modeled using negative binomial regression and extrapolated to the Australian population. RESULTS: The base cost (i.e., the cost for a patient with no modifying factors) over the first 30 days following TJA was $13,060 Australian (AU) (interquartile range $12,126-14,067 AU). The median length of stay was 4 days (range 2-33 days) and 35 patients (4%) were readmitted in the first 30 days following index TJA, the majority of whom had a surgical site infection (SSI) (74%). The following factors were independently associated with increased costs: SSI, preoperative warfarin therapy, American Society of Anesthesiologists score of 3 or 4, hip TJA, increasing operation time, increasing postoperative blood transfusion requirements, other nosocomial infections, postoperative venous thromboembolism (VTE), pressure ulcers, postoperative confusion, and acute urinary retention. Based on data from the present study, the cost of TJA in Australia is estimated to exceed $1 billion AU per year. Preventable postoperative complications were major cost drivers: SSI and VTE added a further $97 million AU and $66 million AU, respectively, to arthroplasty costs in the first 30 days following surgery. CONCLUSION: This unique study has identified important factors influencing TJA costs and providing guidance for future research and resource allocation.

Economic evaluation of micafungin versus caspofungin for the treatment of candidaemia and invasive candidiasis.

BACKGROUND: Micafungin demonstrated non-inferiority to caspofungin as definitive therapy for candidaemia and invasive candidiasis (IC) in a major randomised clinical trial. AIM: The aim of this study was to investigate if micafungin is a cost-saving option compared with caspofungin for treating candidaemia and IC. METHODS: A decision analytical model was constructed to capture downstream consequences of using either agent as initial therapy for candidaemia and IC. The main outcomes were treatment success and treatment failure (i.e. death, mycological persistence, emergent infection, clinical failure but microbiological success). Outcome probabilities and treatment pathways were derived from the literature. Cost inputs were from the latest Australian resources, and resource use was estimated by expert panel. The analysis was from the Australian hospital perspective. Sensitivity analyses using Monte Carlo simulation were conducted. RESULTS: Micafungin (AU$52 816) was associated with a lower total cost than caspofungin (AU$52 976), with a net cost-saving of $160 per patient. This was primarily due to the lower cost associated with alternative antifungal treatment in the micafungin arm. Hospitalisation was the main cost-driver for both arms. The model outcome was most sensitive to the proportion of treatment success in the micafungin arm. Uncertainty analysis demonstrated that micafungin had a 58% chance of being cost-saving compared with caspofungin. CONCLUSIONS: Micafungin was cost-equivalent to caspofungin in treating candidaemia and IC, with variation in drug acquisition cost the critical factor.

Enterococcal bacteraemia: factors influencing mortality, length of stay and costs of hospitalization.

Enterococci are a major cause of nosocomial bacteraemia. The impacts of vanB vancomycin resistance and antibiotic therapy on outcomes in enterococcal bacteraemia are unclear. Factors that affect length of stay (LOS) and costs of managing patients with enterococcal bacteraemia are also unknown. This study aimed to identify factors associated with mortality, LOS and hospitalization costs in patients with enterococcal bacteraemia and the impact of vancomycin resistance and antibiotic therapy on these outcomes. Data from 116 patients with vancomycin-resistant Enterococci (VRE), matched 1:1 with patients with vancomycin-susceptible Enterococcus (VSE), from two Australian hospitals were reviewed for clinical and economic outcomes. Univariable and multivariable logistic and quantile regression analyses identified factors associated with mortality, LOS and costs. Intensive care unit admission (OR, 8.57; 95% CI, 3.99-18.38), a higher burden of co-morbidities (OR, 4.55; 95% CI, 1.83-11.33) and longer time to appropriate antibiotics (OR, 1.02; 95% CI, 1.01-1.03) were significantly associated with mortality in enterococcal bacteraemia. VanB vancomycin resistance increased LOS (4.89 days; 95% CI, 0.56-11.52) and hospitalization costs (AU$ 28 872; 95% CI, 734-70 667), after adjustment for confounders. Notably, linezolid definitive therapy was associated with lower mortality (OR, 0.13; 95% CI, 0.03-0.58) in vanB VRE bacteraemia patients. In patients with VSE bacteraemia, time to appropriate antibiotics independently influenced mortality, LOS and hospitalization costs, and underlying co-morbidities were associated with mortality. The study findings highlight the importance of preventing VRE bacteraemia and the significance of time to appropriate antibiotics in the management of enterococcal bacteraemia.

Cost analysis of debridement and retention for management of prosthetic joint infection.

Prosthetic joint infection remains one of the most devastating complications of arthroplasty. Debridement and retention of the prosthesis is an attractive management option in carefully selected patients. Despite this, there are no data investigating the cost of this management modality for prosthetic joint infections. The aim of this case-control study was to calculate the cost associated with debridement and retention for management of prosthetic joint infection compared with primary joint replacement surgery without prosthetic joint infection. From 1 January 2008 to 30 June 2010, there were 21 prosthetic joint infections matched to 42 control patients. Controls were matched to cases according to the arthroplasty site, age and sex. Cases had a greater number of unplanned readmissions (100% vs. 7.1%; p <0.001), more additional surgery (3.3 vs. 0.07; p <0.001) and longer total bed days (31.6 vs. 7.9 days; p <0.001). In addition they had more inpatient, outpatient and emergency department visits (p <0.001, respectively). For patients with prosthetic joint infection the total cost, including index operation and costs of management of the prosthetic joint infection, was 3.1 times the cost of primary arthoplasty; the mean cost for cases was Australian dollars (AUD) $69,414 (±29,869) compared with $22,085 (±8147) (p <0.001). The demand for arthroplasty continues to grow and with that, the number of prosthetic joint infections will also increase, placing significant burden on the health system. Our study adds significantly to the growing body of evidence highlighting the substantial costs associated with prosthetic joint infection.

Economic implications of obesity among people with atherothrombotic disease.

OBJECTIVE: The purpose of this study was to ascertain the impact of obesity on the cost of disease management in people with or at high risk of atherothrombotic disease from a governmental perspective using a bottom-up approach to cost estimation. In addition, the aim was also to explore the causes of any differences found. METHOD: The health-care costs of obesity were estimated from 2819 participants recruited into the nationwide Australian REACH Registry with established atherothrombotic disease or at least three risk factors for atherothrombosis. Enrollment was in 2004, through primary care general practices. Information was collected on the use of cardiovascular drugs, hospitalizations and ambulatory care services. 'Bottom-up' costing was undertaken by assigning unit costs to each health-care item, based on Australian Government-reimbursed figures 2006-2007. Linear-mixed models were used to estimate associations between direct medical costs and body mass index (BMI) categories. RESULTS: Annual pharmaceutical costs per person increased with increasing BMI category, even after adjusting for gender, age, living place, formal education, smoking status, hypertension and diabetes. Adjusted annual pharmaceutical costs of overweight and obese participants were higher ($7 (P=0.004) and $144 (<0.001), respectively) than those of the normal weight participants. This was due to participants in higher BMI categories receiving more pharmaceuticals than normal weight participants. There was no significant change across the BMI categories in annual ambulatory care costs and annual hospital costs. CONCLUSION: In these participants with or at high risk of atherothrombotic disease, annual pharmaceutical costs were greater in participants of higher BMI category, but there was not such a gradient in the annual hospital or ambulatory care costs. The greater cardiovascular pharmaceutical costs for participants of higher BMI categories remained even after adjusting for a range of demographic factors and comorbidities. Our results suggest that these costs are explained by the higher number of drugs used among people with atherothrombotic disease. Further investigation is needed to understand the reasons for this level of drug use.

A model for predicting the future incidence of coronary heart disease within percentiles of coronary heart disease risk.

BACKGROUND: We present a method (The CHD Prevention Model) for modelling the incidence of fatal and nonfatal coronary heart disease (CHD) within various CHD risk percentiles of an adult population. The model provides a relatively simple tool for lifetime risk prediction for subgroups within a population. It allows an estimation of the absolute primary CHD risk in different populations and will help identify subgroups of the adult population where primary CHD prevention is most appropriate and cost-effective. METHODS: The CHD risk distribution within the Australian population was modelled, based on the prevalence of CHD risk, individual estimates of integrated CHD risk, and current CHD mortality rates. Predicted incidence of first fatal and nonfatal myocardial infarction within CHD risk strata of the Australian population was determined. RESULTS: Approximately 25% of CHD deaths were predicted to occur amongst those in the top 10 percentiles of integrated CHD risk, regardless of age group or gender. It was found that while all causes survival did not differ markedly between percentiles of CHD risk before the ages of around 50-60, event-free survival began visibly to differ about 5 years earlier. CONCLUSIONS: The CHD Prevention Model provides a means of predicting future CHD incidence amongst various strata of integrated CHD risk within an adult population. It has significant application both in individual risk counselling and in the identification of subgroups of the population where drug therapy to reduce CHD risk is most cost-effective.

Cost-effectiveness of prescribing statins according to pharmaceutical benefits scheme criteria.

OBJECTIVES: (i) To analyse how well Pharmaceutical Benefits Scheme (PBS) criteria for prescribing lipid-lowering therapy identify people most at risk of coronary heart disease (CHD); and (ii) to determine the cost-effectiveness of primary prevention therapy with pravastatin according to these criteria in Australia. DESIGN: (i) Analysis of targeting of CHD risk according to PBS criteria; (ii) cost-effectiveness analysis for pravastatin as primary preventive therapy (40 mg/day), with a 20-year projection from 1999. PARTICIPANTS: (i) Men and women aged 25-69 years from the 1989 National Heart Foundation Risk Factor Prevalence Survey; (ii) Australian men and women, aged 25-85 years, excluding those with diabetes and existing CHD. MAIN OUTCOME MEASURES: (i) Proportion eligible for lipid lowering treatment according to PBS criteria within 15-year risk of CHD mortality groups; (ii) average net cost in Australian dollars ($) per year of life saved (YOLS), with 80% uncertainty ranges (UR). RESULTS: (i) PBS criteria do not adequately identify those most at risk of CHD, as only 61% of Australians (aged 25-69 years) with a greater than 10% 15-year risk of CHD mortality were eligible for treatment; and 11% of those at low risk of CHD mortality (< 2.5% over 15 years) were eligible for treatment. (ii) Cost-effectiveness of treatment according to PBS criteria was estimated at $110,000 (80% UR, $96,000-$150,000) per YOLS for men and $87,000 (80% UR, $80,000-$130,000) per YOLS for women. As an indicator of the likely recurrent annual costs, total first-year treatment costs (excluding the costs of non-compliers) were estimated at $940 million. Assuming compliance of 50%, cost-effectiveness of treatment was markedly improved using 32.5% 15-year risk of CHD mortality as a cut-off, with ratios of $31,000 (80% UR, $27,000-$40,000) per YOLS for men and $39,000 (80% UR, $33,000-$53,000) per YOLS for women. First-year treatment costs of $940 million were the same as treating according to PBS criteria, but absolute health impact in terms of deaths averted and years of life saved was more than doubled. CONCLUSIONS: While PBS criteria do target patients at risk of CHD, there is room for improvement in identifying those most at risk of CHD, and treatment according to PBS criteria is not likely to be the most cost-effective. For optimal cost-effectiveness, targeting of therapy for primary CHD prevention needs to be based on population-specific, multivariable risk.