Lamotrigine versus levetiracetam or zonisamide for focal epilepsy and valproate versus levetiracetam for generalised and unclassified epilepsy: two SANAD II non-inferiority RCTs.

BACKGROUND: Levetiracetam (Keppra®, UCB Pharma Ltd, Slough, UK) and zonisamide (Zonegran®, Eisai Co. Ltd, Tokyo, Japan) are licensed as monotherapy for focal epilepsy, and levetiracetam is increasingly used as a first-line treatment for generalised epilepsy, particularly for women of childbearing age. However, there is uncertainty as to whether or not they should be recommended as first-line treatments owing to a lack of evidence of clinical effectiveness and cost-effectiveness. OBJECTIVES: To compare the clinical effectiveness and cost-effectiveness of lamotrigine (Lamictal®, GlaxoSmithKline plc, Brentford, UK) (standard treatment) with levetiracetam and zonisamide (new treatments) for focal epilepsy, and to compare valproate (Epilim®, Sanofi SA, Paris, France) (standard treatment) with levetiracetam (new treatment) for generalised and unclassified epilepsy. DESIGN: Two pragmatic randomised unblinded non-inferiority trials run in parallel. SETTING: Outpatient services in NHS hospitals throughout the UK. PARTICIPANTS: Those aged ≥ 5 years with two or more spontaneous seizures that require anti-seizure medication. INTERVENTIONS: Participants with focal epilepsy were randomised to receive lamotrigine, levetiracetam or zonisamide. Participants with generalised or unclassifiable epilepsy were randomised to receive valproate or levetiracetam. The randomisation method was minimisation using a web-based program. MAIN OUTCOME MEASURES: The primary outcome was time to 12-month remission from seizures. For this outcome, and all other time-to-event outcomes, we report hazard ratios for the standard treatment compared with the new treatment. For the focal epilepsy trial, the non-inferiority limit (lamotrigine vs. new treatments) was 1.329. For the generalised and unclassified epilepsy trial, the non-inferiority limit (valproate vs. new treatments) was 1.314. Secondary outcomes included time to treatment failure, time to first seizure, time to 24-month remission, adverse reactions, quality of life and cost-effectiveness. RESULTS: Focal epilepsy. A total of 990 participants were recruited, of whom 330 were randomised to receive lamotrigine, 332 were randomised to receive levetiracetam and 328 were randomised to receive zonisamide. Levetiracetam did not meet the criteria for non-inferiority (hazard ratio 1.329) in the primary intention-to-treat analysis of time to 12-month remission (hazard ratio vs. lamotrigine 1.18, 97.5% confidence interval 0.95 to 1.47), but zonisamide did meet the criteria (hazard ratio vs. lamotrigine 1.03, 97.5% confidence interval 0.83 to 1.28). In the per-protocol analysis, lamotrigine was superior to both levetiracetam (hazard ratio 1.32, 95% confidence interval 1.05 to 1.66) and zonisamide (hazard ratio 1.37, 95% confidence interval 1.08 to 1.73). For time to treatment failure, lamotrigine was superior to levetiracetam (hazard ratio 0.60, 95% confidence interval 0.46 to 0.77) and zonisamide (hazard ratio 0.46, 95% confidence interval 0.36 to 0.60). Adverse reactions were reported by 33% of participants starting lamotrigine, 44% starting levetiracetam and 45% starting zonisamide. In the economic analysis, both levetiracetam and zonisamide were more costly and less effective than lamotrigine and were therefore dominated. Generalised and unclassifiable epilepsy. Of 520 patients recruited, 260 were randomised to receive valproate and 260 were randomised to receive to levetiracetam. A total of 397 patients had generalised epilepsy and 123 had unclassified epilepsy. Levetiracetam did not meet the criteria for non-inferiority in the primary intention-to-treat analysis of time to 12-month remission (hazard ratio 1.19, 95% confidence interval 0.96 to 1.47; non-inferiority margin 1.314). In the per-protocol analysis of time to 12-month remission, valproate was superior to levetiracetam (hazard ratio 1.68, 95% confidence interval 1.30 to 2.15). Valproate was superior to levetiracetam for time to treatment failure (hazard ratio 0.65, 95% confidence interval 0.50 to 0.83). Adverse reactions were reported by 37.4% of participants receiving valproate and 41.5% of those receiving levetiracetam. Levetiracetam was both more costly (incremental cost of £104, 95% central range -£587 to £1234) and less effective (incremental quality-adjusted life-year of -0.035, 95% central range -0.137 to 0.032) than valproate, and was therefore dominated. At a cost-effectiveness threshold of £20,000 per quality-adjusted life-year, levetiracetam was associated with a probability of 0.17 of being cost-effective. LIMITATIONS: The SANAD II trial was unblinded, which could have biased results by influencing decisions about dosing, treatment failure and the attribution of adverse reactions. FUTURE WORK: SANAD II data could now be included in an individual participant meta-analysis of similar trials, and future similar trials are required to assess the clinical effectiveness and cost-effectiveness of other new treatments, including lacosamide and perampanel. CONCLUSIONS: Focal epilepsy - The SANAD II findings do not support the use of levetiracetam or zonisamide as first-line treatments in focal epilepsy. Generalised and unclassifiable epilepsy - The SANAD II findings do not support the use of levetiracetam as a first-line treatment for newly diagnosed generalised epilepsy. For women of childbearing potential, these results inform discussions about the benefit (lower teratogenicity) and harm (worse seizure outcomes and higher treatment failure rate) of levetiracetam compared with valproate. TRIAL REGISTRATION: Current Controlled Trials ISRCTN30294119 and EudraCT 2012-001884-64. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 75. See the NIHR Journals Library website for further project information.

BACKGROUND AND METHODS: The SANAD II trial was a clinical trial designed to identify the most clinically effective and cost-effective treatment for adults and children aged > 5 years with newly diagnosed epilepsy. There are two main epilepsy types: focal and generalised. In focal epilepsy, seizures start at a single place in the brain (a focus), whereas in generalised epilepsy seizures start in both sides of the brain at the same time. Anti-seizure medications are the main treatment. For people with newly diagnosed epilepsy, the first anti-seizure medication should control the seizures as quickly as possible while avoiding side effects. The first-choice treatments are lamotrigine (Lamictal®, GlaxoSmithKline plc, Brentford, UK) for focal epilepsy and valproate (Epilim®, Sanofi SA, Paris, France) for generalised epilepsy (however, the latter should be avoided in women who could become pregnant). A number of newer anti-seizure medications have been approved for NHS use, but it is unclear whether or not they should be used as first-line treatments. The SANAD II trial focused on the new medicines levetiracetam (Keppra®, UCB Pharma Ltd, Slough, UK) and zonisamide (Zonegran®, Eisai Co. Ltd, Tokyo, Japan). We recruited 1510 people aged ≥ 5 years with newly diagnosed epilepsy: 990 with focal epilepsy and 520 with generalised or unclassified epilepsy. FINDINGS: FOCAL EPILEPSY: People starting treatment with levetiracetam or zonisamide were significantly less likely to have a 12-month remission from seizures than people starting treatment with lamotrigine, unless they were changed to another anti-seizure medication. Side effects that were thought to be caused by anti-seizure medications were reported by 33% of participants starting lamotrigine, 44% of those starting levetiracetam and 45% of those starting zonisamide. The cost-effectiveness analyses showed that neither levetiracetam nor zonisamide is value for money for the NHS when compared with lamotrigine. The SANAD II findings do not support the use of levetiracetam or zonisamide as first-line treatments in focal epilepsy. FINDINGS: GENERALISED AND UNCLASSIFIABLE EPILEPSY: People starting treatment with levetiracetam were significantly less likely to have a 12-month remission from seizures than people starting valproate, unless they were changed to another anti-seizure medication. Side effects that were thought to be caused by anti-seizure medications were reported by 37% of participants starting valproate and 42% of participants starting levetiracetam. The cost-effectiveness analyses showed that levetiracetam is not good value for money for the NHS when compared with valproate. The SANAD II findings do not support the use of levetiracetam as a first-line treatment for newly diagnosed generalised epilepsy. Importantly, our results will inform treatment decisions for women, who may choose a less effective treatment that is safer in pregnancy.

Cost-Effectiveness Analysis of Adalimumab for the Treatment of Uveitis Associated with Juvenile Idiopathic Arthritis.

PURPOSE: To investigate the cost effectiveness of adalimumab in combination with methotrexate, compared with methotrexate alone, for the management of uveitis associated with juvenile idiopathic arthritis (JIA). DESIGN: A cost-utility analysis based on a clinical trial and decision analytic model. PARTICIPANTS: Children and adolescents 2 to 18 years of age with persistently active uveitis associated with JIA, despite optimized methotrexate treatment for at least 12 weeks. METHODS: The SYCAMORE (Randomised controlled trial of the clinical effectiveness, SafetY and Cost effectiveness of Adalimumab in combination with MethOtRExate for the treatment of juvenile idiopathic arthritis associated uveitis) trial (identifier, ISRCTN10065623) of methotrexate (up to 25 mg weekly) with or without fortnightly administered adalimumab (20 or 40 mg, according to body weight) provided data on resource use (based on patient self-report and electronic records) and health utilities (from the Health Utilities Index questionnaire). Surgical event rates and long-term outcomes were based on data from a 10-year longitudinal cohort. A Markov model was used to extrapolate the effects of treatment based on visual impairment. MAIN OUTCOME MEASURES: Medical costs to the National Health Service in the United Kingdom, utility of defined health states, quality-adjusted life-years (QALYs), and incremental cost per QALY. RESULTS: Adalimumab in combination with methotrexate resulted in additional costs of £39 316, with a 0.30 QALY gain compared with methotrexate alone, resulting in an incremental cost-effectiveness ratio of £129 025 per QALY gained. The probability of cost effectiveness at a threshold of £30 000 per QALY was less than 1%. Based on a threshold analysis, a price reduction of 84% would be necessary for adalimumab to be cost effective. CONCLUSIONS: Adalimumab is clinically effective in uveitis associated with JIA; however, its cost effectiveness is not demonstrated compared with methotrexate alone in the United Kingdom setting.

Cost-Effectiveness of Panel Tests for Multiple Pharmacogenes Associated With Adverse Drug Reactions: An Evaluation Framework.

The cost-effectiveness of testing for multiple genes implicated in adverse drug reactions requires the simultaneous assessment of all actionable information, including future prescribing decisions based on incidental findings. We developed methodology for determining the value of pharmacogenetic panel tests, illustrated with a multigene panel, including HLA-A*31:01, HLA-B*15:02, HLA-B*57:01, HLA-B*58:01, HLA-B (158T), and HLA-DQB1 (126Q). If the findings for all alleles are acted upon, regardless of their individual cost-effectiveness, the HLA panel resulted in cost savings of £378 (US $491), and a quality-adjusted life year gain of 0.0069. Based on a stratified analysis and compared with no testing, initial use of the panel was cost-effective in patients eligible for abacavir (HLA-B*57:01), carbamazepine (HLA-A*31:01), and clozapine (HLA-B (158T) and HLA-DQB1 (126Q)), but not for carbamazepine (HLA-B*15:02) or allopurinol (HLA-B*58:01). The methods presented allow for the assessment of the cost-effectiveness of multiple-gene panels.

Societal Preferences for Funding Orphan Drugs in the United Kingdom: An Application of Person Trade-Off and Discrete Choice Experiment Methods.

BACKGROUND: It is unclear whether UK National Health Service (NHS) policies for orphan drugs, which permit funding of non-cost-effective treatments, reflect societal preferences. METHODS: We conducted person trade-off (PTO) and discrete choice experiment (DCE) among 3950 adults selected to be representative of the UK general population. Experimental design was informed by surveys of patients affected by rare diseases, their caregivers, health care staff, and policymakers. Societal preferences were estimated in relation to treating a common disease, increases in waiting lists, or filling of vacant NHS posts. Results of the DCE were applied to recently licensed orphan drugs. RESULTS: On the basis of equal cost, the majority of respondents to the PTO (54%; 95% confidence interval [CI] 50-59) chose to allocate funds equally between patients treated for rare diseases and those treated for common diseases, with 32% (95% CI 28-36) favoring treating rare diseases over treating common diseases (14%; 95% CI 11-17), which this reduced to 23% (95% CI 20-27) when rare disease treatments were 10 times more expensive. When framed differently, more respondents prioritized not increasing waiting list size (43%; 95% CI 39-48) than to treat rare disease patients (34%; 95% CI 30-38). DISCUSSION: The DCE indicated a greater preference for treating a common disease over a rare disease. Respondents agreed with five of 12 positive appraisal recommendations for orphan drugs, even if their list price was higher, but preferred the NHS not to fund the remainder. CONCLUSIONS: The general public does not value rarity as a sufficient reason to justify special consideration for additional NHS funding of orphan drugs. This has implications regarding the appropriateness of operating higher thresholds of cost-effectiveness.

Rare disease prevention and treatment: the need for a level playing field.

Pharmacogenetic tests are being used increasingly to prevent rare and potentially life-threatening adverse drug reactions. For many tests, however, cost-effectiveness is hard to demonstrate, and with the exception of a few cases, widespread implementation remains a distant prospect. Many orphan drugs for rare diseases are also not cost effective but are nonetheless normally reimbursed. In this article, we argue that the health technology assessment of pharmacogenetic tests aimed to prevent rare but severe adverse drug reactions should be on a level playing field with orphan drugs. This is supported by a number of arguments, concerning the severity, rarity and iatrogenic nature of adverse drug reactions, the distribution of benefits and costs and societal preference towards prevention over treatment.

Cost-Effectiveness of Pediatric Central Venous Catheters in the UK: A Secondary Publication from the CATCH Clinical Trial.

Background: Antibiotic-impregnated central venous catheters (CVCs) reduce the risk of bloodstream infections (BSIs) in patients treated in pediatric intensive care units (PICUs). However, it is unclear if they are cost-effective from the perspective of the National Health Service (NHS) in the UK. Methods: Economic evaluation alongside the CATCH trial (ISRCTN34884569) to estimate the incremental cost effectiveness ratio (ICER) of antibiotic-impregnated (rifampicin and minocycline), heparin-bonded and standard polyurethane CVCs. The 6-month costs of CVCs and hospital admissions and visits were determined from administrative hospital data and case report forms. Results: BSIs were detected in 3.59% (18/502) of patients randomized to standard, 1.44% (7/486) to antibiotic and 3.42% (17/497) to heparin CVCs. Lengths of hospital stay did not differ between intervention groups. Total mean costs (95% confidence interval) were: £45,663 (£41,647-£50,009) for antibiotic, £42,065 (£38,322-£46,110) for heparin, and £44,503 (£40,619-£48,666) for standard CVCs. As heparin CVCs were not clinically effective at reducing BSI rate compared to standard CVCs, they were considered not to be cost-effective. The ICER for antibiotic vs. standard CVCs, of £54,057 per BSI avoided, was sensitive to the analytical time horizon. Conclusions: Substituting standard CVCs for antibiotic CVCs in PICUs will result in reduced occurrence of BSI but there is uncertainty as to whether this would be a cost-effective strategy for the NHS.

Cost effectiveness analysis of HLA-B*58:01 genotyping prior to initiation of allopurinol for gout.

Objective: To determine whether prospective testing for HLA-B*58:01, as a strategy to prevent serious adverse reactions to allopurinol in patients with gout, is cost-effective from the perspective of the National Health Service in the UK. Methods: A systematic review and meta-analysis for the association of HLA-B*58:01 with cutaneous and hypersensitivity adverse drug reactions informed a decision analytic and Markov model to estimate lifetime costs and outcomes associated with testing vs standard care (with febuxostat prescribed for patients who test positive). Scenario analyses assessed alternative treatment assumptions and patient populations. Results: The number of patients needed to test to prevent one case of adverse drug reaction was 11 286 (95% central range (CR): 2573, 53 594). Cost and quality-adjusted life-year (QALY) gains were small, £103 (95% CR: £98, £106) and 0.0023 (95% CR: -0.0006, 0.0055), respectively, resulting in an incremental cost-effectiveness ratio (ICER) of £44 954 per QALY gained. The probability of testing being cost-effective at a threshold of £30 000 per QALY was 0.25. Reduced costs of testing or febuxostat resulted in an ICER below £30 000 per QALY gained. The ICER for patients with chronic renal insufficiency was £38 478 per QALY gained. Conclusion: Routine testing for HLA-B*58:01 in order to reduce the incidence of adverse drug reactions in patients being prescribed allopurinol for gout is unlikely to be cost-effective in the UK; however testing is expected to become cost-effective with reductions in the cost of genotyping, and with the future availability of cheaper, generic febuxostat.

A Systematic Review of Economic Evaluations of Pharmacogenetic Testing for Prevention of Adverse Drug Reactions.

BACKGROUND: Pharmacogenetics offers the potential to improve health outcomes by identifying individuals who are at greater risk of harm from certain medicines. Routine adoption of pharmacogenetic tests requires evidence of their cost effectiveness. OBJECTIVE: The present review aims to systematically review published economic evaluations of pharmacogenetic tests that aim to prevent or reduce the incidence of ADRs. METHODS: We conducted a systematic literature review of economic evaluations of pharmacogenetic tests aimed to reduce the incidence of adverse drug reactions. Literature was searched using Embase, MEDLINE and the NHS Economic Evaluation Database with search terms relating to pharmacogenetic testing, adverse drug reactions, economic evaluations and pharmaceuticals. Titles were screened independently by two reviewers. Articles deemed to meet the inclusion criteria were screened independently on abstract, and full texts reviewed. RESULTS: We identified 852 articles, of which 47 met the inclusion criteria. There was evidence supporting the cost effectiveness of testing for HLA-B*57:01 (prior to abacavir), HLA-B*15:02 and HLA-A*31:01 (prior to carbamazepine), HLA-B*58:01 (prior to allopurinol) and CYP2C19 (prior to clopidogrel treatment). Economic evidence was inconclusive with respect to TPMT (prior to 6-mercaptoputine, azathioprine and cisplatin therapy), CYP2C9 and VKORC1 (to inform genotype-guided dosing of coumarin derivatives), MTHFR (prior to methotrexate treatment) and factor V Leiden testing (prior to oral contraception). Testing for A1555G is not cost effective before prescribing aminoglycosides. CONCLUSIONS: Our systematic review identified robust evidence of the cost effectiveness of genotyping prior to treatment with a number of common drugs. However, further analyses and (or) availability of robust clinical evidence is necessary to make recommendations for others.

Multiple imputation of multiple multi-item scales when a full imputation model is infeasible.

BACKGROUND: Missing data in a large scale survey presents major challenges. We focus on performing multiple imputation by chained equations when data contain multiple incomplete multi-item scales. Recent authors have proposed imputing such data at the level of the individual item, but this can lead to infeasibly large imputation models. METHODS: We use data gathered from a large multinational survey, where analysis uses separate logistic regression models in each of nine country-specific data sets. In these data, applying multiple imputation by chained equations to the individual scale items is computationally infeasible. We propose an adaptation of multiple imputation by chained equations which imputes the individual scale items but reduces the number of variables in the imputation models by replacing most scale items with scale summary scores. We evaluate the feasibility of the proposed approach and compare it with a complete case analysis. We perform a simulation study to compare the proposed method with alternative approaches: we do this in a simplified setting to allow comparison with the full imputation model. RESULTS: For the case study, the proposed approach reduces the size of the prediction models from 134 predictors to a maximum of 72 and makes multiple imputation by chained equations computationally feasible. Distributions of imputed data are seen to be consistent with observed data. Results from the regression analysis with multiple imputation are similar to, but more precise than, results for complete case analysis; for the same regression models a 39% reduction in the standard error is observed. The simulation shows that our proposed method can perform comparably against the alternatives. CONCLUSIONS: By substantially reducing imputation model sizes, our adaptation makes multiple imputation feasible for large scale survey data with multiple multi-item scales. For the data considered, analysis of the multiply imputed data shows greater power and efficiency than complete case analysis. The adaptation of multiple imputation makes better use of available data and can yield substantively different results from simpler techniques.

Pharmacogenetic testing prior to carbamazepine treatment of epilepsy: patients' and physicians' preferences for testing and service delivery.

AIM: Pharmacogenetic studies have identified the presence of the HLA-A*31:01 allele as a predictor of cutaneous adverse drugs reactions (ADRs) to carbamazepine. This study aimed to ascertain the preferences of patients and clinicians to inform carbamazepine pharmacogenetic testing services. METHODS: Attributes of importance to people with epilepsy and neurologists were identified through interviews and from published sources. Discrete choice experiments (DCEs) were conducted in 82 people with epilepsy and 83 neurologists. Random-effects logit regression models were used to determine the importance of the attributes and direction of effect. RESULTS: In the patient DCE, all attributes (seizure remission, reduction in seizure frequency, memory problems, skin rash and rare, severe ADRs) were significant. The estimated utility of testing was greater, at 0.52 (95% CI 0.19, 1.00) than not testing at 0.33 (95% CI -0.07, 0.81). In the physician DCE, cost, inclusion in the British National Formulary, coverage, negative predictive value (NPV) and positive predictive value (PPV) were significant. Marginal rates of substitution indicated that neurologists were willing to pay £5.87 for a 1 percentage point increase in NPV and £3.99 for a 1 percentage point increase in PPV. CONCLUSION: The inclusion of both patients' and clinicians' perspectives represents an important contribution to the understanding of preferences towards pharmacogenetic testing prior to initiating carbamazepine. Both groups identified different attributes but had generally consistent preferences. Patients' acceptance of a decrease in treatment benefit for a reduced chance of severe ADRs adds support for the implementation of HLA-A*31:01 testing in routine practice.

Cost-effectiveness of screening for HLA-A*31:01 prior to initiation of carbamazepine in epilepsy.

OBJECTIVE: Carbamazepine causes severe cutaneous adverse drug reactions that may be predicted by the presence of the HLA-A*31:01 allele in northern European populations. There is uncertainty as to whether routine testing of patients with epilepsy is cost-effective. We conducted an economic evaluation of HLA-A*31:01 testing from the perspective of the National Health Service (NHS) in the United Kingdom. METHODS: A short-term, decision analytic model was developed to estimate the outcomes and costs associated with a policy of routine testing (with lamotrigine prescribed for patients who test positive) versus the current standard of care, which is carbamazepine prescribed without testing. A Markov model was used to estimate total costs and quality-adjusted life-years (QALYs) over a lifetime to account for differences in drug effectiveness and the long-term consequences of adverse drug reactions. RESULTS: Testing reduced the expected rate of cutaneous adverse drug reactions from 780 to 700 per 10,000 patients. The incremental cost-effectiveness ratio for pharmacogenetic testing versus standard care was £12,808 per QALY gained. The probability of testing being cost-effective at a threshold of £20,000 per QALY was 0.80, but the results were sensitive to estimated remission rates for alternative antiepileptic drugs (AEDs). SIGNIFICANCE: Routine testing for HLA-A*31:01 in order to reduce the incidence of cutaneous adverse drug reactions in patients being prescribed carbamazepine for epilepsy is likely to represent a cost-effective use of health care resources.

Predictors of self-reported adherence to antihypertensive medicines: a multinational, cross-sectional survey.

BACKGROUND: Nonadherence to antihypertensive medicines limits their effectiveness, increases the risk of adverse health outcome, and is associated with significant health care costs. The multiple causes of nonadherence differ both within and between patients and are influenced by patients' care settings. OBJECTIVES: The objective of this article was to identify determinants of patient nonadherence to antihypertensive medicines, drawing from psychosocial and economic models of behavior. METHODS: Outpatients with hypertension from Austria, Belgium, England, Germany, Greece, Hungary, The Netherlands, Poland, and Wales were recruited to a cross-sectional online survey. Nonadherence to medicines was assessed using the Morisky Medication Adherence Scale (primary outcome) and the Medication Adherence Rating Scale. Associations with adherence and nonadherence were tested for demographic, clinical, and psychosocial factors. RESULTS: A total of 2595 patients completed the questionnaire. The percentage of patients classed as nonadherent ranged from 24% in The Netherlands to 70% in Hungary. Low age, low self-efficacy, and respondents' perceptions of their illness and cost-related barriers were associated with nonadherence measured on the Morisky Medication Adherence Scale across several countries. In multilevel, multivariate analysis, low self-efficacy (odds ratio = 0.73; 95% confidence interval 0.70-0.77) and a high number of perceived barriers to taking medicines (odds ratio = 1.70; 95% confidence interval 1.38-2.09) were the main significant determinants of nonadherence. Country differences explained 11% of the variance in nonadherence. CONCLUSIONS: Among the variables measured, patients' adherence to antihypertensive medicines is influenced primarily by their self-efficacy, illness beliefs, and perceived barriers. These should be targets for interventions for improving adherence, as should an appreciation of differences among the countries in which they are being delivered.

Economic evaluation of a behavior-modifying intervention to enhance antiepileptic drug adherence.

Between 35% and 50% of patients with epilepsy are reported to be not fully adherent to their medication schedule. We aimed to conduct an economic evaluation of strategies for improving adherence to antiepileptic drugs. Based on the findings of a systematic review, we identified an implementation intention intervention (specifying when, where, and how to act) which was tested in a trial that closely resembled current clinical management of patients with epilepsy and which measured adherence with an objective and least biased method. Using patient-level data, trial patients were matched with those recruited for the Standard and New Antiepileptic Drugs trial according to their clinical characteristics and adherence. Generalized linear models were used to adjust cost and utility in order to estimate the incremental cost per quality-adjusted life-year (QALY) gained from the perspective of the National Health Service in the UK. The mean cost of the intervention group, £1340 (95% CI: £1132, £1688), was marginally lower than that of the control group representing standard care, £1352 (95% CI: £1132, £1727). Quality-adjusted life-year values in the intervention group were higher than those in the control group, i.e., 0.75 (95% CI: 0.70, 0.79) compared with 0.74 (95% CI: 0.68, 0.79), resulting in a cost saving of £12 (€15, US$19) and with the intervention being dominant. The probability that the intervention is cost-effective at a threshold of £20,000 per QALY is 94%. Our analysis lends support to the cost-effectiveness of a self-directed, implementation intention intervention for improving adherence to antiepileptic drugs. However, as with any modeling dependent on limited data on efficacy, there is considerable uncertainty surrounding the clinical effectiveness of the intervention which would require a substantive trial for a more definitive conclusion.

Random subspace ensembles for FMRI classification.

Classification of brain images obtained through functional magnetic resonance imaging (fMRI) poses a serious challenge to pattern recognition and machine learning due to the extremely large feature-to-instance ratio. This calls for revision and adaptation of the current state-of-the-art classification methods. We investigate the suitability of the random subspace (RS) ensemble method for fMRI classification. RS samples from the original feature set and builds one (base) classifier on each subset. The ensemble assigns a class label by either majority voting or averaging of output probabilities. Looking for guidelines for setting the two parameters of the method-ensemble size and feature sample size-we introduce three criteria calculated through these parameters: usability of the selected feature sets, coverage of the set of "important" features, and feature set diversity. Optimized together, these criteria work toward producing accurate and diverse individual classifiers. RS was tested on three fMRI datasets from single-subject experiments: the Haxby data (Haxby, 2001.) and two datasets collected in-house. We found that RS with support vector machines (SVM) as the base classifier outperformed single classifiers as well as some of the most widely used classifier ensembles such as bagging, AdaBoost, random forest, and rotation forest. The closest rivals were the single SVM and bagging of SVM classifiers. We use kappa-error diagrams to understand the success of RS.