David Wonderling

Pulling cost-effectiveness analysis up by its bootstraps: A non-parametric approach to confidence interval estimation

The statistic of interest in the economic evaluation of health care interventions is the incremental cost effectiveness ratio (ICER), which is defined as the difference in cost between two treatment interventions over the difference in their effect. Where patient-specific data on costs and health outcomes are available, it is natural to attempt to quantify uncertainty in the estimated ICER using confidence intervals. Recent articles have focused on parametric methods for constructing confidence intervals. In this paper, we describe the construction of non-parametric bootstrap confidence intervals. The advantage of such intervals is that they do not depend on parametric assumptions of the sampling distribution of the ICER. We present a detailed description of the non-parametric bootstrap applied to data from a clinical trial, in order to demonstrate the strengths and weaknesses of the approach. By examining the bootstrap confidence limits successively as the number of bootstrap replications increases, we conclude that percentile bootstrap confidence interval methods provide a promising approach to estimating the uncertainty of ICER point estimates. However, successive bootstrap estimates of bias and standard error suggests that these may be unstable; accordingly, we strongly recommend a cautious interpretation of such estimates.

Cost effectiveness analysis of different approaches of screening for familial hypercholesterolaemia

Abstract Objectives: To assess the cost effectiveness of strategies to screen for and treat familial hypercholesterolaemia. Design: Cost effectiveness analysis. A care pathway for each patient was delineated and the associated probabilities, benefits, and costs were calculated. Participants: Simulated population aged 16-54 years in England and Wales. Interventions: Identification and treatment of patients with familial hypercholesterolaemia by universal screening, opportunistic screening in primary care, screening of people admitted to hospital with premature myocardial infarction, or tracing family members of affected patients. Main outcome measure: Cost effectiveness calculated as cost per life year gained (extension of life expectancy resulting from intervention) including estimated costs of screening and treatment. Results: Tracing of family members was the most cost effective strategy (£3097 (€5066,

Constructing confidence intervals for cost‐effectiveness ratios: an evaluation of parametric and non‐parametric techniques using Monte Carlo simulation

4479) per life year gained) as 2.6 individuals need to be screened to identify one case at a cost of £133 per case detected. If the genetic mutation was known within the family then the cost per life year gained (£4914) was only slightly increased by genetic confirmation of the diagnosis. Universal population screening was least cost effective (£13 029 per life year gained) as 1365 individuals need to be screened at a cost of £9754 per case detected. For each strategy it was more cost effective to screen younger people and women. Targeted strategies were more expensive per person screened, but the cost per case detected was lower. Population screening of 16 year olds only was as cost effective as family tracing (£2777 with a clinical confirmation). Conclusions: Screening family members of people with familial hypercholesterolaemia is the most cost effective option for detecting cases across the whole population.

Non-parametric methods for cost-effectiveness analysis: the central limit theorem and the bootstrap compared.

The statistic of interest in most health economic evaluations is the incremental cost-effectiveness ratio. Since the variance of a ratio estimator is intractable, the health economics literature has suggested a number of alternative approaches to estimating confidence intervals for the cost-effectiveness ratio. In this paper, Monte Carlo simulation techniques are employed to address the question of which of the proposed methods is most appropriate. By repeatedly sampling from a known distribution and applying the different methods of confidence interval estimation, it is possible to calculate the coverage properties of each method to see if these correspond to the chosen confidence level. As the results of a single Monte Carlo experiment would be valid only for that particular set of circumstances, a series of experiments was conducted in order to examine the performance of the different methods under a variety of conditions relating to the sample size, the coefficient of variation of the numerator and denominator of the ratio, and the covariance between costs and effects in the underlying data. Response surface analysis was used to analyse the results and substantial differences between the different methods of confidence interval estimation were identified. The methods, both parametric and non-parametric, which assume a normal sampling distribution performed poorly, as did the approach based on simply combining the separate intervals on costs and effects. The choice of method for confidence interval estimation can lead to large differences in the estimated confidence limits for cost-effectiveness ratios. The importance of such differences is an empirical question and will depend to a large extent on the role of hypothesis testing in economic appraisal. However, where it is suspected that the sampling distribution is skewed, normal approximation methods produce particularly poor results and should be avoided.

Costs and cost effectiveness of cardiovascular screening and intervention: the British family heart study

Cost-effectiveness analyses (CEA) alongside randomised controlled trials commonly estimate incremental net benefits (INB), with 95% confidence intervals, and compute cost-effectiveness acceptability curves and confidence ellipses. Two alternative non-parametric methods for estimating INB are to apply the central limit theorem (CLT) or to use the non-parametric bootstrap method, although it is unclear which method is preferable. This paper describes the statistical rationale underlying each of these methods and illustrates their application with a trial-based CEA. It compares the sampling uncertainty from using either technique in a Monte Carlo simulation. The experiments are repeated varying the sample size and the skewness of costs in the population. The results showed that, even when data were highly skewed, both methods accurately estimated the true standard errors (SEs) when sample sizes were moderate to large (n>50), and also gave good estimates for small data sets with low skewness. However, when sample sizes were relatively small and the data highly skewed, using the CLT rather than the bootstrap led to slightly more accurate SEs. We conclude that while in general using either method is appropriate, the CLT is easier to implement, and provides SEs that are at least as accurate as the bootstrap.

Acupuncture in mainstream health care

Abstract Objective: To measure costs and cost effectiveness of the British family heart study cardiovascular screening and intervention programme. Design: Cost effectiveness analysis of randomised controlled trial. Clinical and resource use data taken from trial and unit cost data from external estimates. Setting: 13 general practices across Britain. Subjects: 4185 men aged 40-59 and their 2827 partners. Intervention: Nurse led programme using a family centred approach, with follow up according to degree of risk. Main outcome measures: Cost of the programme itself; overall short term cost to NHS; cost per 1% reduction in coronary risk at one year. Results: Estimated cost of putting the programme into practice for one year was £63 per person (95% confidence interval £60 to £65). The overall short term cost to the health service was £77 per man (£29 to £124) but only £13 per woman (-£48 to £74), owing to differences in utilisation of other health service resources. The cost per 1% reduction in risk was £5.08 per man (£5.92 including broader health service costs) and £5.78 per woman (£1.28 taking into account wider health service savings). Conclusions: The direct cost of the programme to a four partner practice of 7500 patients would be approximately £58 000. Annually, £8300 would currently be paid to a practice of this size working to the maximum target on the health promotion bands, plus any additional reimbursement of practice staff salaries for which the practice qualified. The broader short term costs to the NHS may augment these costs for men but offset them considerably for women. Key messages Patient specific data from the British family heart study are used in this detailed cost effectiveness analysis The costs of the programme to general practitioners was estimated with reasonable precision: an average four partner practice of 7500 patients will require 1.75 nurse years to implement this programme, costing £58 000 The direct costs of the programme may not be fully reimbursed under the current health promotion banding scheme The broader impact on drug costs and use of other health care resources is uncertain, and larger trials will be needed to estimate these important effects

Economic analysis of a pragmatic randomised trial of home visits by a nurse to elderly people with hypertension in Mexico

Is cost effective for chronic non-specific low back pain and migraine

Cost effectiveness analysis of a randomised trial of acupuncture for chronic headache in primary care

OBJECTIVE To analyse the costs and the effectiveness of an intervention of home visits made by nurses to elderly people versus usual care given by the family medicine units. MATERIAL AND METHODS A sample of 4,777 subjects aged 60 years and over covered by the Mexican Institute of Social Security (Instituto Mexicano del Seguro Social, IMSS) were screened. Those with a systolic and/or diastolic blood pressure level higher or equal than 160/90 mm Hg were randomly allocated to the intervention or control groups. The intervention consisted of visits at home by nurses who gave health and lifestyle advice to the participants. The economic evaluation was considered from a health services and patient perspective. Direct and indirect costs were calculated as incremental. Effectiveness was measured in terms of cost per millimetre of mercury reduced. RESULTS Three hundred and forty five participants were allocated to the intervention group and compared with 338 controls. At the end of the intervention period the difference in the mean change in systolic blood pressure was 3.31 mm Hg (95% CI 6.32, 0.29; p = 0.03) comparing with the control group. In diastolic blood pressure the difference was 3.67 (95% CI 5.22, 2.12; p < 0.001). The total cost of the intervention was 101 901.66 pesos. The intervention cost per patient was 34.61 pesos (US

Screening for hypercholesterolaemia versus case finding for familial hypercholesterolaemia: a systematic review and cost-effectiveness analysis.

3.78), (CI 95% 34.44, 35.46). The cost-effectiveness ratios was 10.46 pesos (US

Acupuncture of chronic headache disorders in primary care: randomised controlled trial and economic analysis

1.14) for systolic (CI 95% 129.31, 5.51) and 9.43 (US