You are here

Article Text

Article menu
Cardiomyopathy
Original article
A cost-effectiveness model of genetic testing for the evaluation of families with hypertrophic cardiomyopathy
  1. Jodie Ingles1,2,
  2. Julie McGaughran3,4,
  3. Paul A Scuffham5,
  4. John Atherton4,6,
  5. Christopher Semsarian1,2,7
  1. 1Agnes Ginges Centre for Molecular Cardiology, Centenary Institute, Sydney, Australia
  2. 2Central Clinical School, University of Sydney, Sydney, Australia
  3. 3Genetic Health Queensland, Royal Brisbane and Women's Hospital, Brisbane, Australia
  4. 4Department of Medicine, University of Queensland, Brisbane, Australia
  5. 5Centre for Applied Health Economics, School of Medicine, Griffith University, Brisbane, Australia
  6. 6Department of Cardiology, Royal Brisbane and Women's Hospital, Brisbane, Australia
  7. 7Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia
  1. Correspondence to Professor Christopher Semsarian, Agnes Ginges Centre for Molecular Cardiology, Centenary Institute, Locked Bag 6, Newtown NSW 2042 Australia; c.semsarian{at}centenary.org.au

Abstract

Background Traditional management of families with hypertrophic cardiomyopathy (HCM) involves periodic lifetime clinical screening of family members, an approach that does not identify all gene carriers owing to incomplete penetrance and significant clinical heterogeneity. Limitations in availability and cost have meant genetic testing is not part of routine clinical management for many HCM families.

Objective To determine the cost-effectiveness of the addition of genetic testing to HCM family management, compared with clinical screening alone.

Methods A probabilistic Markov decision model was used to determine cost per quality-adjusted life-year and cost for each life-year gained when genetic testing is included in the management of Australian families with HCM, compared with the conventional approach of periodic clinical screening alone.

Results The incremental cost-effectiveness ratio (ICER) was $A785 (£510 or €587) per quality-adjusted life-year gained, and $A12 720 (£8261 or €9509) per additional life-year gained making genetic testing a very cost-effective strategy. Sensitivity analyses showed that the cost of proband genetic testing was an important variable. As the cost of proband genetic testing decreased, the ICER decreased and was cost saving when the cost fell below $A248 (£161 or €185). In addition, the mutation identification rate was also important in reducing the overall ICER, although even at the upper limits, the ICER still fell well within accepted willingness to pay bounds.

Conclusions The addition of genetic testing to the management of HCM families is cost-effective in comparison with the conventional approach of regular clinical screening. This has important implications for the evaluation of families with HCM, and suggests that all should have access to specialised cardiac genetic clinics that can offer genetic testing.

  • Hypertrophic cardiomyopathy
  • cost-effectiveness analysis
  • genetic testing
  • cardiomyopathy hypertrophic
  • sudden cardiac death
  • genetics

https://doi.org/10.1136/heartjnl-2011-300368

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Hypertrophic cardiomyopathy (HCM) is a primary genetic disorder of the myocardium characterised by hypertrophy, usually of the left ventricle, in the absence of other loading conditions such as hypertension.1 HCM affects 1 in 500 (0.2%) of the general population, and importantly remains the commonest structural cause of sudden cardiac death (SCD) in people aged <35 years.2 3 HCM shows marked clinical heterogeneity, which includes individuals with incomplete penetrance, making clinical diagnosis of family members problematic.4 To date, at least 13 causative genes have been identified in HCM, and genetic testing of the 10 most common genes yields a genetic diagnosis in about 60% of families.5 6 In the absence of a genetic diagnosis, first-degree relatives of an affected person are resigned to a lifetime of clinical surveillance based on current guidelines.7

    Proband genetic testing in HCM has limitations in clinical practice, including a less than ideal mutation detection rate and a significant cost burden. Despite this, HCM genetic testing is widely regarded as an important component of family management. At present, genotype does not guide therapeutic interventions in patients with HCM, but is primarily used for the relatives of the proband, in particular allowing for early identification of gene-carrier relatives who can be targeted for closer surveillance, particularly prevention of SCD. A new category of patient has emerged, the genotype positive–phenotype negative individual, suggesting that a number of family members may never develop clinical manifestations of the disease. Importantly, these carriers still have a 1 in 2 (50%) risk of transmitting the gene mutation to children, who might otherwise have escaped clinical detection.4 8 Predictive genetic testing is thus the only method of truly determining an individual's genetic status in the absence of a clinical phenotype. Of equal significance, a negative predictive genetic test result in an at-risk individual means they can be released from all future clinical surveillance and their children are likewise no longer at risk. Up to 50% of family members will be genotype negative, alleviating anxiety and lifetime costs of clinical screening in a significant number of people.

    The efficacy of predictive HCM genetic testing in family members at risk of HCM has been shown, and no negative impact on psychological well-being or health-related quality of life reported (HR-QoL).9–12 In fact, genetic testing may impose better HR-QoL and lower anxiety and depression than found in the general population.9 Genetic testing of inherited heart diseases in Australia is offered through clinical genetics departments or specialised cardiac genetic clinics, which adopt a multidisciplinary approach including genetic counselling.11 13

    Despite its clear benefits, genetic testing is not part of routine clinical management for the majority of families. This study sought to identify the incremental costs, effects and the incremental cost-effectiveness ratio (ICER) of a clinical surveillance strategy including genetic testing compared with the traditional method of clinical screening alone in HCM.

    Methods

    Markov decision model

    A probabilistic Markov decision model was used to compare two HCM family management strategies, one with access to genetic testing and the other being the more conventional strategy of periodic clinical surveillance alone based on current guidelines. The model was built using TreeAge Pro Software 2009, and focuses specifically on the way in which clinical surveillance is altered owing to a predictive genetic test result, and SCD prevention in clinically affected HCM individuals. Effectiveness was primarily measured in quality-adjusted life-years (QALYs), and life-years gained (LYG) were examined as a secondary measure.

    The decision node in the model represents whether families have access to proband genetic testing in addition to current clinical surveillance practice (figure 1). In the genetic testing arm, only 63% of families will have a gene mutation identified, therefore 37% will proceed through the model similarly to the no genetic testing arm.6 The Markov models can be viewed in detail in the online supplementary material.

    Figure 1
    Figure 1

    Summary of the Markov decision model. The Markov decision model was constructed to determine whether the addition of proband testing to hypertrophic cardiomyopathy (HCM) family management is more cost-effective than no genetic testing. The decision node (the square) determines what strategies are being compared—that is, genetic testing versus no genetic testing. In the path where genetic testing is available, 63% of probands will have a gene mutation identified thus allowing predictive genetic testing of family members. In the path where the proband receives an indeterminate gene result (no gene mutation identified in the proband) family members do not have access to predictive genetic testing. The Markov chance nodes are highlighted in this figure by a circle with an ‘M’, and can be viewed in more detail in the online supplementary material.

    Model variables and assumptions

    We assumed that individuals would enter the model aged 18 years, as clinically unaffected at-risk relatives. As HCM typically develops in teenage years, this age cut-off point would remove some uncertainty about disease expression, and removes ethical concerns related to predictive genetic testing of children. The cycle length was set at 12 months with half-cycle correction performed. Individuals were tracked through health states until death or age 100 years. Clinical data that informed the transition variables were sought from the Australian National Genetic Heart Disease Registry,14 from available literature and in some cases where data were lacking, expert opinion was used (provided by CS, JA and JM) (table 1). Costs were from a third-party payer perspective, and were sought from the Australian National Hospital Cost Data Collection Report Round 12 (2007–8)22 and Australian Medicare Benefits Schedule.23 All costs are presented in Australian dollars (currently £0.65 and €0.75). Utility weights were based on a recent study specifically looking at Short Form-36 (SF-36), HR-QoL and SF-6D utility weights in Australian patients with HCM and their at-risk relatives (table 2).12 Key variables used in the model are outlined in table 1 and a full list of the variables, including costs, can be found in the online supplementary material.

    View this table:
    Table 1

    Summary of key transition probabilities used in the Markov decision model

    View this table:
    Table 2

    Summary of utility weights used in the Markov decision model

    Clinical surveillance

    Clinical screening involved a consultation with a cardiologist, electrocardiogram (ECG) and a 2D and M-mode transthoracic echocardiogram. In the base case, screening occurred every 2 years from age 18 to 30 years and every three plus years from 30 years and up.7 The sensitivity analyses accounted for clinical screening occurring with a frequency of 1 to 5 yearly.

    Genetic testing

    In the base case, proband genetic testing was assumed to have a pick-up rate of 63% and laboratory cost of $A2000. It was assumed that there would be four relatives for each proband, and one to eight family members were tested in the sensitivity analyses. All genetic testing costs included laboratory test costs and an initial pre-test and follow-up result consultation with a clinical geneticist including genetic counselling. At-risk relatives with a normal phenotype who tested genotype negative required no further clinical surveillance and incurred no further costs, while those who tested positive (genotype positive–phenotype negative) were assessed annually.

    HCM clinical diagnosis and sudden death risk stratification

    In the model, individuals diagnosed with HCM underwent SCD risk stratification every 2 years involving 48 h Holter monitoring and an exercise stress test, in addition to their routine clinical evaluation (clinical review, ECG and echocardiogram). A proportion were deemed at high risk of SCD annually, and this was based on the individual having one or more risk factors as previously described.20 26

    Mortality

    The annual risk of all-cause death is based on Australian life tables from the Australian Bureau of Statistics.27

    Cost-utility and cost-effectiveness analyses

    Future costs and effects were discounted at a rate of 5% (3%–5% sensitivity range).28 The cost-effectiveness of each strategy was expressed as an ICER (cost per QALY gained and cost per LYG). Probabilistic sensitivity analysis was carried out to address the joint parameter uncertainty in the model. Cost parameters were given γ distributions, while utilities and transition probabilities were given β distributions. The lifetime costs and effects of an at-risk relative were simulated 25 000 times using Monte Carlo second-order simulations, and parameter distributions were randomly drawn on. One-way sensitivity analyses were conducted over the plausible ranges of all variables entered in the model, and represented in a tornado plot.

    Results

    Genetic testing of HCM families versus no genetic testing

    The addition of genetic testing to the management of families with HCM was cost-effective compared with the no genetic testing strategy (table 3). Specifically, the genetic testing arm had an incremental cost of $A305, incremental effect of 0.389 QALYS and ICER of $A785 per additional QALY gained (£510 and €587). For the secondary measure, there was an incremental effect of 0.024 LYG and the ICER was $A12 720 per additional LYG (£8261 and €9509), in comparison with clinical screening alone.

    View this table:
    Table 3

    Incremental costs, effects and ICER of adding proband genetic testing to the management of HCM families

    Sensitivity analysis

    The overall findings were generally robust to variation in the parameters tested in the model; however, the incremental costs, effects and ICER were sensitive to some key variables. Figure 2 shows a tornado plot illustrating the variables that have the greatest effect on the ICER in the one-way sensitivity analyses. As there was only a narrow margin of difference in costs and effects between the two strategies, variables that contributed significantly to the costs or utilities of the major health states—namely, the HCM health state, had the biggest impact on the ICER. Specifically, variables that dictate the number of patients who are diagnosed with HCM were important. Major costs associated with patients with HCM came from those deemed at high risk of SCD who would go on to have an ICD. Relatively rare events such as VF arrest and severe neurological impairment as a consequence added very little to the overall incremental costs, despite them being significantly large outlays. The variable that had the greatest effect on the ICER was the probability that a clinically unaffected at-risk relative would test gene positive, whereby the upper range led to an ICER of $A8058 for each QALY gained.

    Figure 2
    Figure 2

    Tornado plot of sensitivity analyses comparing the addition of genetic testing with clinical screening alone. The tornado plot outlines the key variables that affect the model when sensitivity analyses are carried out. GPPN, genotype positive-phenotype negative; HCM, hypertrophic cardiomyopathy; ICD, implantable cardioverter-defibrillator; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-year.

    The cost of proband genetic testing was found to be an important factor in the sensitivity analysis (table 4). The ICER was found to become cost saving when the cost of proband genetic testing fell below $A248 (equates to more than eight family members accessing predictive genetic testing); however, even with the most expensive scenario (where only one family member would undergo predictive genetic testing) the ICER remained well within accepted willingness to pay bounds (ICER $A5047 for each QALY gained). Additional analyses assuming a substantial decrease in the cost of proband genetic testing to $A500 (laboratory costs were $A2000 in the base case), with sensitivity ranges likewise spreading the cost among one to eight family members resulted in genetic testing becoming the dominant strategy when at least three at-risk relatives underwent predictive genetic testing (table 4).

    View this table:
    Table 4

    Sensitivity analyses of proband genetic testing variables

    Frequency of clinical screening of the at-risk group varied from one to five yearly, with the base case analysis reflecting current guidelines in Australia. As clinical screening costs increased (such as with more frequent screening) the genetic testing strategy became less costly.

    The utility weights were examined in one-way sensitivity analysis at upper and lower values as outlined in table 2. The utility of the at-risk health state was found to be important, with the clinical screening approach being the dominant strategy when the at-risk health state utility became >0.90. Similarly, as utilities improved in the group with borderline HCM the ICER became more expensive. The HCM utilities were examined between the ranges of 0.6 and 0.84,12 and this had little impact on the ICER.

    Figure 3 shows the scatterplot of the simulated ICERs from the probabilistic sensitivity analysis. Quadrant II on the cost effectiveness plane represents ICERs that are more costly, but more effective, and after 25 000 simulations 82.1% of the ICERs were in this quadrant. If the willingness-to-pay threshold is set at $A50 000 then there is a 78% probability that the genetic testing strategy will be cost-effective compared with clinical screening alone.

    Figure 3
    Figure 3

    Scatterplot of the outcomes of probabilistic sensitivity analysis was carried out and 25 000 simulations of the model were run. The results are shown on the cost-effectiveness plane, where quadrant I represents incremental cost-effectiveness ratios (ICERs) that are more costly and less effective, quadrant II shows ICERs that are more costly and more effective (inferior), quadrant III are ICERs that are less costly and less effective, and quadrant IV are ICERs that are less costly but more effective (superior). The percentages of ICERs in each quadrant are: quadrant I (14.25%), quadrant II (82.1%), quadrant III (0.41%) and quadrant IV (3.24%). 95% confidence ellipse indicated.

    Discounting was conservatively set at 5%; however, sensitivity analyses altering discounting to 3% altered the ICER to $A136 and $A518 per QALY gained, for cost and utilities respectively.

    Discussion

    The key finding of this study is that the addition of genetic testing to the management of HCM families is highly cost-effective in comparison with the conventional management strategy of clinical screening alone. While the genetic testing strategy remained marginally more expensive, there were increases in both QALYs and LYG. Indeed, with reductions in the cost of proband genetic testing, our model suggests that the genetic testing strategy may be cost saving.

    Genetic testing is widely regarded as an important aspect of clinical management, particularly for the relatives of the proband, but access is often limited owing to the cost of testing. Less than 25% of HCM families enrolled in the Australian National Genetic Heart Disease Registry have had a proband undergo genetic testing (unpublished data), indicating that access in this population is limited. In a healthcare system constantly lacking adequate funding, health economic analyses to guide resource allocation are crucial. Importantly, our findings support a recent study that concluded that HCM genetic testing can be cost-effective compared with clinical screening alone.29 This is an important consideration, as both studies used very different models, and also were based on two different healthcare systems (UK and Australia), but significantly, reached similar conclusions. Despite these similarities, no study to date has investigated a cost per QALY approach to HCM genetic testing, and indeed the use of SF-6D utility scores from a large Australian HCM cohort provides robust data to inform a cost utility evaluation.

    An important limitation of genetic testing in HCM is the causative mutation pick-up rate, which at best is approximately 60%.6 This means that the remainder of families are resigned to a lifetime of clinical surveillance. Sensitivity analysis of the model examining the upper and lower ranges of proband gene mutation pick-up rates led to the ICER becoming less costly as the pick-up rate increased. When the upper limit was increased to a 100% mutation identification rate, as is the ultimate goal of any genetic test, the ICER became $A258 for each QALY gained. Interestingly, even at the lowest limit of 40% detection rate (ICER $A1601 for each QALY gained), genetic testing still falls well within accepted willingness-to-pay thresholds. Lower mutation identification rates are often reported in individuals where the proband appears to have sporadic disease,5 and this result suggests that genetic testing is indicated even in these probands where familial disease cannot be established.

    The cost of genetic testing was identified as an important variable in families with HCM. As genetic technology improves further, particularly in light of recent advances in next-generation sequencing, proband genetic testing costs are expected to decrease substantially30 and for this reason we carried out an additional analysis where we set the cost of proband genetic testing at $A500. As proband genetic testing costs decreased, the ICER also decreased, and when it fell below $A248 the strategy became cost saving. This equates to a cost-saving result when at least three family members were tested (following the method of dividing proband genetic tests by the number of relatives who use predictive genetic testing). In contrast, even the most expensive scenario using the current cost of proband genetic testing with only one family member undergoing predictive genetic testing, results in the ICER still falling well under accepted willingness-to-pay thresholds described in the literature.31 This supports the view that genetic testing should be offered to all HCM families, particularly when there are multiple family members who would probably access predictive genetic testing.

    The prevention of SCD using implantable cardioverter-defibrillator (ICD) therapy was an important focus in our model. We expect predictive genetic testing to lead to earlier HCM diagnoses among at-risk relatives, owing to more frequent clinical surveillance, including further evaluation with 48 h Holter monitoring and exercise testing to determine SCD risk status. We therefore expected that there might be more individuals who would receive ICD therapy for primary prevention. Thus, in our model the increased costs in the genetic testing arm can be attributed to more people requiring ICD implantation, and is consistent with the recent study by Wordsworth et al.29

    The utility scores used to inform the effects of each health state were taken from Australian patients and their family members.12 Overall QALYs were not greatly increased owing to the addition of genetic testing to family management strategies, as most disease states were not greatly impaired. The small gains in QALYs are in the most part due to a number of individuals entering the gene-negative health state where they will continue to accumulate QALYs comparable to those of the general population. Utilities were calculated using the SF-6D, which is most sensitive to small changes in health states at the higher levels of well-being.32

    Importantly, our study provides a basis for government-funded HCM genetic testing in Australia, and this may be applicable to other health systems globally. The Australian Pharmaceutical Benefits Advisory Committee has in the past supported the use of new drugs for which the additional cost per LYG was <$A42 000, making our ICER of $A12 720 per additional LYG a very cost-effective option.31 A government-funded genetic testing programme for HCM would probably improve access and availability of testing such that all HCM families could consider the options of genetic testing and the associated benefits. These findings are applicable to the Australian healthcare system, and also can be extrapolated to other countries where clinical screening and management guidelines are similar, and proband genetic testing costs comparable.

    HCM is a complex heterogeneous genetic disorder, and therefore the model will have some limitations. Specifically, Markov modelling tracks the progress of a single individual; however, a diagnosis of HCM has implications for both the proband and the family members. For simplicity, it was assumed all individuals would undergo predictive genetic testing if it was available, or that they would attend for regular clinical surveillance in the absence of a genetic diagnosis. The issue of determining pathogenicity of a gene mutation is an important one, and we made the assumption that our mutation pick-up estimates included only causative gene mutations and not variants of uncertain significance.

    In conclusion, genetic testing is an important component of the evaluation of individuals and families with HCM. The addition of genetic testing to the management of HCM families is cost-effective in comparison with the conventional approach of regular clinical screening. This has important implications for the evaluation of families with HCM, and suggests that all should have access to specialised cardiac genetic clinics that can offer genetic testing. As the cost of genetic testing falls and the mutation pick-up rate for HCM increases, the cost-effectiveness of genetic testing in HCM is likely to improve even further.

    Acknowledgments

    JI is the recipient of a University of Sydney postgraduate award, and CS is the recipient of a National Health and Medical Research Council (NHMRC) practitioner fellowship.

    References

    1. ↵
      1. Watkins H,
      2. Ashrafian H,
      3. Redwood C
      . Inherited cardiomyopathies. N Engl J Med 2011;364:1643–56.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Maron B,
      2. Gardin J,
      3. Flack J,
      4. et al
      . Prevalence of hypertrophic cardiomyopathy in a general population of young adults: echocardiographic analysis of 4111 subjects in the CARDIA Study. Circulation 1995;92:785–9.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Maron BJ
      . Sudden death in young athletes. N Engl J Med 2003;349:1064–75.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Maron BJ,
      2. Yeates L,
      3. Semsarian C
      . Clinical challenges of genotype positive (+)-phenotype negative (-) family members in hypertrophic cardiomyopathy. Am J Cardiol 2011;107:604–8.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Van Driest SL,
      2. Ommen SR,
      3. Tajik AJ,
      4. et al
      . Yield of genetic testing in hypertrophic cardiomyopathy. Mayo Clin Proc 2005;80:739–44.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Richard P,
      2. Charron P,
      3. Carrier L,
      4. et al
      . Hypertrophic cardiomyopathy: distribution of disease genes, spectrum of mutations, and implications for a molecular diagnosis strategy. Circulation 2003;107:2227–32.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Semsarian C
      . Guidelines for the diagnosis and management of hypertrophiccardiomyopathy. Heart Lung Circ 2011;11:688–90. Epub 31 August 2011.
      OpenUrl
    8. ↵
      1. Maron BJ,
      2. Semsarian C
      . Emergence of gene mutation carriers and the expanding disease spectrum of hypertrophic cardiomyopathy. Eur Heart J 2010;31:1551–3.
      OpenUrlFREE Full Text
    9. ↵
      1. Christiaans I,
      2. van Langen IM,
      3. Birnie E,
      4. et al
      . Quality of life and psychological distress in hypertrophic cardiomyopathy mutation carriers: a cross-sectional cohort study. Am J Med Genet A 2009;149A:602–12.
      OpenUrlPubMed
    10. ↵
      1. Hamang A,
      2. Eide GE,
      3. Nordin K,
      4. et al
      . Health status in patients at risk of inherited arrhythmias and sudden unexpected death compared to the general population. BMC Med Genet 2010;11:27.
      OpenUrlCrossRefPubMed
    11. ↵
      1. Ingles J,
      2. Lind JM,
      3. Phongsavan P,
      4. et al
      . Psychosocial impact of specialized cardiac genetic clinics for hypertrophic cardiomyopathy. Genet Med 2008;10:117–20.
      OpenUrlPubMed
    12. ↵
      1. Ingles J,
      2. Yeates L,
      3. Hunt L,
      4. et al
      . Health status of cardiac genetic disease patients and theirat-risk relatives. Int J Cardiol 2011; Epub ahead of print Sep 17.
    13. ↵
      1. Ingles J,
      2. Semsarian C
      . Sudden cardiac death in the young: a clinical genetic approach. Intern Med J 2007;37:32–7.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Ingles J,
      2. McGaughran J,
      3. Vohra J,
      4. et al
      . Establishment of an Australian National Genetic Heart Disease Registry. Heart Lung Circ 2008;17:463–7.
      OpenUrlCrossRefPubMedWeb of Science
      1. Elliott PM,
      2. Poloniecki J,
      3. Dickie S,
      4. et al
      . Sudden death in hypertrophic cardiomyopathy: identification of high risk patients. J Am Coll Cardiol 2000;36:2212–18.
      OpenUrlCrossRefPubMedWeb of Science
      1. Jennings PA,
      2. Cameron P,
      3. Walker T,
      4. et al
      . Out-of-hospital cardiac arrest in Victoria: rural and urban outcomes. Med J Aust 2006;185:135–9.
      OpenUrlPubMedWeb of Science
      1. Graves JR,
      2. Herlitz J,
      3. Bang A,
      4. et al
      . Survivors of out of hospital cardiac arrest: their prognosis, longevity and functional status. Resuscitation 1997;35:117–21.
      OpenUrlCrossRefPubMedWeb of Science
      1. Hsu JW,
      2. Madsen CD,
      3. Callaham ML
      . Quality-of-life and formal functional testing of survivors of out-of-hospital cardiac arrest correlates poorly with traditional neurologic outcome scales. Ann Emerg Med 1996;28:597–605.
      OpenUrlCrossRefPubMedWeb of Science
      1. Michels M,
      2. Soliman OI,
      3. Phefferkorn J,
      4. et al
      . Disease penetrance and risk stratification for sudden cardiac death in asymptomatic hypertrophic cardiomyopathy mutation carriers. Eur Heart J 2009;30:2593–8.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Maron BJ,
      2. Spirito P,
      3. Shen WK,
      4. et al
      . Implantable cardioverter-defibrillators and prevention of sudden cardiac death in hypertrophic cardiomyopathy. JAMA 2007;298:405–12.
      OpenUrlCrossRefPubMedWeb of Science
      1. Andersen PS,
      2. Havndrup O,
      3. Hougs L,
      4. et al
      . Diagnostic yield, interpretation, and clinical utility of mutation screening of sarcomere encoding genes in Danish hypertrophic cardiomyopathy patients and relatives. Hum Mutat 2009;30:363–70.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      National Hospital Cost Data Collection Cost Report Round 12 (2007-2008). Canberra, Australia: Australian Government, 2007–2008.
    23. ↵
      Medicare Benefits Schedule Book. Canberra: Department of Health and Ageing, Australian Government, 2010.
      1. Post PN,
      2. Stiggelbout AM,
      3. Wakker PP
      . The utility of health states after stroke: a systematic review of the literature. Stroke 2001;32:1425–9.
      OpenUrlAbstract/FREE Full Text
      1. Hawthorne G,
      2. Richardson J,
      3. Osborne R
      . The Assessment of Quality of Life (AQoL) instrument: a psychometric measure of health-related quality of life. Qual Life Res 1999;8:209–24.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Jayatilleke I,
      2. Doolan A,
      3. Ingles J,
      4. et al
      . Long-term follow-up of implantable cardioverter defibrillator therapy for hypertrophic cardiomyopathy. Am J Cardiol 2004;93:1192–4.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      ABS. Life Tables, Australia 2005-2008. Canberra, Australia: Australian Bureau of Statistics, 2008.
    28. ↵
      Australian Government DoHaA. Guidelines for Preparing Submissions to the Pharmaceutical Benefits Advisory Committee (Version 4.3). Canberra, Australia: Department of Health and Ageing, 2008.
    29. ↵
      1. Wordsworth S,
      2. Leal J,
      3. Blair E,
      4. et al
      . DNA testing for hypertrophic cardiomyopathy: a cost-effectiveness model. Eur Heart J 2010;31:926–35.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Bagnall R,
      2. Ingles J,
      3. Semsarian C
      . Molecular diagnostics of cardiomyopathies: the future is here. Circ Cardiovasc Genet 2011;4:103–4.
      OpenUrlFREE Full Text
    31. ↵
      1. George B,
      2. Harris A,
      3. Mitchell A
      . Cost-effectiveness analysis and the consistency of decision making: evidence from pharmaceutical reimbursement in Australia (1991 to 1996). Pharmacoeconomics 2001;19:1103–9.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Walters SJ,
      2. Brazier JE
      . Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D. Qual Life Res 2005;14:1523–32.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • See Editorial, p 603

    • Funding This study was supported, in part, by an NHMRC project grant.

    • Competing interests None.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement Data available from the authors on request.

    Linked Articles