Roslyn Prichard

Purpose: Health related quality of life (HRQoL) is rarely routinely measured in the clinical setting. In the absence of patient reported data, clinicians rely on proxy and informal estimates to support clinical decisions. This study compares clinician estimates (proxy) with patient reported HRQoL in patients with advanced heart failure and examines factors influencing discrepancies.

Methods: Seventy-five patients with heart failure, (22 females, 53 males) completed the EQ-5D-5L questionnaire. Thirty-nine clinicians (11 medical, 23 nursing, 5 allied health) completed the proxy version (V1) producing 194 dyads. Correlation was assessed using Spearman’s rank tests, systematic bias was examined with Bland–Altman analyses. Inter-rater agreement at the domain level, was investigated using linear weighted Kappa statistics while factors influencing the IRG were explored using independent student t-tests, analysis of variance and regression.

Results: There was a moderate positive correlation between clinician HRQoL estimates and patient reported utility (r = 0.38; p < .0005). Mean clinician estimates were higher than patient reported utility (0.60 vs 0.54; p = 0.008), with significant underestimation of reported problems apparent in three of the five EQ-5D-5L domains. Patient sex (female), depressed mood and frailty were all associated with an increased inter-rater gap.

Conclusion: Clinicians in this sample overestimated HRQoL. Factors affecting the inter-rater gap, including sex and depression, support formal HRQoL screening to enhance clinical conversations and decision making. The discrepancy also supports regulatory restriction on the use of expert opinion in the development of QALYs in health economic analysis.

Heart failure is a common, costly and burdensome condition for individuals, health care systems and societies. Multiple innovations, both pharmacological and non-pharmacological, have transformed care for millions around the world. Ventricular assist devices (VADs), as one of these new technologies, have saved many lives and bridged many to transplantation, while improving quality of life. VAD registry and clinical trial data are providing a roadmap for clinicians in their effective and most appropriate use, but the implementation of evidence based interventions requires consideration of patient, provider and system characteristics. Australia supports a system of universal health coverage, yet has challenges in ensuring equity of access to care. Complex and fragmented funding models challenge an integrated understanding of the total costs associated with care pathways. A number of regulatory, and comprehensive health technology assessment processes inform policy and funding, and costing methods specific to local care models, are critical to inform health services planning. As health care systems around the world examine value-based funding models, understanding patient preferences is also important. Many clinical trials do not include comprehensive economic analyses nor consider incentives for, and barriers to achieving access across health systems and funding models. Communicating the value of an intervention to improve quality of life, requires consideration of patient, provider and population preferences. This thesis is comprised of a number of discrete yet linked studies that sought to address the following questions, １. How can linked data be used to establish costs and hospitalisation across different institutions in the years preceding and following VAD implant? ２. What is the relationship between VAD therapy as a bridge to transplant and overall heart failure hospitalisation and management costs compared with medical management? ３. What are the direct costs of managing advanced heart failure and VAD therapy in a quaternary specialist heart failure centre? ４. Could proxy quality of life assessments provided by clinicians for their heart failure patients provide useful utility estimates? Four studies were undertaken, and a brief statement of findings is given below.

𝘚𝘵𝘶𝘥𝘺 1: 𝘔𝘦𝘵𝘩𝘰𝘥𝘴: Linked administrative data provides a useful adjunct source for imputing costs external to the implanting centre, and combined with institutional data can illuminate both the pathways to transplant referral and the hospital activity generated by patients experiencing the terminal phases of heart failure in the year prior to transplant, cf-VAD implant or death.

𝘚𝘵𝘶𝘥𝘺 2: 𝘏𝘰𝘴𝘱𝘪𝘵𝘢𝘭𝘪𝘴𝘢𝘵𝘪𝘰𝘯: Higher pre-implant hospitalisation in VAD patients reflects higher clinical acuity and implant is significantly associated with reduced admissions, and hospitalisation once reconditioning has occurred.

𝘚𝘵𝘶𝘥𝘺 3: 𝘊𝘰𝘴𝘵𝘴: Once discharged home VAD implant stabilises hospital costs compared with accelerating costs in the year preceding implant. A high proportion of the hospital costs in the pre-implant year occur outside the implanting centre and should be considered in economic models assessing the impact of VAD implant.

𝘚𝘵𝘶𝘥𝘺 4: 𝘗𝘳𝘰𝘹𝘺 𝘘𝘶𝘢𝘭𝘪𝘵𝘺 𝘰𝘧 𝘓𝘪𝘧𝘦: Clinicians in our sample tend to overestimate prQoL in heart failure patients, with patient sex, depressed mood and measured frailty all worsening the inter rater gap

𝗜𝗺𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗳𝗼𝗿 𝗽𝗼𝗹𝗶𝗰𝘆, 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗲, 𝗲𝗱𝘂𝗰𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵. This thesis has shown the power of data to enrich clinical decision-making, as well as ensure policy and funding decisions are strategic, evidence based and reflect the needs of patients, families, communities, and clinicians. Progressing value-based health care remains a challenge requiring a commitment to rigorous studies within the framework of economic evaluation. In establishing the comparator populations, future cost effectiveness studies investigating VADs in advanced heart failure, should consider the costs and hospitalisations accrued at institutions other than the implanting centre. Formal screening for quality of life in this cohort would support reimbursement decisions and enhance shared decision-making in heart failure with real world patient reported outcomes data. Such data could impact patient selection, pre- conditioning pathways and the timing of implantation which could all impact the cost and effectiveness of ventricular assist device therapy in advanced heart failure.