Search

In light of increasing attention towards the use of real-world evidence (RWE) in decision making in recent years, this commentary aims to reflect on the experiences gained in accessing and using RWE for comparative effectiveness research as a part of the Innovative Medicines Initiative GetReal Consortium and discuss their implications for RWE use in decision-making.

Details of the rationale used to generate datasets to be used in the simulation. (DOCX 19 kb)

Details of the sensitivity analyses performed and results obtained. (DOCX 770 kb)

Objective: Amyloid-β (Aβ) positron emission tomography (PET) imaging helps estimate Aβ neuritic plaque density in patients with cognitive impairment who are under evaluation for Alzheimer’s disease (AD). This study aims to evaluate the cost-effectiveness of the Aβ-PET scan as an adjunct to standard diagnostic assessment for diagnosis of AD in France, using florbetapir as an example. Methods: A state-transition probability analysis was developed adopting the French Health Technology Assessment (HTA) perspective per guidance. Parameters included test characteristics, rate of cognitive decline, treatment effect, costs, and quality-of-life. Additional scenarios assessed the validity of the analytical framework, including (1) earlier evaluation/treatment; (2) cerebrospinal fluid (CSF) as a comparator; and (3) use of other diagnostic procedures. Outputs included differences in quality-adjusted life years (QALYs), costs, and incremental cost-effectiveness ratios (ICERs). All benefits and costs were discounted for time preferences. Sensitivity analyses were performed to assess the robustness of findings and key influencers of outcomes. Results: Aβ-PET used as an adjunct to standard diagnostic assessment increased QALYs by 0.021 years and 10-year costs by €470 per patient. The ICER was €21,888 per QALY gained compared to standard diagnostic assessment alone. When compared with CSF, Aβ-PET costs €24,084 per QALY gained. In other scenarios, Aβ-PET was consistently cost-effective relative to the commonly used affordability threshold (€40,000 per QALY). Over 95% of simulations in the sensitivity analysis were cost-effective. Conclusion: Aβ-PET is projected to affordably increase QALYs from the French HTA perspective per guidance over a range of clinical scenarios, comparators, and input parameters.

The Risk Adjustment Reform Act of 2001 mandates that a health-status-based risk adjustment mechanism has to be implemented in Germany's Statutory Health Insurance system by January 1, 2007. German parliament decided this as with the existing demographic risk adjustment model, that means there is cream skimming and sickness funds hesitate to engage in managing care for the chronical ill. Four approaches were used to test the feasibility of incorporating use of diagnosis as a proxy measure for health status in a German risk adjustment formula. The first two models used standard demographic and socio-demographic variables. The other two models are separately incorporating a simple binary indicator for hospitilization and Hierarchical Coexisting Conditions (HCCs: DxCG® Risk Adjustment Software Release 6.1) using inpatient diagnosis. Age and gender grouping accounted for 3.2% of the variation in total expenditures for concurrent as well as prospective models. The current German risk adjusters age, sex, and invalidity status account for 5.1% and 4.5% of the variance in the concurrent and prospective models respectively. There are substantial increases in explanatory power, however, when HCCs are added. Age, gender, invalidity status and HCC covariates explain about 37% of the variations of the total expenditures in a concurrent model and roughly 12% of the variations of total expenditures in a prospective model. For high-risk (cost) groups, substantial underprediction remains; conversely, for the low-risk group, represented by enrolees who did not show any health care expense in the base year, all of the models over-predict expenditure.

Utility-based measures for health-related quality of life gain more and more importance in cost-effectiveness analysis. The axiomatic foundation qualifies them as decision weights in use of the QALY concept. But their use is strained for they are loaded with assumptions to make them work. Pliskin et al. (1980) have impressively shown which assumptions might be reasonable to combine quality of life with length of life, those attributes fundamental to the QALY concept. One of those assumptions is the so called "constant proportional tradeoff". It states that people will always sacrifice the same proportion of remaining life years in order to gain better health. This assumption restricts the underlying utility functions for life years to those consistent with constant proportional risk posture, i.e. power, logarithmic and linear function. However, these types of function might be too restrictive for they do not reflect "constant absolut tradeoff". That means people might rather exchange the same number of life years for better health, independent of remaining life expectancy. Pliskin et al. mentioned that case already and suggested the exponential function as a propper function to reflect the underlying constant absolut risk posture. I will deliver its proof. In addition, a survey among Tinnitus patients is mentioned that could further stress the validity of those functions. Nutzen-basierte Maße für gesundheitsbezogene Lebensqualität gewinnen für Kosten-Effektivitäts-Analysen immer mehr an Bedeutung. Ihre axiomatische Fundierung qualifiziert sie im Gebrauch des QALY Konzeptes. Aber die Nutzung ist problematisch, da sie von verschiedenen Annahmen abhängt. Pliskin et al. (1980) haben gezeigt, welche Annahmen plausibel sind, Lebensqualität mit Lebenslänge zu kombinieren, jenen Attributen, die grundlegend für das QALY Konzept sind. Eine ist die sogenannte "konstante proportionale Austausch" Annahme. Sie besagt, dass Menschen immer eine proportionale Anzahl an Lebensjahre für bessere Gesundheit opfern würden. Diese Annahme beschränkt die zugrunde liegenden Nutzenfunktionen für Lebensjahre auf solche, die mit konstanter proportionaler Risikoeinstellung übereinstimmen, dass heisst Potenz- and Logarithmusfunktion oder eine lineare Funktion. Dennoch könnten diese Funktionstypen zu restriktiv sein, da sie nicht "konstanten absoluten Austausch" darstellen können. Damit ist gemeint, dass Menschen eventuell immer dieselbe Anzahl von Lebensjahren aufgeben, unabhängig ihrer verbleibenden Lebenserwartung. Pliskin et al. erwähnten diesen Fall bereits und schlugen die Exponentialfunktion als geeigneten Funktionstyp vor, die zugrunde liegende konstante absolute Risikoeinstellung wiederzugeben. Ich werde den Beweis liefern. Eine Befragung unter Tinnitus Patienten könnte darüber hinaus die Gültigkeit dieser Funktionen unter Beweis stellen.

Most would agree that priority setting is necessary to avoid a financial collapse in the health sector. It is much harder to find criteria how to do it. Discussions lead straight to the principles of decision making. But since all theories depend on assumptions given to make them work, debates on the assumption side are open for any kind of critic. This might be a reason hyprefernce-based methods for evaluations of different health states are not as common and popular as they could be. Indeed, it can be shown that results derived by such methods are severly biased by phenomenons which are summarized in a so-called "Prospect Theory". These biases are quite obvious if one compares data of affected and unaffected people. But this theory offers, as well, a way to get results more accurate. Man ist sich weitgehend einig, dass eine Priorisierung im Gesundheitswesen notwendig ist, will man eine Finanzierungskrise verhindern. Die Kriterien dafür festzulegen, ist bedeutend schwieriger. Diskussionen kreisen immer wieder um die grundlegenden Annahmen, wie Entscheidungen getroffen werden. Da man gezwungen ist, Verhaltensaxiome festzulegen, sind darauf aufbauende, präferenz-basierte Methoden zur Evaluation von Gesundheitszuständen nicht so weit verbreitet, wie es denkbar wäre. Tatsächlich lässt sich zeigen, dass derart erzielte Ergebnisse verzerrt sind durch Phänomene, die von Kahneman und Tversky 1979 in der sogenannten "Prospect Theory" zusammengefasst wurden. Diese Verzerrungen werden deutlich im Vergleich von Evaluationen betroffener und nicht-betroffener Personen. Auf der anderen Seite bietet diese Theorie die Möglichkeit, Ergebnisse entsprechend zu korrigieren.