|
Clare McGrath, Senior Director, Evidence-Based Strategy, Pfizer, Walton on the Hill, Surrey, UK
Patient-Benefit: What Separates Regulatory
(Licensing) From HTA Assessments?
The following is taken from the Second Plenary
Session, “Patient Reported Outcomes: A
European Perspective,” presented at the ISPOR
10th Annual European Congress, 22 October
2007, Dublin, Ireland.
Pharmaceutical companies need to satisfy the
requirements of both regulators and reimbursement
authorities when developing programmes to
assess the patient's experience of disease and
the impact of medicines. These requirements differ
substantially in terms of the scientific standards
for the data to be provided as well as the
weight given to them in decision making. This is
turn has implications for the design and feasibility
of the drug development programme. First, we
will present some definitions.
Patient benefit (as distinct from clinical outcomes)
has been described as a continuum of
parameters ranging from those “proximal to the
individual” such as symptoms, to more distal,
such as economic impact. Symptoms may be
bothersome and impact on daily life, functioning
as assessed by specific Quality of Life measures.
The symptoms may affect areas of life that are
also common to other diseases and can be
assessed using “generic” Health Related Quality
of life measures (some of which also allow one to
assess the relative importance or weight of different
health states). More peripherally still is the
impact on dependency, use of health and social
services which are influenced by external factors,
which can make it more difficult to make the link
from the symptoms through to the economic
impact of disease and interventions.
Patient Reported Outcomes (PROs) as
defined by the FDA comprise all of these
concepts, symptoms, QOL and economic
as long as they are assessed by the
patient directly. Parameters are shown in
Figure 1.
In an ideal world with no constraints, we
would be able to study the often complex
interrelationships amongst these different
levels of patient impact, how each is valued
and contributes to service use and
other health outcomes. However this is
rarely the case with drug development
where we are working against the clock
and competing for resources across the
portfolio of development opportunities.
The decision context for patient benefit assessments
depends on whether the decision is being
made for a label claim or reimbursement.
Decisions by a regulator on what should be in the
drug label are aimed at informing the individual
doctor and patient about the specific benefits and
risks of a medicine so that they can decide on its
suitability in the specific patient situation.
Reimbursement authorities on the other hand are
charged with decisions on which drugs should be
funded at which price and in which patient populations
in a way that makes best use of limited
resources. Some reimbursement authorities wish
to compare patient benefit across diseases so as
to decide on what is cost effective for a population
or group of patients (using cost per QALY
thresholds).
There are marked differences between regulators
and payors in accepted scientific standards and
the decision criteria from the patient reports.
Regulatory agencies require a rigorous approach
that ensures the content and measurement properties
of the instrument reflect what patients say
is important to them and can measure a minimally
important difference due to treatment.
Cost utility analysis uses a cost per QALY threshold
that requires consistency in assessment
across diseases and may trade off a loss of information
or precision in the patient questionnaire in
order to be ale to compare across patient groups.
Utility measures do not meet the FDA scientific
criteria for patient reported outcome often
because they are not derived directly from
patients and not validated to ensure the abovementioned
content and measurement properties
are relevant to the disease in question. The
required utility measure may also be driven by a
jurisdictional preference such as the NICE's stated
preference for the EQ5D. There is much
debate however about whether consistency in
QALY assessment is achievable and indeed
whether comparisons across diseases are actually
made in deciding on the reimbursement of
medicines.
The challenge for drug development then is to
develop a robust method of measuring patient
benefit that is both sensitive enough to detect
small but important effects but is also suitable for
comparison with other diseases that my have a
significantly different impact on the patient. Of the
many measurement systems that exist with varying
levels of dimensions, severity, and sensitivity,
no single measurement covers all patient groups.
The choice of instrument can however have significant
impact on the results of the cost per QALY
values obtained. Too little relevant content in an
instrument can result in a cost per QALY estimate
several times higher than that produced by an
instrument that contains more relevant information
for the research question such as is illustrated
in the wet AMD example in Table 1 and Graph
1. These compare the content and measurement
properties of 4 different descriptor systems
applied to wet AMD, a condition that causes visual
impairment. The ability of the instrument to
measure visual impairment specifically as in the
case of the HUI, and the higher precision of this
rating scale contribute to significantly larger
improvement in quality of life due to treatment
and resulting in a cost per QALY value
that would fall within current thresholds. The
absence of a vision specific measure (as
occurs for the EQ5D and SF8D) results in
QALY thresholds well above the threshold for
recommending this treatment. By having to
adhere to jurisdictional preferences for a
descriptor system in order to achieve consistency,
the ability to measure the important
QOL improvement may be lost and an erroneous
decision about treatment made.
In considering the different options for measurement
systems, we also need to consider
the return on the research investment.
Clearly, the potential to influence decision
making with the patient reported assessment is a key consideration. The FDA label determines
what can be used in promotion. This is relevant in
the US market that uses a high level of direct to
consumer advertising to ensure that information
about patient benefits is robust. The
FDA guidelines mandate more upfront
investment in hypothesis development,
instrument selection, and
validation. This does present challenges
in meeting these requirements
during the development programme.
Early dialogue with FDA staff is critical
to a successful measurement system.
Despite the high level of research
spending in the field of patient reported
outcomes assessment however,
there appears to be a shortage of
robust instruments to meet the FDA
requirements especially for disease
areas of high unmet need where the
up front research investment in instrument
development may not have been
made. This is being addressed
through both FDA and IMI sponsored
initiatives but may not yield instruments
for several years to come.
In Europe however, the importance
given to patient benefit assessments
in regulatory decision-making varies
by country Payers have barriers to
the use of quality of life (QoL) which
is subject to the same scientific standards
as the FDA. However, QALYs
are systematically used in 3 to 5
countries (such as UK Sweden, The
Netherlands). Many uncertainties
around the use of QALYs as a single
decision criterion exist, so QALYs should be used
judiciously in decision making with full knowledge
of the limitations of the method for each disease
and treatment situation.
Summary
Clinical Development needs to meet the very
diverse requirements of licensing and multiplepayer
requirements in the same programme.
This presents challenges in the design of Phase
2 and 3 clinical development programmes within
competitive time frames and with available
resources.
The weight given to patient reported outcomes
data is highest in the US given the role in deciding
what can be used in DTC promotion. The
resulting scientific standards are challenging to
meet but helpful in addressing misperceptions
about the credibility of patient reported data. This
should help in the longer run to improve the
acceptance of PRO data amongst European regulators
and payers.
Caution should be used in interpreting the QALY
assessment. The relevance of the content and
measurement precision of the instrument to the
disease and treatment in question play a large
role in the level of uncertainty in the results
obtained. Therefore, rigorous scientific standards
should be given priority over jurisdictional
preference when deciding on the in the
choice of instrument. In addition, by strengthening
reporting standards for the authorities
requiring QALY assessment to include the
descriptive elements of the QALY and the rationale
for choice could go a long way to improving
their interpretation and the weight given to this
types of data in decision making. |
