Outcomes Research and

Pharmacoeconomics
Chapter 54
 I. GENERAL CONCEPTS
A. Outcomes research (OR)
• It is used to describe the study of health care interventions
(treatment modalities such as drug therapies, surgery, and
palliative therapy), care delivery processes, and health care
quality in order to measure the extent to which optimal and
desirable outcomes can be reached.
• It evaluates effectiveness as compared to efficacy; utilizes
observational studies that balance generalizability with
confounding/bias; is used to supplement rather than replace
randomized, controlled clinical trials; and is more extensive
than “pharmacoeconomics.”
 I. GENERAL CONCEPTS
A. Outcomes research (OR), cont’d
• Methodologies of Outcomes research:
They include retrospective chart review, prospective clinical
trials (attempts), meta-analysis, observational studies, and
computer modeling studies.
Meta-analysis is a method for systematically combining
pertinent qualitative and quantitative study data from
several selected studies to develop a single conclusion that
has greater statistical power.
 I. GENERAL CONCEPTS
A. Outcomes research (OR), cont’d
• Three areas of outcomes identified by Kozma and Reeder (1998) are:
• (a) Economic outcomes include: acquisition costs associated with
care, labor costs associated with care, costs to treat adverse drug
reactions, costs of treatment failure, costs of hospital readmission,
and costs of emergency room and clinic visits.
• (b) Clinical outcomes include: the length of hospital stay, adverse
drug reactions, hospital readmission costs, emergency room/clinic
costs, and death.
• (c) Humanistic outcomes include: patient satisfaction, functional
status as measured by a validated instrument, and quality-of-life
assessment.
 I. GENERAL CONCEPTS
B. Pharmacoeconomics (PE)
• What is Pharmacoeconomics (PE)?
• It is a division of health economics, designed to provide
decision makers with information about the value of the
different pharmacotherapies.
• Pharmacoeconomics balances the costs with the
consequences or outcomes of pharmaceutical therapies and
services.
 II. COST
A. Several concepts of cost:
• 1. Total cost. All expenses directly and indirectly necessary to provide a
product (good or service).
• 2. Average cost. The average cost per unit of output (total cost divided by
quantity of the product). Note that output is another word for the product.
• 3. Fixed cost. Costs that do not vary with the quantity of output for a short-
run production (e.g. rent, fixtures, fixed salary, depreciation, administrative
costs).
• 4. Variable cost. Costs that vary with the level of output (e.g. wages, supplies).
• 5. Marginal cost. The e extra cost of producing one extra unit of output.
• 6. Incremental cost. Additional costs when comparing one alternative or one
activity to another. It is like marginal cost but applied to changes in activities
rather than changes in units of output.
 II. COST
A. Several concepts of cost, cont’d:
• 7. Direct cost. Costs directly related to producing/providing a specific quantity
of output (e.g., salary, drug cost, supply cost for the provision of pharmacy
services).
• 8. Indirect cost. Costs that are allocated to the area(s) that produce/provide a
specific quantity of services or output (e.g., overhead cost).
• 9. Allowable cost. A cost that is eligible to claim for purposes of
reimbursement as necessary and relevant to the delivery of a unit of output.
• 10. Opportunity cost. The cost of the next best alternative forgone in order to
take the current course of action.
• 11. Operating cost. Any cost that supports the operations to provide the
output.
 II. COST
B. Cost and charge
The cost of the service may not be the same as what the patient is
charged. For example:
• Providers may include hospitals, physician offi ces, or ambulatory
surgery centers, and the term cost means the total costs for
providing the specifc service(s).
• Payors (i.e. patients) may include third-party providers, such as
insurance companies, and the term cost means the price that
they have to pay to obtain the service (i.e. charges by providers).
 II. COST
B. Cost and charge, cont’d
• Depending on the contractual terms, many providers receive only
a percentage of the charges for payment. These are called
discounted charges.
• Contractual terms are rarely reviewed and vary with different
insurance carriers. This creates confusion as to the costs
associated for various services or therapies.
• To resolve these problems, one uses the cost to charge (RCC)
ratio for the cost estimation. Multiply the RCC by the patient’s
charge to yield the estimated cost. RCC can be obtained from the
individual hospital’s Centers for Medicare and Medicaid Services
(CMS) yearly cost report.
 II. COST
C. Basic steps in assessing (estimating) cost
• 1. Define the units of service or output.
• 2. Determine the number of service or output units
provided.
• 3. Determine cost drivers of these units of service or output.
• 4. Calculate total costs, direct or indirect, related to the
provision of this output or service.
• 5. Calculate the average cost and incremental cost.
 III. ELEMENTS OF A GOOD STUDY
• A. Sound objective(s). The study has well-defined objective(s) and
answerable question(s).
• B. Perspective(s). There is a defined perspective for the analysis of
the cost-effectiveness of an intervention (e.g., perspective of the
patient, payer, provider, society). Most of the PE guidelines suggest
adding the societal perspective in the analysis.
• C. Patient population chosen must be within the scope of analysis.
Patient selection criteria that are too stringent pose a threat to
external validity. Patient selection criteria that are too liberal are a
threat to internal validity. Consideration is also given to comorbidity
and multiple treatment modalities.
 III. ELEMENTS OF A GOOD STUDY, cont’d
• D. Possible comparators and their effectiveness. All relevant
alternatives for comparison are identified. Chosen comparator(s)
should be reasonable. In some instances, no treatment (placebo) is
considered as an alternative.
• E. Metrics for costs and consequences. Measures chosen for costs
and consequences can affect the results of the analysis. Biases may
occur if units of measures are not clearly defined. This may pose a
problem with a multi country study in the aggregation of the final
costs and consequences because currency exchange rates may
fluctuate from time to time.
 III. ELEMENTS OF A GOOD STUDY, cont’d
• F. Inclusion of relevant costs and consequences in the analysis.
These are based on the perspective chosen. The computation of costs
and consequences should be transparent to readers. The key is
reproducibility. Readers should be able to use the published
computation methods with the local data to validate the findings.
• G. A valid data source. Depending on the design of the study, data
can come from clinical trials, observational studies, health care claim
databases, chart reviews, and/or epidemiological data. Each of these
sources of data was designed for some other purpose than economic
analysis, and therefore the limitations are listed in part VIII below.
 III. ELEMENTS OF A GOOD STUDY, cont’d
• H. Discounting for costs and consequences. What should be
discounted? What is the discount rate? The agreement at this time is
to discount both costs and consequences. The discount rate normally
reflects the opportunity cost of using resources. Many researchers
use the government bond rate. The researcher must justify his/her
chosen discount rate(s).
• I. Incremental analysis is the study of the comparison of costs,
generated from one alternative (activity) to another alternative, and
the additional benefit yielded from the increased cost.
 III. ELEMENTS OF A GOOD STUDY, cont’d
• J. Sensitivity analysis should include all the plausible values and their
justification for the key parameters.
• K. Time horizon of the analysis covers the full duration of treatment
for the disease process. For example, the time horizon for the
cholesterol-lowering agent’s treatment should be the life span from
the start day of treatment to the end of life.
• L. Appropriateness and comprehensiveness of presentation and
discussion of the study results. The presentation of study results
should not be biased. Interventional alternatives and study
limitations must be addressed. Generalizability and applicability are
also discussed.
 IV. PHARMACOECONOMIC METHODOLOGIES
• A. Cost of illness (COI) is the evaluation and assessment, in
dollars, of the resources used in treating an illness. It is
done before the introduction of a new intervention. Like any
PE analysis, the evaluator needs to define the analysis
perspective, because a different perspective will change the
cost structure. For example, from a patient’s perspective, the
cost of illness includes the transportation to and from the
treatment site. Time duration of the disease can be critical in
determining the cost and may be a source of bias. No
comparison is made in this type of analysis.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• B. Cost-benefit analysis (CBA) is a tool used to determine priority
for the resource allocation. The technique can be applied to the
comparison between any programs, health care programs and non-
health care programs (such as social welfare programs). For
example, one can compare the costs and benefits of a coronary risk
factor reduction program and the childhood immunization program
and the domestic violence prevention program. After converting
costs and benefits of the programs into dollars, the streams of costs
and benefits (over years) are discounted to present value at the
selected discount rate and the net benefits of the program are
computed and then then compared to each other.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• C. Cost-minimization analysis (CMA). The underlying assumption
for this type of analysis is that consequences (outputs or outcomes)
are equivalent. Therefore, only cost is compared and the cheapest
intervention will be chosen for implementation. Equivalent
outcomes may not necessarily be equal. One needs to determine
the key outcome of each comparator. For example, two drugs may
have the equivalent therapeutic value but different side effect
profiles. In such cases, consequences may not be equal, and this
technique (CMA) is not appropriate. The CEA method below should
be used instead.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• D. Cost-effectiveness analysis (CEA) is a technique to assist the decision
maker in identifying a preferred choice among possible alternatives within
similar consequences (e.g., same therapeutic category) in terms of health
improvement created (e.g., life year gained, clinical cures, etc.). It is not to
be used to compare different consequences for each alternative, such as
blood pressure reduction to degree of cholesterol lowering. Consequences
can be intermediate outcomes or surrogate outcomes such as the
reperfusion time of the vessel after thrombolytic therapy.
• Generally, the incremental cost of a program or an intervention from a
specified perspective (numerator) is compared to the incremental health
effects (denominator). An example is comparing the cost per unit of blood
pressure reduction with each antihypertensive agent. The results of the
analysis normally are stated in terms of cost per unit of effectiveness.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• D. Cost-effectiveness analysis (CEA), cont’d.
• In calculating CEA, costs refer to the expenses on the intervention, usually in
monetary terms. The measure of effects depends on the intervention being
considered. Examples include the number of people cured of a disease, the mm
Hg reduction in diastolic blood pressure and the number of symptom-free days
experienced by a patient. The selection of the appropriate effect measure should
be based on clinical judgment in the context of the intervention being
considered.
• A special case of CEA is cost utility analysis CUA, where the effects are measured
in terms of years of full health lived, using a measure, such as quality adjusted life
years QALYs or disability adjusted life years.
• Cost-effectiveness is typically expressed as an incremental cost effectiveness ratio
(ICER), the ratio of change in costs to the change in effects.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• E. Cost-utility analysis (CUA).
• CUA is the ratio between the cost of a health-related intervention and
the benefit it produces in terms of the number of years lived in full
health by the beneficiaries. Hence it can be considered a special case
of cost-effectiveness analysis CEA, and the two terms are often used
interchangeably.
• Cost is measured in monetary units. Benefit needs to be expressed in
a way that allows health states that are considered less preferable to
full health to be given quantitative values. Unlike cost benefit analysis
CBA, the benefits do not have to be expressed in monetary terms. It is
usually expressed in quality-adjusted life years (QALYs).
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• E. Cost-utility analysis (CUA), cont’d
• The metric of QALY incorporates both the improvement in quantity of
life, quantity of life, and the preference (utility value) of the health
state. There are three sources of obtaining utility values for health
states in CUA: judgment to estimate the utility values, values from the
literature, or utility values elicited from a sample of subjects.
Common techniques for eliciting utility values are: rating scale (visual
analog scale), standard gamble, and time trade-off. For these three
techniques, see: http://onlinelibrary.wiley.com/doi/10.1111/j.1553-
2712.2003.tb01349.x/pdf
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• E. Cost-utility analysis (CUA), cont’d
• After calculating the number of years gained (number) and the quality of
life weight (percentage), the QALY is calculated as follows:
• If, for example, intervention A allows a patient to live for three more years
than without intervention A, but only with a quality of life weight of 0.6,
then the intervention confers 3 * 0.6 = 1.8 QALYs to the patient. If
intervention B confers two extra years of life at a quality of life weight of
0.75, then it confers an additional 1.5 QALYs to the patient. The net benefit
of intervention A over intervention B is therefore 1.8 – 1.5 = 0.3 QALYs.
• While QALYs are used in the United States, they are not utilized to the same
degree as they are in Europe.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• E. Cost-utility analysis (CUA), cont’d
• The incremental cost-effectiveness ratio (ICER) is the ratio between the
incremental costs and the incremental benefits of two interventions. The
ICER may be stated as (C1 – C0)/(E1 – E0) in a simple example, where C0 and
E0 represent the cost and gain, from taking no health intervention action,
whereas C1 and E1 represent the cost and gain of taking a specific action.
For example, if the incremental costs and incremental gains, respectively,
are $140,000 and 3.5 QALYs, then ICER = 140000/3.5 = $40,000 per QALY.
These values are often used by policy makers and hospital administrators
to determine relative priorities when determining treatments for disease
conditions. It is important to note that CUA measures relative patient or
general population utility of a treatment or intervention. Its results give no
absolute indicator of the value of a certain treatment, but they are used for
comparison. The following explains this.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• E. Cost-utility analysis (CUA), cont’d
• In the United Kingdom, in January 2005, the health authority is believed to
have a threshold of about £30,000 per QALY – roughly twice the mean
income after tax. Thus, any health intervention which has an incremental
cost of more than £30,000 per additional QALY gained is likely to be
rejected and any intervention which has an incremental cost of less than or
equal to £30,000 per extra QALY gained is likely to be accepted as cost-
effective.
• In North America, a counterpart figure of US$50000 per QALY is often
suggested as a threshold ICER for a cost-effective intervention.
• Five circumstances have been summarized when CUA may be the
appropriate technique to apply.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• The five circumstances for applying CUA are:
• 1. When quality of life is the only outcome
• 2. When quality and quantity of life are health outcomes
• 3. When the intervention affects both mortality and morbidity and a
combined unit of outcome is desired
• 4. When the intervention being compared has a wide range of potential
outcomes and a common unit of outcome is needed
• 5. When the objective is to compare a gold standard intervention that already
has the cost per QALY. QALY is calculated by multiplying the utility values
obtained for the specific health state with the quantity of life years spent in
that specific health state. Comparison can then be made for the program and
intervention.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• F. Multiattribute utility theory (MAUT) or analysis (MAUA)
is another technique available in assessing utilities. In this
situation, several attributes can be included, such as clinical,
financial, and quality of life. It is possible to preferentially
weigh the decision based on what the priorities are for the
decision maker, and then apply the weights to identify the
most preferable therapy, service, and so on. As evidenced
from the following three examples, the individual’s
perspective will have a major effect on the final decision
made, based on the levels of priority chosen for evaluation.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• 1. A physician may view clinical outcome to represent 70% of his or
her decision, followed by patient’s quality of life (20%), and last being
therapy costs (10%).
• 2. A hospital administrator may view the financial outcomes (70%) as
a major priority, followed by the clinical outcome (20%), and last, the
quality of life (10%).
• 3. A patient with health insurance might view the clinical outcome
(45%) and quality of life (45%) as the top priorities and have minimal
concern for the financial outcomes (10%) of such a decision, specially
with adequate insurance coverage.
 IV. PHARMACOECONOMIC METHODOLOGIES,
cont’d
• G. Willingness to pay (WTP) technique is used to assess the
perceived value or benefit of a product and service. The WTP values
can be obtained through two approaches:
• 1. Indirect measurement, which examines in actual payments
previous real-world decisions that involve trade-offs between money
and expected outcomes.
• 2. Direct measurement, which uses survey methods to elicit stated
dollar on the perceived benefits. In this second approach, researchers
are seeking to provide sufficient background information to create
within the respondent’s mind a hypothetical market in which the
person provides a judgment of the value of the proposed service.
 V. DECISION ANALYSIS
• It is a systematic approach to decision making under conditions of
uncertainty. It is a tool for helping the decision makers identify options that
are available, predict the consequences and value of each option based on
the probabilities assigned to each option, and choose the option that has
the best payoff (which is the final decision of the process).
Steps in performing a decision analysis are as follows:
• A. Identify the decision. All the ground rules such as analysis perspective,
comparators selection, time span, and decision rules are identified clearly.
• B. Develop a decision tree. The decision maker structures the decision in
the form of a tree with branches (options) from left to right. Each branch is
a segment of a path that leads to an outcome.
 V. DECISION ANALYSIS
• C. Assign probabilities to branches (options). Probabilities can be obtained from
the published literature, an expert panel, or clinical trials.
• D. Value outcomes. For each of the outcome assign a monetary or utility value.
• E. Calculate the expected value by sing the averaging-out and folding-back
method, starting from the right and working backward to the left , and calculate
the expected value by multiplying the outcome value by its probability.
• F. Choose the preferred course of action. Depending on the ground rules set in
the beginning of either maximization (in case of profits, etc.) or minimization (in
case of costs, etc.), choose the best course of action.
• G. Perform a sensitivity analysis. Assign different values to all plausible outcomes
and resolve the decision tree to be sure you get consistently stable results.
 V. DECISION ANALYSIS
• https://youtu.be/jzoDKtnTPpg
Simple + an exercise at the end: (Copy and paste in youtube)
• https://youtu.be/fuBLLLPmUqg
Another simple explanation: (Copy and paste in youtube)
• http://treeplan.com/chapters/introduction-to-decision-
trees.pdf
Very clear and clean pdf explanation for the terms and for the
calculation of terminal values + an exercise at the end of the
chapter
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• PRO refers to any outcomes based on data provided by patient or
patient proxy. It includes health-related quality of life data.
• PRO data can be collected during the clinical trial. Examples of PRO
data include patient satisfaction with treatment and providers,
functional status, psychosocial well-being, treatment compliance
(adherence), and disease symptoms.
• There is a growing amount of interest in adding PRO data into the
drug review and evaluation process.
• Here are some aspects related to PRO:
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• A. Health-related quality of life (HRQOL).
• Although quality of life focuses on all aspects of life, HRQOL focuses
only on a patient’s nonclinical information such as functional status,
well-being, perception of health, return to work from an illness, and
other health outcomes that are directly affected by health status.
• Standardized questionnaires are used to capture HRQOL data in
various research settings. Data are obtained either by telephone
interview, self-administration, personal face-to-face interview,
observation, or mail-in survey.
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• A. Health-related quality of life (HRQOL). Cont’d
• Standardized questionnaires can also be divided into general health
status instruments (for example, Short-Form 36, SF-36, Short-Form
12, SF-10 for Children, SF-8 Health Surveys, sickness impact profiles
SIPs), or disease-specific instrument.
• The general health status instruments measure the global health
status, whereas the disease-specific instruments target the disease-
specific issues.
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• B. Short form surveys are the most frequently used as general health
status instruments. There are a variety of dimensions available,
depending on the chosen short form instrument. The original SF-36
includes physical functions, social functions, emotional role, physical
role, bodily pain, mental health, general health, and vitality. Other
survey instruments may contain only some of these dimensions.
Interested readers can explore the following Web site for more
information:
http://www.qualitymetric.com.
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• C. Rules and Regulations of PPACA.
• Until now, the entire set of rules and regulations of the recently approved,
Patient Protection and Affordable Care Act (PPACA) have not been
finalized. However, a key aspect of the act, besides working to ensure
health care for all U.S. citizens, is to begin creating a way where “quality-
performing” health care providers and institutions will be able to recapture
health care dollars by demonstrating improvements in patient outcomes.
This will include patient satisfaction with the services received and quality
clinical therapies provided as a way to improve outcomes (blockers after
myocardial infarctions, correct antibiotics for designated infections,
appropriate surgical procedures to minimize negative outcomes, etc.).
Health care providers will become incentivized to improve in quality as a
way to capture maximal amounts of reimbursements. Bad performances
will result in poor reimbursements.
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• D. Psychometric properties.
• Before using any instrument, the researcher must understand the
psychometric properties of the chosen instrument.
• The psychometric properties consist of the reliability and validity
information of the instrument. In addition, the sensitivity and
specificity of the instrument are also important.
• Details of this are the following:
 VI. PATIENT-REPORTED OUTCOMES (PRO)
• 1. Reliability is a measure of consistency. Can we reproduce the same score
under the same conditions with the same individual? Statistical methods of
measuring reliability are Cronbach’s α, Pearson’s r coefficient, and the κ
statistic.
• 2. Validity is a measure of accuracy. Is the instrument measuring what it is
supposed to measure? Types of validity are content validity, construct
validity, criterion validity, and convergent/divergent validity.
• 3. Use of the instrument. The psychometric properties preclude “mixing
and matching” sections of established questionnaires or selection of a
section of an established questionnaire for administration without
recalibrating the instrument’s psychometric properties.
 VII. MODELING STUDIES
• Mathematical modeling is widely used today in the economic evaluation of medications
and health care technologies.
• A. The goal for modeling is to assemble the costs and outcomes in a form that can
project long term consequences. This is a great tool for health care decision makers.
• B. Mathematical models provide the cost-consequence estimates that cannot be
revealed by randomized control trials or epidemiological studies because of the duration
required for long-term studies (10 to 20 years).
• C. Results derived from modeling assist decision makers in making informed decisions.
However, the quality of the decision is based completely on the truthfulness of the
projected results, which in turn depends on the input information and assumptions
imposed for each model.
• D. The International Society of Pharmacoeconomics and Outcomes Research (ISPOR)
recommends the following criteria for assessing the quality of models: model structure,
data used as inputs for the model, and model validation.
 VIII. 1997 FDA MODERNIZATION ACT, SECTION 114,
HEALTH
CARE ECONOMIC INFORMATION
• A. “Health care economic information provided to a formulary
committee, or other similar entity, in the course of the committee or
entity carrying out its responsibilities for the selection of drugs for
managed care or other similar organizations, shall not be considered
to be false or misleading under this paragraph if the health care
economic information directly relates to an indication approved . . .
for such drug and is based on competent and reliable scientific
evidence.” (From the FDA’s Modernization Act of 1997,
http://www.fda.gov/cder/guidance/s830enr.txt)
 VIII. 1997 FDA MODERNIZATION ACT, SECTION 114,
HEALTH
CARE ECONOMIC INFORMATION
• B. Health economic information means any analysis that identifies,
measures, or compares the economic consequences including the
costs of the represented health outcomes, or the use of a drug to the
use of another drug, or to another health care intervention, or to no
intervention.
 VIII. 1997 FDA MODERNIZATION ACT, SECTION 114,
HEALTH
CARE ECONOMIC INFORMATION
• C. Key concepts of the act
• 1. A venue for the pharmaceutical industry to provide OR and/or PE
research studies to decision makers.
• 2. Economic information can be provided in the form of CMA, CBA,
CUA, COI, and cost quality of life.
• 3. Competent and reliable scientific information pertaining to an
approved indication.
• 4. Standard of competent and reliable scientific information has not
been addressed.
IX. PRACTICAL ISSUES IN INTERPRETING OUTCOMES
RESEARCH AND PHARMACOECONOMIC STUDIES
• A. Comparisons between economic study and randomized clinical
trials (RCTs)
• 1. Economic studies are carried out in an observational environment,
whereas RCTs depend on rigorous experimental design with strict
inclusion/exclusion criteria.
• 2. RCTs rely on highly controlled and artificial clinical settings to
demonstrate clinical efficacy. Clinical and economic end points of the
study may not be the same. In addition, RCTs tend to have additional
protocol costs (e.g., extra tests) and inflated benefits (e.g., medication
compliance, appropriateness of utilization).
IX. PRACTICAL ISSUES IN INTERPRETING OUTCOMES
RESEARCH AND PHARMACOECONOMIC STUDIES
• 3. Economic studies have large sample sizes, whereas RCTs
are limited to a relatively small sample size.
• 4. Economic studies are generalizable to the broader patient
population, whereas RCTs are limited to those included
within the stringent entry criteria, which might not represent
the typical patient receiving the tested therapy.
• 5. Economic studies are usually involved in evaluating
“effectiveness” as compared to RCTs, which evaluate
“efficacy.”
IX. PRACTICAL ISSUES IN INTERPRETING OUTCOMES
RESEARCH AND PHARMACOECONOMIC STUDIES
• B. Multiple countries’ OR and PE studies
• 1. There are significant differences in physician practice patterns and
care delivery systems among different countries.
• 2. Different methods of funding health care and allocating health
expenditures make it almost impossible to calculate costs.
• 3. Patients’ concerns and beliefs are different.
IX. PRACTICAL ISSUES IN INTERPRETING OUTCOMES
RESEARCH AND PHARMACOECONOMIC STUDIES
• C. Budgetary constraints.
• Decision making should not be solely based on the
information from the PE analysis because most published PE
studies do not impose budgetary constraint as part of the
analysis. Cost-effective does not equal affordable. In
addition, one should also consider the implementation costs
of the program. In many instances, implementation costs
may exceed the benefits or effectiveness of the program.
IX. PRACTICAL ISSUES IN INTERPRETING OUTCOMES
RESEARCH AND PHARMACOECONOMIC STUDIES
• D. Reproducibility
• 1. Often, owing to journal space limitation, lengthy cost computations
are eliminated from the published article. Such practice creates an
impossible auditing mechanism for the derivation and computation of
costs. Critical assessment of this section of the published article is
necessary to ensure the validity and reliability of the results.
• 2. Modeling is an appropriate method when the disease and treatment
in question has a lengthy time span and ethical dilemma of withdrawing
treatment. However, assumptions and input values to these models are
not transparent to readers.
• 3. Both issues make it almost impossible to reproduce the study results
using the local data.
IX. PRACTICAL ISSUES IN INTERPRETING OUTCOMES
RESEARCH AND PHARMACOECONOMIC STUDIES
• E. Limitations of claim data studies.
• Claim data studies are designed for billing purposes. There is
no differentiation between comorbid conditions and
complications in coding data. This can pose a problem in
quality benchmark studies. In addition, coding practice may
be different from one institution to another, a threat to
reliability.