CEA Fundamentals: Valuing Costs

Learning Objectives and Outline

Learning Objectives

• Identify theoretical and methodological differences between different economic evaluation techniques

• Grasp the foundations of cost-effectiveness analysis

• Describe the steps of valuing costs in economic evaluations & identify ways to curate cost parameters

Outline

1. Introduction to economic evaluations
2. Valuing costs

Introduction to economic evluations

So far…

We’ve touched on the basic framework for decision analysis, focusing on:

Decision trees & probabilities

Bayes theorem & probability revision

Constructing decision trees using Amua

Today…

• We will touch on some of the core concepts for representing costs and health benefits within decision problems

Economic Evaluation

• Relevant when decision alternatives have different costs and health consequences.

• We want to measure the relative value of one strategy in comparison to others.

• This can help us make resource allocation decisions in the face of constraints (e.g., budget).

Features of Economic Evaluation

• Systematic quantification of costs and consequences.

• Comparative analysis of alternative courses of action.

Techniques for Economic Evaluation

Type of study	Measurement/valuation of costs	Identification of consequences	Measurement / valuation of consequences
Cost analysis	Monetary units	None	None

Cost analysis

• Only looks at healthcare costs

• Relevant when alternative options are equally effective (provide equal benefits)

  • Rarely the case in reality!

• Costs are valued in monetary terms (e.g., U.S. dollars)

• Decision criterion: often to minimize cost

Techniques for Economic Evaluation

Type of study	Measurement/Valuation of costs	Identification of consequences	Measurement / valuation of consequences
Cost analysis	Monetary units	None	None
Cost-effectiveness analysis	Monetary units	Single effect of interest, common to both alternatives, but achieved to different degrees.	Natural units (e.g., life-years gained, disability days saved, points of blood pressure reduction, etc.)

Cost-Effectiveness Analysis (CEA)

Most useful when decision makers consider multiple options within a budget, and the relevant outcome is common across strategies

• Costs are valued in monetary terms ($)
• Benefits are valued in terms of clinical outcomes (e.g., cases prevented or cured, lives saved, years of life gained)
• Results reported as a cost-effectiveness ratio

Cost-Effectiveness Analysis

• Suppose we are interested in the prolongation of life after an intervention.

• Outcome of interest: life-years gained.

• The outcome is common to alternative strategies; they differ only in the magnitude of life-years gained.

• We can report results in terms of $/Life-years gained

Techniques for Economic Evaluation

Type of study	Measurement/Valuation of costs both alternative	Identification of consequences	Measurement / valuation of consequences
Cost analysis	Monetary units	None	None
Cost-effectiveness analysis	Monetary units	Single effect of interest, common to both alternatives, but achieved to different degrees.	Natural units (e.g., life-years gained, disability days saved, points of blood pressure reduction, etc.)
Cost-utility analysis	Monetary units	Single or multiple effects, not necessarily common to both alternatives.	Healthy years (typically measured as quality-adjusted life-years)

Cost-Utility Analysis

• Essentially a variant of cost-effectiveness analysis.
• Major feature: use of summary measure of health: QALY.
• Quality-Adjusted Life Year (QALY): A metric that reflects both changes in life expectancy and quality of life (pain, function, or both).
• By far the most widely published form of economic evaluation.

Techniques for Economic Evaluation

Type of study	Measurement/Valuation of costs both alternative	Identification of consequences	Measurement / valuation of consequences
Cost analysis	Monetary units	None	None
Cost-effectiveness analysis	Monetary units	Single effect of interest, common to both alternatives, but achieved to different degrees.	Natural units (e.g., life-years gained, disability days saved, points of blood pressure reduction, etc.)
Cost-utility analysis	Monetary units	Single or multiple effects, not necessarily common to both alternatives.	Healthy years (typically measured as quality-adjusted life-years)
Cost-benefit analysis	Monetary units	Single or multiple effects, not necessarily common to both alternatives	Monetary units

Cost-Benefit Analysis

• Also known as Benefit-Cost Analysis
• Relevant for resource allocation between health care and other areas (e.g., education)
• Costs and health consequences are valued in monetary terms (e.g., U.S. dollars)
• Valuation of health consequences in monetary terms ($) is obtained by estimating individuals willingness to pay for life saving or health improving interventions.
  • e.g. US estimate of value per statistical life ~$9 million
• Cost-benefit criterion: the benefits of a program > its costs
  • Notice that we’re not making comparisons across strategies–only comparisons of costs and benefits for the same strategy
• To read more: Robinson et al, 2019

Cost-Benefit Analysis

https://pubmed.ncbi.nlm.nih.gov/28183740/

Cost-Benefit Analysis

https://www.cambridge.org/core/product/identifier/S2194588818000271/type/journal_article

Back to Cost-Effectiveness Analysis!

• Relevant when healthcare alternatives have different costs & health consequences
  

\frac{\text{Cost (Intervention A) - Cost (Intervention B)}}{\text{Benefit (A) - Benefit (B)}}

• Relative VALUE of an intervention in comparison to its alternative is expressed as a cost-effectiveness RATIO (the focus of next lecture!)

1. Cost-effectiveness analysis is relevant when healthcare alternatives between different services/interventions have DIFFERENT costs AND varying health consequences EXAMPLE: Different types of chemotherapy

2. A KEY OUTCOME in cost-effectiveness analyses is what’s called the ICER [or incremental cost-effectiveness ratio], which gives us the relative VALUE of an intervention compared to an alternative.

3. In other words, THE PRICE PER UNIT BENEFIT of one intervention over the other

4. We have 2 full classes on ICERs in the coming weeks & how to interpret these ratios across interventions What we will focus on TODAY is the components that make up the numerator (i.e. COSTS) & the denominator (i.e. BENEFITS)

5. WHAT’S IMPORTANT HERE - By estimating the magnitude of health outcomes and costs of interventions, cost-effectiveness analyses can show the trade-offs involved across interventions AND THEREFORE, hopefully contribute to better decision making

Who uses economic evaluations?

• Health Technology Advisory Committees

  • NICE (The National Institute for Health and Care Excellence, UK)

  • Canada’s Drug and Health Technology Agency

  • PBAC (Pharmaceutical Benefits Advisory Committee in Australia)

  • Brazil’s health technology assessment institute

• Groups developing clinical guidelines

  • WHO

  • CDC

  • Disease-specific organizations: American Cancer Society; American Heart Association; European Stroke Organisation

• Regulatory agencies:

  • FDA (U.S. Food and Drug Administration)

  • EPA (U.S. Environmental Protection Agency)

When we look at how cost-effectiveness is used around the world –

Unlike the US, many country governments have their own institutional bodies that conduct formal assessments on the cost-effectiveness of drugs & devices – some of which include:

1. The UK - which is a single-payer system known as the NHS & has an independent body called NICE that evaluates the cost-effectiveness of health strategies for the NHS – NICE recommends to the NHS what they should cover based on VALUE (i.e. cost-effectiveness).

• & the NHS is legally obligated to fund medicines/treatments recommended by NICE

2. Canada also has a health technology assessment body called the Canadian Agency for Drugs and Technologies in Health (CADTH), which works in a similar way as NICE

3. & likewise, Australia as well as Brazil, Germany, France, Italy, Sweden, the Netherlands, etc. all have similar government institutions

4. Groups developing clinical guidelines around CEA, includes…

CEAs: Identifying Alternatives

Identifying Alternatives

• Decision modeling / economic evaluation requires identifying strategies or alternative courses of action.

• These alternatives could include different therapies / policies / technologies.

• Or, our alternatives could capture different combinations or sequences of treatment (e.g., what dose? what age to start?)

Once we have identified the alternatives, we’ll want to quantify their associated consequences in terms of:

• Health outcomes

• Costs

CEA components

\frac{\text{Cost (Intervention A) - Cost (Intervention B)}}{\text{Benefit (A) - Benefit (B)}}

1. We will now get into the components of cost-effectiveness analyses by first looking at what comprises the NUMERATOR of the cost-effectiveness ratio & then delving into the DENOMINATOR

Valuing Costs

Valuing Costs: Steps

Source: Gold 1996, Drummond 2015, Gray 2012)

1. Identify – Estimate the different categories of resources likely to be required (e.g., surgical staff, medical equipment, surgical complications, re-admissions)
   
   

2. Measure – Estimate how much of each resource category is required (e.g. type of staff performing the surgery and time involved, post-surgery length of stay, re-admission rates)
   
   

3. Value – Apply unit costs to each resource category (e.g., salary scales from the relevant hospital or national wage rates for staff inputs, cost per inpatient day for the post-surgery hospital stay)

We can identify different types of healthcare costs

• Direct Health Care Costs

  • Hospital, office, home, facilities

  • Medications, procedures, tests, professional fees

• Direct Non-Health Care Costs

  • Childcare, transportation costs

• Time Costs

  • Patient time receiving care, opportunity cost of time

• Productivity costs (‘indirect costs’)

  • impaired ability to work due to morbidity?

  • lost economic productivity due to death?

• Unrelated healthcare costs

  • Cumulative trajectory of total healthcare costs over time (unrelated to medical interventions)

Identifying costs (continued)

• In practice, we count what is likely to matter

  • Exclude what is likely to have little effect or equal effects across alternatives

• Any exclusion must be noted & possible bias examined

• We are constrained by what data are available

We can measure costs using different approaches

• Micro-costing (bottom-up)

  • Measure all resources used by individual patients, then assign the unit cost for each type of resource consumed to calculate the total cost

• Gross-costing (top-down)

  • Estimate cost for a given volume of patients by dividing the total cost by the volume of service use
  • Example: Downstream costs (e.g., hospitalization due to opioid overdose)

• Ingredients-based approach (P x Q x C)

  • Probability of occurrence (P)

  • Quantity (Q)

  • Unit costs (C)

1. Micro-costing: most precise (which is a bottom up approach where we measure, for instance, all the lab tests, hospital bed, provider time, imaging tests, equipment used, etc.)

2. Gross-costing: Usually what we have to quantify “downstream costs” – but we can also use this approach for costing out an intervention if we don’t have the data for the micro-costing approach.

• For example, if we know that 10 total knee replacements cost us $300,000 last month, that’s $30,000 per procedure.

3. Ingredients approach: Suppose we know that 10% of surgeries are complicated, and complicated surgeries require 50% more surgeon time, and surgeon time costs what $500/hr

Whose perspective?

Sanders GD, Neumann PJ, Basu A, et al. Recommendations for Conduct, Methodological Practices, and Reporting of Cost-effectiveness Analyses: Second Panel on Cost-Effectiveness in Health and Medicine. JAMA. 2016;316:1093–1103.

Whose perspective?

PERSPECTIVE MATTERS –

Formal Healthcare Sector: Medical costs borne by third-party payers & paid for out-of-pocket by patients. Should include current + future costs, related & unrelated to the condition under consideration

Societal perspective: Represents the wider “public interest” & inter-sectoral distribution of resources that are important to consider - reflects costs on all affected parties

1. PERSPECTIVE really matters when you’re deciding what costs to include in your analysis; we typically conduct CEAs from two different perspectives:

1. The formal healthcare sector perspective includes medical costs borne by third-party payers & paid for out-of-pocket by patients. Modelers should include current + future costs, related & unrelated to the condition under consideration.

• Analysts have mostly used this perspective over the last 20 years
• It is the most useful perspective to decision makers within the healthcare sector
• As mentioned earlier NICE in the UK usually takes the perspective of the healthcare payer, whereas ICER tries to do analysis from both a formal healthcare sector & wider societal perspective

2. Societal perspective includes everything in the formal health sector PLUS other things like time costs of patients seeking and receiving care, time costs of informal (unpaid) caregivers, transportation costs, effects on future productivity and consumption, and other costs and effects outside the healthcare sector.

2. The second panel on cost-effectiveness in health & medicine (the “gold standard” CEA recommendations in the US) recommends that studies be conducted from BOTH healthcare sector + societal perspectives so then we can more fully understand the wider implications of decision making.

Whose perspective?

Healthcare sector perspective

MAMMOGRAPHY (Healthcare Sector):

• Costs associated with the screening itself [mammogram procedure + physician time]

• Costs of follow-up tests for both false-positive & true positive results

• Downstream costs (or savings) associated with cases of breast cancer, such as: Hospitalization + treatment costs

• Costs unrelated to medical intervention/disease; of living longer due to mammography

1. Let’s look at an example - Mammography for breast cancer screening

2. If we were going to cost this out from a Healthcare Sector perspective, we would need to account for:

• Costs associated with the screening itself – accounting for the costs of the procedure + clinician time

• Costs of follow-up tests for both false-positive & true positive results

• Downstream costs (or savings) associated with cases of breast cancer, including hospitalizations & treatment costs

***EXAMPLE: Let’s say that we wanted to examine the difference between routine mammography screening at age 40 versus 50 – - There will be higher upfront costs for Age 40, BUT if we catch more cases of breast cancer than if we waited, then that could potentially save a lot of treatment/hospitalization costs down the line

LISA, BOOK says AVERTED COSTS for downstream costs

3. Costs unrelated to the medical intervention due to living longer (we will discuss in the next few slides)

Whose perspective?

Societal perspective

MAMMOGRAPHY (Societal perspective):

• Costs associated with the screening itself [mammogram procedure + physician time]

• Costs of follow-up tests for both false-positive & true positive results

• Downstream costs (or savings) associated with cases of breast cancer, such as: Hospitalization + treatment costs

• Costs unrelated to medical intervention/disease; of living longer due to mammography

• Patient productivity losses associated with the screening or cancer treatment

• Childcare/transportation costs

Data collection

Two approaches:

(1) Alongside clinical trials
(2) Using secondary data

Costs (secondary data)

International versus US will have different approaches

Costs (International)

1. In country/hospital/donor data registries - key is to get as close to the “true” cost associated with each procedure per patient
   • E.g., “TB healthcare & diagnostics are from official price list of the National Health Laboratory Service in South Africa; Costs for follow-up reflect local clinic and culture-based screening for active-tuberculosis”
2. Review of published literature
3. Tufts CEA Registry
4. DCP3: Disease Control Priorities

Costs (Published Literature)

Costs (Published Literature)

Costs (Tufts CEVR)

https://cevr.tuftsmedicalcenter.org/databases/cea-registry

Costs (Tufts CEVR)

Costs (Tufts CEVR)

Costs (Tufts CEVR)

http://ghcearegistry.org/ghcearegistry/

Costs (Tufts CEVR)

Costs (DCP3)

• The Disease Control Priorities is a multi-year project funded in 2010 by the Bill & Melinda Gates Foundation.
• From 2010 – 2017, it was led by the University of Washington, which worked with local organizations to promote and support the use of economic evaluation (CEAs) for priority setting at both global and national levels
• The DCP3 is now hosted at the London School of Hygiene and Tropical Medicine (LSHTM). It is currently collaborating with a number of pilot countries in the area of priority setting, development and implementation of UHC Essential Packages of Health Services.

Adjustments needed for Valuing Costs

• Adjusting for currency and currency year

• Discounting

Inflation Adjustment

Inflation Adjustment: Motivation

• $100 in 2000 is not equivalent to $100 in 2020

  • $100 could buy a lot more in 2000!

• Important to adjust for the price difference over time, especially when working with cost sources from multiple years

Inflation Adjustment: Example

Inflation Adjustment: Method

• Choose a reference year (usually the current year of analysis)

• Convert all costs to the reference year

Converting cost in Year X to Year Y (reference year):

\textbf{Cost(Year Y)} = \textbf{Cost(Year X)} \times \frac{\textbf{Price index(Year Y)}}{\textbf{Price index(Year X)}}

Inflation Adjustment: Example

Cost of hospitalization for mild stroke in the US was ~15,000 USD in 2016. What if we want to convert this number to 2020 USD?

• PCE (Personal Consumption Expenditure Health Price Index) in 2016: 105.430 (second column of Table 3 (PCE, health)

• PCE in 2020: 112.978

\textbf{Cost(2020)} = \textbf{Cost(2016)} \times \frac{\textbf{PCE(2020)}}{\textbf{PCE(2016)}} \\ = 15,000 \times \frac{112.978}{105.430} \\ = 16,674 \ (\text{2020 USD})

Currency Conversion

Currency Conversion

• Isn’t required for CEA but may be useful in some situations:

  • Example: may need to convert local currency to USD because cost-effectiveness thresholds are often estimated in the unit of USD per DALY.

• How do we convert 1,000 Turkish Liras to USD?

• Current exchange rate in 2024: 1 Turkish Lira = ~0.029 USD

• 1,000 Liras = 29 USD

Discounting

Why discounting?

• Adjust costs at social discount rate to reflect social “rate of time preference”

  • Pure time preference (“inpatience”)

  • Potential catastrophic risk in the future

  • Economic growth/return

Discounting

1. If you have $100 now, this could either be consumed or invested in the most profitable alternative (e.g., risk free government bond)

2. Let’s say the net return on the bond is 2%, then this means that next year the current $100 has grown to $102

3. The “present value” of $102 next year is $100 today

4. Similarly, $100 next year = $98.04 today

Discounting

Inflation: We convert PAST cost to present-day values

Discounting: We convert FUTURE costs to present-day values

How do we discount?

• Present value: PV = FV/(1+r)^t

  • FV = future value, the nominal cost incurred in the future

  • r = annual discount rate (analogous to interest rate)

  • t = number of years in future when cost is incurred

• Reasonable consensus around 3% per year

• May vary according to country guidelines

Adjust for inflation and currency first, then discount

Intuition

• r = 0.03

• Recall that PV = FV/(1+r)^t, and we’re at Year 0:

  • $1 in Year 0 is valued as 1/1.03^0 = \$ 1

  • $1 in Year 1 is valued as 1/1.03^1 = \$0.97

  • $1 in Year 2 is valued as 1/1.03^2 = \$0.94

  • $1 in Year 3 is valued as 1/1.03^3 = \$0.92

  • …

  • In other words, we are converting what a $1 would be in Year 2, for example, to the PRESENT VALUE of today. Today, it will be 0.94.

1. We generally value future costs and effects less than current costs and effects AND their value diminishes the more distant into the future they occur

2. Because we can invest in a dollar today to yield a higher return on investment later

Example

• Assume in year 5, a patient develops disease, and there is a treatment cost of $500
  • This is the future value (FV) of the cost!
• Present value PV = FV/(1+r)^t = 500/(1+0.03)^5 = \$ 431.3

Next up: Benefits! (the denominator)