What is the relative importance of cost-effectiveness information? Results from a Discrete Choice Experiment among Swedish medical decision makers. Sandra.

What is the relative importance of cost-effectiveness information?

Results from a Discrete Choice Experiment among Swedish medical decision makers.

Sandra Erntoft (PhD)Project Manager

The Swedish Institute for Health Economics (IHE)

P.O. Box 2127, 220 02 Lund

+46 46 32 91 21

www.ihe.se

Background

• Previous research suggests that the relative importance of cost-effectiveness information varies between reimbursement-, formulary-, and prescribing - decisions.

• Little research has, however, investigated all three priority setting context simultanously…

• …and often used different methodologies and methods to investigate this question.

• Does the potential differences in the importance influence the threshold values of cost per QALY?

2

Purpose

The purpose of the experiment was to investigate the relative importance of cost-effectiveness information (cost/QALY) compared with four other criteria;

• health status,• expected size of medical effect,• type of medical effect,• budget impact,

AND

which values of a QALY are acceptable to the TLV, formulary committees and prescribing physicians?

3

Methods

• A sample of 996 questionnaires (TLV 53; formulary committee members 362; physicians 581).

• Previous study (Johnson & Backhouse 2006) and focus group consisting of 5 senior experts).

• 5 criteria – three reflecting need and two economics - 3 levels each.

• Two questions; A (ranking – ”forced choice”) and B (decision) in order to identify threshold values.

• 243 possible combinations or approx. 29 000 questions – main effects only + division into three blocks.

• Orthogonal design – iterative computer search algorithms in order to maximize D-efficiency.

• Conditional logit models.

4

Example of a D C question directed towards the TLV

Criteria Treatment A Treatment B

The average health status in patient population

High degree of pain/discomfort

Low degree of pain/discomfort

Type of medical effect Increased QoL Life-sustainment

Expected size of medical effect (effectiveness)

Avoid loss of 1 QALY Avoid loss of 0.2 QALY

Cost per QALY 102 000 € 28 000 €

Budget Impact 280 000 € per 100 000 inhabitants

18 600 € per 100 000 inhabitants

5

A) Which treatment is better? (A is better, B is better)

B) Which treatment do You think TLV should reimburse? (A, B, both A and B, neither A or B)

Formulas

6

A question (ranking):

Uij=αpain*PAIN+αtype_eff*TYPE_EFF+αQALYgain*QALY_GAIN+αcost/QALY*COST_QALY+αbudg.imp.*BUDG_IMP

B question (decision):

Vij=βpain*PAIN+βtype eff* TYPE_EFF+ β QALYgain*QALY_GAIN+ β cost/QALY*COST_QALY+ β budg.imp.*BUDG_IMP

Descriptive statistics

TLV Form. Com. Physicians

Age (mean) 52,2 54,9 47,3

Sex ( % males) 78 62 43

Education:PhysicianEconomistLawyerPharmacistOther/no answer

255000

25

86002

12

83N/aN/aN/a17

HE education (% yes) 58 35 12

Budget/ Operational responsibility (% yes) N/a 22 23

7

Response rate: 21 %

Result 1: Relative importance when ranking pharmaceutical treatments

8

Result 2: Relative importance when making a decision

9

Result 3: Cost-effectiveness threshold values

• 41 cases of statistically significant differences between decision makers.

• In 28 cases the cost-effectiveness threshold values were lower rated by the TLV, than by formulary committee members and prescribing physicians.

Cost per QALY• TLV: Lowest 43 600 € ; Highest 107 500 €.• Formulary committees: Lowest - 5 400 € ; Highest 304 200 €• Physicians: Lowest 4 900 € ; Highest 240 800 €.

1€ = 10,75 SEK (December 2009) ~ 1.3 U.S. $

10

Discussion

• Cost-effectiveness information more important in reimbursement- than in formulary- and prescribing- decisions. Confirms results from previous research.

• Threshold values are lower in reimbursment- than in formulary- and prescribing decisions. Can this be explained by differences in educational backgrounds?

• Higher threshold values in Sweden than in for instance the Netherlands.

• Willingness to reimburse (WTR) rather than willingness to pay (WTP) – social utilities rather than individual utilities.

• The WTR is based on the relative value of the public program (the treatment option rejected) foregone.

11

Conclusions

• Both the relative importance of cost-effectiveness information and the threshold values of the cost/QALY varies between decision makers at national, regional and local level.

• The relatively high threshold values among formulary committee members and prescribing physicians may be a sign of a lack of social learning regarding the necessity of setting priorities due to scarce resources….

• …or a result of the fact that priority setting is more difficult the closer the decision maker is to the patient.

12

Thank you for your attention!

Sandra Erntoft

Email: [email protected]

Phone: +46 46 32 91 21

13