Patient preferences and HIV drugs: what about uncertainty?

Patient preferences and HIV drugs:

what about uncertainty?

Henk Broekhuizen, MSc

; Maarten IJzerman, PhD

;

Brett Hauber, PhD

;

Karin Groothuis-Oudshoorn, PhD

;

(1) University of Twente, dept. Health Technology and Services Research, Enschede, the Netherlands (2) RTI Health Solutions, Research Triangle Park, NC, USA

Poster presenter: Henk Broekhuizen, MSc. PhD candidate at University of Twente

Contact information:

h.broekhuizen@utwente.nl

www.utwente.nl/mb/htsr/Staff/broekhuizen.doc

[1] Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. National Institute of Health. May 2014.

[2] Hauber AB, Mohamed AF, Watson ME, Johnson FR, Hernandez JE. Benefits, risk, and uncertainty: preferences of antiretroviral-naïve african americans for HIV treatments. AIDS Patient Care STDS. 2009;23(1):1–6.

Part-worth utility Clinical outcome

1

2

3

4

5

6

7

8

9

1

2

3

4

5

6

7

8

9

10

Performance estimate of drug, e.g. probability of cure

Preference estimate, e.g. the added utility of 1% more probability of cure

Part-worth utility that is yielded by the performance in (1) for a patient with the preference in (2)

Probability distribution of the performance of the drug

A performance sample from the performance distribution

Probability distribution of the preferences of the patient (population)

A sampled preference representing an individual patient

Part-worth utility that corresponds to the samples in steps (5) and (7)

Repeating this process a large number of times in a Monte Carlo simulation yields a distribution of the part-worth utility of a drug

Steps (1) through (9) are repeated per treatment for each attribute and the

results are summed to obtain the probability distribution of each treatment’s

utility.

In

tegr

ation

me

thodology

Case on

tr

ea

tmen

ts

f

or

HIV

Allergic reaction (events / nonevents)

Bone damage (events / nonevents)

Kidney damage (events / nonevents) Abacavir/Lamivudine 167 / 574 14 / 329 29 / 561 33 / 689 Tenofovir/Emtricitabine 140 / 604 8 / 389 18 / 574 51 / 676 Dolutegravir 85 / 982 1 / 413 0/414 1 / 1066 Efavirenz 339 / 1884 80 / 1261 49 / 1225 11 / 2074 Atazanavir/Ritonavi 70 / 826 27 / 437 21 / 443 6 / 458 Elvitegravir/Cobicistat 43 / 305 0/348 6 / 342 3 / 345 Utility (95%CI) Base case Abacavir/Lamivudine -5.3 (-6.3 to -4.4) Tenofovir/Emtricitabine -5.3 (-6.2 to -4.35) Dolutegravir -2.0 (-2.7 to -1.3) Efavirenz -3.8 (-4.5 to -3.0) Atazanavir/Ritonavir -3.8 (-4.7 to -2.9) Elvitegravir/Cobicistat -2.8 (-3.7 to -2.0) Sc e n ar io : N o p re fe re n ce u n ce rtai n ty Abacavir/Lamivudine -5.3 (-6.0 to -4.8) Tenofovir/Emtricitabine -5.2 (-5.9 to -4.7) Dolutegravir -2.0 (-2.1 to -1.9) Efavirenz -3.7 (-4.1 to -3.5) Atazanavir/Ritonavir -3.8 (-4.4 to -3.3) Elvitegravir/Cobicistat -2.8 (-3.4 to -2.4) Objectives Conclusion

Quantitative patient preferences are increasingly considered for healthcare policy decisions. The objective of this study is to develop a methodology to combine patient preferences with clinical evidence in a multi-criteria framework that takes into account uncertainty in both preferences and clinical evidence. The methodology is illustrated with a case on antiretroviral treatments.

A probabilistic multi-criteria methodology was developed that explicitly combines patient preferences and clinical evidence. The impact of uncertainty in one or both of these on the

treatments’patient-weighted utilities can be assessed. Although limited by the small number of attributes and preference sample, the illustrative case suggests the choice of HAART is highly sensitive to patient preferences.

𝛽 𝑆𝐸(𝛽 ) 𝜎𝑞 𝑆𝐸(𝜎𝑞) Virological failure -0.04 0.01 0.05 0.02 Allergic reaction -0.06 0.01 0.06 0.02 B o n e d am ag e Treatable -0.01 0.02 0.06 0.06 Unknown -0.15 0.03 0.06 0.04 Not treatable -0.21 0.04 0.22 0.06 K id n e y D am ag e Treatable -0.05 0.02 0.12 0.02 Unknown -0.06 0.03 0.03 0.06 Not treatable -0.17 0.04 0.21 0.05