Application of Network Meta-Analytic Approaches to Evaluate the Comparative Effectiveness, Economic Efficiency, and Placebo Effects among Osteoarthritis Therapeutics.
osteoarthritis is a major public health issue with an increasing incidence and
prevalence worldwide. In the absence of effective disease-modifying medical treatments,
a range of symptomatic treatments are available. Given the field's lack of comprehensive
comparative effectiveness studies that examine all modalities, making informed treatment
decisions by weighing benefits, ... read moreharms, and costs is difficult at best. Determining the
relative benefits of these treatments is further complicated by differences in the modes
of delivery of osteoarthritis treatment modalities (e.g., oral, IA, and topical).
Because randomized controlled trials employ differing placebo controls to estimate their
efficacy, a systematic variation in these placebo responses would have important
implications for the interpretation of placebo-controlled trials. Our three objectives
for the present study were: 1) To use network meta-analysis to compare the effectiveness
of different pharmacological interventions for managing knee osteoarthritis; 2) To
determine whether different placebo interventions employed in knee osteoarthritis trials
have different effect sizes, and to quantify the impact of differential placebo effects
on active treatment effect estimates; 3) To evaluate the cost-effectiveness of various
pharmacological interventions used to manage knee osteoarthritis. We systematically
searched all relevant electronic databases and unpublished data for randomized trials of
adults with knee osteoarthritis comparing two or more of the following treatments:
acetaminophen, diclofenac, ibuprofen, naproxen, celecoxib, intra-articular (IA)
corticosteroids, IA hyaluronic acid, oral placebo, and IA placebo. A network
meta-analysis using a Bayesian random-effects model was performed to assess the
comparative effectiveness of the studied interventions. We then quantified the
differences between alternative placebo interventions. Subsequently, we assessed the
implications of ignoring differential placebo effects on apparent treatment effects by
comparing the results of two different network models: (1) differential placebo network
model that treats each placebo as a different intervention; and (2) non-differential
combined placebo network model that combines all placebos into a single group. Finally,
we assessed the cost-effectiveness of these interventions using a decision analytic
model. We evaluated these strategies with and without a proton pump inhibitor PPI for
the oral treatments for the management of knee OA over a period of three months. Our
comparative effectiveness analysis identified 137 studies randomizing 33,243 patients.
For pain, all interventions significantly outperformed oral placebo, with effect sizes
from 0.63 (95% credible interval [CrI], 0.39 to 0.88) for the most efficacious treatment
(hyaluronic acid) to 0.18 (95% CrI, 0.04 to 0.33) for the least efficacious treatment
(acetaminophen). Our differential placebo response analysis identified 147 knee
osteoarthritis trials randomizing 39,022 patients. We found that IA placebo (effect
size, 95% CrI) [0.28 (0.08, 0.48)] and topical placebo [0.20 (0.02, 0.38)] had
significantly greater effect sizes for pain than oral placebo. The relative efficacies
and the hierarchy of the active treatments in the network were substantially changed by
ignoring the differential response to the four placebo types. We found that IA
corticosteroids are a cost-effective strategy for the three-month management of knee OA
with an ICER of $2870 per QALY gained compared with acetaminophen. The ibuprofen + PPI
strategy had an ICER of $23,448 per QALY gained compared to IA corticosteroids. This
study has demonstrated that IA treatments were comparatively more effective than oral
therapies in the short term management of knee osteoarthritis. Additionally, this study
illustrated that differential placebo effects can substantially alter estimates of the
relative efficacies of active treatments. Network meta-analysis provides a novel
framework to explore variations in placebo effects. Furthermore, the cost effectiveness
analysis suggests that in the short-term management of knee OA (i.e., a period of 3
months), IA corticosteroids and ibuprofen + PPI strategies are cost-effective when using
the typical US $50,000 per QALY gained willingness to pay threshold to define
Thesis (Ph.D.)--Tufts University, 2015.
Submitted to the Dept. of Clinical & Translational Science.
Advisor: Timothy McAlindon.
Committee: Christopher Schmid, John Wong, and David Kent.
Keyword: Health care management.read less