5. METHODS
• 2013 Journal of Hepatology
• Meta-analysis informed Markov
model
• 17 studies in meta-analysis
• Base cases estimates from Italian
National Health System and
Medicare payments
• Cost-effectiveness analysis via
EVEREST guidelines
• Estimation of treatment costs
(systemic therapies, transplant,
etc.) were not considered
RESULTS
• Pts undergoing RFA:
• Were significantly older
• Less likely to be in Child-Pugh A
• Had shorter hospital stays
• Pt survival and disease-free survival
were similar for solitary lesion <
2cm
• RFA was more cost-effective for
lesions <2cm or multiple lesions
<3cm
6. METHODS
• 2015 World Journal of Surgery
• Based on a reference case
• Survival benefit in QALY
($50,000/QALY threshold)
• Literature informed model
• Costs derived from single
institution cost-effective analysis
from 1995 (variable costs)*
RESULTS
• Liver transplant not cost-effective
as compared to liver resection or
radiofrequency ablation
• RFA with lower lifetime costs for
all HCCs <5cm
• Only Child-Pugh A patients were
modeled
*Sarasin et al. Partial hepatectomy or orthotopic liver transplantation for the treatment of resectable
hepatocellular carcinoma? A cost-effectiveness perspective. Hepatology 1998; 28(2)-436-442.
7. METHODS
• 2014 JVIR
• Calculated 3 costs:
• Cost of angiography suite
• Cost of recovery
• Cost of consumable material
• 36 patients; 50 TACE procedures
(conventional TACE)
• Costs derived from cost center
reporting and divided into
variable and fixed costs
• Hourly costs derived for each
component
RESULTS
• Angiography suite = $589/hr
• Each square foot of the hospital cost
$18.40/yr
• IR overhead = $110,336
($21.24/hr)
• Recovery hourly rate = $108/hr
• Overall, TACE cost $2223 - $5654
(median = $3269)
• Consumables = 62% of costs
8. METHODS
• 2014 JVIR
• Stage 1A/1B, SEER-Medicare data,
matched by propensity scores
• 2007-2009
• Costs derived from Medicare claims
data (total payments)
• Inpatient, outpatient, carrier files
• Costs calculated from treatment, as
well as 1, 3, 12, 18, and 24 months
after treatment (partitioned inverse
probability weighted estimator)
• Payer perspective costs
RESULTS
• No difference in survival
• $16,105 median difference in costs
between treatments
• Ablation cheaper
• Ablation patients required shorter
length of stay (2 days vs 6 days)
10. METHODS
• 2000 Hepatology
• Developed a decision model
• Assumed same all-cause mortality for
TIPS and endoscopic arms
• Probabilities for complications derived
from MEDLINE search (1990-2000)
• Costs derived from two medical
centers
• Direct costs calculated by activity-
based costing and TSI systems
solutions
• Avg cost of hospitalization for
each diagnosis
RESULTS
• Incremental cost to prevent one episode
of bleeding with TIPS as compared to
sclerotherapy ($8803) and ligation
($12,660)
• $25,000/QALY chosen as threshold
value to change therapies (vs.
$50,000 per life year saved)
• Total annual costs:
• Sclero = $23,459
• Ligation = $23,111
• TIPS = $26,275
• 30% of costs in sclero and ligation groups
were cross-over to TIPS
• Assumed 40% 1-year stenosis rate of TIPS
12. METHODS
• 2008 JVIR
• 28 hospitals, Netherlands (177 pts,
prospective)
• 4 cost categories:
• Direct medical in-hospital costs
• Out-of-hospital costs
• Direct non-medical costs
• Indirect costs (work absence)
• Direct medical costs derived from “true”
economic cost calculations from hospital
management system in the Academic Medical
Center, Amsterdam
• Complications costs collected prospectively
RESULTS
• Mean total costs of UAE were
significantly lower than
hysterectomy
• $6396 less
• Attributable to direct in-
hospital costs and indirect
costs
13. METHODS
• 2014 Journal of Comparative Effectiveness Res
• Markov Model informed by the literature
• 2 cost categories:
• Direct medical costs
• Direct and indirect non-medical costs
• Costs of treatment derived from 2003-2010
Marketscan data
• Procedure costs calculated as the sum of
payments made by patients and insurers
on the day of surgery (outpatient) or
during entire hospitalization (inpatient)
• Societal perspective
RESULTS
• Difference in effectiveness
between myomectomy,
MRgFUS, and UAE was
minimal
• But myomectomy and
MRgFUS both had better
costs/QALY
• MRgFUS found to be more
cost-effective than UAE
15. METHODS
• 2012 JVIR
• Retrospective cohort analysis
• 5% anonymized Medicare files 1999-
2009
• Primary variables = total costs, LOS,
in-hospital mortality
• Total costs = Sum of amounts paid by
Medicare and non-Medicare insurers
for the claim that covered the hospital
admission
• Co-morbidities controlled
RESULTS
• Percutaneous therapy:
• $13,565 total cost saving
• 4.1 fewer day hospitalization
• Twice as likely to be discharged
home
• 10x odds of being discharged to
hospice
• Surgery
• 8x more likely to be discharged
to another inpatient facility
17. METHODS
• 2015 JACR
• Retrospective cohort analysis
• Single-center
• Procedural, short, and long term
complications
• Costs excluded professional service
fees and charges
• Stratified into 4 components:
• Total variable labor
• Total variable supply
• Total variable other
• Total fixed cost
RESULTS
• Mean length of procedure shorter in
IR (37 min vs 69 min)
• Costs of placing port in the OR =
193% more expensive than in IR
• Early and late complications equal
between both groups
• No difference in procedural
complications
18. METHODS
• 2010 JVIR
• Retrospective cohort analysis
• Single-center
• Bacteremia and thrombosis (30-day)
were only complications recorded
• Costs derived from hospital-level cost
databases
• Out-of-pocket travel expenses and
productivity losses using average
income and average cost of travel per
mile
RESULTS
• Mean length of procedure shorter in
IR (84.9 min vs 112.8 min)
• Ward cost = most expensive
• IR insertion less costly ($13,817 less
per 100 patients)
• Significantly fewer complications in IR
vs surgery (p = 0.039)
• Incremental cost-effectiveness ratio of
$9,579 per complication averted per
100 patients
19. Costs of Complications from
Central Venous Access (preview*)
• Central venous stenosis occurs in 5—8%
of children with long-term central venous
access
• Leads to 10-17 more days in the hospital
• $200,000 - $250,000 per patient more
expensive (nearly 100% increase)
• *Kokabi et al; to be presented at SIR 2017
20. What is the current status of
cost research in IR?
• Wide variation in calculating costs from
provider, payer, and patient perspectives
• Wide variation in reported absolute costs
between single-institution studies
• Payments do not seem to be the dominant
surrogate for costs
• In nearly all aspects, IR appears to be cost-
effective