studies

920A AASLD ABSTRACTS HEPATOLOGY, October, 2015
1455
A probabilistic decision model using non-invasive fibrosis
markers in Primary Care NAFLD pathways predicts
increased cirrhosis detection rates and reduced overall
healthcare expenditure
Ankur Srivastava 1,2 , Simcha Jong 3 , Anna Gola 4 , Laura Fenlon 5 ,
Petra Scantlebury 5 , Sudeep Tanwar 1,2 , Hannah Liu 7 , Ruth E.
Gailer 6,7 , Sarah Morgan 6 , Alex Warner 6 , Karen Sennett 7 , Julie
Parkes 1 , James O’Beirne 1,2 , Emmanuel Tsochatzis 1,2 , William M.
Rosenberg 1,2 ; 1 Institute of Liver and Digestive Health, University
College London (UCL), London, United Kingdom; 2 Hepatology,
Royal Free London NHS Foundation Trust, London, United Kingdom;
3 Management Science & Innovation, University College
London, London, United Kingdom; 4 Health Economics, University
College London, London, United Kingdom; 5 Department of
Finance, Royal Free London NHS Foundation Trust, London, United
Kingdom; 6 Primary Care, Camden Clinical Commissioning Group,
London, United Kingdom; 7 Primary Care, Islington Clinical Commissioning
Group, London, United Kingdom
Background: Risk factors for Non-Alcoholic Fatty Liver Disease
(NAFLD) have reached epidemic proportions and whilst only
a minority of at-risk individuals develop significant liver disease
warranting specialist referral, ensuring identification of
this group is a primary care challenge. Non-invasive tests (NIT)
for liver fibrosis enable community-based doctors to identify
patients with NAFLD who have advanced fibrosis and would
benefit from early referral for specialist care and interventions
to limit the complications of cirrhosis. As part of an evaluation
of a novel primary care NAFLD pathway in two London boroughs
(Camden and Islington, C&I), which uses NIT to stratify
patients for advanced liver fibrosis, we have constructed an
analytical model to examine the potential cost implications of
this strategy on an intention-to-treat basis. Methods: A probabilistic
decision analytical model was constructed to explore
the financial dimensions of two strategies in community based
management of NAFLD from a healthcare payer perspective:
1) Standard of care (SOC) comprising physicians’ clinical judgment
based on clinical history, examination, blood tests and
ultrasound 2) SOC plus Non-Invasive Testing for liver fibrosis
using FIB-4, and the ELF test (to resolve indeterminate FIB-4
tests) to guide referral. The model was populated from the literature,
national UK data and expert opinion. A five year time
horizon was applied. Results: The C&I population is 434,958.
It was assumed that 30% of the NAFLD population consult their
general practitioner (GP) annually. >F2 fibrosis prevalence in
NAFLD was set at 5%. Local clinical audit estimated SOC sensitivities
and specificities of 0.35 and 0.70. The model estimated
24% needed ELF test. Utilizing NIT reduced referrals of
≤F2 disease by over 80% and resulted in over 50% increase
in the detection of cirrhosis. There was a significant reduction
in liver transplantation rates by approximately 15%. The
overall expenditure was reduced by a fifth, primarily through
reductions in referrals for low risk cases, costs related to HCC
management and emergency inpatient admissions. Discussion:
The use of NIT led to a reduction in inappropriate referrals and
improvement in the early detection of cirrhosis and its complications.
These changes are likely to result in significant health
gains and economic cost savings. With NAFLD prevalence
likely to increase, policy needs to address these challenges.
Modeling suggests that the Camden and Islington NAFLD pathway
employing FIB-4 and ELF to guide GPs’ risk stratification
of patients will achieve these goals. A full clinical effectiveness
and cost consequence analysis is underway.
Disclosures:
Simcha Jong - Grant/Research Support: Siemens
Julie Parkes - Stock Shareholder: iQur (spouse is shareholder)
William M. Rosenberg - Advisory Committees or Review Panels: Janssen, Merk,
Gilead, Merk, Gilead, GSK; Board Membership: iQur Limited, iQur Limited;
Consulting: siemens; Speaking and Teaching: siemens, Roche
The following authors have nothing to disclose: Ankur Srivastava, Anna Gola,
Laura Fenlon, Petra Scantlebury, Sudeep Tanwar, Hannah Liu, Ruth E. Gailer,
Sarah Morgan, Alex Warner, Karen Sennett, James O’Beirne, Emmanuel Tsochatzis
1456
Impact of Daclatasvir-Sofosbuvir Combination Treatment
on Medical Events and Costs in Patients Infected with
Genotype 3 Hepatitis C Virus
Boris Gorsh 1 , Bruce E. Sill 1 , Shalini Hede 1 , Catherine St-Laurent
Thibault 2 , Divya Moorjaney 3 , Michael Ganz 3 , Anupama
Kalsekar 1 , Yong Yuan 1 ; 1 Bristol-Myers Squibb, Plainsboro, NJ;
2 Evidera, Montreal, QC, Canada; 3 Evidera, Boston, MA
Purpose: Patients infected with genotype (GT) 3 of the hepatitis
C virus (HCV) fail antiviral therapy more often than patients
infected with other GTs. Combination daclatasvir-sofosbuvir
(DCV+SOF) is a significant advancement as it requires 12
rather than 24 weeks for other combination therapies. Our
objective is to estimate the number of medical events and
direct medical costs avoided if, otherwise untreated, GT3 HCV
patients were treated with DCV+SOF. Methods: We used a
published Markov model [i] to estimate the number of medical
events and costs experienced by a hypothetical cohort of
1,000 HCV GT3 patients (mean age of 54 years, 65% male)
who received DCV+SOF or no treatment. The patient distribution
we used was based on McGarry et al (2012). [ii] We
assumed that 89% of the DCV+SOF-treated patients would
achieve a sustained viralogic response based on results of the
ALLY-3 study. Medical events and costs were assessed for three
scenarios: overall HCV population (F0-F4) receives no treatment,
only the F3/F4 subpopulation receives DCV+SOF, and
all patients (F0-F4) receive DCV+SOF. Total medical events
were defined as the sum of compensated cirrhosis, decompensated
cirrhosis, hepatocellular carcinoma, liver transplant, or
liver-related death. Medical costs include monitoring costs and
costs for treating complications and adverse events. Disease
progression and costs were modelled over time horizons of 20
and 50 years. Results: Medical events and direct medical costs
were lowest when all patients were treated with DCV+SOF (76
events; $14,343,786). When only F3/F4 patients received
DCV+SOF, medical events and costs were substantially higher
(335 events; $68,414,213). In the scenario where no one
was treated, events and costs were the highest (729 events,
$129,838,658). Similar results were seen over 50 years.
Conclusion: Substantially fewer medical events and lower
medical costs are accrued when all GT3 patients are treated
with DCV+SOF compared with scenarios in which no one is
treated or treatment is limited to F3/F4 patients. [i] McEwan P,
Ward T, Yuan Y, et al. The impact of timing and prioritization
on the cost-effectiveness of birth cohort testing and treatment
for hepatitis C virus in the United States. Hepatology. 2013
Jul;58(1):54-64. [ii] McGarry LJ, et al. Economic model of a
birth cohort screening program for hepatitis C virus. Hepatology.
2012 May;55(5):1344-55.
Medical Events and Costs (Per Cohort)
Disclosures:
Boris Gorsh - Employment: Bristol-Myers Squibb; Stock Shareholder: Bristol-Myers
Squibb