Healthcare Cost and Utilization in Nonalcoholic Fatty Liver Disease: Real‐World Data From a Large U.S. Claims Database

2230

HEPATOLOGY, VOL. 68, NO. 6, 2018

Healthcare Cost and Utilization in

Nonalcoholic Fatty Liver Disease:

Real-World Data From a Large U.S.

Claims Database

Alina M. Allen,

Holly K. Van Houten,

2,3

Lindsey R. Sangaralingham,

2,3

Jayant A. Talwalkar,

1,2

and Rozalina G. McCoy

2,4,5

The prevalence of nonalcoholic fatty liver disease (NAFLD) is increasing. The health care burden resulting from the mul-

tidisciplinary management of this complex disease is unknown. We assessed the total health care cost and resource utiliza-

tion associated with a new NAFLD diagnosis, compared with controls with similar comorbidities. We used OptumLabs

Data Warehouse, a large national administrative claims database with longitudinal health data of over 100 million indi-

viduals enrolled in private and Medicare Advantage health plans. We identified 152,064 adults with a first claim for

NAFLD between 2010 and 2014, of which 108,420 were matched 1:1 by age, sex, metabolic comorbidities, length of fol-

low-up, year of diagnosis, race, geographic region, and insurance type to non-NAFLD contemporary controls from the

OptumLabs Data Warehouse database. Median follow-up time was 2.6 (range 1-6.5) years. The final study cohort con-

sisted of 216,840 people with median age 55 (range 18-86) years, 53% female, 78% white. The total annual cost of care per

NAFLD patient with private insurance was $7,804 (interquartile range [IQR] $3,068-$18,688) for a new diagnosis and

$3,789 (IQR $1,176-$10,539) for long-term management. These costs are significantly higher than the total annual costs

of $2,298 (IQR $681-$6,580) per matched control with similar metabolic comorbidities but without NAFLD. The largest

increases in health care utilization that may account for the increased costs in NAFLD compared with controls are repre-

sented by liver biopsies (relative risk [RR] = 55.00, 95% confidence interval [CI] 24.48-123.59), imaging (RR = 3.95, 95%

CI 3.77-4.15), and hospitalizations (RR = 1.87, 95% CI 1.73-2.02). Conclusion: The costs associated with the care for

NAFLD independent of its metabolic comorbidities are very high, especially at first diagnosis. Research efforts shouldfo-

cus on identification of underlying determinants of use, sources of excess cost, and development of cost-effective diagnostic

tests. (H 2018;68:2230-2238).

onalcoholic fatty liver disease (NAFLD) is

the most common chronic liver disease in

Western countries, affecting 24%

(1)

to 45%

(2)

of the U.S. population or 64-100 million people.

Most (approximately 80%) patients with NAFLD

have hepatic steatosis without inflammation, which

is associated with a relatively low risk of fibrosis,

(3,4)

but does have a strong correlation with cardiovascu-

lar disease, metabolic complications,

(5)

and increased

mortality compared with the general population.

(6)

The remaining 20% of patients have nonalcoholic

steatohepatitis (NASH), which can lead to cirrho-

sis, hepatocellular carcinoma, and other liver-related

complications.

(7)

Abbreviations: CI, confidence interval; ED, emergency department; ICD, International Classification of Diseases; IQR, interquartile range;

NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; OLDW, OptumLabs Data Warehouse; and RR, relative risk.

Received March 12, 2018; accepted May 8 2018.

Additional Supporting Information may be found at onlinelibrary.wiley.com/doi/10.1002/hep.30094/suppinfo.

A.M.A. is supported by the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, American College of

Gastroenterology Junior Faculty Development Award, and the National Institute of Diabetes and Digestive and Kidney Diseases of the National

Institutes of Health (K23 DK115594). R.G.M. is supported by the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health

Care Delivery and the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (K23 DK114497).

View this article online at wileyonlinelibrary.com.

DOI 10.1002/hep.30094

Potential conflict of interest: Nothing to report.

HEPATOLOGY, Vol. 68, No. 6, 2018 ALLEN ET AL.

2231

Given the increasing prevalence of NAFLD, the

economic burden is undoubtedly considerable, but real-

world data are scarce. U.S. health care expenditures have

steadily increased over the last decades and are pro-

jected to account for 20% of the economy by 2024.

(8,9)

The NAFLD epidemic wave could hasten this

increase; therefore, assessment of its contribution to

the economic burden and the major health care utiliza-

tion drivers is imperative. In a recent study, Younossi et

al. used Markov modeling to estimate the annual direct

healthcare costs at $1,612 per NAFLD patient.

(1)

However, as the authors acknowledged, the models

were constructed based on assumptions of NAFLD

epidemiology, fibrosis progression rate, and incident

complications, some of which were imputed from hep-

atitis C studies, resulting in uncertainty around many

inputs. Another study conducted among NAFLD

Medicare beneficiaries in 2010 estimated annual total

medical charges per patient to be $3,608 for outpa-

tient

(10)

and $36,289 for inpatient care.

(11)

Although

this provided direct cost data, it included an older pop-

ulation, whereas NAFLD is most prevalent in the mid-

dle age group (45-64 years).

Moreover, as the clinical care of NAFLD subjects is

directed not only by liver disease but also by the coex-

istent comorbidities, such as diabetes, hypertension or

cardiovascular disease, previous studies did not isolate

the specific contribution of NAFLD to the healthcare

burden from that of other metabolic diseases.

We therefore assessed the total health care cost and

use of patients with NAFLD, compared with a control

population with similar comorbidities, among commer-

cially insured and Medicare Advantage beneficiaries,

using a large administrative claims database. The data

set used in this study, OptumLabs Data Warehouse

(OLDW), is uniquely suited to study NAFLD burden

as it includes over 100 million people across the United

States, with greatest representation in the South, where

the prevalence of NAFLD is highest. It includes adults

of all ages, thereby updating and completing previously

published data that focused on Medicare beneficiaries.

The estimation of direct costs and utilization offers

better understanding about the financial implications

of NAFLD for patients and the health care system and

helps identify areas in need of better resource alloca-

tion, standardized management, and greater efficien-

cies in delivered care.

Methods

DATA SOURCE

This was a retrospective analysis of medical and

pharmacy claims data from the OLDW, a large national

administrative claims database that includes longitudinal

health data of more than 100 million individuals enrolled

in private and Medicare Advantage health plans since

1994,

(12,13)

which offers an excellent platform to trend

the cost of care

(14-16)

and private health insurance. The

population is diversely distributed in age, race, and geo-

graphical location in all 50 states. The database includes

deidentified enrollee information (sex, age, race/ethnic-

ity, region of residency, insurance plan), medical claims

(including diagnosis and procedure codes, site of service

codes, provider specialty codes, and total paid amounts),

and pharmacy claims. The study involved analysis of

preexisting de-identified data; thus, it was exempt from

Institutional Review Board approval.

STUDY POPULATION

We identified all patients with a first medical claim

for NAFLD using the International Classification of

Diseases (ICD) 9-CM 571.5 (cirrhosis of the liver

ARTICLE INFORMATION:

From the

Division of Gastroenterology and Hepatology, Department of Medicine, Mayo Clinic, Rochester, Minnesota;

Robert D.

and Patricia E. Kern Center for the Science of Health Care Delivery,Mayo Clinic, Rochester, Minnesota;

OptumLabs, Cambridge,

Massachusetts;

Division of Primary Care Internal Medicine, Department of Medicine,Mayo Clinic, Rochester, Minnesota;

Division of

Health Care Policy & Research, Department of Health Sciences Research,Mayo Clinic, Rochester, Minnesota.

ADDRESS CORRESPONDENCE AND REPRINT REQUESTS TO:

Alina M. Allen, M.D.

Division of Gastroenterology and Hepatology, Mayo Clinic

200 First Street SW

Rochester, MN 55905

E-mail: allen.alina@mayo.edu

Tel: +1-507-284 -3917

Fax: +1-507-284-0938

ALLEN ET AL. HEPATOLOGY, December 2018

2232

without mention of alcohol), 571.8 (other chronic non-

alcoholic liver disease), and 571.9 (unspecified chronic

liver disease without mention of alcohol) between

2010 and 2014. From this cohort, we excluded subjects

diagnosed with other liver diseases, including viral,

alcoholic, and cholestatic liver disease (ICD 9-CM

codes in Supporting Table S1). Subjects were clas-

sified as NAFLD cases if no alternative liver disease

was identified prior to the index NAFLD diagnosis or

during follow-up. This diagnostic algorithm correctly

identified true NAFLD cases with 85% accuracy in a

previously published retrospective population-based

cohort.

(5)

The service date of the first observed claim

for NAFLD was defined as the index date for patients

in the NAFLD cohort.

A control cohort was assembled by identifying

patients with at least one medical claim for an office

visit during 2010 to 2014 and no medical claims with

diagnosis codes for NAFLD or other liver diseases

during the study period. The controls were matched

1:1 on age, sex, race, diabetes mellitus, hypertension,

dyslipidemia, cardiovascular disease, length of fol-

low-up, year of diagnosis, geographic region, and insur-

ance type. The index date for the control cohort was

assigned to a randomly chosen office visit during the

identification period.

All subjects were continuously enrolled in the health

plan with medical and pharmacy benefits for at least 1

year before and 1 year after their index date. The sub-

jects were followed until disenrollment from the health

care plan or study end date ( June 2016). Supporting

Fig. S1 illustrates the study scheme.

COVARIATES AND OUTCOMES OF

INTEREST

Comorbidities associated with NAFLD including

diabetes mellitus, hypertension, dyslipidemia, and car-

diovascular disease were identified using the diagnos-

tic codes listed in Supporting Table S2. The NAFLD

subjects and controls were matched on these comor-

bidities at the index date to maximize the association

of cost and utilization with NAFLD and not with its

comorbidities. Outcomes of interest were direct costs

and health care use, such as office visits, hospitaliza-

tions, emergency department (ED) visits, as well as

tests and procedures attributable to liver disease: liver

biopsy, imaging (ultrasound, abdominal computed

tomography, and magnetic resonance imaging), and

laboratory tests (Supporting Tables S3 and S4). The

outcomes were measured at three different time points

in reference to the index date of NAFLD diagnosis or

matching: 1 year before, 1 and 5 years after.

STATISTICAL ANALYSIS

Patient characteristics (age, sex, race, census region,

year of diagnosis, comorbidities, and insurance type)

were described using mean (SD) or count (percentage)

as appropriate. Unadjusted utilization rates and total

costs of care were compared between NAFLD cases

and controls for 1 year prior to diagnosis date, 1 year

following diagnosis date, and 5 years following diagno-

sis date. Total cost of care included both medical (inpa-

tient and outpatient) claims and outpatient pharmacy

claims. Total health care costs were reported per patient

and were inflation-adjusted to 2015 U.S. dollars using

the Consumer Price Index.

(17)

Health care resource

use was identified as rates (number of events per 1000

patients) and rate ratios between 1 year post-index

date versus 1 year pre-index date and NAFLD versus

controls. Data were analyzed separately for privately

insured and Medicare Advantage subjects. Statistical

analyses were performed in SAS version 9.4 (SAS

Institute, Cary, NC).

Results

We identified 350,406 people with a first diagnosis

of NAFLD between 2010 and 2014, of which 165,281

were excluded for lack of medical and pharmacy coverage

at least 1 year prior to and 1 year after the index NAFLD

diagnosis. Additionally, 33,061 people were excluded

due to concurrent liver diseases other than NAFLD.

From the remaining cohort of 152,064 people with inci-

dent NAFLD, 108,420 were matched 1:1 by age, sex,

metabolic comorbidities, length of follow-up, year of

diagnosis, race, geographic region, and insurance type to

non-NAFLD contemporary controls from the OLDW

database. We were unable to match all NAFLD patients

to controls due to the multitude of matching variables.

The final study cohort consisted of 216,840 people with

median age 55 (range 18-86) years, 53% female and 78%

white (Table 1). Median follow-up time was 2.6 (range

1-6.5) years for both NAFLD and controls.

HEALTHCARE COSTS IN NAFLD

Figure 1 shows the annual total health care costs

of NAFLD subjects compared with matched con-

trols, in reference to the date of index (first) diagnosis

HEPATOLOGY, Vol. 68, No. 6, 2018 ALLEN ET AL.

2233

or matching, respectively. We show the total annual

costs starting 1 year prior to the index date, to allow

comparisons within the peridiagnosis period (1 year

before versus 1 year post), as well as long-term annual

costs,reflective of disease monitoring and management

of comorbidities. For both NAFLD subjects and con-

trols, the costs of care for Medicare Advantage enroll-

ees were considerably higher than for subjects with

private insurance.

The costs were highest during the first year follow-

ing the index NAFLD diagnosis, likely reflecting the

costs of diagnosis and initial evaluation for NAFLD

and its comorbidities. Specifically, among patients with

commercial insurance the median cost of medical care

during the year following NAFLD diagnosis increased

by 72%, from $4,547 (interquartile range [IQR]

$1,648-$11,661) to $7,804 (IQR $3,068-$18,688).

The median costs for Medicare Advantage enrollees

with NAFLD increased by 38%, from $6,566 (IQR

$3139-$14,787) during the year prior to NAFLD

diagnosis to $9,062 (IQR $4,313-$20,765) during the

year after diagnosis. For reference, the annual health-

care costs of non-NAFLD matched controls increased

only by 5% to 10% after the index date, in line with the

expected increase in annual rates.

The annual health care costs in the subsequent years

were lower than the immediate peri-diagnosis period.

Nevertheless, the annual costs for NAFLD patients

remained considerably higher than those for matched

controls. Specifically, at 5 years following NAFLD

diagnosis, the median annual health care cost was

$3,789 (IQR $1,176-$10,539) per NAFLD patient

with commercial insurance and $2,298 (IQR $681-

$6,580) per control. Among the Medicare Advantage

population, the median annual health care cost was

$5,363 (IQR $2,402-$12,515) per NAFLD patient

and $4,111 (IQR $1,677-$9,958) per control.

Consequently, the median cumulative healthcare

costs 5 years following the index NAFLD diagnosis

TABLE 1. Characteristics of NAFLD Patients and Matched

Controls

n = 108,420

NAFLD

n = 108,420

Age (years)

Median (IQR) 55 (45-65) 55 (45-65)

Age groups (years)

18-34 9341 (8.6%) 9341 (8.6%)

35-54 43,599 (40.2%) 43,599 (40.2%)

55-64 28,147 (26.0%) 28,147 (26.0%)

≥65 27,333 (25.2%) 27,333 (25.2%)

Gender

Female 57,167 (52.7 %) 57,167 (52.7 %)

Male 51,253 (47.3%) 51,253 (47.3%)

Index year

2010 19,663 (18.1%) 19,663 (18.1%)

2011 19,890 (18.3%) 19,890 (18.3%)

2012 22,538 (20.8%) 22,538 (20.8%)

2013 22,111 (20.4%) 22,111 ( 20.4%)

2014 24,218 (22.3%) 24,218 (22.3%)

Region

Midwest 27,230 (25.1%) 27,230 (25.1%)

Northeast 12, 8 26 (11.8%) 12,826 (11.8 %)

South 55,269 (51.0%) 55,269 (51.0%)

West 13,095 (12.1%) 13,095 (12.1%)

Race

White 84,613 (78.0%) 84,613 (78.0%)

Asian 2742 (2.5%) 2742 (2.5%)

Black 8508 (7.8%) 8508 (7.8%)

Hispanic 11,051 (10.2%) 11,051 (10.2%)

Unknown 1506 (1.4%) 1506 (1.4%)

Comorbidities

Hypertension 66,064 (60.9%) 66,064 (60.9%)

Hyperlipidemia 69,549 (64.1%) 69,549 (64.1%)

Cardiovascular disease 33,418 (30.8%) 33,418 (30.8%)

Diabetes mellitus 30,906 (28.5%) 30,906 (28.5%)

Insurance type

Commercial 76,697 (70.7%) 76,697 (70.7%)

Medicare Advantage 31,723 (29.3%) 31,723 (29.3%)

FIG. 1. Annual total health care costs of NAFLD patients

compared with matched controls in reference to the date of

index (first) NAFLD diagnosis or matching, respectively.

Abbreviations: NAFLD MA, NAFLD patients with Medicare

Advantage; NAFLD commercial, NAFLD patients with

commercial insurance; controls MA, matched controls with

Medicare Advantage; controls commercial, matched controls

with commercial insurance.

ALLEN ET AL. HEPATOLOGY, December 2018

2234

for an individual with commercial insurance were

nearly 80% higher than a control with similar age

and comorbidities: $30,994 (IQR $14,688- $64,972)

versus $17,345 (IQR $7,198-$38,713). The median

cumulative 5-year costs for a NAFLD individual with

Medicare Advantage were 42% higher than controls:

$39,588 (IQR $20,950-$71,226) versus $27,777 (IQR

$14,192-$54,666).

HEALTH CARE UTILIZATION IN

NAFLD

To explore what health care utilization parameters

may account for higher cost of care in NAFLD, we

assessed several utilization indices at similar time-

frames used for the cost estimates: peri-diagnosis and

at 5 years after the index diagnosis. In reference to the

year prior to the index date, most utilization parame-

ters during the following year increased slightly among

controls, as expected with the passage of time and aging,

but the rise was markedly higher among patients newly

diagnosed with NAFLD. The largest increase in use

(rate per 1000 patients) after NAFLD diagnosis was

liver biopsy from 5.5 to 28.8, followed by liver-related

imaging and all-cause hospitalizations. There were

smaller, but consistent, increases in laboratory testing

episodes, ED visits, and office visits. Figure 2A demon-

strates the relative change in utilization rates among

commercially insured patients with NAFLD when

compared with controls. Patients with NAFLD expe-

rienced substantial increases in use of imaging (relative

risk [RR] = 2.52, 95% confidence interval [CI] 2.49-

2.56), hospitalizations (RR = 1.69, 95% CI 1.64-1.75),

and laboratory tests (RR = 1.30, 95% CI 1.29-1.32)

when compared with controls, in whom the relative

increases were minimal. Among the most commonly

used imaging modalities, magnetic resonance imag-

ing showed the highest increase after diagnosis (RR =

3.42, 95% CI 3.20-3.66), followed by ultrasound (RR

= 2.77, 95% CI 2.71-2.82) and computed tomography

(RR = 2.57, 95% CI 2.52-2.62) (Supporting Table S5).

The trends were similar among the 63,442 subjects

with Medicare Advantage insurance, in whom the

largest increases in utilization after NAFLD diagnosis

were due to increased rates of liver biopsy, imaging, and

hospitalizations (Fig. 2B and Supporting Table S5).

Longitudinal follow-up data at 5 years following

NAFLD diagnosis/matching were available in a subset

of 20,840 individuals. The cumulative healthcare uti-

lization remained significantly higher among patients

with NAFLD compared with controls (Table 2).

Among commercially insured beneficiaries, liver biop-

sies continued to account for the largest difference in use

in NAFLD compared with controls (RR = 55.00, 95%

CI 24.48-123.59), followed by imaging (RR = 3.95,

95% CI 3.77-4.15) and hospitalizations (RR = 1.87,

95% CI 1.73-2.02). Among the Medicare Advantage

beneficiaries, the largest differences in health care use

with NAFLD were due to liver biopsies, imaging, and

ED visits.

The average cumulative rate of overall outpatient

office visits at 5 years following diagnosis was 40%

higher among patients with NAFLD compared with

controls: 31,079 versus 22,244 visits per 1000 patients.

Only 4.6% of these visits were to gastroenterology spe-

cialists, whereas 46% were to primary care (Fig. 3). The

proportion of other specialty visits, such as endocrinol-

ogy and cardiovascular diseases, was similar between

NAFLD and controls (2.4% and 3%, respectively),

reflective of robust matching by comorbidity status

FIG. 2. Relative change in utilization rates after a new diagnosis

of NAFLD compared with matched controls. (A) Commercial

insurance enrollees. (B) Medicare Advantage enrollees. The

bars represent utilization rate ratios (rates 1 year after diagnosis/

matching/rates 1 year prior to diagnosis/matching). The

corresponding absolute rates are presented below the bars.

HEPATOLOGY, Vol. 68, No. 6, 2018 ALLEN ET AL.

2235

during cohort selection. In the Medicare Advantage

cohort, gastroenterology visits represented 3.0% of all

visits among patients with NAFLD compared with

1.6% of all visits among controls.

Discussion

Using real-world data from a large nationwide

medical claims database, we show that the long-term

cumulative health care cost of a NAFLD patient is

80% higher than that of a non-NAFLD control of

similar age and metabolic comorbidities. The highest

annual costs occur around a new diagnosis of NAFLD,

reaching $7,804 and $9,062 per individual with pri-

vate insurance and Medicare Advantage, respectively.

Annual costs for long-term management decrease

to $3,789 and $5,363 per individual with private

insurance and Medicare Advantage, respectively, but

remain considerably higher than controls. The largest

increases in health care utilization, which may account

for the increased costs in NAFLD, are represented

by liver biopsies, imaging, and hospitalizations. The

large burden of NAFLD is managed predominantly

by primary care physicians, whereas subspecialty vis-

its in gastroenterology represent only 3 to 4.6% of

the total office visits. These data highlight that, as the

NAFLD burden will continue to increase,

(18)

solutions

are needed to promote innovative health care delivery

TABLE 2. Cumulative Use Rates per 1000 Patients at 5 Years After NAFLD Diagnosis

Controls NAFLD Rate Ratio (95% CI)

A. Commercial insurance

n 7464 7464

Liver biopsy 0.8 44.2 55.00 (24.48, 123.59)

Imaging 567.3 2243.2 3.95 (3.77, 4.15)

Ultrasound 144.0 762.7 5.30 (4.95, 5.66)

Computer tomography 287.5 1085.6 3.78 (3.54, 4.03)

Magnetic resonance imaging 19.0 116.0 6.10 (4.94, 7.53)

Transient elastography 0.0 0.4 —

Laboratory tests 8,517.0 12,380.5 1.45 (1.41, 1.49)

Hospitalizations 263.1 492.4 1.87 (1.73, 2.02)

Outpatient visits 22,243.7 31,078.8 1.40 (1.36, 1.43)

ED visits 328.5 402.7 1.23 (1.16, 1.30)

B. Medicare Advantage

n 2956 2956

Liver biopsy 1.0 25.0 24.67 (7.77, 78.34)

Imaging 1167.1 3297.0 2.82 (2.64, 3.02)

Ultrasound 182.3 879.2 4.82 (4.36, 5.33)

Computer tomography 651.6 1758.4 2.70 (2.49. 2.92)

Magnetic resonance imaging 36.2 126.5 3.50 (2.69, 4.55)

Transient elastography 0.0 0.0 -

Laboratory tests 14,740.2 17,169. 5 1.16 (1.10, 1.23)

Hospitalizations 1041.6 1489.5 1.43 (1.32, 1.54)

Outpatient visits 35,903.9 45,885.0 1.28 (1.23, 1.32)

ED visits 1724.0 2732.1 1.58 (1.48, 1.70)

FIG. 3. Average cumulative rate of overall outpatient office visits

5 years after diagnosis/matching and distribution by medical

specialties of interest.

ALLEN ET AL. HEPATOLOGY, December 2018

2236

platforms to reduce costs and to provide primary care

physicians with the necessary strategies and resources

to optimally manage this complex patient population.

The disease characteristics and the enormous clin-

ical burden of NAFLD pose considerable challenges

to the medical community, which extend beyond the

hepatology field. In this cohort, a strikingly low pro-

portion of the outpatient visits were represented by

gastroenterology and hepatology. The overwhelming

clinical burden of NAFLD is supported by general

practitioners, who have a key role in the identification,

risk stratification, and timely referral for specialty care

in NAFLD, butmay be unfamiliar with the intricacies

of the disease.

(19)

The American Association for the

Study of Liver Diseases guidelines suggest vigilance

for NAFLD, but do not provide well-defined screen-

ing recommendations for primary care providers and

cost-effective methods of disease severity assess-

ment.

(20)

The lack of clear guidelines is due to uncer-

tainties surrounding cost-effectiveness of diagnostic

tests and long-term benefits of screening, which are

areas in significant need of further research in the

hepatology community. The current state of NAFLD

diagnosis and disease severity assessment is based on

combinations of several available tests that include lab-

oratory studies, ultrasound, cross-sectional imaging,

elastography and liver biopsy, the use of which is sub-

ject to individual practice patterns.

Although the most cost-effective modality to esti-

mate disease severity in NAFLD remains to be estab-

lished,

(21)

these data offer a much-needed synopsis of

the real-world practice. The total costs soar by 72% in

the first year after the initial NAFLD claim and reach

exorbitant levels when compared with non-NAFLD

controls. Increases in utilization corresponding to

these costs were noted among all diagnostic modalities,

but were dominated by imaging tests (with costs that

vary between $200 and $3000

(22)

), which increased

2.5-fold (1215 per 1000 patients). It has been recog-

nized that using ultrasound to detect hepatic steatosis

is not cost-effective because clinically relevant fibro-

sis is present in no more than 11% of cases.

(23,24)

The

utility of other modalities, including elastography,

which is potentially more effective but costlier, has not

yet been proven. Although liver biopsy is required to

diagnose NASH, only patients at high risk require this

evaluation. In this cohort, liver biopsy, with a cost that

generally varies between $1500 and $3000

(22)

had the

highest relative increase in use (5-fold), although the

absolute rate of utilization remained low (29 per 1000

patients). The use of labs for diagnosis (the least expen-

sive but also least reliable alternative) increased by 30%.

The exorbitant costs of care around the first diagnosis

of NAFLD in this cohort underline the acute need of

more cost-effective methods of screening and disease

severity assessment.

The annual health care costs for NAFLD remained

extremely high beyond the initial peri-diagnosis period.

The long-term annual costs of NAFLD management

are almost double those of the matched cohort ($3789

versus $2298 per subject). Over the 5 years following

the index diagnosis, NAFLD patients are subjected

to abdominal imaging 4-fold more frequently than

matched controls. Similarly, the rate of blood testing

and outpatient visits is 45% and 39% higher, respec-

tively. The relative cost difference between NAFLD

and control patients was higher among the commer-

cially insured (younger) population than it was among

Medicare Advantage enrollees (in which cost is largely

driven by multimorbidity), suggesting that diagnosis

of NAFLD at earlier age in the context of increas-

ing NAFLD incidence in children and young adults

leads to a higher-cost differential at initial diagnosis.

Moreover, diagnosis at an earlier age leads to longer

follow-up time and monitoring for fibrosis progression

or surveillance for hepatocellular carcinoma. These

data highlight the need for cost-effective measures

to identify patients at high risk of disease progression

(i.e., differentiating patients with NASH and/or fibro-

sis from simple steatosis).

Although robust direct comparisons of long-term

costs in other liver diseases are not available, inferences

from hepatitis C models estimating $90,127 life-time

cost per patient treated with direct acting antiviral

agents

(25)

allow estimations that the cost of NAFLD

care is likely to surpass that of hepatitis C, especially in

light of upcoming NASH therapies.

These data are an essential benchmark for future

cost analyses in NAFLD, as several findings cover

important gaps in the existing literature. First, direct

costs are estimated from a large, nationally representa-

tive medical claims database. The annual direct cost per

NAFLD patient is approximately 5-fold higher than

previous estimates from the United States ($1612.18)

and European countries (€354-€1163) that relied on

Medicare data or derived from statistical modeling.

(1,11)

This is in part due to our ability to capture costs for

commercially insured adults who have heretofore been

excluded from NAFLD studies, despite being most of

the patients affected by the disease. Second, by using

HEPATOLOGY, Vol. 68, No. 6, 2018 ALLEN ET AL.

2237

a matched cohort with similar metabolic comorbidi-

ties as reference, we can differentiate liver-related costs

from those related to metabolic complications. Third,

we evaluate the costs at multiple time points and show

that the costs vary in reference to a new diagnosis.

Finally, we identify health care utilization in NAFLD

management, which is an important benchmark for

future cost-effectiveness analyses.

However, patients with Medicaid health coverage,

the uninsured, or those with NAFLD that remains

undiagnosed are not captured in OLDW; thus, prev-

alence estimates should not be extrapolated from this

study. Similarly, societal costs, derived from absentee-

ism and caregiver burden, certainly add even further

to the overall health care burden of NAFLD. As an

inherent limitation of large claims databases, we did

not have the opportunity to distinguish among clini-

cally appropriate and redundant use of tests, impact on

patient outcomes, and sources of excess costs. Further

work is needed to identify underlying determinants of

use, how to avoid high use of low-value services, and

insufficient use of high-value services that can drive

inefficient allocation of resources.

(26)

The care of NAFLD patients is expensive. As diag-

nostic methods and therapies for NAFLD become

increasingly available, early detection of the millions of

patients in the primary care setting, adequate risk strat-

ification, subspecialty referral and monitoring, while

taking into account cost-effectiveness, remains an

enormous challenge. Research efforts should focus on

development of high-value diagnostic tests to monitor

for liver fibrosis progression at appropriate intervals, in

a selected at-risk population, with the ultimate goal to

improve quality of care for the individual patient, while

being mindful of the effects on health care use.

REFERENCES

1) Younossi ZM, Blissett D, Blissett R, Henry L, Stepanova M,

Younossi Y, et al. The economic and clinical burden of non-

alcoholic fatty liver disease in the United States and Europe.

H 2016;64:1577-1586.

2) Williams CD, Stengel J, Asike MI, Torres DM, Shaw J,

Contreras M, et al. Prevalence of nonalcoholic fatty liver disease

and nonalcoholic steatohepatitis among a largely middle-aged

population utilizing ultrasound and liver biopsy: a prospective

study. Gastroenterology 2011;140:124-131.

3) Singh S, Allen AM, Wang Z, Prokop LJ, Murad MH, Loomba

R. Fibrosis progression in nonalcoholic fatty liver vs nonal-

coholic steatohepatitis: a systematic review and meta-analysis

of paired-biopsy studies. Clin Gastroenterol Hepatol 2015;13:

643-654.

4) Rinella ME. Nonalcoholic fatty liver disease: a systematic re-

view. JAMA 2015;313:2263-2273.

5) Allen AM, Terry TM, Larson JJ, Coward A, Somers VK,

Kamath PS. Nonalcoholic fatty liver disease incidence and im-

pact on metabolic burden and death: a 20 year-community study.

H 2018;67:1726-1736.

6) Adams LA, Lymp JF, St. Sauver J, Sanderson SO, Lindor KD,

Feldstein A, et al. The natural history of nonalcoholic fatty liver

disease: a population-based cohort study. Gastroenterology

2005;129:113-121.

7) Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu

YC, McCullough AJ. Nonalcoholic fatty liver disease: a spec-

trum of clinical and pathological severity. Gastroenterology

1999;116:1413-1419.

8) Fuchs VR. How and why US health care differs from that in

other OECD countries. JAMA 2013;309:33-34.

9) Hartman M, Martin AB, Espinosa N, Catlin A, The National

Health Expenditure Accounts Team. National health care spend-

ing in 2016: spending and enrollment growth slow after initial

coverage expansions. Health Aff (Millwood) 2018;37:150-160.

10) Younossi ZM, Zheng L, Stepanova M, Henry L, Venkatesan C,

Mishra A. Trends in outpatient resource utilizations and out-

comes for Medicare beneficiaries with nonalcoholic fatty liver

disease. J Clin Gastroenterol 2015;49:222-227.

11) Sayiner M, Otgonsuren M, Cable R, Younossi I, Afendy M,

Golabi P, et al. Variables associated with inpatient and outpa-

tient resource utilization among medicare beneficiaries with

nonalcoholic fatty liver disease with or without cirrhosis. J Clin

Gastroenterol 2017;51:254-260.

12) Optum. https://www.optumlabs.com/. Accessed February 6, 2018.

13) Wallace PJ, Shah ND, Dennen T, Bleicher PA, Crown WH.

Optum Labs: building a novel node in the learning health care

system. Health Aff (Millwood) 2014;33:1187-1194.

14) Thayer S, Bell C, McDonald CM. The direct cost of manag-

ing a rare disease: assessing medical and pharmacy costs asso-

ciated with Duchenne muscular dystrophy in the United States.

J Manag Care Spec Pharm 2017;23:633-641.

15) Brady BL, Tkacz J, Meyer R, Bolge SC, Ruetsch C. Assessment

of rheumatoid arthritis quality process measures and associated

costs. Popul Health Manag 2017;20:31-40.

16) Weaver J, Sajjan S, Lewiecki EM, Harris ST, Marvos P.

Prevalence and cost of subsequent fractures among U.S. pa-

tients with an incident fracture. J Manag Care Spec Pharm

2017;23:461-471.

17) Department of Labor Bureau of Labor Statistics. Consumer

Price Index: chained consumer price index for all urban consum-

ers (C-CPI-U). Washington, DC; 2017.

18) Estes C, Razavi H, Loomba R, Younossi Z, Sanyal AJ. Modeling

the epidemic of nonalcoholic fatty liver disease demonstrates

an exponential increase in burden of disease. H

2018;67:123-133.

19) Polanco-Briceno S, Glass D, Stuntz M, Caze A. Awareness of

nonalcoholic steatohepatitis and associated practice patterns

of primary care physicians and specialists. BMC Res Notes

2016;9:157.

20) Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella

M, et al. The diagnosis and management of nonalcoholic fatty

liver disease: practice guidance from the American Association

for the Study of Liver Diseases. H 2018;67:328-357.

21) Tapper EB, Hunink MG, Afdhal NH, Lai M, Sengupta N.

Cost-effectiveness analysis: risk stratification of nonalcoholic

fatty liver disease (NAFLD) by the primary care physician using

the NAFLD fibrosis score. PLoS One 2016;11:e0147237.

ALLEN ET AL. HEPATOLOGY, December 2018

2238

22) Costhelper. http://health.costhelper.com. Accessed January 9,

2018.

23) Wright AP, Desai AP, Bajpai S, King LY, Sahani DV, Corey

KE. Gaps in recognition and evaluation of incidentally identified

hepatic steatosis. Dig Dis Sci 2015;60:333-338.

24) Koehler EM, Plompen EP, Schouten JN, Hansen BE, Darwish

Murad S, Taimr P, et al. Presence of diabetes mellitus and ste-

atosis is associated with liver stiffness in a general population: the

Rotterdam study. H 2016;63:138-147.

25) Younossi ZM, Park H, Saab S, Ahmed A, Dieterich D, Gordon

SC. Cost-effectiveness of all-oral ledipasvir/sofosbuvir regimens

in patients with chronic hepatitis C virus genotype 1 infection.

Aliment Pharmacol Ther 2015;41:544-563.

26) Elshaug AG, Rosenthal MB, Lavis JN, Brownlee S, Schmidt H,

Nagpal S, et al. Levers for addressing medical underuse and over-

use: achieving high-value health care. Lancet 2017;390:191-202.

Supporting Information

Additional Supporting Information may be found at

onlinelibrary.wiley.com/doi/10.1002/hep.30094/suppinfo.