Malnutrition Diagnoses in Hospitalized Patients: United States, 2010

Abstract

Malnutrition is common among hospitalized patients in the United States, and its coded prevalence is increasing. Malnutrition is known to be associated with increased morbidity, mortality and healthcare costs. Although national data indicate that the number of malnutrition diagnoses among hospital discharges has been steadily rising, an in-depth examination of the demographic and clinical characteristics of these patients has not been conducted. We examined data from the 2010 Healthcare Cost and Utilization Project (HCUP), the most recent nationally-representative data describing U.S. hospital discharges. Using ICD-9 codes, we constructed a composite variable indicating a diagnosis of malnutrition. Based on our definition, 3.2% of all U.S. hospital discharges in 2010 had this diagnosis. Relative to patients without a malnutrition diagnosis, those with the diagnosis were older, had longer lengths of stay and incurred higher costs. These patients were more likely to have 27 of 29 comorbidities assessed in HCUP. Finally, discharge to home care was twice as common among malnourished patients, and a discharge of death was more than 5 times as common among patients with a malnutrition diagnosis. Taken together, these nationally representative, cross-sectional data indicate that hospitalized patients discharged with a diagnosis of malnutrition are older and sicker and their inpatient care is more expensive than their counterparts without this diagnosis.

Full text

Journal of Parenteral and Enteral
Nutrition
Volume 38 Number 2
February 2014 186 –195
© 2013 American Society
for Parenteral and Enteral Nutrition
DOI: 10.1177/0148607113512154
jpen.sagepub.com
hosted at
online.sagepub.com
Special Report
Introduction
Malnutrition is common in hospitalized patients in the United
States (U.S.), and it is associated with unfavorable outcomes
including higher infection rates, poor wound healing, longer
lengths of stay, and higher frequency of readmission. Not
unexpectedly, these outcomes are associated with increased
costs.
1-13
Although there are compelling historical data sug-
gesting a disproportionate frequency of poor outcomes among
hospitalized patients with malnutrition, this information is
derived from convenience samples or non-representative sam-
ples of hospitalized patients. Methodological issues such as
inadequate power, potentially biased patient populations,
inconsistent ascertainment of malnutrition, and the absence of
comprehensive clinical profiles suggest the need to more reli-
ably describe both the prevalence and clinical correlates of
hospitalized patients with a malnutrition diagnosis.
Since publication of The Skeleton in the Hospital Closet by
Charles Butterworth four decades ago,
14
there has been contin-
ued interest in understanding malnutrition among hospitalized
patients. Despite this consistent focus, rates of malnutrition in
hospitalized patients have changed little since seminal work by
Bistrian and Blackburn was published in the mid-1970s.
2,3
In
these early reports, the prevalence of protein-calorie malnutri-
tion was 44% in hospitalized patients in general medical wards
512154PEN
XXX10.1177/0148607113512154Journal of Parenteral and Enteral NutritionCor
kins et al
research-article
2013
From
1
University of Tennessee Health Science Center/Le Bonheur
Children’s Hospital, Memphis, Tennessee;
2
American Society for
Parenteral and Enteral Nutrition, Silver Spring, Maryland;
3
College
of Nursing and Health Professions, Drexel University, Philadelphia,
Pennsylvania;
4
Department of Nutritional Sciences, Pennsylvania State
University, University Park, Pennsylvania;
5
Mt. Carmel West Hospital,
Columbus, Ohio;
6
Skaggs College of Pharmacy, University of Montana,
Missoula, Montana;
7
Division of Trauma, Burn and Surgical Critical
Care, Brigham and Women’s Hospital, Boston, Massachusetts; and
8
Nationwide Children’s Hospital, Columbus, Ohio.
Financial disclosure: This research received no specific grant from any
funding agency in the public, commercial, or not-for-profit sectors but
was funded by A.S.P.E.N.
Received for publication September 29, 2013; accepted for publication
October 7, 2013.
This article originally appeared online on November 18, 2013.
Corresponding Author:
Peggi Guenter, PhD, RN, American Society for Parenteral and Enteral
Nutrition, Silver Spring, MD, USA.
Email: peggig@aspen.nutr.org
Malnutrition Diagnoses in Hospitalized Patients:
United States, 2010
Mark R. Corkins, MD, CNSC, FAAP
1
; Peggi Guenter, PhD, RN
2
;
Rose Ann DiMaria-Ghalili, PhD, RN, CNSC
3
; Gordon L. Jensen, MD, PhD
4
;
Ainsley Malone, MS, RD, CNSC
5
; Sarah Miller, PharmD, MS, BCNSP
6
;
Vihas Patel, MD, FACS, CNSC
7
; Steve Plogsted, PharmD, BCNSP, CNSC
8
;
Helaine E. Resnick, PhD, MPH
2
; and the American Society for Parenteral
and Enteral Nutrition
Abstract
Malnutrition is common among hospitalized patients in the United States, and its coded prevalence is increasing. Malnutrition is known
to be associated with increased morbidity, mortality and healthcare costs. Although national data indicate that the number of malnutrition
diagnoses among hospital discharges has been steadily rising, an in-depth examination of the demographic and clinical characteristics of
these patients has not been conducted. We examined data from the 2010 Healthcare Cost and Utilization Project (HCUP), the most recent
nationally-representative data describing U.S. hospital discharges. Using ICD-9 codes, we constructed a composite variable indicating a
diagnosis of malnutrition. Based on our definition, 3.2% of all U.S. hospital discharges in 2010 had this diagnosis. Relative to patients without
a malnutrition diagnosis, those with the diagnosis were older, had longer lengths of stay and incurred higher costs. These patients were more
likely to have 27 of 29 comorbidities assessed in HCUP. Finally, discharge to home care was twice as common among malnourished patients,
and a discharge of death was more than 5 times as common among patients with a malnutrition diagnosis. Taken together, these nationally
representative, cross-sectional data indicate that hospitalized patients discharged with a diagnosis of malnutrition are older and sicker and
their inpatient care is more expensive than their counterparts without this diagnosis. (JPEN J Parenter Enteral Nutr. 2014;38:186-195)
Keywords
nutrition assessment; nutrition; outcomes research/quality; nutrition support practice; public policy
Download a QR code reader on your smartphone,
scan this image, and listen to the podcast for this
article instantly. Or listen to this and other JPEN
podcasts at http://pen.sagepub.com/site/misc/
Index/Podcasts.xhtml.
by guest on October 13, 2015pen.sagepub.comDownloaded from

Corkins et al 187
and more than 50% in general surgical wards.
2,3
A more recent
study using a standard screening and assessment protocol
focused on malnutrition in hospitalized patients and suggested
that 33% were malnourished. The same report showed that
51%-54% of hospitalized patients were characterized as mal-
nourished using other methods.
15
In a 1999 report, Sullivan
found that 21% of elderly hospitalized patients took in less
than half of their nutritional requirements.
16
Another study of
400 hospitalized patients found that 53% had malnutrition,
1
while a study in tertiary care children’s hospitals suggested that
24% of children were malnourished.
17
Numerous studies both
in the U.S. and abroad offer a heterogeneous picture of the
prevalence of malnutrition in hospitalized patients.
1-9,18-21
Differences in patient populations, ascertainment methods,
malnutrition definitions, healthcare settings and other factors
may explain some of this heterogeneity. Thus, these findings
cannot be compared across studies, nor can they be used as part
of national public health or advocacy efforts in the U.S. Indeed,
the true prevalence of malnutrition in hospitalized patients is
unknown.
The objectives of the present study are to define the preva-
lence of malnutrition diagnoses among patients discharged
from U.S. hospitals and to compare and contrast demographic
and clinical characteristics of these patients with their counter-
parts who do not have a diagnosis of malnutrition.
Methods
Data in this report are from the Healthcare Cost and Utilization
Project (HCUP), a family of databases sponsored by the
Agency for Healthcare Research and Quality (AHRQ). HCUP
databases include the National Inpatient Sample (NIS), which
contains patient-level data on hospital inpatient stays. Several
studies have been conducted using these data to examine nutri-
tion status in populations with gastrointestinal disease.
22-24
The
2010 NIS contains discharge data from 1,051 hospitals in 45
states, approximating a 20% stratified sample of U.S. hospi-
tals. When weighted appropriately, estimates from HCUP rep-
resent all U.S. hospitalizations. The 2010 NIS contains up to
twenty-five ICD-9 diagnosis codes for each patient. Using
these codes, we identified all patients who were diagnosed
with malnutrition during their hospital stay. This current analy-
sis aims to capture under-nutrition, and therefore excluded
obesity as an indicator of malnutrition. Individual nutrient
deficiency diagnostic codes were also excluded. However,
obese individuals were coded as malnourished if their records
contained another qualifying diagnosis of malnutrition. Table 1
shows the ICD-9 codes that were used to construct the dichoto-
mous malnutrition variable used in this report.
Patients with a diagnosis of malnutrition were contrasted to
those without this diagnosis across a number of demographic
and clinical factors, as well as factors associated with the hos-
pitals where they were treated. In addition to basic demo-
graphic information such as age and race, additional variables
describing income (estimated from the median household
income quartiles for the patient’s zip code), the rural/urban sta-
tus of patient’s residence as well as the expected source of pay-
ment were examined. Rural/urban status is defined by rural
urban commuting areas (RUCA) which are assigned by zip
codes using population and commuting information from the
U.S. Census.
25
Hospital characteristics were also examined
according to presence or absence of a malnutrition diagnosis.
These variables included the size of the hospital, ownership,
rural/urban status, in what region of the country the hospital
was located, and whether it was a teaching hospital or part of a
multi-hospital system. In addition, the most common principal
diagnoses among people with and without a malnutrition diag-
nosis were examined, as well as the prevalence of major oper-
ating room procedures that occurred during the hospital stay.
Administrative variables such as length of stay (LOS), cost,
and the circumstances of both admission and discharge were
examined in relation to presence or absence of a malnutrition
diagnosis. Costs were calculated using the charges as provided
in the HCUP data and converted to costs using hospital-spe-
cific cost-to-charge ratios.
To understand the burden of comorbidity among people
with and without a diagnosis of malnutrition, a series of 29
conditions was examined in relation to the dichotomous mal-
nutrition variable. These 29 conditions were created based
on the presence of secondary diagnoses using HCUP’s
Comorbidity Software, Version 3.7.
26
Finally, the frequency
of receipt of parenteral nutrition (PN) and enteral nutrition
(EN) therapies was examined. Procedure codes 96.6 and
99.15 were used to identify patients who received EN and
PN, respectively, and these variables were cross-tabulated
with the malnutrition diagnosis variable. Several variables
were created for these analyses: two variables indicating
patients who received either therapy and a variable indicat-
ing patients who had received both therapies during their
hospital stay.
Comparisons were conducted using SAS survey statistical
procedures (SAS Institute Inc., Cary, NC: www.sas.com).
Table 1. ICD-9-CM Malnutrition Diagnosis Codes Used to
Identify Malnutrition.
260 Kwashiorkor
261 Nutritional marasmus
262 Other severe protein-calorie malnutrition
263.0 Malnutrition of moderate degree
263.1 Malnutrition of mild degree
263.2 Arrested development following protein-calorie malnutrition
263.8 Other protein-calorie malnutrition
263.9 Unspecified protein-calorie Malnutrition
579.3 Other and unspecified postsurgical nonabsorption
764.12-764.19 Light-for-date with signs of fetal malnutrition ( 500-2500g+)
764.20 Fetal malnutrition without mention of “light-for-dates”
unspecified weight
764.21-764.29 Fetal malnutrition without mention of “light-for-dates”
(<500g-2500g +)
764.90 Fetal growth retardation, unspecified
764.91-764.99 Fetal growth retardation (<500-2500g+)
995.52 Child neglect (nutritional)
995.84 Adult Neglect (nutritional)
by guest on October 13, 2015pen.sagepub.comDownloaded from

188 Journal of Parenteral and Enteral Nutrition 38(2)
These analyses accounted for the clustered design (multiple
patients from the same hospital) that characterizes the NIS.
Additional information on HCUP and the NIS are available.
27
For this report, reliability was assessed using guidelines from
the 1996 NHANES Analytic and Reporting Guidelines.
28
Estimates that do not meet reliability criteria are denoted with
an asterisk and should be interpreted with caution. An estimate
and its standard error were deemed to be reliable if they met
the following two criteria: for means and proportions based on
commonly occurring events (where 0.25 < P < 0.75), the rela-
tive standard error (standard error/estimate) must be less than
or equal to 30%; for proportions that are based on uncommon
or very common events (P ≤ 0.25 or P ≥ 0.75), the denominator
(n) must be sufficiently large that the minimum of nP and n(1-
P) is equal to or greater than 8 times the design effect.
Results
Time Trends in Malnutrition Diagnosis
Codes
HCUP data for the 18-year period between 1993 and 2010
offer a context for the current analyses. Figure 1 shows malnu-
trition diagnoses over this time period as a percentage of all
hospital discharges. While patients diagnosed with malnutri-
tion remained a modest proportion of the total number of hos-
pital discharges, there was an almost three-fold increase in the
proportion of patients with a malnutrition diagnosis during this
period. The most frequently observed malnutrition code among
hospitalized patients in 2010 was 263.9, Protein-Calorie
Malnutrition Not Otherwise Specified. This code was present
in 752,303 discharges, a figure that accounts for approximately
53% of all discharges with a malnutrition diagnosis.
Patient Demographics
In the United States in 2010, a total of 1,248,680 hospital dis-
charges (3.2%) had a malnutrition diagnosis. Patients with a
malnutrition diagnosis were significantly older than their
counterparts without this diagnosis (64.8 versus 47.8 years,
P < .0001; Table 2). The distribution of race/ethnicity was sig-
nificantly different between discharges with and without a
malnutrition diagnosis (P < .0001). Non-Hispanic whites were
predominant in both patient groups with slightly more in the
malnourished group (69.8 and 65%). Females comprised more
than half of both patient groups, although significantly fewer
women had a malnutrition diagnosis (53.1 versus 57.8%, P <
.0001). The distribution of income quartile categories differed
significantly between discharges with and without a malnutri-
tion diagnosis, (P = .0004), with more patients with a malnutri-
tion diagnosis falling below the 50th percentile of income
(57.9 vs 55.0%). The two groups did not differ with regard to
their residence in urban or rural areas. Consistent with observed
differences in age, the primary source of expected primary
payer differed between patients with and without a malnutri-
tion diagnosis (P < .0001), with Medicare being the primary
payer among those with the diagnosis.
Hospital Characteristics
The prevalence of malnutrition diagnoses did not differ by the
number of hospital beds in a given institution, whether the
hospital was located in a rural location or whether it was a
teaching hospital (Table 3). However, the prevalence of
patients with these diagnoses differed significantly across cat-
egories of hospital ownership (P = .002), with malnourished
patients being discharged more often from private for-profit
hospitals. Malnutrition diagnoses also differed significantly
by region (P = .0003), with more diagnoses occurring in
Midwestern and Southern hospitals. Patients were also more
likely to have a malnutrition diagnosis if they were discharged
from a facility that was part of a multi-hospital system (71.4
versus 66.5%, P = .005)
Admission and Discharge Characteristics
Patients with a malnutrition diagnosis had a significantly longer
length of stay (LOS) (12.6+.5 versus 4.4+.1 days, P < .0001)
and higher costs ($26,944 versus $9,485, P < .0001); see
Table 4. The circumstances of admission differed significantly
by whether patients had a diagnosis of malnutrition or not (P <
.0001). For example, while nearly 80% of patients with a mal-
nutrition diagnosis were admitted to the hospital emergently or
urgently, only 65.0% of those without this diagnosis were
admitted under these circumstances. Not surprisingly, discharge
disposition categories also differed significantly according to
whether patients had a malnutrition diagnosis. Only 28.8% of
patients with a malnutrition diagnosis had a routine discharge,
compared to 72.6% of routine discharges among patients with-
out a malnutrition diagnosis. Further, discharge to home care
was twice as common among malnourished patients, and a dis-
charge of death was more than 5 times as common among
patients with a malnutrition diagnosis.
0
0.5
1
1.5
2
2.5
3
3.5
% diagnosed
Figure 1. Percentage of hospital discharges with malnutrition
diagnoses, by year, United States.
by guest on October 13, 2015pen.sagepub.comDownloaded from

Corkins et al 189
Comorbid Conditions, Principal Diagnosis,
Surgeries and Diagnostic Code Ranking
Of the 29 comorbid conditions shown in Table 5, 27 occurred
significantly more frequently among patients with a malnutri-
tion diagnosis. Chief among these were weight loss (91.7%
versus 0.6%), and fluid and electrolyte disorders (52% versus
16.9%). The five most common diagnoses among hospitalized
patients with a diagnosis of malnutrition were septicemia,
pneumonia, acute renal failure, aspiration pneumonitis and
interstitial emphysema (Table 6). Patients with a malnutrition
diagnosis underwent fewer major operations during their hos-
pitalizations (24.1 versus 29.3%, P = .0001) but they had
almost twice as many diagnostic codes listed on their records
Table 2. Demographic Characteristics of Discharged Patients With and Without a Diagnosis of Malnutrition, United States, 2010.
Malnutrition Diagnosis No Malnutrition Diagnosis
Characteristic Estimate 95% CI Estimate 95% CI P Value
Total NA
Weighted N 1,248,680 37,759,618
Percentage 3.2 96.8
Mean age 64.8 64.0-65.7 47.8 47.1-48.5 <.0001
Age (%)
<1 yr 2.8 2.4-3.2 11.8 11.2-12.3 <.0001
1-17 1.5 1.0-2.0 4.6 3.9-5.3
18-44 9.8 9.1-10.4 25.4 24.8-26.0
45-64 27.7 26.9-28.4 24.9 24.3-25.5
65-84 41.9 41.0-42.8 25.6 24.9-26.2
85+ 16.4 15.6-17.2 7.7 7.4-8.0
Race/ethnicity (%)
White 69.8 67.5-72.2 65.0 62.567.6 <.0001
Black 17.2 15.3-19.1 16.1 14.4-17.7
Hispanic 7.2 5.9-8.4 12.4 10.7-14.0
Asian/Pacific Islander 2.3 1.8-2.8 2.7 2.1-3.2
Native American 0.6 0.4-0.8 0.8 0.5-1.1
Other 2.9 2.2-3.7 3.1 2.5-3.7
Female (%) 53.1 52.5-53.6 57.8 57.4-58.2 <.0001
Income quartile (%)
$1-40,999 31.5 29.4-33.7 29.6 27.5-31.8 .0004
$41,000-50,999 26.4 24.7-28.0 25.4 24.0-26.9
$51,000-66,999 23.6 22.1-25.0 24.0 22.7-25.4
$67,000+ 18.5 16.4-20.6 20.9 18.5-23.3
Location (%)
Central counties of metropolitan area
≥1 million
28.3 24.4-32.2 28.4 24.7-32.2 .1128
Fringe counties of metropolitan areas of
≥1 million
22.5 19.3-25.6 24.4 20.9-27.8
Counties in metropolitan areas of
250,000 - 999,999
20.7 16.9-24.5 19.7 16.2-23.2
Counties in metropolitan areas of
50,000 - 250,000
8.7 6.8-10.6 8.9 6.9-10.8
Micropolitan counties
a
12.5 10.9-14.1 11.2 10.1-12.3
Not metropolitan or micropolitan 7.3 6.3-8.3 7.5 6.7-8.3
Expected primary payer (%)
Medicare 62.8 61.0-64.5 36.5 35.5-37.6 <.0001
Medicaid 12.5 11.6-13.5 21.5 20.4-22.7
Private, including HMO 18.9 17.9-20.0 32.4 31.1-33.7
Self-pay 3.2 2.8-3.7 5.6 4.9-6.2
No charge 0.3 0.2-0.4 0.5 0.3-0.8
Other 2.2 1.9-2.6 3.4 3.0-3.8
a
A nonmetropolitan county with an area of 10,000 or more population
by guest on October 13, 2015pen.sagepub.comDownloaded from

190 Journal of Parenteral and Enteral Nutrition 38(2)
than their counterparts without a malnutrition diagnosis (14.8
codes versus 7.9, P < .001). Although only 1.1% of all patients
with a malnutrition diagnosis had this malnutrition code listed
first in their records, more than one-fifth (22.3%) of these
patients had it listed as the second diagnosis (Figure 2).
Parenteral and Enteral Therapy
Among patients with a malnutrition diagnosis, 13.4% received
either PN or EN during their hospital stay (Table 7). A majority
of these patients received PN (8.9%), while a smaller proportion
Table 3. Hospital Characteristics of Discharged Patients With and Without a Diagnosis of Malnutrition, United States, 2010.
Malnutrition Diagnosis No Malnutrition Diagnosis
Characteristic Estimate 95% CI Estimate 95% CI P Value
Bed size (%)
a
Small 13.9 12.1-15.7 11.9 11.0-12.7 .0724
Medium 24.0 21.3-26.6 23.6 22.4-24.8
Large 62.1 59.0-65.2 64.5 63.0-66.1
Ownership (%)
Government/nonfederal 12.4 9.0-15.8 14.3 11.1-17.6 .0002
Private, nonprofit 70.5 66.4-74.7 72.8 69.2-76.3
Private, for-profit 17.0 14.4-19.7 12.9 11.2-14.6
Rural location (%) 11.6 9.7-13.6 12.2 11.1-13.3 .5028
Region (%)
Northeast 15.0 12.7-17.3 19.6 18.1-21.0 .0003
Midwest 25.3 21.9-28.6 22.6 21.1-24.0
South 41.9 38.6-45.3 38.3 36.5-40.1
West 17.8 15.4-20.2 19.5 18.2-20.9
Teaching (%) 48.4 44.8-52.0 48.2 46.3-50.1 .9029
Member multihospital system(%) 71.4 66.1-76.8 66.5 61.6-71.4 .0050
Source: 2010 Healthcare Cost and Utilization Project.
a
Bed size categories are based on hospital beds, and are specific to the hospital’s location and teaching status. Bed size assesses the number of short-term
acute beds in a hospital. Hospital information was obtained from the AHA Annual Survey of Hospitals.
Table 4. Admission and Discharge Characteristics of Discharged Patients With and Without a Diagnosis of Malnutrition, United States, 2010.
Malnutrition Diagnosis No Malnutrition Diagnosis
Characteristic Estimate 95% CI Estimate 95% CI P Value
Length of stay (mean days) 12.6 12.1-13.1 4.4 4.3-4.5 <.0001
Total costs (mean $) 26,944 25,355-28,533 9,485 9,144-9,826 <.0001
Admission type (%)
Emergency 59.9 57.1-62.7 46.2 44.3-48.0 <.0001
Urgent 19.6 17.2-22.0 18.8 17.0-20.5
Elective 17.4 15.3-19.5 24.5 23.3-25.7
Newborn 2.2 1.9-2.5 9.9 9.4-10.5
Trauma center 0.9 0.6-1.2 0.6 0.5-0.8
Other 0.0* 0.0-0.0 0.0* 0.0-0.0
Discharge disposition (%)
Routine 28.8 27.7-29.9 72.6 71.7-73.4 <.0001
Transfer to short-term hospital 3.6 3.3-3.9 2.1 2.0-2.2
Other transfers 38.1 37.2-39.0 12.2 11.8-12.6
Home health care 19.8 19.0-20.6 10.4 9.8-11.0
Against medical advice 0.6 0.5-0.7 1.0 0.9-1.1
Died 8.8 8.5-9.1 1.7 1.6-1.7
Discharged alive, destination unknown 0.3* 0.1-0.5 0.0 0.0-0.1
Source: 2010 Healthcare Cost and Utilization Project.
*Asterisk indicates that estimates failed reliability test
by guest on October 13, 2015pen.sagepub.comDownloaded from

Corkins et al 191
received EN (5.2%). Only 0.7% of patients with a malnutrition
diagnosis had procedure codes indicating receipt of both thera-
pies during their hospital stay. While parenteral and enteral
nutrition therapy was administered primarily to those patients
with a diagnosis of malnutrition, it is notable that 1.2% of
patients without this diagnosis also received one or both of these
therapies.
Discussion
This report presents the first nationally-representative profile
of hospitalized patients with a coded diagnosis of malnutrition.
Our data indicate that, in 2010, 3.2% of patients who were dis-
charged from hospitals had a diagnosis of malnutrition.
Although relatively modest in absolute terms, the proportion of
patients with a diagnosis code for malnutrition has tripled since
1993, suggesting that malnutrition in hospitalized patients may
be on the rise. This increase may be due to a number of factors,
including aging of the population, increased patient acuity,
improved awareness on the part of clinicians regarding identi-
fication of this condition and an appreciation of its signifi-
cance. As well, there may be a greater emphasis on the part of
hospitals for coding malnutrition. Regardless of the reasons for
the increase in the occurrence of patients with this diagnosis,
the trend reflecting growth in the number of hospitalized
patients with a malnutrition diagnosis is unequivocal.
In addition to the increase in malnutrition diagnoses in hos-
pitalized patients, the data presented in this report suggest that
patients with this diagnosis are vulnerable. A clear finding in
this report is that a diagnosis of malnutrition is particularly
Table 5. Comorbid Conditions of Discharged Patients With and Without a Diagnosis of Malnutrition, United States, 2010.
Malnutrition Diagnosis No Malnutrition Diagnosis
Comorbid Condition (%) Estimate 95% CI Estimate 95% CI P Value
Weight loss 91.7 91.1-92.3 0.6 0.5-0.6 <.0001
a
Fluid and electrolyte disorder 52.0 50.8-53.2 16.9 16.5-17.3 <.0001
a
Hypertension 47.0 45.7-48.3 39.3 38.5-40.1 <.0001
Deficiency anemia 34.9 33.4-36.3 12.9 12.5-13.3 <.0001
Chronic pulmonary disease 23.3 22.4-24.1 14.5 14.1-14.9 <.0001
Renal failure 18.2 17.5-19.0 8.5 8.3-8.8 <.0001
Diabetes without chronic
complications
17.2 16.5-17.9 15.2 14.8-15.5 <.0001
Congestive heart failure 16.2 15.5-16.8 6.2 6.0-6.4 <.0001
Other neurological disorders 12.5 12.1-13.0 5.9 5.7-6.0 <.0001
Depression 11.7 11.2-12.2 8.4 8.0-8.7 <.0001
Hypothyroidism 11.7 11.2-12.2 8.5 8.3-8.8 <.0001
Coagulopathy 11.3 10.6-12.0 3.3 3.1-3.4 <.0001
Peripheral vascular disease 8.0 7.6-8.5 4.2 4.0-4.4 <.0001
Alcohol abuse 6.5 6.1-6.8 3.7 3.5-3.8 <.0001
Metastatic cancer 6.5 6.2-6.9 1.7 1.4-1.9 <.0001
Paralysis 6.0 5.7-6.4 2.0 1.9-2.1 <.0001
Obesity 5.9 5.5-6.4 8.1 7.8-8.4 <.0001
Diabetes with chronic
complications
5.7 5.3-6.0 3.2 3.1-3.4 <.0001
Psychosis 5.5 5.2-5.8 3.7 3.5-3.8 <.0001
Liver disease 5.2 4.8-5.6 2.1 2.0-2.2 <.0001
Solid tumor without metastasis 4.5 4.3-4.6 1.5 1.5-1.6 <.0001
Valvular disease 4.5 4.2-4.8 2.6 2.5-2.7 <.0001
Pulmonary circulation disease 4.2 4.0-4.5 1.4 1.4-1.5 <.0001
Rheumatoid arthritis 3.2 3.0-3.3 2.0 1.9-2.1 <.0001
Drug abuse 2.9 2.7-3.2 3.4 3.1-3.6 .0012
Chronic blood loss anemia 2.2 2.1-2.3 2.0 1.9-2.1 .0478
Lymphoma 1.5 1.5-1.6 0.6 0.6-0.6 <.0001
Acquired immune deficiency
syndrome
0.4 0.4-0.5 0.2 0.2-0.3 <.0001
Peptic ulcer disease 0.1 0.1-0.1 0.0 0.0-0.0 <.0001
Source: 2010 Healthcare Cost and Utilization Project.
a
Using logistic regression.
by guest on October 13, 2015pen.sagepub.comDownloaded from

192 Journal of Parenteral and Enteral Nutrition 38(2)
common in older adults: 58.3% of hospitalized patients with a
malnutrition diagnosis in our nationally-representative sample
were over the age of 65 years, whereas only 33.3% of patients
without this diagnosis were in the same age group. Although
the aging of the population in the United States and other
developed countries is well-known, the finding that a diagnosis
of malnutrition occurs more frequently in hospitalized older
adults suggests that, as the healthcare system continues to
adapt to the needs of an aging society, added emphasis may
need to be placed on identification and management of older
malnourished patients. Our data show that both lengths of stay
and costs were nearly 3 times higher among patients with a
malnutrition diagnosis, underscoring the intense resource utili-
zation associated with treating this patient population. The fact
that Medicare was the major payer for these patients further
highlights the public policy implications of our findings.
Our data extend the previous research concerning the asso-
ciation between malnutrition and infection. It has been estab-
lished that malnutrition elevates infection risk, a relationship
that is of particular concern given the advanced age of this
patient population.
6-9,12,16,17
Our data support the previous
findings and advance the existing knowledge by showing that
infections (eg, septicemia, pneumonia and pneumonitis) were
among the top 5 diagnoses for hospitalized patients with a
malnutrition diagnosis. Our data analysis also demonstrates
that, relative to those without a malnutrition diagnosis, a larger
proportion of patients with a coded diagnosis of malnutrition
were admitted emergently, and had non-routine discharges,
Table 6. Principal Diagnoses, Operating Room Procedures, and Number of Listed Diagnoses Among Discharged Patients With and
Without a Diagnosis of Malnutrition, United States, 2010.
Characteristic Malnutrition Diagnosis No Malnutrition Diagnosis P Value
Most common principal
diagnoses
1. Septicemia, unspecified (0389) 1. Single liveborn, without
cesarean (V3000)
2. Pneumonia (486) 2. Single liveborn, with
cesarean (V3001)
3. Acute renal failure, unspecified
(5849)
3. Pneumonia (486)
4. Pneumonitis due to imbalance of
food or vomitus (5070)
4. Chronic atherosclerosis
of native coronary artery
(41401)
5. Interstitial emphysema (5181) 5. Septicemia, unspecified
(0389)
Estimate 95% CI Estimate 95% CI
Major operating room
procedure (%)
24.1 22.9-25.3 29.3 28.6-29.9 <.0001
Number of diagnostic codes 14.8 14.4-15.2 7.88 7.73-8.02 <.0001
1.1
22.3
16.7
14.2
10.7
8.2
6.3
5.5
3.7
2.8
1.8
2.0
1.9
1.1
1.5
0.5
0.4
0.3
0.2
0.1
0.1
0.1
0.1
0.00.0
0
5
10
15
20
25
12345678910111213141516171819202122232425
Posion of malnutrion diagnosis on diagnosc code list
%
Figure 2. Position of malnutrition diagnosis codes on the diagnostic code list among hospital discharges with a malnutrition diagnosis,
United States, 2010.
by guest on October 13, 2015pen.sagepub.comDownloaded from

Corkins et al 193
including discharges to home care. Taken together, these
results suggest an extremely frail population of post-acute
patients that is at high risk of additional healthcare resource
utilization.
The data in this report show that although hospital size and
urban/rural location did not impact the likelihood of having a
malnutrition diagnosis, certain other facility-level characteris-
tics did. Private for-profit hospitals, hospitals that are members
of multi-hospital systems, and those in the Midwest and
Southern United States were all more likely to have patients
with a discharge diagnosis of malnutrition, although in some
cases, the absolute differences were small. Nonetheless, these
findings raise questions about practice patterns, whether proto-
cols for coding malnutrition diagnoses are better developed in
private for-profit hospitals, or whether hospitals that are cen-
trally managed by a hospital system may have more effective
protocols in place for identification and coding of these
patients. The lack of standardization in ascertaining malnutri-
tion in hospitalized patients arises from a number of factors
including the diversity of tools used to identify this condition,
the absence of a formal definition of malnutrition, and limited
education and training of practitioners. Regional differences in
the likelihood of having a malnutrition diagnosis may be asso-
ciated with larger demographic patterns. For instance, greater
numbers of older adults live in the South, thus increasing the
likelihood that they will be hospitalized in southern states.
29
Our data show that 13.4% of patients with a malnutrition
diagnosis received nutrition support therapy during their hos-
pital stay. Although it is encouraging to note that these patients
received therapy, the question of whether EN or PN had a
favorable impact on patient outcomes remains to be clarified.
It is useful to note that although many malnourished patients
received EN or PN therapy, others may not require it. Future
research using HCUP could address questions related to why
some patients with a malnutrition diagnosis receive nutrition
support therapy and others do not. An intriguing finding related
to these therapies is that 1.2% of patients without a coded mal-
nutrition diagnosis received either PN or EN during their hos-
pital stay. This translates into provision of nutrition support
therapy to many thousands of patients, and raises questions
related to assessment and coding of malnutrition as well as the
proactive use of these therapies to prevent malnutrition in at
risk patients in the absence of a malnutrition diagnosis.
Although it is generally accepted that malnutrition is under-
diagnosed in the acute care setting, weight loss and fluid and
electrolyte disorders were the top comorbid conditions in the
malnourished group, and these conditions were much less
common among those not coded for malnutrition (91.7versus
0.6%, and 52 versus 16.9% respectively).
Despite the fact that HCUP provides a nationally-represen-
tative snapshot of malnutrition diagnoses in hospitalized
patients in the United States, the data have a number of impor-
tant limitations. By definition, this data set is cross-sectional, a
design feature that precludes analysis of research questions
related to cause and effect. Thus, the data highlighted in this
report should be interpreted as associations rather than causal
relationships. A second limitation of these data relates to how
malnutrition cases are ascertained. Our case definition is based
on ICD-9 codes that appear in hospital discharge records. As a
result, our ability to identify a case of malnutrition is only as
good as the malnutrition coding practices in the sampled hos-
pitals. Given the well-known absence of a standardized defini-
tion of malnutrition as well as research suggesting that
malnutrition is under-diagnosed even using existing tools and
definitions, it is highly likely that some malnourished patients
were not captured by ICD-9 coding and were therefore catego-
rized as being non-malnourished. This issue is of particular
importance because of the difficulties associated with accu-
rately pinpointing this condition with valid and reliable diag-
nostic characteristics and markers.
30-32
It is also likely that
some patients coded as malnourished may not have been iden-
tified as such for these same reasons. It is useful to emphasize
that while our data show that 3.2% of hospitalized patients had
a malnutrition diagnosis on discharge, previous papers have
put the same figure at anywhere between 21% and 54%,
depending on the assessment instrument and patient popula-
tion. The profound differences between our results and those of
previous studies using standardized approaches to assessing
malnutrition in hospitalized patients suggest that the 3.2% of
patients with a malnutrition diagnosis in this 2010 national
sample likely represents only a small fraction of the true preva-
lence of hospitalized patients with this condition. Given the
less favorable clinical profile of the discharges with a diagnosis
of malnutrition, and if the misclassification of this condition is
Table 7. Receipt of Enteral and Parenteral Nutrition Among Discharged Patients With and Without Malnutrition Diagnoses, United
States, 2010.
Malnutrition Diagnosis No Malnutrition Diagnosis
Characteristic Estimate 95% CI Estimate 95% CI P Value
Received enteral nutrition (%) 5.2 4.1-6.2 0.6 0.4-1.1 <.0001
Received parenteral nutrition (%) 8.9 8.0-9.7 0.7 0.6-0.8 <.0001
Received either enteral or parenteral (%) 13.4 11.9-14.8 1.2 1.0-1.3 <.0001
Received both enteral or parenteral (%) 0.7 0.4-0.9 0.2 0.0-0.3 <.0001
Source: 2010 Healthcare Cost and Utilization Project.
by guest on October 13, 2015pen.sagepub.comDownloaded from

194 Journal of Parenteral and Enteral Nutrition 38(2)
as extensive as our data suggest, there is a compelling argu-
ment to enhance efforts to define and implement a standardized
definition of malnutrition. It should also be noted that our data
focused on comorbidities that were defined using HCUP vari-
ables and software. Other approaches to examining the issue of
comorbidity such as the Charlson Index, are also available. We
elected not to focus on this index because it has traditionally
been used to predict 10-year mortality, data that are not avail-
able in HCUP.
Despite the potential impact that under-identification of
malnutrition may have on defining the prevalence of malnutri-
tion in hospitalized patients, this methodological consideration
would have resulted in making the two groups more similar,
thereby pushing statistical inferences toward the null. From a
statistical point of view, findings that are biased toward the null
are considered conservative errors. Thus, in a setting with bet-
ter classification of malnutrition, we would expect that the
associations we observed here to be stronger as a result of
increased statistical power associated with larger numbers of
patients in the malnourished group.
HCUP’s sampling design makes it a unique data resource
with which broadly generalizable inferences can be made
about the prevalence of malnutrition diagnoses in U. S. hospi-
tals. These data can therefore serve as a reference point for
future analyses focusing on the prevalence of malnutrition in
this patient population, and they may be useful as a vehicle to
assess improvements in malnutrition assessment and coding in
response to implementation of a standardized malnutrition
definition.
There is currently no national benchmarking of malnutri-
tion in acute care hospitals in the United States. Implementation
of a formal malnutrition benchmarking program requires that
standard malnutrition screening and assessment tools be used
to track and monitor malnutrition rates by age and disease cat-
egory, as well as the ability to detect the extent to which
patients identified as malnourished are appropriately classified
for reimbursement purposes by medical record coders. If
implemented, a program of this type not only has the potential
to generate accurate data on annual rates of malnutrition, but it
could also enhance quality of patient care and address pressing
public policy issues such as rehospitalization. Until a prospec-
tive national benchmarking program of hospital malnutrition is
established, annual rates of malnutrition in U.S. hospitals will
remain unclear. However, in the interim, understanding the
extent to which patients are coded for malnutrition using ICD-9
codes like those used in HCUP can contribute to our general
understanding of malnutrition in acute-care hospitals.
Conclusion
Analysis of the 2010 NIS confirmed the increasing prevalence
of malnutrition diagnoses, particularly undernutrition, as
defined by ICD-9 coding. Strong associations between malnu-
trition and more than two dozen comorbid conditions were
observed, as were clear associations with increased cost and
length of stay. The breadth of HCUP data permits exploration
of additional malnutrition-related issues as well as analyses
that are stratified by key factors such as age and discharge dis-
position. Analyses focused on malnutrition coding and receipt
of PN or EN are underway, as well as age-specific data briefs.
Improved characterization of this vulnerable subgroup of hos-
pitalized patients will facilitate development of a more
informed and better defined public advocacy platform.
Development of new malnutrition definition paradigms and
assessment tools should provide opportunities to validate bet-
ter approaches to malnutrition diagnosis and treatment in the
hospital setting.
33,34
References
1. Somanchi M, Tao X, Mullin GE. The facilitated early enteral and dietary
management effectiveness trial in hospitalized patients with malnutrition.
JPEN J Parenter Enteral Nutr. 2011;35:209-216.
2. Bistrian BR, Blackburn GL, Vitale J, Cochran D, Naylor J. Prevalence of
malnutrition in general medical patients. JAMA. 1976;235:1567-1570.
3. Bistrian BR, Blackburn GL, Hallowell E, Heddle R. Protein status of gen-
eral surgical patients. JAMA. 1974;230:858-860.
4. Braunschweig C, Gomez S, Sheean PM. Impact of declines in nutritional
status on outcomes in adult patients hospitalized for more than 7 days. J
Am Diet Assoc. 2000;100:1316-1322.
5. Chima CS, et al. Relationship of nutritional status to length of stay, hos-
pital costs, and discharge status of patients hospitalized in the medicine
service. J Am Diet Assoc. 1997; 97: 975-978.
6. Callahan CM, Wolinsky FD. Hospitalization for pneumonia among older
adults. J Gerontology. 1996;51A:M276-M282.
7. Sullivan DH, Bopp MM, Roberson PK. Protein-energy undernutrition
and life-threatening complications among the hospitalized elderly. J Gen
Intern Med. 2002; 17: 923-932.
8. Sullivan, D. H. Risk factors for early hospital readmission in a select pop-
ulation of geriatric rehabilitation patients: the significance of nutritional
status. J Am Geriatr Soc. 1992; 40: 792-798.
9. Schneider SM, Hebuterne, et al. Malnutrition is an independent factor asso-
ciated with nosocomial infections. Brit J Nutrition. 2004; 92: 105-111.
10. Kassin MT, Owen RM, Perez SD, et al. Risk factors for 30-day hos-
pital readmission among general surgery patients. J Am Coll Surg.
2012;215:322-330.
11. Krumholz HM. Post-hospital syndrome--an acquired, transient condition
of generalized risk. N Engl J Med. 2013;368:100-102.
12. Kirkland KB, Briggs JP, Trivette SL, Wilkinson WE, Sexton DJ. The
impact of surgical-site infections in the 1990s: attributable mortality,
excess length of hospitalization, and extra costs. Infect Control Hosp
Epidemiol. 1999;20:725-730.
13. Elixhauser A, Steiner C. Readmissions to US hospitals by diagnosis, 2010.
Healthcare Cost and Utilization Project Statistical Brief #153. April 2013.
http://www.hcup-us.ahrq.gov/reports/statbriefs/sb153.pdf
14. Butterworth C. The skeleton in the hospital closet. Nutrition Today.
1974;9:4-8.
15. Robinson MK, Trujillo EB, Mogensen KM, Rounds J, McManus K,
Jacobs DO. Improving nutritional screening of hospitalized patients: the
role of prealbumin. JPEN J Parenter Enteral Nutr. 2003;27:389-395.
16. Sullivan DH, Sun S, et al. Protein-energy undernutrition among elderly
hospitalized patients: a prospective study. JAMA. 1999;281:2013-2019.
17. Pawellek I, Dokoupil K, Koletzko B. Prevalence of malnutrition in paedi-
atric hospital patients. Clin Nutr. 2008;27:72-76.
18. Correia MI, Campos AC; ELAN Cooperative Study. Prevalence of hospi-
tal malnutrition in Latin America: the multicenter ELAN study. Nutrition.
2003;19:823-825.
by guest on October 13, 2015pen.sagepub.comDownloaded from

Corkins et al 195
19. Russell A, Elia M on behalf of British Association for Parenteral
and Enteral Nutrition (BAPEN). Nutrition Screening Survey in the
UK and Republic of Ireland 2010 British Association for Parenteral
and Enteral Nutrition (BAPEN) www.bapen.org Accessed Jan. 24,
2013.
20. Barker LA, Gout BS, Crowe TC. Hospital malnutrition: prevalence, iden-
tification and impact on patients and the healthcare system. Int J Environ
Res Public Health. 2011;8:514-527
21. Waitzberg DL, Caiaffa WT, Correia MI. Hospital malnutrition: the
Brazilian national survey (IBRANUTRI): a study of 4000 patients.
Nutrition. 2001;17:573-580.
22. Sam J, Nguyen GC.Protein-calorie malnutrition as a prognostic indicator
of mortality among patients hospitalized with cirrhosis and portal hyper-
tension. Liver Int. 2009;29:1396-1402.
23. Nguyen GC, Munsell M, Brant SR, et al. Racial and geographic disparities
in the use of parenteral nutrition among inflammatory bowel disease inpa-
tients diagnosed with malnutrition in the United States. JPEN J Parenter
Enteral Nutr. 2009;33:563-568.
24. Nguyen GC, Munsell M, Harris ML. Nationwide prevalence and prog-
nostic significance of clinically diagnosable protein-calorie malnutrition
in hospitalized inflammatory bowel disease patients. Inflamm Bowel Dis.
2008;14:1105-1111.
25. Healthcare Cost and Utilization Project (HCUP). HCUP PL_RUCA
- Patient location: Rural-Urban Commuting Area (RUCA) Codes.
January 2013. Agency for Healthcare Research and Quality, Rockville,
MD. http://www.hcup-us.ahrq.gov/db/vars/siddistnote.jsp?var=pl_ruca
(Accessed January 16, 2013)
26. Healthcare Cost and Utilization Project (HCUP). HCUP Comorbidity
Software. January 2013. Agency for Healthcare Research and Quality,
Rockville, MD www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comor-
bidity.jsp. (Accessed January 16, 2013)
27. HCUP Databases. Healthcare Cost and Utilization Project (HCUP). February
2013. Agency for Healthcare Research and Quality, Rockville, MD. www.
hcup-us.ahrq.gov/nisoverview.jsp. (Accessed February 18, 2013)
28. National Center for Health Statistics Centers for Disease Control and
Prevention Analytic and Reporting Guidelines: The Third National
Health and Nutrition Examination Survey, NHANES III (1988-94)
Hyattsville, Maryland http://www.cdc.gov/nchs/data/nhanes/nhanes3/
nh3gui.pdf (accessed January 17, 2013)
29. US Census Bureau. 2010 Population Estimates. Office on Aging. www.
aoa.gov/Aging_Statistics/Profile/2011/docs/2011profile.pdf
30. Upadhyaya C, Mmaneni PV. Editorial: ICD-9 and complications. J.
Neurosurg:Spine. 2011:14;14-15.
31. Campbell PG, Malone J, Yadla S, Chitale R, Nasser R, Maltenfort MG,
et al. Caompariaon of ICD-9-based, retrospective, and prospective assess-
ments of perioperative complications; assessment of accuracy in report-
ing. J. Neurosurg:Spine. 2011:14;16-22.
32. Dang ABC, Garfin SR. Can statisics alone add cliical meaning to non-
specific billing databases? J Bone Joint Surg Am. 2011;93:e128(1-2).
33. Jensen GL, Mirtallo J, Compher C, et al. Adult starvation and disease-
related malnutrition: A proposal for etiology-based diagnosis in the
clinical practice setting from the International Consensus Guideline
Committee. JPEN J Parenter Enteral Nutr. 2010;34:156-159.
34. White JV, Guenter P, Jensen G, Malone A, Schofield M. Consensus
statement: Academy of Nutrition and Dietetics and American Society
for Parenteral and Enteral Nutrition: characteristics recommended for the
identification and documentation of adult malnutrition (undernutrition).
JPEN J Parenter Enteral Nutr. May 2012;36(3):275-283.
by guest on October 13, 2015pen.sagepub.comDownloaded from