Medical Complications Among Hospitalizations for Ischemic

Stroke in the United States From 1998 to 2007

Xin Tong, MPH; Elena V. Kuklina, MD, PhD;
Cathleen Gillespie, MS; Mary G. George, MD, MSPH, FACS

Background and Purpose—The common medical complications after ischemic stroke are associated with increased
mortality and resource use.
Method—The study population consisted of 1 150 336 adult hospitalizations with ischemic stroke as a primary diagnosis
included in the 1998 to 2007 Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project. Multiple
logistic regression analyses were used to examine changes between 1998 to 1999 and 2006 to 2007 in the prevalence
of acute myocardial infarction, pneumonia, deep venous thrombosis, pulmonary embolism, or urinary tract infection,
in-hospital mortality, and length of stay.
Results—In 2006 to 2007, the prevalence of hospitalizations with a secondary diagnosis of acute myocardial infarction,
pneumonia, deep venous thrombosis, pulmonary embolism, and urinary tract infection was 1.6%, 2.9%, 0.8%, 0.3%, and
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017

10.1%, respectively. The adjusted ORs for a hospitalization in 2006 to 2007 complicated by acute myocardial infarction,
deep venous thrombosis, pulmonary embolism, or urinary tract infection, using 1998 to 1999 as the referent, were 1.39,
1.68, 2.39, and 1.18, respectively. The odds of pneumonia did not change significantly between 1998 to 1999 and 2006
to 2007. In-hospital mortality was significantly lower in 2006 to 2007 than in 1998 to 1999. Despite the overall length
of stay decreasing significantly from 1998 to 1999 to 2006 to 2007, it remained the same for hospitalizations with acute
myocardial infarction, pneumonia, deep vein thrombosis, and pulmonary embolism.
Conclusion—Although in-hospital mortality decreased over the study period, 4 of the 5 complications were more common
in 2006 to 2007 than they were 8 years earlier with the largest increase observed for deep venous thrombosis and
pulmonary embolism. (Stroke. 2010;41:980-986.)
Key Words: cerebral infarct 䡲 embolism 䡲 epidemiology 䡲 outcomes 䡲 stroke care 䡲 venous thrombosis

A lthough the annual stroke death rate in the United States

fell 29.7% during the last decade, and although the
actual number of stroke deaths declined 13.5% during the
same period, stroke remains a leading cause of death and
serious long-term disability in this country.1 Medical compli-
cations are common in patients with stroke,2 but the impact of
these complications has not been well studied. The compli-
cations of pneumonia (PN), deep vein thrombosis (DVT),
pulmonary embolism (PE), and urinary tract infection (UTI)
after an acute ischemic stroke are recognized as negative
indicators of the quality of stroke care,3,4 and acute myocar-
dial infarction (AMI) is not uncommon after an acute ische-
mic stroke.5 All of these complications are largely prevent-
able or treatable and may adversely affect the patient’s
outcome and extend the length of stay (LOS). Even so, there
have been few population-based estimates of these compli-
cations and very little on their recent trends. Prevention of
venous thromboembolic disease (DVT and PE) and PN are a
focus of several programs for improving the quality of stroke
care, including The Joint Commission’s Primary Stroke
Center program, American Heart Association’s Get With The
Guidelines–Stroke program, and the Centers for Disease
Control and Prevention’s Paul Coverdell National Acute
Stroke Registry. Additionally, in October 2008, the Centers
for Medicare and Medicaid instituted efforts to reduce
catheter-associated UTIs in hospitalized patients.
Recently, a need for initiating surveillance of noncommu-
nicable disease at the national and global levels has been
increasingly recognized (World Health Organization web
site: www.who.int/ncd_surveillance/en/).6 However, cur-
rently no comprehensive surveillance system is available to
monitor changes in patient care. Previous experience with the
development of surveillance systems in other areas at the
national level shows that the standardized nationwide hospital
discharge data sets and standardized definitions can be used
for this purpose.7 In the present report, we examined hospi-

Received January 11, 2010; accepted January 13, 2010.

From the Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease
Control and Prevention, Atlanta, Ga.
This paper is part of an oral presentation at the 2010 International Stroke Conference, San Antonio, Texas, February 24-26, 2010.
Correspondence to Elena V. Kuklina, MD, PhD, Division for Heart Disease and Stroke Prevention, Centers for Disease Prevention and Control, 4770
Buford Highway, MS-K47, Atlanta, GA 30341. E-mail ekuklina@cdc.gov
© 2010 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.110.578674

980
 Tong et al Medical Complications Among Hospitalizations for IS
talizations of adults (age ⱖ18 years) whose primary diagnosis
was ischemic stroke as revealed by International Classifica-
tion of Diseases, Ninth Revision, Clinical Modification (ICD-
9-CM) codes. The 3 primary objectives were to (1) examine
trends in AMI, PN, DVT, PE, and UTI as secondary diag-
noses, in-hospital mortality, and LOS; (2) investigate the
associations of AMI, PN, DVT, PE, and UTI with in-hospital
mortality; and (3) compare LOS for hospitalizations with and
without AMI, PN, DVT, PE, and UTI diagnosis. We drew on
the 1998 to 2007 Nationwide Inpatient Sample (NIS) of the
Healthcare Cost and Utilization Project of the Agency for
Healthcare Research and Quality. For this analysis, we have
chosen to use the term “complications” for these secondary
diagnoses rather than comorbid conditions, but we realize that
we were not able to distinguish them as present on arrival or
developing during the admission.
Methods
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017
Data Sources
The NIS is a stratified, cross-sectional sample that includes approx-
imately 20% of all community (nonfederal) hospital discharges in the
United States. The NIS database collects yearly data from approxi-
mately 1000 hospitals stratified by region, location/teaching status,
number of beds, and ownership and the sampling unit represents a
hospitalization, not a person or patient (the same patient may be
included in ⱖ2 hospitalizations). The NIS is the largest all-payer
inpatient care database in the United States. The details of the NIS
sampling design, data elements, and estimates can be found at the
Healthcare Cost and Utilization Project web site (www.hcup-us.ahrq.
gov/).
Study Variables
Complications were identified by searching the secondary diagnostic
codes (ICD-9-CM codes) for AMI, PN, DVT, PE, and UTI (the
detailed ICD-9-CM coding is presented in the Appendix to this
article). In addition to medical complications, in-hospital mortality
and LOS were also assessed in this analysis. Nine comorbidity
conditions (congestive heart failure, peripheral vascular disease,
hypertension, paralysis, other neurological disorders, chronic pulmo-
nary disease, diabetes with chronic complications, renal failure, and
coagulopathy) were chosen from the list of comorbidities proposed
by the Agency for Healthcare Research and Quality based on the
significance of associations with in-hospital mortality reported in
the previous studies8 and these conditions did not overlap with the
complications focused on in this report. Furthermore, we examined
the association between the outcome variables and the following
demographic features: age (⬍65 years versus ⱖ65 years), sex (male
versus female), payment type (Medicare, Medicaid, private, and
other), and hospital location (urban versus rural).
Statistical Methods
The unit of analysis was the hospital discharge. To develop national
estimates from our findings, we used discharge weights; which were
developed to extrapolate NIS sample discharges to the total dis-
charges. Five time intervals were selected in this study: 1998 to
1999, 2000 to 2001, 2002 to 2003, 2004 to 2005, and 2006 to 2007.
The prevalence of medical complications (AMI, PN, DVT, PE, and
UTI) as well as in-hospital mortality was assessed across 5 time
intervals. Multiple logistic regression analyses were conducted to
examine whether there was a linear trend across the 5 time periods
in the prevalence of medical complications after adjusting for age,
sex, payment type, and hospital location. To assess the changes over
time in in-hospital mortality, multiple logistic regression modeling
was used adjusting for age, sex, payment type, the 9 comorbidities
mentioned, and hospital location as well as the complications of
interest in this report. A linear regression model was used to assess
981
the changes in LOS across the 5 time periods after adjusting for age,
sex, payment type, and hospital locations. The orthogonal polyno-
mial coefficients were obtained in both the logistic and linear
regression models for testing the linear trend. In addition, we
examined the changes in prevalence of medical complications and
in-hospital mortality between 1998 to 1999 and 2006 to 2007 by
using multiple logistic regression analysis after the adjustments
described previously.
To evaluate the patient and hospital factors that might be associ-
ated with medical complications and in-hospital mortality, we
applied the same logistic regression model within 1998 to 1999 and
2006 to 2007. Adjusted ORs and 95% CIs were obtained from the
multiple logistic regression models. Lastly, we used t tests to assess
the difference in LOS between hospitalizations with and without
medical complications between 1998 to 1999 and 2006 to 2007.
Because of the larger sample size, 2-tailed probability values ⬍0.01
were considered to be significant throughout the analyses. All
statistical analyses were performed with the statistical software
package SUDAAN (Release 9.2; Research Triangle Institute, Re-
search Park, NC) to account for the complex sample design.
Results
During 1998 to 2007, there were 1 150 336 adult hospitaliza-
tions with ischemic stroke as a primary diagnosis, represent-
ing 5 635 404 hospitalizations across the United States in this
time period. The weighted proportions of hospitalizations in
which the patient was aged ⱖ65 years decreased from 77.6%
in 1998 to 1999 to 76.5% in 2000 to 2001 to 74.5% in 2002
to 2003, 72.7% in 2004 to 2005, and 71.3% in 2006 to 2007
(Table 1), indicating earlier age of onset over time. Women
comprised more than half of the stroke hospitalizations, and
the majority of hospitalizations represented patients with
Medicare coverage. The proportion of urban hospitals in-
creased from 82.5% in 1998 to 1999 to 86.2% in 2006 to 2007
(P for linear trend ⬍0.001).
Medical Complications
The prevalence of hospitalizations with a secondary diagnosis
of PN did not change during the study period (Table 1). In
contrast, the prevalence of hospitalizations with a secondary
diagnosis of AMI, DVT, PE, or UTI increased significantly
across the study period (P⬍0.001). The prevalence of hospi-
talizations with at least 1 complication (AMI, PN, DVT, PE,
or UTI) increased from 12.8% in 1998 to 1999 to 14.1% in
2006 to 2007 (P for linear trend ⬍0.001).
The prevalence of hospitalizations with complication in
2006 to 2007 and 1998 to 1999 was 1.58% versus 1.17% for
AMI; 2.93% versus 2.97%, PN; 0.79% versus 0.46%, DVT;
0.27% versus 0.11%, PE; and 10.08% versus 9.16%, UTI.
The hospitalizations with secondary PN did not change
significantly between 1998 to 1999 and 2006 to 2007 (Table
2). In contrast, the hospitalizations with the other 4 compli-
cations all rose significantly during the time period of
interest. After adjustment for age, sex, payer, and hospital
location, the OR for prevalence of AMI, DVT, PE, and UTI
was 1.39 (95% CI: 1.30 to 1.48), 1.68 (95% CI: 1.52 to 1.86),
2.39 (95% CI: 2.04 to 2.81), and 1.18 (95% CI: 1.14 to 1.22),
respectively, in 2006 to 2007 when the 1998 to 1999 period
was used as the referent. For hospitalization with at least 1
complication, the adjusted OR was 1.17 (95% CI: 1.13 to
1.21) in 2006 to 2007 by using 1998 to 1999 as the referent
(data not shown).
 982 Stroke May 2010

Table 1. Characteristics of Hospitalizations for Ischemic Stroke, 1998 to 2007, United States
Weighted Percent (SE)

1998 –1999 2000 –2001 2002–2003 2004 –2005 2006 –2007

Characteristic (N⫽241 452) (N⫽237 412) (N⫽235 411) (N⫽217 076) (N⫽218 985)
Demographic or clinical feature
Age ⱖ65 years 77.60 (0.23) 76.51 (0.24) 74.49 (0.25) 72.68 (0.25) 71.30 (0.28)
Female 53.45 (0.15) 53.72 (0.16) 53.08 (0.16) 52.32 (0.16) 51.49 (0.16)
Payer
Medicare 71.84 (0.37) 72.01 (0.38) 71.99 (0.33) 70.44 (0.35) 68.99 (0.33)
Medicaid 4.25 (0.14) 4.42 (0.17) 4.69 (0.15) 4.97 (0.15) 5.20 (0.17)
Private 20.04 (0.34) 19.56 (0.36) 18.98 (0.31) 19.37 (0.32) 19.90 (0.28)
Other 3.87 (0.16) 4.01 (0.18) 4.34 (0.17) 5.23(0.20) 5.91 (0.20)
Urban hospital 82.50 (0.52) 82.63 (0.53) 83.40 (0.54) 85.23 (0.49) 86.18 (0.51)
Outcome measures
Mean length of stay in days (SE)* 5.56 (0.05) 5.35 (0.05) 5.12 (0.04) 4.92 (0.04) 4.77 (0.04)
AMI* 1.17 (0.03) 1.40 (0.03) 1.50 (0.03) 1.52 (0.04) 1.58 (0.04)
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017

PN 2.97 (0.05) 2.90 (0.05) 3.01 (0.05) 3.07 (0.05) 2.93 (0.06)
DVT* 0.46 (0.02) 0.49 (0.02) 0.60 (0.02) 0.72 (0.02) 0.79 (0.03)
PE* 0.11 (0.01) 0.13 (0.01) 0.14 (0.01) 0.21 (0.01) 0.27 (0.01)
UTI* 9.16 (0.10) 8.67 (0.10) 9.07 (0.11) 9.97 (0.11) 10.08 (0.11)
At least 1 of the 5 complications* 12.82 (0.12) 12.46 (0.11) 13.05 (0.12) 14.01 (0.13) 14.13 (0.14)
In-hospital mortality† 5.79 (0.07) 5.55 (0.08) 5.17 (0.07) 4.84 (0.07) 4.35 (0.07)
*P for linear trend ⬍0.001 after adjusting for age, sex, payer, and hospital location.
†P for linear trend ⬍0.001 after adjusting for age, sex, payer, hospital location, AMI, PN, DVT, PE, and UTI
complications and 9 comorbidity conditions.

Patient age (ⱖ65 years) was significantly associated with
AMI, PN, and UTI (but not DVT or PE) in both 1998 to 1999
and 2006 to 2007 in adjusted analyses (Table 2). Hospital-
izations of women were significantly more likely to include
UTI and significantly less likely to include PN in both time
periods. Female sex was associated with an increased risk of
AMI in 2006 to 2007 but was not associated with this risk in
1998 to 1999. In contrast, female sex was associated with an
increased risk DVT in 1998 to 1999 but was unrelated to
DVT risk in 2006 to 2007. Insurance status and hospital
location were also significantly associated with the compli-
cations of interest (Table 2).
In-Hospital Mortality
Overall, the in-hospital mortality rate decreased signifi-
cantly across the study period after adjustment for age, sex,
payer, hospital location, complications, and comorbidities
(P⬍0.001). Using 1998 to 1999 as the referent, the adjusted
OR for in-hospital mortality was 0.75 (95% CI: 0.72 to 0.79)

Table 2. Adjusted Odds of Type of Medical Complications in 2006 to 2007 Compared With 1998 to 1999 and Selected Characteristics
on Type of Medical Complication Within 1998 to 1999 and 2006 to 2007
AMI PN DVT

1998 –1999 2006 –2007 1998 –1999 2006 –2007 1998 –1999 2006 –2007
Adjusted Reference 1.39 (1.30 –1.48)* Reference 1.00 (0.95–1.05) Reference 1.68 (1.52–1.86)*
Values of individual covariates (age,
female sex, payer, location) reflected
adjustment for each of the others
Age ⱖ65 years 1.50 (1.30 –1.73)* 1.66 (1.46 –1.89)* 1.83 (1.65–2.02)* 1.71 (1.56 –1.87)* 0.87 (0.71–1.07) 0.96 (0.82–1.12)
Female sex 0.97 (0.90 –1.05) 1.17 (1.09 –1.26)* 0.69 (0.65– 0.72)* 0.79 (0.75– 0.83)* 1.17 (1.03–1.33)* 1.08 (0.98 –1.19)
Payer
Medicare 1.04 (0.91–1.19) 1.09 (0.95–1.23) 1.28 (1.16 –1.41)* 1.29 (1.17–1.43)* 0.95 (0.78 –1.15) 0.94 (0.79 –1.12)
Medicaid 1.07 (0.84 –1.37) 1.65 (1.39 –1.96)* 2.03 (1.77–2.34)* 1.98 (1.75–2.24)* 1.62 (1.25–2.10)* 1.47 (1.18 –1.82)*
Private Reference Reference Reference Reference Reference Reference
Other 1.17 (0.92–1.49) 1.41 (1.20 –1.67)* 1.19 (1.02–1.40)* 1.24 (1.08 –1.42)* 0.79 (0.55–1.14) 1.04 (0.83–1.30)
Urban hospital 1.05 (0.93–1.19) 1.11 (0.98 –1.25) 1.21 (1.11–1.31)* 1.14 (1.03–1.27)* 1.47 (1.19 –1.82)* 1.71 (1.37–2.14)*
*Statistically significant. Values are OR with 95% CIs.
 Tong et al Medical Complications Among Hospitalizations for IS 983

in 2006 to 2007 (Table 3). The unadjusted OR for in-hospital Table 3. Adjusted ORs for In-Hospital Mortality in 2006 to
mortality was 0.74 (95% CI: 0.71 to 0.77). 2007 Compared With 1998 to 1999 and Selected Characteristics
Age ⱖ65 years and female sex were associated with on In-Hospital Mortality Within 1998 to 1999 and 2006 to 2007
significantly higher in-hospital mortality in both 1998 to 1999 Adjusted OR (95% CI)
and 2006 to 2007 (Table 3). In terms of payment source,
when private insurance was used as the referent, hospitaliza- 1998 –1999 2006 –2007
tions with Medicare had lower odds of in-hospital mortality Adjusted Reference 0.75 (0.72– 0.79)*
in 2006 to 2007 (OR⫽0.69, 95% CI: 0.60 to 0.79) but not in Values for the 9 covariates listed
1998 to 1999. However, in both time periods, stays covered subsequently reflected adjustment
by neither Medicare nor Medicaid (“other” in Table 3) had for each of the other 8 as well as
the 9 comorbidity conditions
significantly higher odds of in-hospital mortality than the
reference group (private insurance). Additionally, hospitaliza- Age ⱖ65 years 1.86 (1.71–2.02)* 2.54 (2.26–2.86)*
tions in urban areas had significantly lower odds of in-hospital Female sex 1.21 (1.17–1.26)* 1.29 (1.23–1.35)*
mortality than hospitalizations in rural areas (adjusted OR 0.8 in Payer
both 1998 to 1999 and 2006 to 2007). Medicare 0.92 (0.84–1.01) 0.69 (0.60–0.79)*
All 5 complications except UTI in 2006 to 2007 were Medicaid 1.09 (0.96–1.24) 1.11 (0.96–1.28)
independent predictors of in-hospital mortality in the multi- Private Reference Reference
variable analysis in both time periods with PN, PE, and AMI
Other 1.40 (1.22–1.61)* 1.53 (1.30–1.81)*
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017

being the strongest predictors of in-hospital mortality (Table 3).

Urban hospital 0.85 (0.80–0.90)* 0.82 (0.76–0.88)*
Length of Stay AMI 3.90 (3.50–4.35)* 3.69 (3.31–4.10)*
LOS declined significantly across the study period with the PN 6.32 (5.94–6.72)* 6.11 (5.67–6.59)*
mean value (in days) dropping from 5.56 in 1998 to 1999 to DVT 1.52 (1.22–1.90)* 1.29 (1.06–1.55)*
5.35 in 2000 to 2001, then 5.12 in 2002 to 2003, then 4.92 in PE 4.13 (2.91–5.86)* 3.04 (2.33–3.95)*
2004 to 2005, and finally to 4.77 in 2006 to 2007 (Table 1). UTI 1.10 (1.03–1.16)* 0.98 (0.91–1.05)
The probability value for trend was ⬍0.001 after adjustment
*Statistically significant.
for age, sex, payment type, and hospital location. Univariate Adjustments of 9 comorbidity conditions: congestive heart failure, peripheral
analyses found that LOS deceased significantly between the vascular, hypertension, paralysis, other neurological disorders, chronic pulmonary
first and last time periods among hospital stays without AMI disease, diabetes with chronic complications, renal failure, and coagulopathy.
or PN or DVT or PE, stays with or without UTI, or stays with
at least 1 complication and without any complication (Fig- who died was 7.21 in 1998 to 1999 and 6.78 in 2006 to 2007
ure). However, stays with complications of AMI, PN, DVT, (P⫽0.001). The mean LOS (in days) for patients who died
or PE remained the same between these 2 time periods. The without any medical complications was 5.77 in 1998 to 1999
mean LOS with a complication was significantly longer than
and 5.25 in 2006 to 2007, whereas the mean LOS (in days) for
those without the complication in both 1998 to 1999 and 2006
patients who died and with at least 1 medical complications
to 2007 (P⬍0.0001, data not shown). In 2006 to 2007, the
was 10.35 in 1998 to 1999 and 9.55 in 2006 to 2007
mean LOS (in days) for hospitalizations with complication
(P⬍0.001 for both). The adjustment for LOS had no effect on
and those without was 9.3 versus 4.7 for AMI; 13.4 versus
our reported unadjusted and adjusted OR for in-hospital
4.5, PN; 12.7 versus 4.7, DVT; 13.7 days versus 4.7, PE; and
mortality in 2006 to 2007 compared with 1998 to 1999.
8.3 versus 4.4, UTI. The mean LOS (in days) for all patients

Table 2. Continued Discussion

Consistent with results from other studies,9,10 we found that
in-hospital mortality for ischemic stroke decreased signifi-
PE UTI
cantly during the period of study (1998 to 1999 to 2006 to
1998 –1999 2006 –2007 1998 –1999 2006 –2007 2007). Unfortunately, we found increasing trends in the
Reference 2.39 (2.04 –2.81)* Reference 1.18 (1.14 –1.22)* prevalence of 4 of the 5 complications we examined with the
largest increases observed for DVT and PE.
Among all the medical complications evaluated in this
study, UTI was the most common, and estimates for this
0.80 (0.52–1.25) 0.86 (0.67–1.11) 1.88 (1.77–2.00)* 1.91 (1.81–2.02)* event were comparable to other studies.11,12 The nationwide
1.12 (0.87–1.44) 1.14 (0.96 –1.34) 2.80 (2.71–2.91)* 3.05 (2.95–3.16)* prevalence of AMI, PN, DVT, PE, and UTI among hospital-
izations with ischemic stroke has been reported by Qureshi
0.94 (0.63–1.40) 0.83 (0.63–1.08) 1.38 (1.31–1.46)* 1.44 (1.36 –1.53)* and colleagues,9 but it is difficult to compare their results with
2.09 (1.27–3.43)* 1.66 (1.22–2.27)* 1.90 (1.76 –2.06)* 1.99 (1.84 –2.15)* ours because of the differences in the ICD-9-CM codes used
Reference Reference Reference Reference to identify the complications. Regardless, the increasing
1.13 (0.60 –2.14) 0.95 (0.66 –1.37) 1.13 (1.01–1.26)* 1.38 (1.26 –1.51)* trends in the prevalence of DVT reported by Qureshi and
0.94 (0.66 –1.33) 1.27 (0.96 –1.68) 0.94 (0.89 – 0.99)* 0.93 (0.88 – 0.99)* coworkers were seen in our study as well. Moreover, the
increasing trends we found in DVT and PE are parallel to
 984 Stroke May 2010
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017
Figure. LOS of patients with medical complications and those without between 1998 to 1999 and 2006 to 2007, United States. The
asterisk (*) indicates a significant difference between 1998 to 1999 and 2006 to 2007.
trends reported in other studies as well as reports in other
groups of hospitalized patients.7,13,14 These trends may reflect
a true increase in the prevalence of these conditions or they
may indicate improved detection due to the increasing avail-
ability of new and improved diagnostic procedures. Finally,
changes in hospital coding practices in the later years of the
targeted analysis period compared with the earlier years could
contribute to the trends reported by us.
Recommendations have been developed to improve the
quality of acute stroke care by several professional and
scientific organizations.15,16 For example, the guidelines of
the American Heart Association Science Advisory and Co-
ordinating Committee recommend early mobilization, general
supportive care, and treatment of acute complications such as
AMI, PN, and UTI and prevention of venous thromboembolic
disease in patients with ischemic stroke.15 Prophylactic low-
dose subcutaneous heparin or low-molecular-weight heparins
are recommended for patients with acute ischemic stroke with
restricted mobility.16 Currently, the extent to which preven-
tive measures for DVT and PE among patients with ischemic
stroke are implemented in the United States is not known.
However, our results underline the importance of preventing
venous thromboembolic disease and PE in these patients and
the need to collect and evaluate information on the imple-
mentation of these preventive measurements using registries
and other related clinical data sets.
We found that all 5 of the complications except UTI in
2006 to 2007 were significant predictors of in-hospital
mortality after adjustment of comorbidities with AMI, PN,
and PE the strongest predicators. Several studies have found
a significant association between both PN and PE and early
mortality (ⱕ30 days) among patients hospitalized for acute
stroke.17–19 In fact, the German Stroke Registers study group
found that PN was the complication with the highest attrib-
utable proportion of deaths in the entire stroke population,
accounting for 31% of all deaths.19 Although stroke as a
serious complication among patients with AMI has been
documented,20 –22 few studies have focused on associations
between in-hospital AMI and outcomes among patients with
stroke. A recent study published by the Canadian Stroke
Network, however, found that AMI was an important medical
complication after acute ischemic stroke and was indepen-
dently associated with 1-year mortality or severe dependence
at discharge.5 We were unable to distinguish between early
and late mortality in our study, but our results confirmed that
AMI, PN, and PE are strong predictors of in-hospital mortal-
ity. In addition, although the overall LOS decreased signifi-
cantly from 1998 to 1999 to 2006 to 2007, it remained the
same among the hospitalizations with AMI, PN, DVT, or PE
during this study period. Our study confirmed the findings
that patients hospitalized with complications had significantly
longer LOS than those without.11,12 These results suggest the
longer LOS among hospitalizations with complications may
contribute significantly to acute treatment cost.23,24 Future
studies are needed to examine the impact of decreased LOS
on cost, quality of care, and long-term outcomes among
ischemic stroke hospitalizations without complications.
The results of this analysis should be interpreted in light of
several limitations. First, the unit of the analysis was the
hospital discharge, not the patient. We were unable to identify
whether patients were admitted to the hospital multiple times
in any 1-year period, and thus the results might be overesti-
mated. Second, medical complications were identified by
using ICD-9-CM codes abstracted from discharge records.
Although any complication should be coded if this condi-
tion interferes with the patient management,25 bias associ-
ated with coding is possible. Unavailability or inconsistent
use of standardized diagnostic criteria for some conditions
 Tong et al Medical Complications Among Hospitalizations for IS
among physicians, differences in ICD-9-CM coding at the
hospital level, and decreased likelihood of coding for
nonthreatening conditions among patients with severe
illness may contribute to this.26 In addition, we were
unable to identify if secondary conditions defined by
ICD-9-CM codes were pre-existing or developing during
the hospitalization. Currently, except for 4 states (Califor-
nia, New York, Florida, and Wisconsin), the Healthcare
Cost and Utilization Project data did not include a pres-
ent-on-admission indicator. Thus, the time of onset of
secondary diagnoses cannot be established from this ad-
ministrative data set and pre-existing conditions and com-
plications cannot be distinguished. Third, although we
used payment source as a proxy of socioeconomic status, a
lack of information on race precluded us from assessing
this confounder. Fourth, it is possible that an increase in
discharge to a hospice could account for some decrease in
in-hospital mortality. Early discharge to home, rehabilita-
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017
tion, or chronic care could account for the decrease of
LOS. We were unable to examine the impact of discharge
destination such as a hospice or rehabilitation on mortality
and LOS due to lack of detailed information on discharge
destination collected in Healthcare Cost and Utilization
Project data. Finally, clinical (stroke severity, history of
DVT, body mass index, use of oral contraceptives or
estrogens, and so on)27 and healthcare factors (anticoagu-
lation therapy, long immobilization, recent major surgery,
and so on) and their contribution to the trends could not be
examined because of the absence of clinical details in our
database.
In conclusion, given our findings on the 5 complications
of interest, there is clearly a need for further investigation
of the sociodemographic, clinical, and healthcare factors
associated with these conditions during hospitalizations
with ischemic stroke. Our results also underscore the need
for continued efforts to reduce complications from acute
stroke through active quality improvement programs. Ad-
ditionally, these results suggest the need for further studies
to investigate AMI as a complication of stroke hospital-
izations. Finally, we believe that our results can be applied
to extend surveillance for quality improvement initiatives
at the state and hospital level as well as internationally and
provide valuable information for comparisons across many
settings.
Disclosures
None.
References
1. Lloyd-Jones D, Adams R, Carnethon M, Simone GD, Ferguson TB,
Flegal K, Ford E, Furie K, Go A, Greenlund K, Haase N, Hailpern S, Ho
M, Howard V, Kissela B, Kittner S, Lackland D, Lisabeth L, Marelli A,
McDermott M, Meigs J, Mozaffarian D, Nichol G, O’Donnell C, Roger
V, Rosamond W, Sacco R, Sorlie P, Stafford R, Steinberger J, Thom T,
Wasserthiel-Smoller S, Wong N, Wylie-Rosett J, Hong Y. Heart disease
and stroke statistics—2009 update: a report from the American Heart
Association Statistics Committee and Stroke Statistics Subcommittee.
Circulation. 2009;119:e21– e81.
2. Johnson KC, Li JY, Lyden PD, Hanson SK, Feasby TE, Adams RJ,
Faught RE Jr, Haley EC. Medical and neurological complications of
ischemic stroke: experience from the RANTTAS trial. RANTTAS Inves-
tigators. Stroke. 1998;29:447– 453.
985
3. Schwamm LH, Fonarow GC, Reeves MJ, Pan W, Frankel MR, Smith EE,
Ellrodt G, Cannon CP, Liang L, Peterson E, LaBresh KA. Get With the
Guidelines–Stroke is associated with sustained improvement in care for
patients hospitalized with acute stroke or transient ischemic attack.
Stroke. 2009;119:107–115.
4. Lindsay MP, Kapral MK, Gladstone D, Holloway R, Tu JV, Laupacis
A, Grimshaw JM. The Canadian Stroke Quality of Care Study: estab-
lishing indicators for optimal acute stroke care. CMAJ. 2005;172:
363–365.
5. Liao J, O’Donnell MJ, Silver FL, Thiruchelvam D, Saposnik G, Fang J,
Gould L, Mohaamed N, Kapral MK. In-hospital myocardial infarction
following acute ischaemic stroke: an observational study. Eur J Neurol.
2009;16:1035–1040.
6. Goff DC Jr, Brass L, Braun LT, Croft JB, Flesch JD, Fowkes FG, Hong
Y, Howard V, Huston S, Jencks SF, Luepker R, Manolio T, O’Donnell C,
Robertson RM, Rosamond W, Rumsfeld J, Sidney S, Zheng ZJ. American
Heart Association Council on Epidemiology and Prevention; American
Heart Association Council on Stroke; American Heart Association Council
on Cardiovascular Nursing; American Heart Association Working Group on
Quality of Care and Outcomes Research; American Heart Association
Working Group on Atherosclerotic Peripheral Vascular Disease. Essential
features of a surveillance system to support the prevention and management
of heart disease and stroke: a scientific statement from the American Heart
Association Councils on Epidemiology and Prevention, Stroke, and Cardio-
vascular Nursing and the Interdisciplinary Working Groups on Quality of
Care and Outcomes Research and Atherosclerotic Peripheral Vascular
Disease. Circulation. 2007;115:127–155.
7. Kuklina EV, Meikle SF, Jamieson DJ, Whiteman MK, Barfield WD,
Hillis SD, Posner SF. Severe obstetric morbidity in the United States:
1998 –2005. Obstet Gynecol. 2009;113:293–299.
8. Zhu H, Hill MD. The Elixhauser index for comorbidity adjustment of
in-hospital case fatality. Neurology. 2008;71:283–287.
9. Qureshi AI, Suri MF, Nasar A, Kirmani JF, Ezzeddine MA, Divani AA,
Giles WH. Changes in cost and outcome among US patients with stroke
hospitalized in 1990 to 1991 and those hospitalized in 2000 to 2001.
Stroke. 2007;38:2180 –2184.
10. Arboix A, Cendros V, Besa M, Garcia-Eroles L, Oliveres M, Targa C,
Balcells M, Comes E, Massons J. Trends in risk factors, stoke
subtypes and outcome. Nineteen-year data from the Sagrat Cor
Hospital of Barcelona stroke registry. Cerebrovasc Dis. 2008;26:
509 –516.
11. Ovbiagele B, Hills NK, Saver JL, Johnston SC. Frequency and deter-
minants of pneumonia and urinary tract infection during stroke hospital-
ization. J Stroke Cerebrovasc Dis. 2006;15:209 –213.
12. Tirschwell DL, Kukull WA, Longstreth WT Jr. Medical complications of
ischemic stroke and length of hospital stay: experience in Seattle, Wash-
ington. J Stroke Cerebrovasc Dis. 1999;8:336 –343.
13. Stein PD, Beemath A, Olson RE. Trends in the incidence of pulmonary
embolism and deep venous thrombosis in hospitalized patients. Am J
Cardiol. 2005;95:1525–1526.
14. Cai B, Xu L, Guo S. Trends of underline diseases of pulmonary embolism
from Peking Union Medical College Hospital [in Chinese]. Zhonghua Jie
He He Hu Xi Za Zhi. 2001;24:715–717.
15. Adams HP Jr, del Zoppo G, Alberts MJ, Bhatt DL, Brass L, Furlan A,
Grubb RL, Higashida RT, Jauch EC, Kidwell C, Lyden PD, Morgenstern
LB, Qureshi AI, Rosenwasser RH, Scott PA, Wijdicks EF. Guidelines for
the early management of adults with ischemic stroke: a guideline from the
American Heart Association/American Stroke Association Stroke
Council, Clinical Cardiology Council, Cardiovascular Radiology and
Intervention Council, and the Atherosclerotic Peripheral Vascular Disease
and Quality of Care Outcomes in Research Interdisciplinary Working
Groups: the American Academy of Neurology affirms the value of this
guideline as an educational tool for neurologists. Circulation. 2007;115:
e478 – e534.
16. Albers GW, Amarenco P, Easton JD, Sacco RL, Teal P. Antithrombotic
and thrombolytic therapy for ischemic stroke: American College of Chest
Physicians Evidence-Based Clinical Practice Guidelines (8th Edition).
Chest. 2008;133(suppl):630S– 669S.
17. Katzan IL, Cebul RD, Husak SH, Dawson NV, Baker DW. The effect of
pneumonia on mortality among patients hospitalized for acute stroke.
Neurology. 2003;60:620 – 625.
18. Silver FL, Norris JW, Lewis AJ, Hachinski VC. Early mortality following
stroke: a prospective review. Stroke. 1984;15:492– 496.
19. Heuschmann PU, Kolominsky-Rabas PL, Misselwitz B, Hermanek P,
Leffmann C, Janzen RW, Rother J, Buecker-Nott HJ, Berger K. Predi-
 986 Stroke May 2010
cators of in-hospital mortality and attributable risks of death after ische-
mic stroke: the German Stroke Registers Study Group. Arch Intern Med.
2004;164:1761–1768.
20. Saczynski JS, Spencer FA, Gore JM, Gurwitz JH, Yarzebski J, Lessard D,
Goldberg RJ. Twenty-year trends in the incidence of stroke complicating
acute myocardial infarction: Worcester Heart Attack Study. Arch Intern
Med. 2008;168:2104 –2410.
21. Saczynski JS, McManus D, Zhou Z, Spencer F, Yarzebski J, Lessard D,
Gore JM, Goldberg RJ. Trends in atrial fibrillation complicating acute
myocardial infarction. Am J Cardiol. 2009;104:169 –174.
22. Witt BJ, Ballman KV, Brown RD Jr, Meverden RA, Jacobsen SJ, Roger
VL. The incidence of stroke after myocardial infarction: a meta-analysis.
Am J Med. 2006;119:354.e1– e9.
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017
23. Christensen MC, Valiente R, Silva GA, Lee WC, Dutcher S, Rocha MSG,
Massaro A. Acute treatment costs of stroke in Brazil. Neuroepidemiology.
2009;32:142–149.
24. Christensen MC, Munro V. Ischemic stroke and intracerebral hemor-
rhage: the latest evidence on mortality, readmissions and hospital costs
from Scotland. Neuroepidemiology. 2008;30:239 –246.
25. Falen TJ. Learning to Code With ICD-9-CM for Health Information
Management and Health Services Administration 2009. Philadelphia:
Lippincott Williams & Wilkins; 2009.
26. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures
for use with administrative data. Med Care. 1998;36:8 –27.
27. Roth EJ, Lovell L, Harvey RL, Heinemann AW, Semik P, Diaz S.
Incidence of and risk factors for medical complications during stroke
rehabilitation. Stroke. 2001;32:523–529.
 Medical Complications Among Hospitalizations for Ischemic Stroke in the United States
From 1998 to 2007
Xin Tong, Elena V. Kuklina, Cathleen Gillespie and Mary G. George
Downloaded from http://stroke.ahajournals.org/ by guest on October 7, 2017

Stroke. 2010;41:980-986; originally published online March 4, 2010;

doi: 10.1161/STROKEAHA.110.578674
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2010 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/41/5/980

Data Supplement (unedited) at:

http://stroke.ahajournals.org/content/suppl/2016/03/31/STROKEAHA.110.578674.DC1

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at:

http://stroke.ahajournals.org//subscriptions/
 Abstract
中国城市缺血性卒中的二级预防
Secondary Prevention of Ischemic Stroke in Urban China
Jade W. Wei, BPharm; Ji-Guang Wang, MD; Yining Huang, MD; Ming Liu, MD; Yangfeng Wu, MD; Lawrence K.S. Wong,
MD; Yan Cheng, MD; En Xu, MD; Qidong Yang, MD; Hisatomi Arima, MD; Emma L. Heeley, PhD; Craig S. Anderson,
MD; for the ChinaQUEST Investigators

背景及目的：本文旨在介绍中国城市实施缺血性卒中二级预防的情况。
方法：本研究为一项前瞻性、多中心研究，研究对象为中国2006年入院登记的4782个急性缺血性卒中病例。在出院前以
及卒中后3 个月和12个月时评估二级预防方案的效果。使用Logistic回归分析以明确各种基线变量与住院期间降血压、降
血糖和降血脂治疗的关系，以及12个月后变量与继续治疗之间的关系。
结果：住院期间使用降压(63%)、抗血小板(81%)和降脂(31%)治疗与入院前就已使用这些治疗及并存的心血管危险因素密
切相关，而与卒中的严重性关系并不密切。降压治疗在随访期间一直坚持使用，但抗血小板及降脂治疗的使用在卒中后
12个月下降(分别为66%和17%；P<0.001)，这与病人及医生没有坚持使用有关。
解释：从这次对中国城市住院病人全国性的抽样调查看，大部分病人实施了缺血性卒中的二级预防。但是，抗血小板治
疗及降脂治疗在出院后的使用率大幅度下降，这与医生和病人对潜在的疾病风险认识不足有明显关系。

关键词：中国，预防，卒中
(Stroke. 2010;41:967-974. 何莎 译 张苏明 校)

美国 1998-2007 年住院治疗缺血性卒中的并发症
Medical Complications Among Hospitalizations for Ischemic Stroke in the United
States From 1998 to 2007
Xin Tong, MPH; Elena V. Kuklina, MD, PhD; Cathleen Gillespie, MS; Mary G. George, MD, MSPH, FACS

背景及目的：缺血性卒中常见并发症与死亡率增加以及医疗资源的利用有关。
方法：研究人群来自于1998-2007年全国住院样本(NIS)的医疗费用和资源利用项目(HCUP)中1 150 336名第一诊断为缺血
性卒中的住院病人。使用多元回归分析的方法来检测急性心肌梗死、肺炎、深静脉血栓、肺栓塞、尿路感染的发病率、
住院期间死亡率及住院天数在1998-1999年期间和2006-2007年期间相比的变化。
结果：在2006-2007年期间，第二诊断为急性心肌梗死、肺炎、深静脉血栓、肺栓塞，以及尿路感染的发病率分别为
1.6%、2.9%、0.8%、0.3%和10.1%。以1998-1999年作为参照，2006-2007年住院期间患急性心肌梗死、深静脉血栓、肺
栓塞，以及尿路感染并发症的OR校正值分别为1.39、1.68、2.39、1.18。肺炎的OR值在1998-1999年和2006-2007年没有显
著的变化。2006-2007年住院期间的死亡率较1998-1999年有明显的下降。尽管从1998-1999年到2006-2007年住院总天数有
了明显的下降，但是并发急性心肌梗死、肺炎、深静脉血栓以及肺栓塞的患者住院总天数却没有明显变化。
结论：尽管在研究阶段，住院期间死亡率有所下降，但2006-2007年五种并发症中有四种较8年前更为常见，尤其是深静
脉血栓和肺栓塞发病率有很高的增长。

关键词：脑梗塞，栓塞，流行病学，结果，卒中护理，静脉血栓形成
(Stroke. 2010;41:980-986. 何莎 译 张苏明 校)

62