PIN21 THE CLINICAL AND ECONOMIC BURDEN OF COMPLICATED SKIN AND SKIN STRUCTURE INFECTIONS DUE TO STAPHYLOCOCCUS AUREUS: FINDINGS FROM A NATIONAL DATABASE


A165Abstracts

Analyses were conducted from both the societal and health care
system perspective. Sensitivity analyses were performed.
RESULTS: Reduction in disease outcomes, disease sequelae and
cost-of-illness by health state was observed in the time period
post-Prevnar®, across all age groups. The total cost of the vac-
cination program to the Canadian health care system (including
herd immunity effects), from a payer perspective amounted to
$74,682,790; this decreased to $46,197,274 from a societal per-
spective. The total number of illnesses avoided was 86,164. The
incremental cost-effectiveness ratio (ICER) was $28,551 and
$17,661 per additional QALY from the health system and soci-
etal perspectives, respectively. When herd immunity effects were
excluded from the analysis, the ICER increased to $166,560 and
$115,995 per QALY, respectively. Sensitivity analysis indicated
that total cost and ICER results were most sensitive to changes
in the epidemiology and cost of otitis media. However, these
changes did not considerably impact the results, indicating a
robust model. CONCLUSION: Consistent with previous find-
ings, vaccination with Prevnar® is cost-effective. Administration
of Prevnar® results in a substantial reduction in pneumococcal
disease in vaccinated children and unvaccinated adults.

PIN19
COST-EFFECTIVENESS OF GARGLING FOR PREVENTION OF
UPPER RESPIRATORY TRACT INFECTIONS
Shimbo T1, Omata K1, Takahashi Y1, Satomura K2, Kitamura T3,
Kawamura T2
1International Medical Center of Japan, Shinjyuku-ku, Tokyo, Japan,
2Kyoto University, Kyoto, Kyoto, Japan, 3Osaka Saiseikai Senri Hospital,
Suita, Osaka, Japan
OBJECTIVES: To investigate the cost effectiveness of gargling to
prevent upper respiratory tract infections (URI) from a societal
perspective. METHODS: The effectiveness of gargling for pre-
venting URI has been demonstrated in a randomized controlled
trial in which the participants recorded the frequency of gargling,
incidence and severity of URI and duration of daily medicine.
Costs of gargling, visiting physicians, medicine, and lost pro-
ductivity were considered. The cost of gargling was estimated as
the opportunity cost of the time required. The utility of severe
and moderate URI was also considered. Average costs and utility
during 60 days of observation in the trial were estimated as the
sum of the average daily cost and utility of the participants
remaining staying in the trial. The incremental cost effectiveness
ratio (ICER) of gargling when compared with the absence of gar-
gling was calculated, and bootstrap sampling generated an
acceptability curve. RESULTS: The estimated unit cost of gar-
gling was 49.2 yen. Assigned participants gargled 4.5 times per
day on average. The gargling group had higher costs and utility
than the group that did not gargle. The incremental cost and
effectiveness for 60 days were 4750 yen and 0.43 quality-
adjusted life days respectively. The gargling group required 8020
yen more for gargling, but saved 3270 yen by preventing URI
for 60 days. This showed that the ICER of gargling was 4.07
million yen/QALYs (34,400 US$/QALYs). The acceptability
curve showed 67.1% was less than 6 million yen/QALYs, and
88.7% less than 12 million yen/QALYs. CONCLUSION:
Although it can prevent URI, gargling is more costly than not
gargling because the cost of gargling exceeded the savings
derived from URI prevention. However, the ICER of gargling
was comparable with that of other widespread medical 
technologies.

PIN20
PHARMACOECONOMIC ANALYSIS OF SEVERE COMMUNITY-
ACQUIRED PNEUMONIA TREATMENT
Kulikov A, Krysanov I
Moscow Medical Academy, Moscow, Russia
OBJECTIVES: Selection of the most cost-effective treatment
regimen of severe community-acquired pneumonia. METHODS:
Direct medical expenditure on courses of treatment with lev-
ofloxacin and ceftriaxone were evaluated during this trial. “Cost
minimization” analysis was chosen as a pharmacoeconomic
method, and a “lost opportunities” index was calculated when
using less cost-effective drugs. RESULTS: Direct medical costs in
the group of patients who received levofloxacin amounted to
16,097.99 rubles, and in the group of patients who were treated
with ceftriaxone they totaled 32,573.47 rubles per patient. They
were made up of levofloxacin and ceftriaxone antibacterial drug
treatment costs and the costs of patients’ hospital stay. The cost
of the drug treatment course amounted to 3,997.99 rubles for
the first group (levofloxacin) and to 19,073.47 rubles for the
second group (ceftriaxone); the cost of hospital stay amounted
to 12,100 rubles and 13,500 rubles respectively. In the break-
down of expenditure on treatment of community-acquired pneu-
monia with levofloxacin, patients’ hospital stay accounted for
75% of expenses, whereas drug treatment accounted for only
25% thereof; when treating with ceftriaxone, the expenditure on
patients’ hospital stay amounted to 40% and that on drug treat-
ment—to 60%. The “lost opportunities” index equaled one and
thus indicated that when using a more cost-effective drug (lev-
ofloxacin) for the treatment of one patient compared to a less
cost-effective drug (ceftriaxone) it is possible to theoretically
treat an additional patient, taking into account the difference in
the costs of treatment with the drugs compared, provided the
profile of antibiotic resistance is congruent with that under the
conditions of the clinical study used herein. CONCLUSION:
Antibacterial treatment of severe community-acquired pneumo-
nia with levofloxacin is more cost-effective, enabling the reduc-
tion of costs by 16,475 rubles per patient compared to treatment
with ceftriaxone owing to lower expenditure on drugs.

PIN21
THE CLINICAL AND ECONOMIC BURDEN OF COMPLICATED
SKIN AND SKIN STRUCTURE INFECTIONS DUE TO
STAPHYLOCOCCUS AUREUS: FINDINGS FROM A 
NATIONAL DATABASE
Marton JP1, Menzin J2, Sussman M2, Friedman M2, Philburn RT1,
Rothermel CD1
1Pfizer Global Pharmaceuticals, Pfizer Inc, New York, NY, USA,
2Boston Health Economics, Inc, Waltham, MA, USA
OBJECTIVES: Complicated skin and skin structure infections
(cSSSIs) are a common complication among hospitalized patients.
There are limited national data on the costs of cSSSIs due to
Staphylococcus aureus, a common hospital-acquired pathogen.
METHODS: This retrospective cohort study used data from the
2004 Health care Cost and Utilization Project Nationwide Inpa-
tient Sample (HCUP-NIS). Patients with S. aureus cSSSIs were
identified based on ICD-9-CM diagnosis codes and compared to
patients without skin infections. Excess mortality, length of stay
(LOS), and costs were estimated for both groups. Multivariate
models (with log transformation) were used to adjust costs for
potential confounding factors, including age, gender, mortality,
hospital region, and comorbidity. RESULTS: We identified 55,585
hospitalized patients with cSSSIs due to S. aureus. The compari-
son cohort consisted of 7,618,776 patients without skin infec-
tions. The mortality rates were similar for the S. aureus cSSSI and
comparison cohorts (3.9% and 2.0%, respectively). For com-



A166 Abstracts

parison purposes, the mortality rate of all inpatients with S. aureus
infections was 7.3%. Patients with S. aureus cSSSIs were older
(mean age 54 years vs. 46 years for the comparison group) and
were more likely to have congestive heart failure, diabetes with
chronic complications and bacteremia/septicemia. Relative to the
comparison group, patients with S. aureus cSSSIs had significantly
(P < 0.0001) longer mean length of stay (9.7 vs. 4.4 days) and
higher average costs per stay ($16,941 vs. $9,154). After con-
trolling for potentially confounding factors, the excess mean costs
associated with S. aureus cSSSIs were estimated to be $3,396.
CONCLUSION: Our findings suggest that the clinical and eco-
nomic burden of complicated skin and skin structure infections
(cSSSIs) due to Staphylococcus aureus among hospitalized
patients is substantial.

PIN22
IMPACT OF S. AUREUS INFECTIONS ON EXPENDITURES AND
LENGTH-OF-STAY IN U.S. HOSPITALS
Noskin G1, Rubin R2, Schentag J3, Kluytmans J4, Hedblom E5,
Jacobson C5, Smulders M6, Gemmen EK7, Bharmal M7
1Northwestern University Feinberg School of Medicine, Chicago, IL,
USA, 2Georgetown University, Washington, DC, USA, 3University at
Buffalo School of Pharmacy, Buffalo, NY, USA, 4Amphia Hospital and
VUmc Medical University, Amsterdam, 4800 RK, The Netherlands,
53M Medical, St Paul, MN, USA, 6Quintiles, Falls Church, VA, USA,
7Quintiles Strategic Research & Safety, Falls Church, VA, USA
OBJECTIVES: Evaluate the incremental impact of S. aureus
infection stays on hospital charges and length-of-stay in U.S.
hospitals in 2003. METHODS: The 2003 Nationwide Inpatient
Sample data were analyzed. Hospital discharges were classified
as either a S. aureus–related infection stay or a non-S. aureus-
related infection stay using a combination of several ICD-9
codes. Incremental effect of S. aureus infection on hospital
charges and length-of-stay was estimated using multivariate
regression models adjusting for hospital fixed effects and patient
variables including age, gender, race, payer, diagnosis-related
grouping and concomitant conditions including diabetes, dialy-
sis and lung disease. RESULTS: S. aureus infection was reported
as a discharge diagnosis for 1.0% of all hospital inpatients, or
389,963 stays, in 2003. S. aureus infection hospital stays were
significantly more likely among male, older patients, stays that
were paid by Medicare, white or non-Hispanics and hospital
stays among individuals with diabetes, lung disease or dialysis.
After adjusting for covariates, the mean incremental impact of
S. aureus infection on hospital charges and length-of-stay was
$37,251 (95% Confidence Interval (CI): $34,678–$39,823) and
8.2 days (95% CI: 7.9–8.5) among all inpatient stays, $40,637
(95% CI: $37,683–$43,591) and 9.2 days (95% CI: 8.8–9.6)
among surgical stays, $83,952 (95% CI: $75,853–$92,052) and
16.8 days (95% CI:15.7–17.9) among invasive cardiovascular
stays, $34,202 (95% CI: $29,612–$38,791) and 9.6 days (95%
CI: 9.0–10.2) among invasive orthopedic stays and $119,292
(95% CI: $106,209–$132,374) and 19.8 days (95% CI:
17.5–22.2) among invasive neurosurgical stays. CONCLUSION:
S. aureus infections present a considerable economic burden to
U.S hospitals. Based on the prevalence of S. aureus infection and
its incremental impact, the total economic impact of S. aureus
among all hospital admissions was estimated at $14.5 billion in
2004 U.S. dollars.

PIN23
DETERMINANTS OF TOTAL HOSPITAL COSTS AMONG
INPATIENTS WITH CANDIDEMIA
Lee SP1, Bunz T2, Gagne JJ2, Maio V2, Goldfarb NI2, Horn DL2,
Swanson R3
1Thomas Jefferson Univeristy, Philadelphia, PA, USA, 2Thomas Jefferson
University, Philadelphia, PA, USA, 3Pfizer Inc, New York, NY, USA
OBJECTIVES: To identify factors associated with total hospital
costs among patients with candidemia in a large, urban, tertiary
care, teaching hospital. METHODS: This retrospective cohort
study examined hospitalizations during calendar years 2004 and
2005 at Thomas Jefferson University Hospital in Philadelphia,
PA, using data from various hospital systems. Candidemia cases
were identified via the microbiological laboratory information
system as those patients with at least one confirmed positive
blood culture for any Candida species at any point during the
study period. Demographic, economic, and clinical data, includ-
ing length-of-stay (LOS), payer types, total costs, and Diagnosis
Related Group (DRG) assignments were collected from the hos-
pital cost accounting system. Pharmacy data (i.e. medications
administered and associated costs) were retrieved from the phar-
macy electronic information system. A multivariate regression
analysis, using the natural logarithm of total hospitalization
costs as the dependent variable, was conducted. Independent
variables included demographic information, relative DRG
weights, and Candida species. RESULTS: Among 68,526 total
hospitalizations during the study period, 287 cases were con-
firmed positive for candidemia. The mean age of cases was 58
years, 52% were female, and 60% were Caucasian. The mean
LOS was 43.3 days and the average inpatient cost for candidemia
cases was $130,759 (SD = $116,560; median, $97,869). The
most common Candida species was C. albicans (n = 127; 44%).
The most commonly used antifungal treatment was fluconazole
(N = 176, 61%). Age, and relative DRG weights (p < 0.05) were
significantly positively correlated with total hospital costs. Older
patients with higher relative DRG weights were associated with
the higher total costs. Race, gender, and marital status were not
associated with total costs. CONCLUSION: Relative DRG
weights, as well as age are associated with total hospital costs
among patients with candidemia. Candidemia is expensive to
treat and results in lengthy hospital stay. Early detection and
treatment may significantly reduce resource use as well as
improve outcomes.

PIN24
THE ECONOMIC IMPACT OF METHICILLIN RESISTANCE IN
STAPHYLOCOCCUS AUREUS BACTEREMIA IN KOREA
Park EJ1, Lee EK2, Chae S3
1Korea Institute for Health and Social Afffairs, Seoul, South Korea,
2Sookmyung Women’s University, Seoul, South Korea, 3Korea Institute
for Health and Social Affairs, Seoul, South Korea
OBJECTIVES: The objective of this study is to examine the eco-
nomic outcome of Methicillin resistance in Staphylococcus
aureus bacteremia in Korea, where MRSA is endemic in major-
ity of hospitals. METHODS: We conducted retrospective case-
control study of patients admitted to three university-based
teaching hospitals in Seoul, Korea in 2005. Cases were defined
as patients with Methicillin-resistant S. aureus (MRSA) bac-
teremia and controls were Methicillin- susceptible S. aureus
(MSSA) bacteremia selected according to a priori matching cri-
teria. 58 cases and 58 controls were identified. Hospital charges
were collected from hospitals’ billing system. RESULTS: The
median hospital charge after the development of bacteremia was
higher for cases with MRSA bacteremia ($8245) than for con-
trols with MSSA bacteremia ($6569). The median hospital