1524
Brief Reports
9.
10.
11.
12.
13.
14.
15.
Figure 5. In situ hybridization for Epstein-Barr virus (EBV) EBER1 mRNA of a kidney tissue specimen from a patient with acute renal
failure, secondary to EBV-induced infectious mononucleosis; positive
nuclear staining of rare interstitial lymphocytes (arrows) is shown (bar,
25 mm).
16.
agent in the pathogenesis of lymphoproliferative and aproliferative diseases in immune deficient patients. Int Rev Exp Pathol 1985; 27:113–83.
Sainsbury R, Smith PK, LeQuesne G, Davidson GP, Jureidini KF, Moore DJ.
Gallbladder wall thickening with infectious mononucleosis hepatitis in an
immunosuppressed adolescent. J Pediatr Gastroenterol Nutr 1994; 19:123–5.
Yang CW, Pan MJ, Wu MS, et al. Leptospirosis: an ignored cause of acute
renal failure in Taiwan. Am J Kidney Dis 1997; 30:840–5.
Dharnidharka VR, Rosen S, Somers MJG. Acute interstitial nephritis presenting as presumed minimal change nephrotic syndrome. Pediatr Nephrol
1998; 12:576–8.
Ito M, Hirabayashi N, Uno Y, Nakayama A, Asai J. Necrotizing tubulointerstitial nephritis associated with adenovirus infection. Hum Pathol
1991; 22:1225–31.
Shashaty GG, Atamer MA. Hemolytic uremic syndrome associated with
infectious mononucleosis. Am J Dis Child 1974; 127:720–2.
Joh K, Kanetsuna Y, Ishikawa Y, et al. Epstein-Barr virus genome-positive
tubulointerstitial nephritis associated with immune complex-mediated
18.
3.
4.
5.
6.
7.
8.
Burden of Aspergillosis-Related Hospitalizations
in the United States
In the United States in 1996, there were an estimated
10,190 aspergillosis-related hospitalizations (95% confidence interval [CI], 9000–11,380); these resulted in 1970
deaths (95% CI, 1659–2280), 176,272 hospital days (95%
CI, 147,163–206,275), and $633.1 million in costs (95% CI,
$492.0–$780.2 million). The average hospitalization lasted
17.3 days (95% CI, 16.1–18.6) and cost $62,426 (95% CI,
Reprints or correspondence: Dr. Erik J. Dasbach, Merck Research
Laboratories, 10 Sentry Pkwy., BL2-3, Blue Bell, PA 19422 (erikdasbach
@merck.com).
Clinical Infectious Diseases 2000; 31:1524–8
q 2000 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2000/3106-0035$03.00
17.
19.
20.
21.
22.
23.
24.
CID 2000;31 (December)
glomerulonephritis in chronic active EB virus infection. Virchows Arch
1998; 432:567–73.
Muso E, Fujiwara H, Yoshida H, et al. Epstein-Barr virus genome-positive
tubulointerstitial nephritis associated with Kawasaki disease–like coronary aneurysms. Clin Nephrol 1993; 40:7–15.
Ramelli GP, Marone C, Truniger B. Acute kidney failure in infectious mononucleosis. Schweiz Med Wochenschr 1990; 120:1590–4.
Iwama H, Horikoshi S, Shirato I, Tomino Y. Epstein-Barr virus detection in
kidney biopsy specimens correlates with glomerular mesangial injury. Am
J Kidney Dis 1998; 32:785–93.
Friedman BI, Libby R. Epstein-Barr virus infection associated with rhabdomyolysis and acute renal failure. Clin Pediatr (Phila) 1986; 25:228–9.
Woodroffe AJ, Row PG, Meadows R, Lawrence JR. Nephritis in infectious
mononucleosis. Q J Med 1974; 43:451–60.
Arm JP, Rainford DJ, Turk EP. Acute renal failure and infectious mononucleosis. J Infect 1984; 9:293–7.
Mayer HB, Wanke CA, Williams M, Crosson AW, Federman M, Hammer SM.
Epstein-Barr virus–induced infectious mononucleosis complicated by acute
renal failure: case report and review. Clin Infect Dis 1996; 22:1009–18.
Grefer J, Santer R, Ankermann T, Faul S, Nolle B, Eggert P. Tubulointerstitial
nephritis and uveitis in association with Epstein-Barr virus infection. Pediatr Nephrol 1999; 13:336–9.
Lopez-Navidad A, Domingo P, Lopez-Talavera JC, Rabella N, Verger G.
Epstein-Barr virus infection associated with interstitial nephritis and
chronic fatigue. Scand J Infect Dis 1996; 28:185–7.
Bao L, Zhang Y, Zheng X. Detection of Epstein-Barr virus DNA in renal
tissue from patients with interstitial nephritis. Chung Hua Nei Ko Tsa
Chih 1996; 35:542–4.
Kopolovic J, Pinkus G, Rosen S. Interstitial nephritis in infectious mononucleosis. Am J Kidney Dis 1988; 12:76–7.
Michel DM, Kelly CJ. Acute interstitial nephritis. J Am Soc Nephrol 1998;9:
506–15.
Patton DF, Wilkowski CW, Hanson CA, et al. Epstein-Barr virus–determined
clonality in posttransplant lymphoproliferative disease. Transplantation
1990; 49:1080–4.
Becker BN, Fall P, Hall C, et al. Rapidly progressive acute renal failure due
to acyclovir: case report and review of the literature. Am J Kidney Dis
1993; 22:611–5.
Becker BN, Schulman G. Nephrotoxicity of antiviral therapies. Curr Opin
Nephrol Hypertens 1996; 5:375–9.
Frenia ML, Long KS. Methotrexate and nonsteroidal anti-inflammatory
drug interactions. Ann Pharmacother 1992; 26:234–7.
$52,670–$72,181). Although aspergillosis-related hospitalizations account for a small percentage of hospitalizations in the United States, patients hospitalized with the
condition have lengthy hospital stays and high mortality
rates.
Aspergillosis is an opportunistic focal or invasive fungal infection that can occur in the lung, brain, or sinuses of patients
who are immunocompromised or have pulmonary disease, cancer, or HIV infection, or have undergone transplantation [1].
The incidence of invasive aspergillosis in developed countries
has increased because of increases in the use of chemotherapy
for cancer and transplant surgery and as immunosuppressive
therapy for autoimmune disorders [2, 3]. The mortality rate
among patients with untreated invasive aspergillosis has been
estimated to be near 100% [4]. However, national estimates of
CID 2000;31 (December)
Brief Reports
the overall mortality rate, incidence, number of hospitalizations, length of stay, and costs associated with invasive aspergillosis have not been reported for the United States since 1976
[5, 6]. This analysis describes the burden of aspergillosis-related
hospitalizations in the United States for 1996 on the basis of
data from a sample (∼20%) of discharges from community hospitals in the United States.
We estimated the number of discharges, mortality rate, length
of stay, and costs for aspergillosis-related hospitalizations on
the basis of the Nationwide Inpatient Sample (NIS) Release 5
[7]. NIS is part of the HCUP sponsored by the Agency for
Healthcare Research and Quality. NIS Release 5 includes data
on a sample (∼20%) of discharges from community hospitals
in the United States in 1996 from a stratified probability sample
of hospitals and uses sampling probabilities proportional to the
number of US community hospitals in each stratum [7]. Data
in NIS Release 5 were sampled from 19 states in the United
States. The data set contains information on inpatient stays
from 906 hospitals, totaling 6.5 million records in 1996.
We defined an aspergillosis-related hospitalization using the
classifications of International Classification of Diseases, 9th
Revision, Clinical Modification (ICD-9-CM), as follows: any
hospital discharge with a primary or secondary diagnosis of
aspergillosis and infection due to Aspergillus species (diagnosis
code 117.3) or pneumonia in aspergillosis (diagnosis code
484.6). We classified principal diagnoses using the Clinical Classifications for Health Policy Research [8], which combines individual ICD-9-CM codes into broad diagnosis and procedure
groups for statistical analysis and reporting. Total hospital
charges were used to estimate the total costs of hospital care.
The total cost of aspergillosis-related hospital care in 1996 was
estimated as the sum of the total hospital charges incurred for
all hospitalizations for which aspergillosis was listed as a diagnosis. To estimate whether hospitalizations with aspergillosis
as a secondary diagnosis were associated with greater mortality,
longer lengths of stay, and higher costs than hospitalizations
without a secondary diagnosis of aspergillosis, we stratified the
sample into 4 major disease groups associated with aspergillosis:
pneumonia, cancer or leukemia, other respiratory infections,
and HIV infection. These 4 groups accounted for most of the
primary diagnoses in hospitalizations for which aspergillosis
was a secondary diagnosis.
The categories of infection were as follows: “pneumonia”
included patients with a primary diagnosis of pneumonia; “HIV
infection” included patients with a primary diagnosis of HIV
infection; “other respiratory infections” included all patients
discharged with a primary diagnosis of respiratory disease, excluding pneumonia; “cancer or leukemia” included all patients
hospitalized with a primary diagnosis of neoplastic disease; all
the remaining patients hospitalized for other unspecified reasons were grouped into the category “other.”
Weighted means, totals, and percentages were estimated by
use of SAS software [9] and SUDAAN software [10]. Estimates
1525
of SEs and 95% CIs for the weighted means, totals, and percentages were computed by use of SUDAAN release 7.5.2 [10].
The SUDAAN software took into account the sample design
and the sample weighting scheme of the HCUP to produce
appropriate estimates of variability that were based on the study
design. In instances in which data were missing from the estimates of population totals, estimates were adjusted on the basis
of the fraction of missing data in the sample for each variable.
The average age of patients hospitalized with aspergillosis
was 53 years; most (58%) of the patients were men. The demographic characteristics of patients with aspergillosis-related
hospitalizations are summarized in table 1. We estimated that
there were 10,190 aspergillosis-related hospitalizations (95% CI,
9000–11,380) in the United States in 1996. Of those patients
hospitalized with an aspergillosis-related diagnosis, 19.3% or
1970 (95% CI, 1659–2280) died during hospitalization. We also
estimated that there were 176,272 aspergillosis-related hospital
days of care (95% CI, 147,163–206,275) costing $633.1 million
(95% CI, $492.0–$780.2 million) in the United States in 1996.
The mean length of stay and mean total costs per hospitalization for aspergillosis-related hospitalizations were 17.3 days
(95% CI, 16.1–18.6) and $62,426 (95% CI, $52,670–$72,181),
respectively.
Twenty-seven percent (2704; 95% CI, 2368–3041) of the aspergillosis-related hospitalizations were classified as having aspergillosis as the primary diagnosis (ICD-9-CM diagnosis code
117.3). Hospitalizations with an ICD-9-CM diagnosis code of
484.6 could not be reliably estimated given the small number
of cases in the data set (i.e., 42) and thus were not included in
our analyses. Of those patients hospitalized with a primary
diagnosis of aspergillosis, 14.7% or 397 (95% CI, 297–497) died
Table 1. Characteristics of patients with aspergillosis-related hospitalizations in the United States in 1996.
Characteristic
Male
Age, y
<17
18–64
>65
Race
White
Black
Hispanic
Asian
Other
Payer
Medicare
Medicaid
Private insurance or HMO
Other
Region
Northeast
Midwest
South
West
NOTE.
% of patients
Hospitalization rate
per 100,000 population
58
4.5
6
59
35
0.9
3.7
10.4
75
16
6
2
1
3.5
4.9
2.2
2.2
5.8
41
14
39
6
12.0
4.5
2.1
1.4
14
24
36
26
2.8
3.9
4.1
4.2
HMO, health maintenance organization.
1526
Brief Reports
during hospitalization. We also estimated that hospitalizations
with a primary diagnosis of aspergillosis required 36,145 hospital days of care (95% CI, 31,100–41,190) costing $112.5 million (95% CI, $84.8–$140.2 million) in the United States in
1996. The mean length of stay per hospitalization was 13.4
days (95% CI, 12.6–14.2) and the mean total cost per hospitalization was $41,611 (95% CI, $33,275–$49,948). The characteristics of patients hospitalized with primary and secondary
diagnoses of aspergillosis are compared in table 2. Patients hospitalized with aspergillosis as a primary diagnosis tended to be
older, have shorter lengths of stay, have less mortality, and incur
fewer total costs than did patients hospitalized with aspergillosis
as a secondary diagnosis.
Seventy-three percent (7486; 95% CI, 6539–8432) of the aspergillosis-related hospitalizations were classified as having aspergillosis as a secondary diagnosis. Respiratory, neoplastic,
and HIV diseases accounted for most (63%) of the primary
diagnoses in patients who had aspergillosis as a secondary diagnosis. The proportion of patients who died during hospitalization as well as the mean age, length of stay, total costs, and
number of diagnoses by principal diagnosis for those patients
with aspergillosis as a secondary diagnosis are shown in table
3. Hospitalizations with aspergillosis as a secondary diagnosis
were associated with longer hospital stays, higher total costs,
and greater hospital mortality than were hospitalizations with
similar primary diagnoses but without any diagnosis of aspergillosis. These findings are also stratified by primary diagnosis
category in table 3. For example, patients with cancer or leukemia as a primary diagnosis and aspergillosis as a secondary
diagnosis utilized 26 more hospital days, incurred $115,262
more in total costs, and had 4 times the rate of mortality during
hospitalization than did patients with a primary diagnosis of
cancer or leukemia and no secondary diagnosis of aspergillosis.
In 1996, there were an estimated 10,190 aspergillosis-related
discharges from community hospitals in the United States.
These discharges represent 3.8 hospitalizations per 100,000 US
population or 0.03% of all the community hospital discharges
in the United States in 1996. Although aspergillosis-related hospitalizations account for only a small portion of the hospitalizations in the United States, the burden of the disease at the
Table 2.
Burden of aspergillosis-related hospitalizations in the
United States, stratified by principal or secondary aspergillosis diagnosis.
Aspergillosis diagnosis
Primary
(n p 2704)
Variable
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
Mean
Secondary
(n p 7486)
SE
Mean
SE
56.4
7.5
0.9
0.1
51.6
8.8
0.8
0.2
14.7
13.4
41,611
1.4
0.4
4169
21.0
18.8
70,003
1.0
0.8
5859
CID 2000;31 (December)
Table 3. Comparison of hospitalizations in the United States of patients with a secondary diagnosis of aspergillosis with those without
any diagnosis of aspergillosis, stratified by primary diagnosis group.
Hospitalizations of patients with
Secondary diagnosis
of aspergillosis
Primary diagnosis, variable
Cancer or leukemia
No. of hospitalizations
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
Pneumonia
No. of hospitalizations
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
Other respiratory infections
No. of hospitalizations
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
HIV infection
No. of patients
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
Other infections
No. of hospitalizations
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
Overall
No. of hospitalizations
Age of patients, y
No. of diagnoses
Death in hospital, % of
hospitalizations
Length of stay, d
Total costs, $US
Mean
a
1311
52.5
10.5
31.3
33.4
132,405
936
54.4
8.4
19.9
15.0
45,670
1618
58.4
8.2
19.7
14.8
47,941
857
38.5
8.9
21.0
13.9
43,108
2765
50.4
8.5
17.3
17.0
69,723
7486
51.6
8.8
21.0
18.8
70,003
No diagnosis of
aspergillosis
SE
Mean
SE
—
1.3
0.4
1,652,144
62.7
5.3
!0.05
3.4
1.7
15,573
7.3
7.1
17,143
—
1.9
0.2
1,233,702
59.1
5.7
2.2
1.1
7275
6.3
6.6
12,164
—
—
!0.05
0.1
0.2
396
—
0.4
!0.05
0.1
0.1
172
1.0
0.2
1,941,770
52.5
4.9
1.7
0.9
4111
5.6
5.7
11,730
0.1
0.1
251
0.8
0.2
127,906
37.6
6.7
—
0.2
0.1
2.4
1.2
4388
10.9
9.3
18,170
0.3
0.2
516
—
—
—
0.6
!0.05
0.8
0.2
29,902,691
45.5
4.5
—
0.3
!0.05
1.3
0.7
5719
1.9
4.8
10,127
!0.05
!0.05
162
0.8
0.2
34,864,214
47.1
4.6
—
0.3
!0.05
1.0
0.8
5859
2.5
5.0
10,647
!0.05
—
!0.05
165
a
Sample size is an estimate for the US population. The overall number of
hospitalizations may not add up precisely due to rounding error and missing
data.
patient level is significant. For example, we found that 19.3%
of those patients hospitalized with an aspergillosis diagnosis
died during hospitalization. In contrast, the overall mortality
rate among those patients hospitalized without a diagnosis of
aspergillosis in the United States was 2.5%. Patients hospitalized with aspergillosis also required on average 17.3 days of
CID 2000;31 (December)
Brief Reports
care incurring $62,426 in costs. However, patients without an
aspergillosis diagnosis required on average 5.0 days of care
incurring $10,647 in costs. Thus, an approximation of the excess
length of hospitalization and cost due to aspergillosis would be
12.3 days and $51,779 per patient or 125,337 hospital days and
$528 million in costs per year, respectively.
The most recent estimates of the number of aspergillosisrelated hospitalizations for the United States are from 1976 [5].
In the 1976 analysis, Fraser et al. [5] reported that the incidence
of aspergillosis-related hospitalizations increased 12-fold from
1970 to 1976; these researchers used discharge abstracts from
1875 hospitals participating in the Professional Activity Study
of the Commission on Professional and Hospital Activities.
Fraser et al. [5] also suggested that the incidence would continue
to increase in the ensuing years. Although the sample of hospitals in the Commission on Professional and Hospital Activities data set differs from the sample of hospitals in the HCUP
data set, our 1996 estimates for the United States appear to be
consistent with this forecast, in that the numbers of aspergillosis-related hospitalizations and hospital days appear to have
increased ∼8-fold in the 20 years since the report of Fraser et
al. [5]. This increase was relatively larger (i.e., approximately
double) for those patients hospitalized with aspergillosis as a
secondary diagnosis than for those patients hospitalized with
a primary diagnosis of aspergillosis.
Overall, the increase in hospitalizations for aspergillosis also
appears to be consistent with other epidemiological studies of
fungal disease in general for the 1980s. According to the National Nosocomial Infections Surveillance system, the incidence
of nosocomial fungal infections escalated from 1980 to 1990
[11, 12]. Moreover, an analysis of data from the National Hospital Discharge Survey from 1980 to 1994 found that the incidence of hospitalizations due to fungal disease increased at
an annual rate of 5.7% per year [13]. Another measure that
appears to be increasing since 1976 is the mortality rate associated with an aspergillosis-related hospitalization. Specifically, the mortality rate reported for 1976 (14.3%) increased
31% to 19.3% in 1996. This rise is consistent with a descriptive
study of mortality due to infectious diseases in the United
States, which reported that the mortality rate associated with
fungal infections in general increased in the United States from
1980 to 1992 [14].
Given the lack of recent estimates on the number of aspergillosis-related hospitalizations in the United States, the NIS
Release 5, an administrative data set, offered a unique opportunity to update the previous estimates. The strength of the
HCUP database is that it can indicate the scope of the problem
across the spectrum of hospitals in the United States. This is
useful for understanding the scope, magnitude, and temporal
aspects of a disease on a population basis; institution-based
studies, in contrast, more narrowly characterize the burden.
However, as with any administrative database, there are recognized limitations [15, 16]. One limitation is that the data only
1527
reflect what happens during a hospitalization. Hence, one can
only assess the impact of a disease on mortality during the
hospitalization, which make it difficult to compare results with
mortality rates observed in studies of similar patients followed
for longer than the period of hospitalization. Similarly, because
patients are not individually identifiable, one cannot follow
patients longitudinally and determine whether the admission is
a readmission or interhospital transfer for the same medical
condition.
A second limitation is the lack of clinical data in the database,
which can affect the interpretation of the data in a number of
ways. The lack of clinical data makes it difficult to adjust for
severity differences when comparing outcomes among different
patient groups. For example, we estimated that the crude unadjusted excess length of hospitalization and cost due to aspergillosis was fairly substantial. Unfortunately, these estimates
may be confounded by other underlying differences not available in the data set at both the individual and institutional level
between those patients hospitalized with and without a diagnosis of aspergillosis. As a result, we did not attempt to estimate
a more precise difference. Although the sensitivity of ICD-9CM coding is generally high for serious uncommon illnesses,
the lack of precision may also affect the specificity of an aspergillosis diagnosis assigned for the hospitalization. In other
words, some patients may be classified as having aspergillosis
on the basis of culture evidence only, without tissue specimens
documenting invasive disease; we cannot even be certain that
such patients have clinical manifestations of disease. Hence, the
burden of disease for these patients may be significantly less
than that for patients classified on the basis of clinical evidence
of aspergillosis. This may partially account for the mortality
rates we determined, which are lower than the rates reported
in other studies.
Another limitation of administrative data sets is the accuracy
of the coding, because administrative data are collected for
financial and administrative purposes and not for research initiatives. Hence, diagnostic coding can be affected by incentives
to maximize hospital payments. The coding can also be affected
by whether the hospital is attuned to infectious diseases such
as aspergillosis or by the rigor of the process used to verify the
coding. Finally, all of these factors affecting coding can vary
from hospital to hospital. The overall effect of coding accuracy
on the under- or overreporting of aspergillosis is unfortunately
unknown. Nonetheless, given the lack of recent information on
the burden of aspergillosis in the literature, we believe that our
estimates provide a useful update from which to understand
the magnitude and changes of the burden of aspergillosis in
patients in the United States.
The number of aspergillosis-related hospitalizations in the
United States appears to have significantly increased during the
past 20 years. Although aspergillosis-related hospitalizations
account for a small percentage of hospitalizations in the United
States, patients hospitalized with aspergillosis continue to have
1528
Brief Reports
lengthy hospital stays and high mortality rates. Development
and implementation of more rapid and accurate diagnostic tests
as well as safe and effective therapies may reduce the burden
of this condition.
Erik J. Dasbach, Glenn M. Davies,
and Steven M. Teutsch
Merck Research Laboratories, Blue Bell, Pennsylvania
References
1. Fekety R. Immunologic disorders and infectious diseases: aspergillosis. In:
Thoene JG, Coker NP, eds. Physicians’ guide to rare diseases. 2d ed.
Montvale, NJ: Dowden Publishing Co., 1995.
2. Denning DW. Therapeutic outcome in invasive aspergillosis. Clin Infect Dis
1996; 23:608–15.
3. Manuel RJ, Kibbler CC. The epidemiology and prevention of invasive aspergillosis. J Hosp Infect 1998; 39:95–109.
4. Denning DW. Invasive aspergillosis. Clin Infect Dis 1998; 26:781–805.
5. Fraser DW, Ward JI, Ajello L, Plikaytis BD. Aspergillosis and other systemic
mycoses: the growing problem. JAMA 1979; 242:1631–5.
6. Centers for Disease Control and Prevention, NCIDOD, DBMD disease listing: aspergillosis. Available at: www.cdc.gov/ncidod/dbmd/diseaseinfo
/aspergillosis_t.htm. Accessed May 1998.
7. Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sam-
CID 2000;31 (December)
ple (NIS) Release 5 1996. Rockville, MD: Agency for Health Care Policy
and Research, 1999.
8. Elixhauser A, McCarthy EM. Clinical classifications for health policy research, version 2: hospital inpatient statistics. Healthcare Cost and Utilization Project (HCUP-3) Research Note Series. Rockville, MD: Agency
for Health Care Policy and Research, 1996 AHCPR publication no 960017.
9. SAS release 6.12. Cary, NC: SAS Institute, Inc., 2000.
10. SUDAAN release 7.5. Research Triangle Park, NC: Research Triangle Institute, 2000.
11. Centers for Disease Control and Prevention NNIS System. National Nosocomial Infections Surveillance (NNIS) report, data summary from October 1986-April 1997. Am J Infect Control 1997; 25:477–87.
12. Beck-Sague C, Jarvis WR. Secular trends in the epidemiology of nosocomial
fungal infections in the United States, 1980–1990. National Nosocomial
Infections Surveillance System. J Infect Dis 1993; 167:1247–51.
13. Simonsen L, Conn LA, Pinner RW, Teutsch SM. Trends in infectious disease
hospitalizations in the United States, 1980-1994. Arch Intern Med
1998; 158:1923–8.
14. Pinner RW, Teutsch SM, Simonsen L, et al. Trends in infectious diseases
mortality in the United States. JAMA 1996; 275:189–93.
15. Iezzoni LI. Assessing quality using administrative data. Ann Intern Med
1997; 127:666–74.
16. Paul JE, Tilson HH. Use and opportunities for administrative data bases in
pharmacoeconomic research. In: Spilker B, ed. Quality of life and pharmacoeconomics in clinical trials. Philadelphia: Lippincott-Raven, 1996.