|
Items in cart [0] |
Questions? Call 888-624-8373 |
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 322
For-Profit Enterprise in Health Care. 1986.
National Academy Press, Washington, D.C.
Hospital Ownership and
Comparative Hospice Costs
Craig G. Coelen
INTRODUCTION
Proprietary hospitals have become increas-
ingly common in the United States in the past
decade, and this trend has provoked a debate
about the relative cost of patient care in pro-
pnetary versus nonprofit hospitals. Of partic-
ular concern have been the implications of the
growth of proprietary chains, through con-
struction and acquisition. Advocates of pro-
prietary ownership point to the advantages of
economies of scale and the profit-related in-
centive for efficiency. Detractors raise con-
cerns about high costs of capital, high charge/
cost ratios, and limitations on access for charity
cases and Medicaid beneficiaries.
Previous empirical studies of comparative
costs among different types of hospitals do not
provide uniform conclusions. Ermann and Ga-
be} (1986) review a dozen studies and conclude
that "the consensus . . . is that [proprietary
and nonprofit chains] increase the cost of care
frelative to independent hospitals]. The con-
clusion holds whether costs are measured as
hospital expenses, revenues, or charges, on a
per admission or per diem basis." Because half
of the studies indicate that proprietary hos-
pitals and/or proprietary chains are less ex-
pensive or, at worst, no more expensive than
nonprofit hospitals, it is our view that there is
no consensus in the existing evidence.
The 12 studies reviewed by Ermann and
Gabel diner widely with respect to measures
of cost, research samples, analytic techniques,
and controls for possible differences in case
max. Most studies compare costs among hos-
pitals in only one or a few states and warn
readers about the danger of generalizing their
Dr. Coelen is with Abt Associates, Inc., Cambridge,
Massachusetts.
322
results to other areas. In most cases, the sam-
ple sizes are so small that multivariate statis-
tical techniques cannot be used to account for
the influence of cost determinants other than
ownership. Because case mix is not accounted
for explicitly in any of the studies, the reader
is left with conclusions of the form: "Unless
the patients admitted to proprietary hospitals
are sicker (less sick) than those admitted to
nonprofit hospitals, proprietary hospitals are
more expensive (less expensive) than similar
nonprofit institutions."
Reexamining the relationship between hos-
pital ownership and the cost of hospital care,
we hope to improve upon earlier research in
several ways. First, we use data for a gener-
alizable sample of community hospitals in the
48 contiguous states and the District of Co-
lumbia. Second, by presenting the dispersion
of cost per patient among hospitals within each
group and then comparing the distributions
across groups, we are able to assess the relative
homogeneity of different groups. Third, by ex-
plicitly adjusting for cross-sectional differences
in case mix in our analysis, we are able to limit
the degree to which we must qualif y our con-
clusions about comparative costs.
We confine our discussion of interhospital
differences to the following set of indicators of
hospital performance:
· Level and annual rate of change of ex-
pense per discharge
· Average length of stay
· Level and annual rate of change of num-
ber of discharges
· Total, routine, and ancillary charges per
discharge and
· Margins on patient revenue and total rev-
enue.
Although we do not provide detailed statis-
tical results, we have examined other indica
OCR for page 323
HOSPITAL OWNERSHIP AND COMPARATIVE COSTS
tors during our analysis and will refer to the
findings as relevant.
DATA
Data are drawn from information compiled
for a large evaluation of hospital prospective
reimbursement programs for the Health Care
Financing Administration (HCFA). The prin-
cipal facts about our data are the following:
· Measures of hospital expense, utilization,
and margins are drawn from hospitals' Medi-
care cost reports, although information from
American Hospital Association (AMA) annual
surveys has been used extensively to test for
errors in coding and keypunching.
· Measures of total, routine, and ancillary
charges per discharge are drawn from a 20
percent sample of Medicare patients with se-
lected medical problems.2 Comparing total
charges per patient from this source with hos-
pital-wide average charges per patient from
cost reports, we find that Medicare-only charges
accurately reflect differences in gross patient
service revenue across groups of hospitals.
~ Hospital expense has been divided by a
uniquely derived measure of adjusted dis-
charges. The adjustment for outpatient ser-
vices and inpatient days in skilled nursing facility
(SNF) units is accomplished by regressing the
annual rate of change of expense on annual
rates of change of acute care discharges, length
of stay, SNF days per acute care discharge,
outpatient visits per acute care discharges, and
a variety of variables measuring the charac-
teristics of hospital catchment areas. Regres-
sion coefficients from this short-term cost
function are then used to convert SNF days
and outpatient visits to equivalent values of
acute care discharges.
RESEARCH SAMPLE
Unlilce the samples used in most other stud-
ies of comparative hospital costs, our sample
is representative of the 48 contiguous states
and the District of Columbia. The sample was
selected as follows:
~ A 25 percent simple random sample of
community hospitals was selected. Specialty
hospitals, those run by federal or state agen
323
cies, and those with median values of average
length of stay from 1969 to 1981 in excess of
15 days were excluded.
~ All remaining eligible hospitals in He 15
states with prospective reimbursement pro-
grams of interest to HCFA were added to the
proportional sample. The states were Arizona,
Colorado, Connecticut, Indiana, Kentucky,
Maryland, Massachusetts, Minnesota, Ne-
braska, New Jersey, New York, Pennsylvania,
Rhode Island, Washington, and Wisconsin.
· For this study, hospitals run by local gov-
ernment agencies and hospitads with mem-
bership in the Council of Teaching Hospitals
were excluded.
· For analyses of hospital expense, utili-
zation rates, and charges, data are available for
each year from 1975 to 1981. The analyses of
hospital margins are based on data from 1975
to 1979.
Table 1 displays sample size by type of hos-
pital and year. Because the sample over-rep-
resents states with prospective reimbursement
programs (primarily in the northeast), all ma-
jor analyses have been redone for a strictly
proportional sample. Because findings about
comparative cost across types of hospitals did
not change when the proportional sample was
used, only results for the complete sample are
reported here.
ANALYTIC METHODS
lbe method used to estimate costs, charges,
utilization rates, and margins among groups of
hospitals is multiple regression analysis. This
technique computes intergroup differences that
are adjusted for geographic location, catch-
ment area charactenstics, presence/absence of
regulatory programs (prospective reimburse-
ment, certificate-of-need programs, and bind-
ing professional standards review organization
review [PSRO]), bed size, and case mix. Once
the influence of these other potential deter-
minants of hospital behavior has been taken
into account, residual~ifferences among groups
of hospitals can be reliably attributed to dif-
ferences in management.
Table A. 1 in the appendix provides a list of
all dependent and explanatory variables in-
cluded in multiple regression models and pre
OCR for page 324
324
FOR-PROFIT ENTER[R1SE IN HEALTH CARE
TABLE 1 Sample Size by Type of Hospital and Yeara
Proprietary Hospitals Nonprofit Hospitals
_ All Four
Year Chain Independent Chain Independent Types
_
975 92 188 419 1,083 1,782
976 96 170 419 1,073 1,758
977 100 157 424 1,065 1,746
978 100 141 424 1,046 1,711
979 104 132 423 1,020 1,679
980 108 128 424 1,004 1,664
981 108 129 418 982 1,637
ache sample excludes hospitals that opened after 1977; sample size by 1981 is
lower by about 60 hospitals third would otherwise be the case. The sample size
decreases over time as the result of hospital closures.
sents means and standard deviations for the
set of observations used in analyses. The ex
planatory variables are
· Six dummy variables for the year to which
data apply; these constitute a more flexible
specification of secular influences than a sim
ple linear time trend (dummy variable for 1975
is omitted).
· Three dummy variables for type of own
ership; since a dummy variable for indepen
dent (nonchain) nonprofit hospitals is omitted,
all three coefficients for ownership variables
measure differences with respect to this omit
ted, and most common, category of hospital.
· The logarithm of the hospital's Medicare
case-mix index for 1982, propagated to all other
years, as a rough measure of differences in case RESULTS
type across hospitals (but not across years).
· Four dummy variables representing ur
ban, nonurban, and regional location of hos
pitals;
· Three dummy variables indicating pres
encelabsence of regulatory programs for each
hospital year prospective reimbursement,
binding PSRO utilization review, and certifi
cate of need; and
~ Sixteen continuous variables to control for
relevant characteristics of the catchment areas
(counties) served by each hospital socioeco
nomic characteristics of the population, eco
nomic conditions, insurance coverage, and
availability of physicians and nursing home
beds.
For selected measures of hospital perfor
mance we present comparative histograms to
indicate differences in distributions across the
four groups of hospitals. Although multiple
regression analyses provide estimates of dif-
ferences in adjusted means and indicate the
statistical significance of Hose differences, they
Lo not provide an adequate indication of dis-
persion within a group and relative dispersion
across groups. Ibe comparative histograms
complement the results of regression analyses
by indicating the degree to which statistically
significant differences in means are typical of
most hospitals or are the result of large dif-
ferences between a minority of hospitals in
each group (so called outliers).
Table 2 presents estimated differences across
the four groups of hospitals for 10 indicators
of performance. These estimates, derived from
multiple regression analyses reported in the
appendix, are adjusted for differences in lo-
cation, catchment area characteristics, pres-
ence/absence of regulatory programs directed
at hospitals, bed size and case mix. Figure 1,
comparing the dispersion of case mix among
hospitals across the four groups, illustrates the
degree to which these potential determinants
of hospital behavior vary across We four groups.
Proprietary chain hospitals, for example, are
concentrated primarily in southern and west-
ern states, whereas other types of hospitals are
more equally distributed geographically. In-
dependent proprietary hospitals tend to have
OCR for page 325
HOSPITAL OWNERSHIP AND COMPARATIVE COSTS
TABLE 2 Estimated Differences in Cost, Utilization, Charges, and Margins Among
Hospitals: Summary of Results from Regression Analyses
Difference Relative to Independent Nonprofit Hospitals
325
Proprietary Proprietary Nonprofit
Performance Measurea Chains Independents Chains
Total expense per adjusted
discharge (% difference) s.8b- 4.2b1.gb
Annual rate of change of
expense per case
(percentage point
difference in rate of
growth) -0.4- 1.2b-0.2
Average length of stay
(% difference) 0.22.2b1.3b
Number of discharges
(% difference) - 7. 7b-2. 32. 2b
Annual rate of change of
number of discharges
(percentage point
difference in rate of
growth l.ob0.3o.5b
Medicare charges/case
(% difference) 1g.obll.2b3.5b
Medicare routine charges/
case (% difference) -1.27.4b2.2b
Ancillary charges per case
(% difference) 33 gb14.5b5.1b
Margin on patient revenue
(percentage point
difFerence) 5.5b3.3b- o.6b
Margin on total revenue
(percentage point
difFerence) 1.8bl.lb-o.6b
CEstimates in each of the 10 rows of the table are derived from Appendix tables A.2 through A. 11. All measures
except Medicare charges per case apply to the entire hospital and all patients.
b probability that measured differences are due to chance is less ~an 5 percent.
fewer beds than other types of hospitals and
a higher frequency of low-intensity case mix.
Only those differences in performance indi-
cators across groups that are not associated
with differences, location, case mix, and other
factors are re~ected in our measures of di~er-
ences due to ownership.
Experience per Discharge
Total hospital expense per adjusted dis-
charge varies by as much as 10 percent among
the average hospitals in each ofthe four groups,
or by as much as $250 per case in 1981 dolIars.2
Independent hospitals are less expensive than
chainmperated hospitals, and a~nong indepen-
dent institutions, proprietary hospitals are
percent less expensive than nonprofit hospi-
tals. Nonprofit chains are 2 percent more ex-
pensive, on average, than independent
nonprofit institutions. Proprietary chains are
the most expensive of all four groups 6 per-
cent more costly than nonprofit independents
and 10 percent more expensive than proprie-
tary independents. Our results clearly dispute
the presumed advantages of chain-operated
hospitals in general, and proprietary chains in
OCR for page 326
326
FOR-PROFIT ENTERPRISE IN HEALTH CARE
40
en
0 30
o
At
mu
C:
20
10
o
~ Proprietary Chains
_ .
Proprietary Independents
Non-Profit Chains
Non-Profit Independents
it
-
Be_
_
<90 90-95 95-1 00
_ , ,
100-105 105-110 >110
PERCENT OF NATIONAL AVERAGE
E]GURE 1 Comparative dispersion of Medicare case-mix index (year 1980~.
particular, from economies of scale and profit-
related incentives for minimization of cost.3
In addition to providing care at the most
economical cost, independent proprietary hos-
pitals have been more effective than other hos-
pitals in slowing the annual rate of increase of
costs. There are no statistically significant dif-
ferences in annual rates of increase of expense
per case among average hospitals in the other
three of the four groups. Among independent
proprietary hospitals, average annual in-
creases were 1.2 percentage points lower than
was the case among over hospitals. Over a 7-
year period (1975 through 1981, the span of
our data), this lower annual rate of grown pro-
duces a cumulative compound savings of 8 per-
cent, large enough by itself to account for the
4 percent average difference in the level of
cost between proprietary independents and
nonprofit independents.
Figure 2 presents the comparative disper-
sion of expense per ease among the four groups
of hospitals. The comparative histograms are
not adjusted for differences in location, case
mix, and other factors among groups of hos-
pitals. Relative to similar comparisons to be
presented later, the distributions of expense
per case among hospitals in each group are
quite similar. The relatively high cost per case
of proprietary chains is not readily apparent
from the histograms and appears only as a re-
sult of multivanate adjustment for differences
in case mix and bed size.4 The moderate cost
advantage of independent proprietary hospi-
tals appears to result from a relatively high
frequency of very low-cost hospitals within this
group (22 percent with cost per case 30 percent
or more below the national average).5
Utilization Rates
After differences in case mix and other fac-
tors have been taken into account, relatively
small, if any, differences in average length of
stay exist across the four groups of hospitals.
Independent proprietary and chain-operated
nonprofit hospitals have longer adjusted av-
erage stays than other hospitals, but the dif-
ferences represent only one to three additional
days of stay for every 20 patients. Propne~y
chains treat slightly fewer patients, on aver-
age, and independent nonprofit hospitals
slightly more patients, on average, than other
hospitals. Occupancy rates are lowest for the
average proprietary chain and slightly above
average for the average nonprofit chain.6
The average chain-operated hospital ap-
pears to be growing slightly faster than the
average independent institution. Annual dis-
charges grew by a halfa percentage point faster
OCR for page 327
HOSPITAL OWNERSHIP AND COMPARATIVE COSTS
40
35
J
6
~ 25
I
us
o
z
LU
a:
10
5
o
40
35
u) 30
J
~ 25
o
I
o
at
UJ
UJ
10
o
40
35
30
U.
6
cn 25
I
o
z
C)
CI:
CL
15
10
i _
5 _
o
A. Proprietary Chains as. Non-Profit Independents
=3 Propriesery Chains
Non-Prof it I ndependen~s
30
20
,
<70 70~9 90-110 llt-130 131-150
PERCENT OF NATIONAL AVERAGE
B. Proprietary Independents vs. Non-Protit Independents
~3 Proprietary Independena
_ Non-Profit Independents
,,, ~
L'
<70
70~9 90-1 10 1 1 1-130 131-150 > 150
PERCENT OF NATIONAL AVERAGE
C. NomProtit Chains vs. Non-Profit Independents
=3 Non-Profit Chains
Non-Profit Independents
L
L
150
PERCENT OF NATIONAL AVERAGE
FIGURE 2 Dispersion of hospital expense per discharge among hospitals (years 1975 through 1981~.
327
OCR for page 328
328
per year for the average nonprofit chain and
by a full percentage point faster per year for
the average proprietary chain than for inde-
pendent nonprofit institutions. Compounded
over the 7-year span of our data, these differ-
entials in annual growth account for about 300
to 400 extra patients, respectively, for non-
profit and proprietary chains.
Charges per Discharge
Earlier studies by Lewin et al. (1981) and
by Pattison and Katz (1983) indicated that, at
least for hospitals in California, Florida, and
Texas, hospital charges per discharge are much
higher for the average proprietary hospital than
for the average nonprofit hospital. The differ-
ence, according to these studies, is due to very
large differentials for ancillary charges; routine
charges per case differ by very moderate
amounts for average hospitals in each group.7
Our results, shown in Table 2, condemn the
findings of those earlier studies. Total charges
per case, from our data, are substantially higher
among proprietary hospitals (19 and 11 per-
cent, respectively, for chain-operated and in-
dependent facilities) and slightly higher among
chain-operated nonprofit hospitals (3.5 per-
cent) than among independent nonprofit hos-
pitals. Routine charges per case are about the
same, on average, across three of the four
groups, but are moderately above average for
independent proprietary hospitals. For ancil-
lary charges per case, however, differences are
quite large: compared to nonprofit indepen-
dents, proprietary chains bil134 percent more
per patient for ancillary services on average;
proprietary independents bill 15 percent more;
and nonprofit chains bill 5 percent more.
Figure 3 compares the dispersion of expense
and ancillary charges per case among hospitals
in each of the four groups. For both types of
proprietary hospitals, a very distinctive pat-
tern is obvious a very high frequency (30
percent of hospitals) of very high ancillary
charges per case despite a very low frequency
(2 to 3 percent of hospitals) of very high ex-
pense per case. In contrast, the dispersions of
expense and ancillary charges among nonprofit
hospitals are quite similar. Lacking data on
units of ancillary services and charge/cost ra-
tios by type of service, we cannot evaluate the
FOR-PROFIT ENTERPRISE IN HEALTH CARE
finding of Pattison and Katz, for California hos-
pitals, that high ancillary charges per case among
proprietary hospitals are due to a combination
of high utilization and high markups.
Margins on Revenue
Using 1978 data for hospitals in California,
Florida, and Texas, Lewin et al. (1981) found
that proprietary hospitals achieved margins on
patient revenue about 9 percentage points on
average above those of nonprofit hospitals (+ 6.5
percent versus -2.7 percent) and after-tax
margins on total revenue about 2.5 percentage
points above those of nonprofit hospitals (3.7
percent versus 1.3 percent). Using 1980 data
for hospitals only in Florida, Sloan and Vraciu
(1983) found that proprietary hospitals achieved
slightly lower after-tax margins on total reve-
nue than those of nonprofit hospitals (5.4 per-
cent versus 5.6 percent).
Sloan and Vraciu argue that after-tax mar-
gins on total revenue provide a more equitable
basis for comparison of the "profitability" of
proprietary and nonprofit hospitals than do
pretax margins on patient revenue. The for-
mer measure resects the gain to nonprofits
from contributions and the burden to proprie-
taryhospitals offederal and state taxes on prof-
its.
Our results, based on a nationally represen-
tative sample of hospitals for the years from
1975 to 1979 and shown in Table 2, indicate
that differences in margins across groups of
hospitals are not as large as those reported by
Lewin et al. or as small as reported by Sloan
and Vraciu. Margins on patient revenue are 3
to 5 percentage points higher among proprie-
tary hospitals than among nonprofit hospitals,
and margins on total revenue are 1 to 2 per-
centage points higher for proprietary hospi-
tals.8 All intergroup differences are statistically
significant.
Figure 4 presents the comparative disper-
sions of margins on patient revenue among
hospitals in each of the four groups. Between
20 an<123 percent of proprietary hospitals have
operating losses of more than 3 percent, com-
pared to 31 to 35 percent of nonprofit hospi-
tals.
At the other end of the spectrum, 34 percent
of proprietary chains have operating profits of
OCR for page 329
HOSPITAL OWNERSHIP AND COMPARATIVE COSTS
more than 8 percent, compared to 12 percent
of independent proprietary hospitals and 3
percent of nonprofit hospitals. Differentials in
relative frequency of operating profits of more
than 8 percent do not change very much when
one examines margins on total revenue (not
shown in Figure 4~: 33 percent of proprietary
chains versus 14 percent of proprietary inde-
pendents and 8 to 9 percent of nonprofits. Data
on comparative distributions are valuable be-
cause they indicate that, despite modest dif-
ferences among averages for the four groups,
two to four times as high a proportion of pro-
prietary hospitals achieve margins on total rev-
enue of more than 8 percent than do nonprofit
hospitals.
DISCUSSION
Our statistical results indicate that the truth
lies somewhere in between the two polar po-
sitions taken in the debate about the advan-
tages and disadvantages of proprietary
ownership of hospitals. On the one hand, our
results tend to disprove the claim that pro-
prietary chains can operate more cost effec-
tively than the traditional independent nonprofit
hospital as a result of economies of scale. A1-
though independent proprietary hospitals op-
erate at a 4 percent cost advantage relative to
independent nonprofit institutions, proprie-
tary chains operate, on average, at a 6 percent
cost disadvantage.
On the other hand, proprietary hospitals
achieve smaller differential profit rates than
some previous studies indicate. If one accepts
the argument of Sloan and Vraciu that perfor-
mance should be compared in terms of after-
tax margins on total income, proprietary hos-
pitals achieve a very small 1 to 2 points dif-
ferential in net income as a percentage of
revenue.
Other aspects of our comparisons indicate
areas in which proprietary hospitals have taken
some additional advantage of the reimburse-
ment system for hospital care. Despite our
measures being adjusted for differences in case
mix and other factors, proprietary hospitals op-
erate with very high ancillary charges per pa-
tient. Higher-than-average ancillary charges
are most common among chain-operated pro-
prietary hospitals.
329
Even though they operate with higher cost
and higher markup of revenue over cost than
other hospitals, proprietary chains managed to
achieve a higher growth of admissions during
the years from 1974 to 1981. This latter point
illustrates the insensitivity of the market for
hospital care to differences in cost and reim-
bursement rates during the late 1970s and early
1980s.
Until recently, the reimbursement system
for hospital care has not encouraged efficient
behavior. Cost-based reimbursement for
Medicare, Medicaid, and in many locations,
Blue Cross imposed a penalty for cost econ-
omy. The mix of cost-based reimbursement for
the large payers and charge-based rennburse-
ment for others, however, tended to encour-
age the type of differential pricing reflected in
above-average ancillary charges among pro-
prietary hospitals. The advent of Medicare's
Prospective Payment System and various com-
petitive strategies (preferred provider orga-
nizations and competitive contracting) can be
expected to provide stronger incentives for cost
economy.
Looking at data for the later years of the
1980s, one should not be surprised to see, in
Me type of comparisons we have provided here,
proprietary hospitals using different strate-
gies, achieving lower than average costs, and
deriving larger differentials on margins than
they have obtained to date.
NOTES
Patients admitted for elective surgery and urgent
care (e.g., acute myocardiac infarction or congestive
heart failure) are those included. See Gaumer (1986)
for a description of the selection of types of patients.
Presuming a 12 percent annual inflation rate be-
tween 1981 and 1985 and no change in relative costs,
one would find an average difference of $400 per case
today.
3When admissions, length of stay, SNF unit days,
and outpatient visits are added to multiple regression
models, to approximate a cost fimction specification,
the estimated differences in cost per case among groups
of hospitals remain virtually unchanged. Therefore, the
hypothesis that chain-operated hospitals benefit from
improved economic and technical efficiency is rejected
by our results.
4Expense per case among proprietary chains is only
2 to 3 percent higher than among nonprofit indepen
OCR for page 330
330
FOR-PROFIT ENTERTRISE IN HEALTH CARE
40 _ A. Proprietary Chains
35 _
30
Cat
In
o
I
o
Z 15
G
UJ
?51 _
20
10
o
5 _
40 _
35 _
30
J
Co 25
I
o
z
C'
10
20
5
o
if// Ancillary Charges
_ Expense
"R
_
~ _
~ g
1
_
'/',
~ 11
_ i} ~
-,,,_
ALL
<70 70~9
90-110 111-130
PERCENT OF NATIONAL AVERAGE
B. Proprietary Independents
~//~1 Ancillary Charges
_ Expense
__
_a
__
_.
~7/~11111
~_
~//~_111
~//~
150
An:
111-130
PERCENT OF NATIONAL AVERAGE
1 31 -150
>150
FIGURE 3 Comparative dispersion of expense and ancillary charges per discharge (years 1975 through 1981~.
dents when no adjustment is made for bed size or case
miss. Adjustment for regional location and catchment
area characteristics is not sufficient to change the mag-
nitude or statistical insignificance of differences com-
puted from completely unadjusted data. The 6 percent
differential reported in Table 2 results, therefore, pri-
m=;ly from adjustment for case mix and bed size.
Seethe cost advantage of proprietary independents rel-
ative to independent nonprofits actually declines as the
result of adjustment for bed size and case mix, from
10 percent to the 4 percent differential shown in Table
2.
6Because number of beds appears as an explanatory
variable in the equations for length of stay and number
of discharges, the sum of differences shown in Table 2
for length of stay and discharges can be interpreted as
the average difference in occupancy rates across groups
of hospitals.
OCR for page 331
HOSPITAL OWNERSHIP AND COMPARATIVE COSTS
40
30
35 _
J
6
c,' 25
I
IL 20 _
o
~ 15 _
Cal ~1 ~ _
s _
35 [
15
70
o
40
30
con 25
o
I
o
FIGURE 3 Continued
20
15
10
5
o
7Pattison and Katz, using data not available from
Medicare cost reports, attribute differences with re-
spect to ancillary charges per case to a combination of
high markups over cost on ancillary services by pro-
pnetary hospitals and to high utilization of ancillary
services by patients served in those institutions.
331
C. Non-Profit Chains
~3 Ancillary Charges
_ _
16 ~ Expense
_ -
, ....
131-150 >150
<70
_
70~89 90-110 111-130
PERCENT OF NATIONAL AVERAGE
D. Non-Profit Independents
A// Ancillary Charges
_ Expense
_ ~
<70 70~9 90-1 10 111-130 131-150
>150
-
~_
-
-
PERCENT OF NATIONAL AVERAGE
four data do not permit taxes on profits to be broken
out as a separate element of nonoperating expense. In
fact, it is likely that some proprietary hospitals did not
include profit tax in their computation of net income
on total revenue (see Lewin et al. [1981:58] on this
point). Hence, our measure of margin on total revenue
may overstate after-tax margins for proprietary hospi-
tals.
OCR for page 332
332
FOR-PROFIT ENTERPRISE IN HEALTH CARE
50
40
20
10
o
~3 Proprietary Chains
Proprietary Independents
Non-Profit Chains
Non Profit Independents
_
LEA
~ .
>8
<-8
-8 to-3
-3 to O Oto3 3to8
MARGIN ON PATIENT REVENUE (%)
FIGURE 4 Comparative dispersion of margins on patient revenue (years 1975 through 1979~.
REFERENCES
Ermann, Dan, and Jon Gabel (1986) Investor-owned
multihospital systems: A synthesis of research findings.
Ibis volume.
Gaumer, Gary (1986) Medicare patient outcomes and
hospital organizational mission. This volume.
Lewin, Lawrence S., Robert A. Derzon, and Rhea
hearties (1981) Investor-owned and nonprofits differ
in economic performance. Hospitals July 1:52-58
Pattison, Robert V., and Hallie M. Katz (1983) Inves-
tor-owned and not-for-profit hospitals. New England
Journal of Medicine 309:347~3.
Sloan, Frank A., and Robert A. Vraciu (1983) Inves-
tor owned and not-for-profit hospitals: Addressing some
issues. Health Affairs (Spring):25~7.
OCR for page 343
343
a: Z
0 ~Lit
2~ ~J _. ~
Z- ~a:Z Z ~IS
ISO O_
z Is ~ ~ ~ Cal ~In
a: O Z ~ ~ ~"0 ~ ~
L" - 4 ~ ~ ~Met
~ Z a: ~ Z IS _ · ~ S ~'
0 1~ ~ X ~ 01 Z ~= J 1 ~
"OtaJO C! Z on =~ - ~zz~socc
~ 1~ c ~3 ~ A ~ to ~3 ~ ~ ~ ~
o ~ Z ~ IS ~ ~ CL ,~ a, - 4 ·
IS Z to: ~ L&1 Cal Z ~ Z A: Z ~ZZ 2 (a
O ~ ~) ~ ~ ~ O aS ~ O O O
X ~ t ~ ~J ~ ~ ~ O O 1~ ~ 1
"015 - ~: =" - - a ~z - ~ - ~
z ~n ~ ~ ~ · =) c~ ~ ~ c~ ~ ~S ~ cr
~J U, ~S ~L~ ~ ~1 ~ 1~ Z O "J ~ J J J L.J
CD ~ ~ L~ C_~ ~ ~ ~ ~ Z ~ ~ ~ =) ~
= ~ ~ ~s cn z ~ ~ o ~ _ ~ z == c~ c~ ~:
O p ~ ~) ~ ~ ~S ~ ~ ~ O t~ 2 t~ ~ ~ ~ ~ ~ ~ ~, _
_ _ _ ~ ~ t~- L~) · (~ o o ~ ,,, . ~ =, ~ ~ Z U
~J ~ C) ~ ~J ~J ~ ~ ~ ~ ~S ~ J-=. ~ ~ ~ ~-
1 ~ =. ~ ~ ~ _ ~ ~ ~ ~ ~ ~ ~ S O ~ ~ ~ ~ ~ ~
~ O: 1~ Z eS ~ ~ Z ~ ~S ~ ~ O tD ~ ~) ~ ~ L&J ~ ~ ~ O O ~ ~ 2
O000w <~01~10 ~---
Z ~ Z C~ ~ ~ ~ ~ C~ ~ · ~
c] ~ O I_e z ~ ~ I_e ~ ~ ~ ae ~e ~e ~ ~ce ~ |_e l_e I_e 3~ ~J Cte
[~ t&~! _~ ~ Z L~J l-e J ~ ~ ~ ~ O l~e C~ Z ~ Z t_~ ~ C_~ ~ ~ ~
~J ~ ~ ~ ~ e ~ ~ e~ ~S ~ ~ C_~ ~S ~J ~ ~ ~S 3 CLe ~ ~ I~1 Cle ~ ~e ~e ~ CLe U~ Z
e __ ~e I I >[e ~J ~J L~ ~~ |- ~ I I t IC] I I I t | I i I I I
t | |- ~e ~ ~e "e ~ ~S ~ I ~t ~ ~ ~ ~I
~e e~ ~ · ~ z |_e i_ ~_e U~ ~ a e ~e ~ e ~ ~ ~ e ~e
0 ~ ~ ~ t!~ C!~ =~ ~e ~ ~e O C~ Z ~ ~ ~C~ t a) ~ C ~) ~ t~
, ~ 0 ~ 0 0 0 0 0 0 "e O O O O O O O ~ O O O O O O O O O O
U) Z 0_ ~ (_7 J J J _' _e _~l J _ _' J
, ~ ~ ~ 1- 0 ~ 0 ~ |~ ~ ~ ~ - C0 - ~D ~ ~ ~ ~e O0 ~ ~ ~ O O _
-~ O "' O ~ ~e 0 ~ O C3O t0 "e O ~ ~ ~e ~ ~ ", ~I C~ 0 0 ~ ~ ~ ~
\0 ~ ~ ~ ~ ~ O U, ~ U, _ ~ ~e ~ ~e ~ ~ ~J ~e t" ~ "e ~ "e ~ |`
0 O, o o {~ ~ ~ _ ~ ~ {" I_ ~ ~ ~ ~ ~ O' ~ 0 |`e
e ~ _ ~g~ a~ 0 _ ~ ~ ~e y~ ~ _ O ~ C~ ~ ~ ~e a) 0 a~ de _ ~ (~e
l Go =~ ~,o u~ ~e _ ~~ ~ ~ ~ ~ ~ e|e ~ ~2e _ ~ _ _ ~ e~ ~ ~e L~ ~e
~e ~ e ~ e e
00000000000000000000000000000
_ <~ "e ~ _ ~ ~ ~ ~ U, _ ~ C0 ~ ~ ~ ~ ~J _ t~ C~ ~ ~ _ O O0 ~e ~ I_
o ~ ~ a) 0 0 ~ 0` =;e ~ ~ ~ ~e _ _ _ C~ q~ O ~ ~ 0 ~ =` C0 ~ U, ~ _
0 _ t~ 0 0 ~l 0 O~ kD - - ~ I~e O O _ ~- ~ O ~ ~ ~ ~ a~ O ~ ~1 LO
0 0 ~ 0 0-~ ~ -0 ~o-~ 0 a, 0-~ 0 0-0 0 ~J 0-0-c~l
00000000000000000000000000000
". 0 ~ 0 ~ O--O ~D ~D ~ M0--r~ ~ ~n ~ ~ ~ I-1` O ~--~-
U' O ~ ~ ~ ~ ~J O ~ ~ O ~ ~ ~ ~ O ~ ~ ~ ~ ete ~ ~ ~ U, O ~
~ ~ t~D a] ~ - o ~ ~e ~ ~ ~ C~ ~ U~ ~ O @~ ~ ~ t~ ~ O ~ ~ ~ ~
· · · · · · · · · ~ ~ · ~ ~ . ~ · ~ ~ ~ ~ ~ ~ ~ · · · ~
~D C~ O ~ de _ 0 ~_ _ ~ ~ _ #_ d_ _ ~ ~
1 1 1 ~ ~ 1 1 1 1 ~ 1 1 1 1
tY el ~ ~ O ~ ~ a ~ ~e 610 U~ ~ ~.D ~ - ~7 U~ ~ ~e ~e a~ ~e ~ g0 ~ ~e ~e
O ~ ~ -3 de l~ tD O ~ ~ O ~J O ~ ~ co _ ~ O ~ C%1 -) tD C~ 0 _ ^1 0 ~
60 0 ~' ~ ~ kD ~ ~ 0 - SD - {%J ~ e _ "e ~e ~ ~ ~ l~e O ~ |`e O - 1 ~ I~ ~J
U1 0 d. ~ ~d t" ~e ~ ~ ~e de ~ O ~ ~ (q ~e ~ de O ~e ~ `9 ~ ~ U,
1= ~ |~e ~J - ~ ~e ~ ai ~ U~ ~ ~ le ~e ~ ~ 0 0 ~ e~ ~ ~ ~ C~ ~e O ~e
a, 0 _ 1`e 0 ~ ~ ~ ~ {~ a, 30 d ~ ~ O 60-O-~ O ~ ~ ~ ~ O ~ ~
"e _ · _ ~ ~ ~ "e _ _ ~ _ ~ ~ ~e _ · ~ `0 · ~ (S\ ~
000000 00 0 00000000000
e~ ~ `0 ~ ~ ~ ~ e~ ~ ~D ~ ~ _ r~ e" ~ ~ ~ _ ~ r~ ~ ~ ~ ~ r~ ~ ~D-
I~ ~ 0 ~ _ ~ ~---co a~-~ ~ 0 ~ ^1 ~ 0 ~ ~ 0 ~ ~ ~ ~ ~-
u, ~ ~ _ ~ _-~ ~ (q ~ e" ~ ~ ~-~ ~ ~ O `0 ~D r" _ -O `0 C~
0 ~ c~ ~q ~ _ 0 t~ q c~ ~ ~ _ c~ ~ ~o ~ ~ \o eo ~ ~ 0\ ~ o0
_ ~ o. c" ~ ~ ~ ~ c~ _ ~ ~-~ ~ ~n _ _ ~ 0 tD ~ r~ ~ ~ `0 ~ ~-
0 ~ ~ ~ 0 c~-O ~ 0 ~ ~ 0 ~-0 0 _ ~ ~ ~ _ ~ oo r" ~ ~ ~ ~
·~+ ·000+ a00-00 · ·0 ·-O ·0m 0~0
000000000000000000000000 00000
1 000 .000 1 0 ·0 · · 1 1 0 0 · ·0 00000
· · · O O O e · O · O O · · O O · · · · O
000 1 · ·0 0 1 0 1 1 0 0 1 1 0 000 ·0
0 1 1 ~1~1 1
S 2 S X ~) ~ ~ ~ ~ C)
=) t~ ~ O ~ ~ ~ ~ ~ Q ~0
I l I e~ ~` ~ ~ ~ t~ O ~ ~ C. ~ O ~ Z Z 0 - ~C Z e~ ~ ~ ~ ~
~L~w~n~s~z ~m~3= 1C:-C~-=~-z
_ ~ C~ O ~I e~ ~ =} ~ (~ ~ Z l l I I ~ Z ~I
