- 1 -
A STUDY OF THE HOSPITAL UTILIZATION
LABORATORY TESTS
OF CLINICAL
MARTIN HOWARD FLAX
A.B., Cornell University
(1946)
Ph.D., Columbia University
(1953)
M.D., University of Chicago
(1955)
and
MICHAEL JOHN DINNIS BRAND
B.Sc., University of Bristol
(1964)
Ph.D.,
Imperial College of Science and Technology
(1968)
SUBMITTED IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE
DEGREE OF
MASTER OF SCIENCE
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
Signature of Author(s)
Date
Sloan School of ManagcS ntg,
.
Certified by .............
.
a .
Accepted by ........
"" UK;HIVES2
Lt~i.: ·:
Date
*i•a
The•sis Supervisor
- 2 A STUDY OF THE HOSPITAL UTILIZATION
OF CLINICAL LABORATORY TESTS
by
MARTIN HOWARD FLAX
and
MICHAEL JOHN DINNIS BRAND
Submitted to the Sloan School of Management
on May 11, 1979 in partial fulfillment of the requirements
for the Degree of Master of Science
ABSTRACT
A study has been made of the utilization of laboratory testing in
a major teaching hospital. Two measures of laboratory utilization were
considered:
1) the total number of tests performed in the chemistry
laboratory over the past ten years and 2) tests used in the management
of a group of patients with a diagnosis of uncomplicated acute myocardial
infarction over a period of fifteen years. In each case, test volume
data was compared with technological, environmental and cultural change
within the hospital and laboratory. Both groups of data showed an
approximately exponential growth in chemistry test utilization over the
past decade. Over 90% of the total test volume in each year comprised
a group of 23 tests. Although there was a dramatic increase in the number of different tests utilized, from 100 to over 600, during the period
studied, new tests contributed little to the increase in test volume.
Significant changes in laboratory technology occurred during the period
studied, in particular the introduction of automated and multichannel
analyzers. There was evidence that these technological changes caused
a small increase in laboratory test growth rate, but the effect was predominantly in the facilitation of increased demand. Clear evidence was
obtained of a change in physician behavior; there was a marked reduction
in the average interval between tests and a greater tendency to use tests
for monitoring as opposed to diagnosis in more recent years.
Name and Title of Thesis Supervisor: Edward B. Roberts, David Sarnoff
Professor of the Management of Technology
- 3 -
TABLE OF CONTENTS
Page No.
Chapter
1.
INTRODUCTION
1.1
The Cost of Health Care
1.2
Laboratory Medicine .
1.3
The Clinical
1.4
Laborato ry
1.3.1
Clinical
Labor ato ry
1.3.2
Clinical
Labor ato ry Cost
Factors Contributing
Clinical
1.5
I
to
ea sea
Chemistry Te
...........
Objectives ..........
Chap ter 2.
............
METHODOLOGY
2.1
The Hospital
2.2
Study of Total
........
2.3
Myocardial
2.4
Technological
Chemistry
Chem istry
Clinic al
Laboratory Testing
2.5
................
(1969-1 978)........
Infarction Study
Change Within
Laboratory
.....
(1963-1978).
the
..
e.........
Enviro nment
2.5.1
Chapter 3.
Malpractice Costs
TOT AL CHEMISTRY LABORATO
3.1
Total
3.2
Distri bution of Tests
3.3
Chemis try Test Costs
ZAT
Tests Performed .....
...
.....
•...
......
-4Page No.
Chapter 4.
MYOCARDIAL
INFARCTION STUDY
4.1
Chemistry Test Utilization .................
52
4.2
Intervals Between Tests
.............
57
4.3
Frequency of Abnormal
Tests ................
62
4.4
Test-Ordering Patterns
4.4.1
4.5
.......
(Grouping)
Frequency of Test Combinations
62
Comparison of the Frequency of Diagnostic
and Monitoring Functions
Chapter 5.
.................
66
................
73
DISCUSSION
5.1
Laboratory Test Utilization
5.2
Factors
5.3
......
Influencing Laboratory
Utilization
5.2.1
Patient Factors....................... 75
5.2.2
Technological
5.2.3
Physician Related Factors
Factors
...............
...........
76
77
Determination of Appropriate Levels of
Laboratory Utilization
..................
REFERENCES .......................................
83
. 87
- 5 -
Chapter
1.1
The Cost
I
INTRODUCTION
of Health Care
Health care expenditure
creased exponentially
in 1975.
the United States has
from $4 billion
in 1940 to $118 billion
The increase in Gross Na tional
same interval
has been from $100
Produ ct during the
b illion to $1 520 billion.
Thus the fraction of GNP spent on health care
creasing, as shown
in Table
1.1.
increase has been a technological
medicine which has
an d scientific
life itself.
as health care costs continue to r ise,
(1).
in-
re volution
in
thr ee decades.
Increased costs were justified bec ause of the
between medical
has been
A major caus e o f this cost
taken place in the last
placed on health, and even
in-
ver y high value
It is ap parent that
choices mu st be made
care and other des irable goods an d services
As a result,
in recent years the costs a nd resulting
benefits of health ca re have come under cri tical examina t ion
and are a major publi c policy
most other
issue in the
United States and
industrial ized nations.
Health care con sumers annually sp end about 25%
total expenditures on pharmaceuticals, medical
other products; the
other non-product
remaining 75%
items as shown
suppl
their
and
is spent on wages, fees and
in Figure
1.1
(2).
Hospital
care and physicians services are the two largest sectors of
health care costs and together account
for almost two thirds
-6
TABLE
-
1. I1
HEALTH CARE EXPENDITURE AS A FRACTION OF GNP
Year
Health Care Expenditure
Fraction of GNP
$ billions
1940
3.86
3.87
1950
12.03
4.22
1960
25.86
5.13
1965
38.89
5.69
1966
42.11
5.61
1967
47.88
6.03
1968
53.77
6.22
1969
60.62
6.52
1970
69.20
7.08
1971
77.16
7.31
1972
86.69
7.49
1973
95.38
7.37
1974
104.03
7.36
1975
118.50
7.82
Source:
Social Securi ty Administration
FIGURE
FIGURE
1.1
Sectoral Analysis of Health Care Expenditures, 1975
Product Analysis of Health Care Expenditures, 1975
(Consumer Expenditures)
1.2
Billion $
$10.0
3.5
2.5
2.
1.2
0.9
9.0
25%
Dental Car
$ 7.5 Billic
Medicine
$ 10.6 Bil
Total: $ 117.6 Billion
Source:
Arthur D. Little, Inc. estimates
Source:
Division of Health Insurance Studies,
Security Administration
Social
- 8 of total expenditures
Not unexpectedly, hospital
(Figure 1.2).
services have shown the fastest rising costs
care are a and it has
hospital costs
(3).
policy
in the health
been suggested that the
is now the
central
rapid growth of
problem of U.S.
The rapi d ris e in hospital
costs
health care
reflects
the
in-
creasing sophisticat ion a nd complexity of the
care provid ed.
The cost of clinical
among the
labo ratory services was
ing elem ents of thes e acc elerating hospital
c osts.
1ead-
Patie nts
admitted to hospital s toda y expect and
receive a battery o f
diagnost ic tests and
are als o used to mo ni-
proc edures which
tor thei r progress d uring
Neither patients nor the
been satisfied
medical
with less
the course of their
physicians who care
hospital
st ay.
for them hav e
than technical excellence
in
care.
1.2 Laboratory Medicine
The practice of medicine has changed substantially during this century, largely as a result of diverse and continuing technological
innovation
(4).
It can be argued that
physicians' ability to alter the outcome of the disease was
almost non-existent until
the development of twentieth
century pharmaceuticals.
Successful application of these
powerful
therapeutic agents
is dependent on accurate diagnosis
which is increasingly achieved through sophisticated technological means.
Diagnostic laboratory methods used
in modern medicine
- 9 include:
(a) visualization:
ultrasonic imagin
endoscopy, x-ray, nuclear and
g, (b) phy sio
l
ical me asurements
pulmon-
(c)
inic al laboratory
hematol-
ary function, EKG
EEG and
ogy, microbiology
immunolo g Y, biochemistry.
Laboratory
medicine can be c haracteriz ed by I) high imp act
2) increasing use
and 3) incre asing costs.
is a growing
not surprising th at there
of
benefit analysis
interest
laborat ory medicine
is focus ed on the
studies have been
cli nical
it
is, therefore,
inte re st in the cost/
Much of this
(5)
laborat o0 y although
.g . x-ray
made of some other areas,
There are a pproximate ly 15,000 clinic a
the U.S.
(7,8).
not
incl uding smal I p hysicians' of f
while the remaind er are
1.3
are hospital
in
ce
laboratories
laboratories,
independent.
The Clinical L aboratory
Laboratory medicine has participated
the technological revolution.
transformed
this
itive "cottage
Significant
industry"
of producing anal ytical
Among the
innovations
have
capacity, capable
data at a very rapid rate and quan-
innovations that have brought about this
(9):
The introduction of quality control
to the clinical
in
into a highly sophisticated, highly
transformation are the following
(1)
in a major way
branch of medicine from a relatively prim-
automated system of impressive analytical
tity.
(6).
aboratories
f
About hal f of the total
of technology,
laboratories
in the
1950s
procedures
in-
providing the essen-
-
tial
10
-
element of reliability and confidence
(2)The introduction of multichannel
in the early
1960s permit ted
in results
automatic analyzers
the rapid generation of large
volumes of re latively rel iable data with minimal
manpower
(10).
tec hnical
(11)
(3)The
itating the
int roduction of electronic data process ing facilinte rnal
from the clini cal
hand ling and transfer of labora tory data
laborato ry to the clinicians.
1.3.1-Clinical Laborat ory Development
These technolog ical devel opments have permitted the
clinical
laboratories to
clinical
demands
for
remarkable increase i
over the past 20 year
the period
respon d effectively to the
aboratory services.
the util ization of
(7,8); clinical
atory tests.
laboratory services
Accord ing to national
rate
of 13.8% in
surveys,
hospital
Figures from Engl and and Wales
pat terns.
tests were performed while
In 1961,
this
laboratories
labor-
indicate similar
21 million
laboratory
figure rose to 46 million
and reached over 90 mi llio n by
accelerating rate of growth
for
laboratory tests
chemi tries com prise about 30% of all
laboratory growth
1971
There has been a
1970-1975, the n umbe r of clinical
increased at an annua
increased
1976,
in
showing again an
Laboratory utilization at the
teaching hospital of the Un iver sity of Minnesota tripled in
volume from 1963-1976 , desp ite little change in the number of
patients or other patient-related factors.
In another
study
-
II -
from the teach ng hospital of the University of Rochester in
1971,
Griner
ly 40%
( 2) found chemistries accounted for approximate-
of all
bill; between
atory tests
aboratory charges and
965 and 1970,
However,
incr ease
in growth
in use of laboratory tests masks
for patients w th different diseases.
have also demonstrated accelerated
cardial
conditi ons
increase
in laboratory
for acute my
infarction bet wee n 1939 and 1969 sho wed exponential
as well
as x-ray a nd b•ctprinlnni
examinations witho ut significant changes
of hospitalization or pat ient mortality
There are also
tient populations
treated
in the duration
(13)
sign ificant differences
utilization by diffe rent
patients
in hospitals
A study of patients hospitaliz ed
increased chemistry te sts
cal
hospital
the number of chemistry labor-
studies of se lected medical
utilization.
of the total
in reased by 95%.
The overall
variability
10%
in
aboratory
physicians treating com arable
in a cl inic setting
(14),
for
imilar
in diffe rent teaching hospi tals
(15)
a teaching hospital compa red to a community hospital
1.32 Clinical
pa-
and
(16).
Labora tory Costs
As indicated a bove,
increasing clinic al
have been a major contributor to the
hospital services.
This does not
laboratory costs
increasing costs of
reflect
labor atory
ine ffic-
iencies, but rathe r the greater utilization of services and
expansion of new s ervices; only 50%
of the
13%
increase
in
-
costs
in 1975 was due to
12-
increased unit costs.
increase in the cost of services
probably reflects
is remarkably
This 6.5%
low, and
laboratories
increased effici ency of the
related to automation.
1.4
Factors Contributing to
Testing
Increa sed Clinical
These factors may be divided
Chemistry
into three major
categor-
ies:
Laboratory factors
Patient factors
Physician factors
I. Laboratory factors:
(a) Increased capability and reliability of results,
with development of automated multichannel
(b) Advances
in medical
knowledge,
analysis.
particularly in the
discovery of new relationships between biochemical alterations
and disease.
There has been a remarkable
able laboratory tests from 100 tests
at present.
increase
in 1950 to more than 600
One contributing factor is that old technology
is replaced but not completely
increased test pool
retired, resulting
in an
size.
(c) Perceived need to keep up with the
ogy,
in avail-
in part necessitated by the
latest technol-
increased demands on the
laboratories by the clinicians.
2. Patient factors:
(a) The method of payment has changed, with
increase
in
-
13 -
third party payments for health care
through private
ance carriers, Medicare and Medicaid
providing broader
ance coverage.
insurinsur-
These are associated with decreased out-of-
pocket expenses,
particularly as a consequence of their
ability as a fringe benefit as well
as
avail -
the deductibility of
health insurance costs from income taxes.
(b) Local
pital.
factors such as "ca se mix" and type of hos-
The introduction of
increas ingly complex technology
and sophisticated services, such as intensive care units has
also stimulated
laboratory utilizat ion.
3. Physician factors:
(a) The type and location of training of the physicians
has a major
influence on laboratory utilization.
Most physi-
cians are trained in university tea ching hospitals, which
tend to utilize the laboratories
community hospitals
ical
research
(16),
into the
laboratory tests.
to a greater extent than
and which are the location of clin-
introduction and evaluation of new
Furthermore thei r role-models are spec-
ialists who, as a group, tend to us e laboratories to a
greater extent;
tended
physicians who were found
to be members of groups
to be high users
in which the
leadership exhib-
ited this characteristic.
(b) Pressure by superiors and peers on house officers
to do comprehensive work-ups and avoid missing diagnosis,and
to avoid failure to monitor and detect changes
in the
-
14 -
patient's condition.
The type of practice also has
(c)
an
influence.
With
increasing specialization, the attending physicians become
less comfortable with medical problems ou tside of their area
of expertise and are more reluctant to cr iticize and constrain
laboratory usage by house officers
in these
relatively
unfamiliar areas.
(d)
Increase
response to
in the practice of defensive medicine,
less
sign ificant
in radiology.
concept of frequent,
to the
hea lthy individual,
regu ar systematic evaluation
including
multiphasic laboratory
in which chemistry testing plays a significant
screen ing
(f)
clinical
is much
The increased emphasis on preventative medicine has
(e)
of the
tests
laboratory areas such as
in the clinical
clinical chemistry, than
led
The con-
increased threat of malpractice suits.
sequent increased tendency to order additional
in
Relatively
inadequate physician education
laboratory medicine, giving rise to an
role.
in
insufficient
awareness of particular test characteristics such as reliability, or
the diagnostic
implications of various
levels of
test sensitivity and specificity.
(g)
hospitals.
The initiation of quality assurance programs in
These seek to establish minimal
standards of care
-
15
-
for various diseases, and monitor conformance with these
standards.
these programs are not a
In quantitative terms
in recent
significant factor
tory utilization,
incr eases
since they are
in clinical
labora-
concerned with minimal
standards of laboratory usage rat her than the
issue of
excess utilization.
1.5
Objectives
The objective of the present study was to examine
depth the
increase
teaching hospital.
excellence and
n laboratory utilization
This hospital
in
in a university
has a reputation for
serv e s as a model for the examination of
factors contributin g to
increasing
laboratory test volumes.
The approach we have taken is to examine both the
laboratory as a whole and as utilized
in a well-defined
group of patients having a common principal
each case we have reviewed the changes
over an extended
diagnosis.
in laboratory testing
period, ten years for the
laboratory as a
whole, and fifteen years for the patient group.
technological,
hospital
environmental
In
In addition,
and cultural changes within the
have been documented where possible.
This set of
data allows us to evaluate the relative importance of factors
affecting the utilization of laboratory tests.
-
Chapter 2
16 -
METHODOLOGY
2.1 The Hospital
The hospital
with an
is a 452-bed volunta ry,
international
reputation in the
non-profi t hospital
fields of ca rdiology,
oncology, endocrinolog y, hematology and k i dney d isease, among
others.
It pioneered miany of the radiolog y techn ques
standard use throughou t the world, and ha s been
pediatrics and in the
to
insure survival
teaching hospital
medical
development of immu n ologic
now in
leader in
techniques
of transplanted organs
It i s a major
in w hich patient care is
integ rated with
student, gradu ate and post-gradua t e phys ician train-
ing and health-related research.
The hospital
adm its
14,000 inpatien ts annu ally, with an
additional
ambulatory and emergency case
patients.
Al 1 major specialities and sub-specia lities with
the exception of ma ternity
The med ical
physicians.
post-doctoral
load of 235,000
inpatient servi ces are available.
staff has a core of 240
full-t ime salaried
They are augmented by 400 ho use off icers and
fell ows and more
iate, consult ing,
than 500 m embers of the assoc-
special and scientific
staffs, and over 750
nurses. The n ursin g complement comprises registe red nurses,
licensed pract ical nurses, aides,
technici ans and
hundred fifty
students rotate from the
medi cal
and dental
others.
Four
two
health school s eac h year, and mo re than 800 students participate annual ly
in a broad range of associated health profes-
sions training programs.
-
17 -
Much of the daily activity at the hospital
the health-related research vital
medical
organizations, over 300
to
to the continued growt h of
knowledge and improved patient care.
largely by government agencies and
is tied
Supported
private foundations a nd
research programs budgeted at
$8
million are conducted annually.
During the period of
hospital
building was completed
96 adult beds,
which
the current study, one major new
in 1973, adding approxim ately
including a four-bed coronary care unit
became operational
in March
2.2 Study of Total Clinical
( CCU),
1973.
Chemistry Laboratory Testing
F1969-1978)
The data on the numb ers of tests
ical
performed by the clin-
chemistry laboratory was obtained from monthly computer
print-outs which enumerate d the frequency of all
formed during that period.
initiated in 1972,
Aggre gate data on the total
chemistry tests beginning
source;
This computer program was first
so that earlier data was not available
through this source.
hospital
tests per-
number of
in 1969 was available from another
these fig ures for total
tests
from 1972 on-
ward corresponded very wel 1 with figures obtained from the
computer print-outs.
The
monthly frequency of each of 23
major tests was determined for the period
tests constitute over 90%
the c linical
1972-1978.
These
of the total testing performed by
chemistry labora tory.
Data on admissions, tota 1 patient days and length of
18
-
-
stay were obtained from summary monthly reports of hospital
activity.
The
interpretation of the aggregate 1 aboratory utili-
zation data
is somewhat complicated by chan ges
which could
con tribute to shifting patterns of laboratory
utili zation. Su ch shifts are evident
Table 2.1 which
to service
in t:his hospi tal,
in recent year s.
in 1973,
ther e have also been
r"elative o ccupancy levels on different services.
For example, thiere has b een a progressive
patients:
adul t surgery ,
in 1970-71 to 25,60 0
increased only
in surgical
surgery
grew from abo ut 15,600 patient
in 1977-79, while
adult medicine
from 31,4 00 to 35,360 during
Orthopedic surg ery also grew significantly
2.3 Myocardial
i ncrease
thoracic surgery and dental
(combined in ea rl ier rec ords)
days
While there has
increa se in patient days, associated with
expansion of thie bed cap acity
in the
from a n examination of
indicate s the number of pat ient days according
been some overa Jll
shifts
in patient mix
Infarctio n St udy (1963
-
the same period.
during this period.
1978)
The purpo se of thi s po,rtion of the study was to compare
the utilization of clini cal
chemistry laboratory testing,
a comparable gr oup of pa tien ts,
during the last
in
15 years.
We
selected acute myocardia 1 infarction because these patients
generally requi re a subs tant ial
and have benefi tted from
volume of laboratory testing,
inc reasingly sophisticated and spec-
ialized care. Patients w ith uncomplicated myocardial
tion were selected
infarc-
from among a larger group of patients with
0
0
TABLE
2.1
THE DISTRIBUTION OF PATIENT DAYS ACCORDING TO CLINICAL AREAS
DEPARTMENT
1970-71
ADULT
Medicine
31,430
Surgery
Thoracic Surgery715,627
Dental
Otolaryngology
1,349
1971-72
1972-73
1973-74
1974-75
1975-76
1976-77
1977-78
29,168
33,021
35,842
15,037
17,086
19,278
38,305
19,676
3,674
1,729
6,880
3,291
7,913
4,415
4,288
507
2,089
7,473
2,576
9,757
4,461
5,299
431
35,048
15,939
3,207
532
1,999
7,469
2,714
12,211
4,669
5,038
447
1,547
7,963
3,188
11,918
5,777
4,913
469
34,465
21,535
4,778
491
1,797
6,825
3,127
12,413
5,668
5,063
317
35,361
20,328
4,625
652
2,151
6,495
2,525
11 ,446
5,774
5,184
400
89,273
98,016
96,479
94,941
14,818
832
3,131
3,300
1,338
421
146
2,164
4,633
15,426
767
2,813
3,104
1 ,282
469
15,637
14,197
1,335
Gynecology
5,738
Urology
Orthopedics
Neurosurgery
Neurology
Opthalmology
3,622
7,488
2,241
3,137
414
1,703
5,893
3,167
8,606
2,924
4,171
362
72,046
71,031
79,130
87,206
13667
283
14,229
13,536
14,752
2,086
3,039
5,204
536
168
2,194
2,920
2,098
1,057
226
4,743
4,348
4,391
427
165
386
164
1,949
1,899
3,726
228
1,582
1,063
257
441
180
2,186
4,385
1,072
937
230
28,353
27,530
29,047
31,613
Adult Total
PEDIATRIC
Medicine -13667
Psychiatr
Infectious Disease
Surgery5,0
Thoracic Sur ger
Otolaryngology
Dental
Urology
Orthopedics
Neurosurgery
Neurology
Ophthalmology
Pediatric Total
3,341
1,389
1,059
586
80
1,088
2,137
3,758
1,503
389
93
2,888
4,161
1,419
335
105
274
4,863
1,901
4,336
243
2,693
4,155
1,151
700
187
226
2,034
164
32,836
32,827
32,871
32,753
940
870
898
905
978
-
that discharge diagnosis.
20 -
For the period 1963 through 1968
the study group was finally selected from among those with
myocardial
Starting
infarcts by direct
review of hospital charts.
in 1970, the discharge diagn oses
(in the form of
International Classification of Disea se Adapted -ICDA-8)
stored
in a computer program, so that
were
a print-out of all
patients with the discharge diagnosis
of acute myocardial
infarction could be obtained.
From t hese data, cases with
uncomplicated acute myocardial
infarc ts were selected and
their hospital
charts were reviewed.
cardiac arrythmias, congestive heart
Only cases without
failure or other compli-
cations were selected for further ana lysis of laboratory
utilization.
For each patient studied, the hospi tal
ined and all
c linical
on a daily bas is.
ting
2.3.
chemistry determina tion s were t abulated
This yielded a comprehensive chart
all chem istry tests
hospital
day.
for the admission according
Examples are
tests were also
tests were des ignated as abnorm al
limits set by the
cated with an asterisk
[*]
to
if the values
laboratory.
desi gnation
recor ded;
fell
in the study.
outside
(These ar e indi-
in the example s.)
Tables 2.4a and b compare some of the general
of the patients
tabula-
illustrated on Tables 2.2 and
The freq uency of abnormal
of the normal
record was exam-
features
The average age of the pat
ents
and the frequency of males to fema les did not change sign
f-
icantly throughout the study.
on
The re was a gradual
reduct
TABLE
TEST
8/
TEST
7 8
2 .2
--
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
SMA-6 - 1 (6 tests)
+
+
+
Sodium
Potassium
Chloride
+
CO2
BUN
+
Creatinine
+
(3 tests)
Enzymes
(4 tests)
Patient:
+'
SGOT
+
+
+,
Glucose
Bloodgases
+-
+
+
+
+'
+9
+
Age:
SGPT
+, +,
Dates of
Hospitalization:
CPK
Isoenzymes
LDH
CPK
8/7/63 - 8/25/63
19 Days
II (6 tests)
Total Tests - 20
Calcium
Phosphate
Total protein
+ Test(s) reported
* Abnormal result
Albumin
Bilirubin
Cholesterol
+
Alkaline phosphatase
Others
Digoxin
Pronestyl
TOTAL
79
Hospital #158-110
LDH
SMA-6 -
G,M.
4
3---3
-
2
1
-
- -
-
6
-
1
-
-
-
22 -
TABLE
2.3
,Patient: A,E.
Age 60
Hospital #99 87 39
Dates of Hospitalization:
2/16/77 - 3/2/77 - 15 days
Total tests - 118
TEST
16 17 18 19 20 21 22 23 24 25 26 27 28
3/
1 2
+
+
2/
SMA-6 - I (6 tests)
Sodium
Potassium
Chloride
C02
BUN
Creatinine
Glucose
Bloodgases
Enzymes
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+,
+ +
+ +
+,
+ +
+
+
+* ,
+
+
+
+,
+
-
,
+
-,
+
(3 tests)
(4 tests)
SGOT
++ +* +*
SGPT
++ +
++ +,
LDH
CPK
+
+,
++ +, +,
Isoenzymes
LDH
++ +, +,
CPK
SMA-6 -
II
(6 tests)
Calcium
Phosphate
Total protein
Albumin
Bilirubin
Cholesterol
Alkaline phosphatase
Others
+
+
+
Uric Acid
+. +-
Amylase
+
T5 Uptake
T4
Free T4
TOTAL
+
+
,
,
5
5
+, +
+
+
+
+
27 23 12
6
4
1
4
6
7
5
6
5
-
23 -
TABLE 2.4a
YEAR
NAME
SEX
1963
C,R.
C,A.
D,M.
F,A.
AGE (years)
LENGTH OF
STAY
M
54
15 days
M
48
16
F
F
56
72
17
27
G,A.
F
69
24
G,B.
G,M.
L,F.
M
F
F
63
79
70
19
19
23
M,F.
F
62
21
"
32
M,D.
M
68
33
"
71
S,R.
M
62
21
"
37
P,J.
W,A.
F
F
52
45
22
22
"
"
19
28
Y,J.
M
69
29
"
23
Average
1967
H,H.
M,J.
M,F.
59
73
41
Average
1968
1970
"
22 days
19 days
18 "
17 "
58
72
59
15 days
27
21
G,J.
M
57
23
H,J.
L,J.
M,E.
M
M
F
60
68
55
22
25
19
S,A.
F
61
23
Z,L.
M
76
24
Average
63
41
36
60
34
67
41
"
32
56
22 days
52
74
32
54
52
61
43
55
25 days
19
20
28
23 "
18 "
16 "
60 yrs.
47
47
67
64
"
63 yrs.
49
50
81
68
40
72
60
30
18 days
M
F
M
M
M
M
F
M
F
F
95
52
20
53
B,V.
F,A.
G,J.
B,F.
C,N.
F,F.
M,A.
P,S.
S,G.
V,C.
26
23
67
58 yrs.
Average
12
"
62 yrs.
M
M
M
TOTAL
TESTS
21 days
- 24 TABLE 2.4b
AGE (years)
LENGTH OF
STAY
F
80
15 days
M
64
11
G.E.
G,E.
H,H.
L,M.
L,G.
F
M
F
M
M
70
54
67
75
65
16
15
21
15
25
N,M.
M
64
16
R,M.
S,H.
W,A.
M
M
F
71
72
66
16
24
12
YEAR
NAME
SEX
1973
B,F.
F,J.
Average
1975
M
M
F
F
F
68
73
76
78
92
P,C.
M
55
Average
1977
F
F
M
M
M
M
M
60
60
55
69
58
59
34
W,H.
M
70
Average
1978
Average
24
90
106
59
66
95
"
121
51
106
76
17 days
14 days
21 "
15 "
14 "
18 "
21
88
17 days
15 days
19
16
18
19
22
12
17
63
75
58
17 days
20
10
M,M.
F
61
15
M,R.
M,M.
M
F
70
62
17
13
yrs.
84
17 days
M
M
M
65
87
118
179
286
143
80
160
69
"
59 yrs.
A,L.
E,A.
G,E.
75
76
87
101
75
92
"
74 yrs.
A,E.
F,H.
F,R.
L,G.
M,F.
P,S.
S,K.
39
"
62 yrs.
B,H.
D,L.
G,E.
G,R.
K,B.
TOTAL
TESTS
139
165
179
85
"
139
207
245
15 days
170
-25in the average
leng th of stay of these patients
(from about
21-22 days to about 15-17 days) and a remarkable increase
in total
1963
clinical
c hemistry
to the present
in utilization will
laboratory determinations from
The detailed analysis of the changes
be provided
in Chapter 3, and constitute
a major element of the study.
2.4
Technological Change Within the Chemistry Laboratory
The chemistry laboratory of the
progressively
study.
incre ased automation du ing the period of this
The timing of the
instruments
is outl ined
were performed
individual
introductio
man ual ly
before
blood urea n itrogen
introduced
(TP),
(BUN),
(AA) were introduced for
for the determination of to tal
(Alb),
calcium (Ca),
and alkal ine phosphatase
analyses using an Instrumentation
In the same year
(1972),
in automation were
SMA 4+2 was
NA, K, Cl,
carbon dioxide
(Creat.) and
serum protein
phosphate
(Alk. Phos.).
(P04 ) ,
The next
i n 1968 of blood gas
La bor atgry,
This was supplement ed by a Corning
changes
(CI),
c rea tinine
major change was the introduction
1972.
determinations
In 1 964, additional aut oanalyzers were
serum albumin
uric acid
All
1963; at which time dedicate d
Technico n AutoAnalyzers
glucose (GI).
of various new
in Table 2.5.
the determination o f serum chloride
(C02),
hospital underwent
165
Inc.
(IL-ll3).
i nstrument
oth er si gnificant
introduced.
A Technicon
introduced for the simultaneous determination of
C0 2 , BUN and Creatinine.
Serum glucose and the
0
TABLE
Year
63
Na
67
4
#
K'
Cl02
AA
CO2
A
AA
DBlkl
"""
MANUAL
0-- -
MANUAL
10
SMA 4+2
TP
-4
Alb.
-o MAN
SMA 4+2
)Pp-
P_.,i
SMA 4+2
SMA 4+2
__
_
ABA- 100
4
Bilirubin -(
SMAklIIAI
.
__
Amylase
SGOKAAL
SGPT
LDH
q
-
4
-•
SGPT'
4
-
CPKL
'"
ALK. Phos•-*- MAN
SNA
-C
I-1.L 113
Il
.
SSMA.
MANUAL
ABG
II
SMA
.. ~
Uric Acid -MAN 0'
SMA
"''"
SMA
"e
-- MAN -M
P04
SCMA
Pý]
4
MAN -I--
,•
-SM
-
MAN -P- -
0
>
SMA 4+2
AA
Glucose-'
)-
SMA 4+2
•-
AA-
4
Ca
75
"AA
-
CREAT
2.5
--
CORNING 165
-.-
AA
po
S*)
MANUA
ABA-100
--4-
ABA-100
ABA-100 --
MANUAL
.- *--
ABL-2
ABA 100.-
-
MANUALMNLA
)----~
-- ABA-100
GEMSAEC
*-
GEMSAEC --GEMSAEC-)--
MANUAL
AIMAN "AL
'~""'
-
..0-
.
A,
ill
ABA-100 --- ~
GEMSAEC
--ABA-100
0---GEMSAEC •-.
4
-
major enzymes
-
27
(SGOT, SGPT, LDH and
utilizing the Abbott ABA-100.
determined with this
CPK*) were determined
Alka line phosphatase was
instrument
in 1973.
The next major technological
improvement occurred in
1976 when a second Technicon SMA-6 was
introduced for the
determination of TP, Alb, Ca,
P0 4 , uric acid and
A GEMSAEC was also introduced
for a 11
at this time; serum cholesterol was
1977, and amylase
in 1978.
hilirhin
enzyme determination
determined by ABA-100
Finally , arterial
blood gases
were determined by the technologica lly superior Radiometer
ABL-2
in 1978.
A brief description of some of the main
these
a.
features of
instruments follows:
AutoAnalyzer (Technicon
Instruments,
The AutoAnalyzer (AA),
Inc.)
developed by Skeg gs
sents a first attempt to introduce automation
mechanization)
into the clinical
chemical
(17)
(or at
,
rep re-
least
labo ratory.
I
The
AutoAnalyzer utilized the principle of continu ous flow:
a
pump moves the sample blood serum and necessar y reagents
along a series of plastic tubes continuously.
methods are used
* SGOT
to separate serum proteins, dilute the
is the abbreviation
transminase;
minase;
Appropriate
for serum glutamic oxaloacetic
SGPT stands for serum glutamic pyruvic transLDH is lactic acid dehydrogenase, and CPK is creati-
nine phosphokinase.
-
sample and add
reagents
-
28
in controlled proportions.
Further
details of AutoAnalyzer methodology a re given elsewhere
Generally, the method of measurement
other detectors can also be used.
ments,
reage nts
product with
the
are added
is photometric, although
In photometric measuresample to produce a color
req uired
of the color is measured
component, and the
in a photoelectric
intens
colorimeter.
rate of
Samples are proces sed by an AutoAnalyzer at the
60 per hour.
Each sampl e is placed
which
is loaded
in turn
plastic cup
The
instrument
pre-
is calibrated
replacing samples wit h reference standards of known comThe output of the analyzer is either a chart
position.
recording of sample peak
color
heights
(which are proportional
to
intensities) or a printed table of sample concentra-
tions.
Once the serum s amples a
the analyzer can
able saving
b. SMA-6
run una ttended.
loaded
hich
into the turntable,
represents a consider-
in operator time over earlier manual
(Technicon Inst ruments,
The SMA 6/60
(19).
Inc.)
performs six different chemical
tests per sample at a ra te ot bU
combinations of the six
but the most common
methods
(Sequiential Mu Itiple Analyzer) is a contin-
uous flow analyzer which
Cl,
in a small
into a turntable for sequential
sentation to a sampling probe.
by
(18).
tests
(t
amples per hour.
test profile)
•
-
are possible
is t he four serum electrolytes
C0 2 ) plus two others
(BUN, Creat.).
Ditterent
,
(Na,
K,
Here, Na and K are
-
29 -
measured with a flam e photometer whil e the other tests utilize
a photoelectric colo rimeter, as
yzer.
Another commo n combination
6 in this hospital)
bumin,
in a single channel
bilirubin and
AutoAnal-
(in use with the second SMA-
is calcium, phosp hate,
total
protein, al-
cholesterol.
Operation of t he SMA-6 is not s ubstantiall y different
from a single channe 1 AutoAnalyzer.
into cups
Serum sampl es are loaded
in a sampl er turntable and
ment operates automa tically.
thereafter the
on a chart
recorder
recorder output for
which
is sufficient for
One tec hnician
routine operation of the instrument.
instru-
The result s are obtained
is designed to be easi ly read;
a single patient sample
is
s hown
the
in
Figure 2.1.
c. ABA-100 (Abbott LIaboratories,
Inc. I)
The ABA-100 is an automated dis crete sampl e clinical
analyzer comprising a spectrophometer , a pipetto r-dilutor,
and an electronic da ta-processor and
appropriate reagents are dispensed
plastic multi-cuvet mounted
The
is measured for each
sample in tu rn by the spectrophotometer; color
proportional to required analytical
time interval
intensity is
concentration.
The
in-
repeat measurements at fixed
intervals of up to 20 minutes.
eration and
Th e required
reagents to produce a color-
intensity of this color
strument can be programmed to
S amples and
into wells of a disposable
in a wate r bath.
sample compo nent reacts with the
ed product.
printer.
Depending on the mode of op-
selected,
the
instrument
is capable
j
'-~-~--'~~~~
~
-
30
-
FIGURE 2.1
SMA 6/60 SERUM ANALYSIS
Na±
meq/liter
160
+
K
meq/liter
10
CI-
COn
meq/liter
9 o' dr
meq/liter
40--
-
CHART
BUN
mg/dl
100
Creat.
mg/dL
15
14
150
9-
120-
90
-
140
13
130
8-
110-
80
-
12
-
30
120
11
7-
110-
70
100
10
100
6-
90-
60-
-
9
90
B_
8
20--
80-
5
80-
50
-
7
-
B-
70-
-
40"
-
70-
60
6
-
5
10-
50
3-
30
--
4
40
30
2.
20
1
10
3
20
C-
10
0
Na
+
0
K+
n.
-0
-
CI-
CO,
BUN
BUN
0
Creat.
-
of up to
31
-
150 measurements per hour, about 5-20 times the rate
for equivalent manual
d. GEMSAEC
procedures.
(Electro-Nucleonics, Inc.)
The GEMSAEC anal yzer, which was developed by Anderson
(20), rep resents an alt ernate
approach to laboratory
instrumen-
tation f rom the ubiqui tious continuous flow analyzers.
GEMSAEC is one of five comme rcially available centrifugal analyzers which operate
by pho tometric measurement of color
samples contained in s eparat e chambers
rapidly
rotating disc.
Befo
commenci
and reag ents are preme asured
nto appro
into the disc.
ted and
During an an a lytical
centrifugal
develops
in the periphery of a
accommod ated
in
mples
te well
chined
in the sample.
reagents outwar ds
co ncentration of t he
Sixteen samples can be
a disc and a photometric ab sorption
is made on each sample
is rot a-
Here mixing occurs and color
in proportion to th e analytical
required component
nalysis
cycle, the disc
fo rce mov es samples and
into a c orresponding c uvette
in
reading
durin g each revoluti on of the disc.
The meth od offers particular advantages
in kinetic assays
which are u sed to determine enzyme activities
in blood
serum
Kinetic met hods of analysis relate the rate of formation of
color to an alytical concentration.
Centrifugal analyzers
generate la rge amounts of data - a consequence of the high
speed
rotat ion of the disc.
effect data
reduction.
A computer is therefore used to
.
- 32 e.
Blood
Gas Analyzers
Arterial
blood gas
analysis (ABG) conventionally re-
quires measurement of pH and the partial
(PC0 2 ) dissolved in the sample.
(P02 ) and carbon dioxide
Practical
pressures of oxygen
electrochemical sensors had been d eveloped for
measurement of these thre e parameters by the
late 1950s.
The
represents an
IL 113
(Instrumentati on Laboratory
Inc.)
early example of a blood gas analyzer.
It c omprises two units:
a thermostat bath contain ing the three elect rode sensors and
an analog meter provided with three differen t scales for each
measurement.
scale
in
In operatio n, it was necessary to select each
turn by means of
The Corning Model
a switch.
1 65 represents a second generation of
It incorporates a d igital
blood gas anal yzers.
in addition to the basi c pH
also calculate total
P0 2 and PCO
dial.
This
requires a knowle dge of the sample
tion, which
measurements, will
2
ca rbon dioxide conce ntration, bicarbonate
ion concentrat ion and b ase excess.
meter
display and,
is entered
into the Model
latter derived parahemoglobin concentra-
165
through a calibrated
Push bu ttons are pro vided to allow any one of the six
parameters to b e displayed.
The ABL- 2 (Radiometer) is a second generation of the first
computerized
blood pH,
made.
b l ood gas analyzer.
P0 2 a nd PCO
2
, a direct measurement of hemoglobin
Derived parameters are:
dioxide, base excess,
In addition to measuring
bicarbonate,
standard base excess,
total carbon
standard
bicar-
is
- 33 bonate and oxygen saturation.
of pH,
P0 2 and PCO
displays.
All
2
The three bas ic measurements
are shown simultaneously
on three digital
measured and derived parameters are printed on
a strip printer.
The ABL-2 also di rectly measures ambient
barometric pressure which is necess ary for automatic calibration of the instrument, another uni que feature.
calibration occurs every two hours or at
This
feature is particularly effect ive
the users'
initiation.
in improving the
reliability of blood gas analysis - earlier
instruments had
required the technician to obtain a barometric
atmospheric pressure, calculate the
Automatic
partial
reading of
pressures of the
standard gas mixtures and then perf orm several
calibration
adjustments to the analyzer.
2.5
Environment
2.5.1.Malpractice Insurance Rate Changes
As a consequence of
the cost of malpractice
ably high rate
changes
increased malpractice litigation,
insurance has
in recent years.
in premium rates
between
increased at a remark-
Table 2. 6 indic ates the
1965 and
classes of physicians in Massachusetts
While some rate
increase s began between
ticularly in higher risk classes),
went
into effect
in 1972
increase through 1974.
1977,
for five
(100/300
limits).
1968 and
the major
1972
(par-
in creases
and continued at this h igh rate of
Further increases contin ued up to
TABLE 2.6
MALPRACTICE COSTS
PHYSICIANS & SURGEONS RATES 100/300 LIMITS
CLASS
YEAR
A
B
C
E
D
1965
$ 87.00
$109.00
$ 208.00
$ 276.00
1966
103.00
130.00
272.00
408.00
1968
95.00
167.00
284.00
396.00
494.00
1969
119.00
210.00
357.00
497.00
622.00
1972*
138.00
242.00
522.00
726.00
907.00
1972
209.00
368.00
794.00
1,104.00
1,380.00
1973
288.00
506.00
1,092.00
1,518.00
1,898.00o
1974
403.00
708.00
1,529.00
2,125.00
2,657.00
1975
469.00
824.00
1,757.00
2,447.00
3,060.00
1976
548.00
,250.00
2,081.00
3,010.00
3,810.00
1977
552.00
,425.00
2,556.00
3,644.00
4,158.00
*
Rates -
25/75
Limits
---
- 35 the present, but not at the approximately 30%
increase seen
in 1972-1974.
rate of
- 36 TOTAL CHEMISTRY LABORATORY UTILIZATION
Chapter 3.
3.1
Performed
Total Tests
Durin g the period
performed p er year
creased
1969-78, the total num ber of tests
in the analyzed chemistry 1 aboratory in-
fro m 270 thousand to over 930 thousand .
volumes for
intermediate years are shown
trend curve
for test volumes
is shown
Total
in Ta ble 3.1
in Figur e 3.1.
test
and a
Total
test volume s grew exponentially from 1969 unti 1 1974.
rate slowed briefly
recurred
in 1976 and
Table 3.1
s hows
in 1975, but the
rate
in 1978.
in tests as a % ove r the previous
year and as an absolute rate
volume with
increased growth rate
1977, moderating somewhat
the growth
The
(first derivative of total
respect to time).
test
It may be noted that the growth
is slightly higher during the period
1972-74 than during
1975-77.
Changes
changes
in total
laboratory utilization may reflect
in the utilization of the hospital
in its entirety,
as well as utilization of chemistry tests in patient care.
We,
therefore, calculated the average number of tests per
hospital
admission and
per patient day of admission for
each year during the period
compared with the total
3.2 and Table 3.2.
1969 to
1978.
These results are
number of tests per year
The general
in Figure
shape of the curves showing
tests per day and tests per admission is very similar to
that of the total
tests.
We conclude that hospital
- 37 TABLE
GROWTH
Year
3. 1
IN TOTAL CHEMISTRY TESTS
Total
Tests
% Increase
Over prior year
(thousands)
Absolute
Growth Rate
(Tests/year)
(thousands)
1969-70
268.9
1970-71
272.3
1
20.3
1971-72
309.5
13
85.2
1972-73
442.80
38
139.3
1973-74
588.20
33
98. 1
1974-75
639.00
9
83.8
1975-76
755.70
18.2
126.7
1976-77
892.30
18
89.5
1977-78
934.77
5
- 38 -
FIGURE
3.1
LAB TOTALS
Ir\~h
r--
900
ALL
TESTS
1
800
700
-
600
-
500
-
400
-
300
200
100
SMA6-I
(LBC)
-
-
-
I
1970
I
71
I
72
I
73
I
74
YEAR
I
75
I
76
I
I
77
78
- 39 -
FI GURE 3.2
Total Chemical Tests Performed per Year
t•'0"
I r
I,UUU0UUU,
Ihr\
F
IuU
TOTAL
TESTS
90
80
-
TESTS/
ADMISSION
70
TESTS/
,TIENT DAY
- zo0 60
(,
C,
-03
500,000
N
50
E
C,,
-,
w
40
30
200,000
-
20
10
-
-
9
YEAR
- 40 TABLE 3.2
GROWTH IN TOTAL CHEMISTRY TESTS PER ADMISSION OR PATIENT DAY
Total
Admissions
Tests/
Admission
Total
Patient
Days
Tests/
Pt. Day
1969-70
10,671
25.2
117,280
2.30
1970-71
11,089
24.7
114,041
2.39
1971-72
11,608
26.7
112,941
2.74
1972-73
11,459
38.5
121,828
3.63
1973-74
12,295
47.8
132,839
4.42
1974-75
13,426
47.7
135,786
4.71
1975-76
14,012
54.0
147,184
5.14
1976-77
13,164
67.6
142,879
6.24
1977-78
13,366
69.9
140,952
6,56
Year
-
- 41
total
occupancy has had some effect on
tests ordered,
However,
importance.
that other factors are of greater
but
interpretation of these results must be made cautiously because, as will
later in the myocardial
be shown
is not uniform over the
study, the distribution of tests
length of hospital
stay.
Figure 3.
1974-75.
yearly
total
is a greater test use immedi-
There
immediately prior to discharge.
ately following admission than
The
test data
shows an anomalously
presented
in Table 3.1 and
low growth rate during
In order to obtain furt he r in
ht
reasons fo r this variance, we exa mi ned
into the
monthly total
1976 and compared
chemistry test performed from 197 3 until
them with the monthly hospital oc cupancy r ate
Monthly va riations were in some c ases extr eme,
of 10-20%.
and total
(patient days).
in the range
Some correlation betw een month ly patient days
test volume was apparen t;
Table
correlatio n coefficient for each year.
correlatio ns exist for 1973,
but for 1975,
was found.
infarction
1974
and
3.3 shows
the
Mo derate positive
1976
(r = 0.55 to 0.85)
a moderate negative correlat ion
(r = -0.44)
Little information is obtained by graphical
examinatio n of the data until
it is smooth ed.
Figure 3.3
shows 3-mo nth moving averages as a functio n of time for
mont hly to tal
are
1975
noted
(by
tests and patient d ays.
narrow arrows)
Pos itive correlation
in mid 1974 a nd at the end of
S In the first quarter of 19 75,
there
is a peak in
TABLE 3.3
CORRELATION BETWEEN MONTHLY TOTAL TESTS AND PATIENT DAYS
Year
-
-
Average Length
of Stay ± 1 s.d.
(days)
Average
Monthly Total Tests
(thousands)
Average Monthly
Patient Days
Correlation
Coefficient
1973
10.71
+
0.28
36.8
10.3
0.553
1974
10.79
+
0.32
49.3
11.1
0.845
1975
10.19
+
0.57
51.1
11.5
-0.435
1976
10.74 ± 0.27
63.1
12.1
0.743
- 43
FIGURE
3.3
Monthly Total Chemistries
and Occupancy (Patient Days)
12
MONTHLY
PATI ENT
DAYS
(000's)
-IV
A
ýj
•
InLVUV
I
9
-,
,
7U
60
MONTHLY
TOTAL
LAB TESTS 50
(000's)
r-
)-
E
40
E
30
E-
I
I
I
I
I
I
MONTH: 3s
YEAR: 1973
1974
1975
* All data points are 3-month
moving averages.
1976
- 44-
TABLE 3.4
AVERAGE LENGTH OF STAY IN THE HOSPITAL UNDER STUDY
Month
Year
1973
Jan
1974
1975
1976
11.50
11.13
10.52
Feb
11.07
10.89
10.63
10.4
Mar
10.57
10.72
10.48
10.68
Apr
10.39
11.08
9.92
10.5
May
10.47
10.63
9.60
10.46
June
10.42
10.55
9.30
10.77
July
11.00
10.63
9.57
10.79
Aug
10.68
10.72
9.8
10.8
Sept
10.53
10.98
10.18
10.9
Oct
10.66
10.55
10.5
11.4
Nov
11.24
10.95
10.2
10.9
Dec
10.75
10.22
10.9
10.8
- 45patient days of ho spital
with the tot al
occupancy which correlates
tes ts performed
(thick arrow).
the average length of stay of a patient
significantly lower than in 1974 or
Also,
negatively
in 1975,
in the hospital
1976
(Tables 3.3 and 3.4).
It is concluded that there was a short-term change
in hospital
utilization during this period, particularly from April
August,
to
1975; associated with the admission of a group of
length of hospitalization.
patients with reduced
3.2
is
Distribution of Tests
The frequency of use of 23 major tests over the period
1972 to
Here,
in Table 3.5.
1978 is shown
grouped according to the technology used
year,
tests have been
in 1978.
In each
these tests account for over 90% of the tests performed.
The percentage fraction of total
tests not
included in this
set of 23 tests shows some variation from year to year, but
there is no clear growth trend.
While the
introduction of
new tests undoubtedly contributes to the total
in the
it is not a significant factor
increase
test volume,
in chemistry
laboratory utilization.
The
BUN,
so-called "LBC"*
group of
tests
(Na,
K, Cl,
C0 2,
Creatinine) represents slightly over half of the total tests
performed.
This fraction has been
1971 onwards, as
shown
in Table 3.6.
parallels the growth in total
in LBCs
remarkably constant from
The growth
in LBCs
tests performed;
see Figure
during the period 1972-74
is greater
3.1
Growth
*LBC
is an acronym standing for
'Lectrolytes, BUN, Creatinine
0
TABLE
0
3.5
TESTS
1972
1973
1974
1975
1976
1977
1978
TOTAL
309.5
442.8
590.2
639.8
755.7
892.3
934.8
SMA6-1*
166.0
236.7
337.7
370.6
437.3
506.3
524.8
Enzymes
38.2
49.6
59.8
67.0
75.0
91.4
94.8
SMA6-11 +
40.4
55.59
66.5
75.9
81.5
94.5
95.0
Glucose
21.4
29.2
34.8
36.3
45.3
47.8
49.0
Alkaline
Phosphatase
8.6
10.9
13.4
15.9
16.6
21.3
20.8
Blood Gases
8.1
16.6
23.9
26.3
35.6
43.01
44.6
Cholesterol
3.4
4.0
5.4
5.9
5.9
7.52
Amylase
3.0
4.79
5.94
7.3
8.1
12.5
14.0
851.0
Subtotal
% of Total Tests
Unaccounted
289.2
407.4
547.5
605.2
705.3
824.3
6.6
8.0
7.2
5.4
6.7
7.6
*SMA6-1: Na,K,C1,CO2,BUN,Creatinine
8.4
9.0
+SMA6-11: Ca,P04,Total Protein,Albumin,Bilirubin,Uric Acid
0
0
TABLE 3.6
GROWTH IN LBC AND ENZYMES
YEAR
TOTAL
"LBC"
1971-72
166,030
.52
38,190
.12
1972-73
236,700
.53
49,640
.12
1973-74
337,690
.57
59,830
.10
1974-75
370,600
.58
65,980
.11
1975-76
437,330
.58
74,960
.10
1976-77
506,310
.57
91,380
.10
1977-78
524,880
.56
94,770
.10
% INCREMENT
OVER PRIOR YEAR
LBC/TOTAL
TOTAL
ENZYMES
% INCREMENT
OVER PRIOR YEAR
ENZYMES
TOTAL
- 48 than during
There wa
1975-77.
labo ratory technology for LBCs
was
a substantial
adva nce accelerated t he growth
the two years followi ng
its
in 1972 when the first SMA-6
in LBC
include
the
labora tory was
promptly.
No formal attempt w as made
of the SMA-6.
main ing
tests for whi ch annual
growth
Figure
in enz ymes and
but
SMA-6 -
realization by hous e
to publicize the
3.4 show s trends
arri val
reg ion between 1974
Mechanisms for
report in g LBC data more
staf f that
b lood gases,
tests performed duri ng
introducti on.
the technology diffus ion could
glucose and
in
it is poss ible tha t this technological
introduced;
anomalo us
change
re-
were available.
totals
and
in the
1975
is apparent for
1ess so for enzymes.
II tests
The
The
is accelerated from
1976 to 1977, again following major te chnological advances
in the
3.3
laboratory for both test methodologies.
Chemistry Test Costs
The relative direct cost of the
was determined
summarized
From
for each year of the study;
(FTE)
test volumes
the data are
in Table 3.7.
1972 to the present, there has been an increase
in the productivity of the technical
laboratory.
increased
Total
increased
tests/full
from 12.1
staff of the
chemistry
time equivalent empl oyees
in 1972 to 20.7
in 1978, an
increase
-
49
-
FIGURE 3.4
LAB TOTALS
IuU
MA6 - II
NZYMES
90
80
70
60
50
+
m
;LUCOSE
3LOOD GAS
40
30
ALKALINE
PHOSPHATASE
-AMYLASE
20
10
- CHOLESTEROL
I
I
I
I
I
I
....
1972
73
74
75
I
YEAR
I
76
~-''I
77
78
0
0
0
·
TABLE 3.7
RELATIONSHIP OF CHANGES IN TOTAL CHEMISTRY TESTING TO STAFF AND COST DATA
COSTS
STAFF
Year
Total
Tests
x 103
FTE*
Tests/FTE
Payroll
$ x 103
NonPayroll
$ x 103
Ratio
Total Tests
Total Costs
69
1.21
.83
Rat io
Cost ($)
Tests
1972
320
26.5
12.08
196
1973
443
27.5
16.11
263
108
1.19
.84
1974
588
33.2
17.71
322
129
1.30
.77
1975
615
34.2
17.98
404
152
1.11
.90
1976
756
39.0
19.38
540
240
0.97
1.03
1977
892
44.4
20.09
664
324
0.90
1.11
1978
935
45.4
20.69
746
370
0.84
1.19
*FTE = Full Time Equivalent Employees
- 51
of 71%.
At the same
time, there has been a very significant
increase in both payroll
a net
cost
reduc tion
in the
(from an average o
tests/dolla r in 1978.
inflation
and non-payroll costs resulting
mber
of tes ts produced per total
1.21
tests/ dollar in 1972 to .84
is c ange reflects a significant
and
in both payro
non-pa yroll
costs.
For
has
ncreas ed from $7400 in 1972
in 1978 refl
ting
increa ses
fr inge benefit
(inc
uding
example, th e cost/F.T.E
to $15,270
tion and
-
retirement) .
social
security and
Cos s show ed an even more remark-
Non-payro
able six-fo Id increase
in direct compensa-
ring this p eriod,
v
volume incr eases and hia her reagent
cated and expensive tests.
predominantly reflects the
The net
reflecting both
costs for more sophisti
increase
in cost/test
inflationary factors, and
real dollars probably reflects a modest increase
ductivity related to automation.
in
in pro-
-
Chapter 4.
4.1
MYOCARDIAL INFARCTION STUDY
Chemi stry Test
Utiliza tion
The a verage num ber of
in the study grou ps
from 1963
52 -
to 1978
chemi st ries
performed on patients
(described in s ection 2.3),
is illustrated
at
in Figure 4.1
intervals
The
& 4.2.
creases
in total
chemistries seems b roadly divisible into
several
phases.
An early growth ph a se from 1963 to 1967
in-
(phase I) during which the average n umber of tests per hospital
stay increased from about 30 to 50 t ests,
an
ximately 66% duri ng this four year
(or an average
exponential
There was
rate
little
p eriod
of growth of appro x imately 14%
growth
the
averages
in 1977.
rate of increase
of approximately
in 1975, and
to
1 6% per year durina
v
I
•
is even m ore rema rkab e
considering that
the average length of hosp ital
has
this period.
reduced over
(phase 2)
This co n stituted an exponential
aDDroximatelv 140
This gro wth rate
1973
rose to a p proximately 75
tests/admission in 1973, about 90 t e sts
period.
per year).
in testing f rom 1967 through 1970,
but a second majo r growth spurt is e vident by
during which
increase of appro-
These data are
stay
this
i
1
- 53
4.1
FIGURE
Average Number of Tests per Hospital Stay
('95%confidence limits)
200
+
150
E-
U)
U)
LL
0 IO
G 100
H
z
50
B~tr
_
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1963 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
YEAR
- 54 -
FIGURE 4.2
Average Number of Tests per Hospital Stay
(- 95% confidence limits)
200
100
50
20
191
191
YEAR
*
SEMI-LOG PLOT
- 55 -
increase
indicative of a marked
laboratory usage
in
for the
nations also
increase parallels the
is no t
Figure
"LBC"
group
These
almos t ident ical
is not
to
surpris
stitu te a ve ry
larg
1972
to
1977
infarction
the myocardial
t
determina tions
of
the early
in
hospital-
in the
evident
4.3 summari zes
creat inine)
This
as
from
platea u observed
The anom a ous
study
wide
rise
phase 2 of
(corresponding to
study).
a rel ati ve plateau
a rapid
followed
1970s
Those determi-
as a whole.
hospital
sho wed
intensity of
these patients.
The overal l pattern of
behavior
the
in
BUN,
(electrolye
data
revea 1 a pattern
that
seen
ng
e so-called
e data for
for
f growth
chemistries.
tot
the
and
since t his group o
tests con-
to
1 clinical
proportio n of the
In contrast,
the changes
chemi stry te sts per
ormed.
the a verage numbers
of serum enzyme determinations did
not
s how
the
case of the
same m arked
LBCs..
in the
evident
marke dly different
in this
group
of patients,
testing waIs not
under
the
indicates that,
enzymes
This
growt h rate
same
pattern
the
growth
serum
stimulating
_I
r
-
56
FIGURE
4.3
Average Number "LBC s" and Serum Enzymes
per Hospital Stay
I
^'
II
10
"LBC"
---SERUM ENZYMES
-
9
8
7
6
5
4
2
I
\9
\9~96:
.Ik
\41'
:1\
- 57 influences observed with the LBC or in the hospital
population at
and might provide some
increased
intriguing observation,
This is an
large.
insight into the causes of the
laboratory utilization, since,
in this group
of patients, the enzymes are used for diagnostic purposes
to the monitoring functions of the
(in contrast
LBC determinations - see
In the period
below).
before the application of serum iso-
enzyme techniques in 1977
(which permit the differenti-
ation of isoenzyme patterns of serum enzymes
that are
characteristic of different diseases and which might
cause elevations
and
in their overall
levels),
multiple
serial enzyme determinations were required
clearly establish the characteristic sequential
to
pattern
of serum enzyme increase and decline associated with
acute
myocardial
infarction.
Thus,
the pattern and
frequency of serum enzyme analysis was not as susceptible
to variation as were other varieties of tests, particularly the "LBCs".
4.2
Intervals Between Tests
Ordering patterns of clinical
tests have undergone
significant changes during the period of study.
A major
- 58 increase in the frequency of testing resulting
change is an
in progressively shorter intervals between consecutive tests.
is evident
This shift
cy of various
from Figure 4.4 which
shows the frequen-
intervals between consecutive tests, for electro-
BUN and creatinine determinations. Although the most
lytes,
intervals between tests
frequent
in 1963
is one day, very sig-
nificant numbers of tests were ordered at
some with
intervals,
intervals greater than 6 days. As a consequence the
average interval
between tests in 1963 was 4.2 days. The
tervals between consecutive tests fell
the subsequent periods of study,
with
requests for daily determinations. By
were requested at daily
interval
in-
progressively during
increasingly
frequent
1978, almost all
intervals, with an additional
tests
signifi-
requested on the same day. The
cant number of repeated tests
average
longer
between tests
is reduced to .87 days
in 1978.
The patterns obtained for the serum enzymes exhibited some
differences
(Fig.4.5). The tendency to orde r these tests pre-
dominantly at daily or 2-day
intervals was
1963, but there was a slower progression
already evident
in
in the direction of
daily ordering of these tests. The average
interval
between
individual
1.7 days
in 1963,
but the
serum enzyme determinations was
interval shortened to 1.1
It should be noted that the
days by
increase
1978 (Table 4.1).
in daily testing
was not solely a consequence of the availab ility of automated
instrumentation, since a significant number of instances of
i--
- 59 4.4
FIGURE
INTERVALS
BETWEEN
CONSECUTIVE
TESTS
Electrolytes, BUN, Creatinine
la-CE
1963
)
w
O.,
z
C
L 40
WJ 40
40
1ir
ni-
ci 30
u-LL 30
Li
> 20
- z 20
1-J
0
l1,
LLJto
I 2 3 4 5 6 >6
INTERVAL (DAYS)
0
1973
£o
70 I--
,)
oz 50
Lu
w
C3
Lu
w
I-
L-
C',
u, 30
ir
1978
60
z
M
I 2 3 4 5 6 >6
INTERVAL (DAYS)
40
+
-
w
Lu
I-.
Ui
-J
W
ui-
B
IwU
IC
W
0
I 2 3 4 5 6 >6
INTERVAL (DAYS)
II
I1
0
-U-
I 2 3 4 5 6 >6
INTERVAL (DAYS)
FIGURE 4.5
INTERVALS BETWEEN CONSECUTIVE TESTS
Serum Enzymes (SGOT, SGPT, LDH, CPK)
-rr`
IOU
--
r-
IOU
7
1963
1973
1968
100
00
50
50
1978
100
E-
50
0I
II.
2 34
-
m
5
6>6
0
.
I.
1234
INTERVAL (DAYS)
SI
56>6
I
0
1 2 3
4 5
6>6
- 61
-
TABLE 4. 1
WEIGHTED INTERVALS BETWEEN TESTS
Year
1978
"LBC"
(Electrolytes, BUN, Creatinine)
Intervals (Days)
.87
Serum Enzymes
Intervals (Days)
1.1
1975
1.6
N.D.
1973
1.7
1.5
1970
2.9
N.D.
1968
2.9
2.0
1963
4.2
1.7
-
to this
-
observed in 1963 (admittedly less frequen-
daily testing was
tly than
62
in 1973).
Thus although the laboratory could respond
intensity of demand
in 1963, these daily requests were
significantly less common before 1973.
4.3 Frequency of Abnormal Tests
number of tests that were found
The percent of total
be abnormal
(Table 4.2).
to
was determined at each of the intervals studied
With respect to the serum electrolytes,
creatinine, there was a generally progressive
percent of abnormal
tests between
1963 and
BUN and
reduction in the
1977.
Comparable
changes were not observed for the serum enzyme determinations,
since myocardial
infarction
results
of serum enzyme abnormalities and
in predictable patterns
is consequently not as open
to such fluctuation.
4.4 Test-Ordering Patterns
(Grouping)
With the development of
4.4.1-Frequency of test combinations:
multichannel
could
instruments such as
the Technicon SMA-6, which
provide simultaneous determinations of multiple differ-
ent laboratory tests from a given sample, the question arises
whether such
instruments stimulate the physicians to request
groups of tests (according
instruments)
merly).
rather than
to their availability from these
individual
tests
(as was common for-
We approached this question by examining the fre-
quency with which the combinations of the 6 "LBC"
bined
in the SMA 4+2) had been ordered
tests
(com-
simultaneously, during
- 63 -
TABLE 4.2
Percent Abnormal Tests - Electrolytes, BUN, Creatinine
Yea r
% Abnormal
1963
1968
1970
1973
1975
1977
48
37
39
34
24
28
- 64 the period prior to the acquisition of that
and whether the frequency
instrument
increa sed after this
time.
the ordering patte rns
Figure 4.6 illustrates
1972,
fo r ele c-
trolytes, BUN and creatinine wit hin the study g roup a t
ent periods
in the study.
The f requency with w hich g roups of
either 1-3,4,5 or 6 of these tes ts were ordered
is plotted for each year of the
study.
indicate that, although the
simul taneo usly
(These figure s are
in each sam ple.
adjusted for the number of patie nts
data
dif fer-
)
The
frequency of o rderin g all
inc reased followin g the
the six tests simultaneously
sition of the multiple determina tion
of
acqui
instrument , this
patt ern
was already evident as early as 1967 and was we 1 es tablished
by
1970,
two years before the SM A-4+2 was opera
iona I in the
laboratory.
In 1963, the majority of test
tions of 1 to 3 of the
for sing le te sts.
in
six
By 1968,
requests were for combina-
tests; many of these were
requests
a majority of tests was o rdered
group s of 6, with a sign ificant frequency of combi nations
of 4 or
5 tes ts.
By 1970,
the trend has progressed
further, with combinat ion o f 4 or more tests now the
and the
combi nation of all
six becoming more common.
the prop ortio n of comb inati ons of six tests had
further with
categori es.
only smal I num bers of test
B y 1978, a Imost
test battery, with
tests;
all
relative ly
requests
s till
rule,
By 1973,
incre ased
in o ther
tests are requested a s the six
infrequent groups of I or 2
groups of 4 or 5 tes ts had become uncommon.
FIGURE
4.6
Group Ordering of "LBC s"
CV
15
0
LL
a
U-
10
5
h.
1963
Ij
1967
1968
l
1970
YEAR
1973
1975
1977
1978
- 66 This clearly
indicates that,
in this sample, the prac-
tice of ordering the combination of six tests
dated the availability of the multichannel
sion
("LBC") ante-
SMA 4+2.
to offer a combination of these 6 tests seems
an attempt by the
instrument manufacturers to be
The decito reflect
responsive
to a pre-existent propensity to order a combination of these
tests.
While the availability of this combination in a single
instrument may have encouraged the ordering of tests
bination, the
this
instrumentation did not
in com-
initiate the trend,
in
instance.
4.5 Comparison of the Frequency of Diagnostic and Monitoring
Functions
The clinical
varies within
utility
and purpose of laboratory tests
the hospital stay.
During the
initial
period of
hospitali zatio n, the laboratory tests are used predominantly
to assist the
clinician in the d iagnosis of a
In the ca se of acute myocardial
certain
infarction,
serum enzyme abnormaliti es
is
an
patient's
illness.
evidence of
important
factor
in
the diagn osis, and serial determ inations during the first 3
hospital days
is common.
Other tests are also useful
different ial d iagnosis as well
a s evaluation of the function
of other organ systems which may
generally
orde
red
within
the
fi
in the
be impaired;
these
also are
ew days of hospitalization.
In contra st to this diagnostic
zation, tests obtained
during the
ization tend to serve a
"monitoring
remainder of the hos
function", assistin
the eva luation of the
- 67 response to therapy.
patient's current condition and
We have analyzed
the laboratory uti
ization data
respect to the incidence of diagnost c or monitoring
with
tests ; the data are presented in Figu re 4. 7.
of th is comparison, diagnostic
all
tests
For the purpose
are d efined to include
t ests ordered within th e first 3 days of hospital ization,
as we 11 as any test ordered
3-day period.
All
for the f irst time after this
others a re termed "moni toring"
tests.
As
can b e seen, the numbers of both moni torin g and diagnostic
tests
per hospitalization rose throug hou t
Howev er,
the proportion of monitoring
the test period.
test s rose
from 1975 to the present, d espite the drop
significantly
in length of stay.
Whi le
the ratio of diagnost ic to moni torin g tests varied
1.0 -
1.2 from 1963 to 1973 , the rati o declined
and f inally to .60
the
i ncreased
in 1977
and
1978.
The data
to .75
from
in 1975
indicate that
intensity of monitoring activities constitutes
a maj or element
in the
incr eased labo ratory utilization
in
recent years.
Still
another demonstratio n of this phenomenon
is evi-
dent by comparison of the averag e number of tests performed
during the first and
4.3).
In general,
the number of tests
present.
tests
1978.
last 5 days of the hospital
stay
(Table
the data show a progressive increase
in these two
Considering the
in
intervals from 1963 to the
initia 1 5-day period,
increased from an average of 3.8
the number of
in 1968 to 21.8
in
Similarly, the number of tests ordered during the
last
;___
68 FIGURE
4.7
Average Number of Chemistries per Patient
as Diagnostic or Monitoring Functions
eUU
DIAGNOSTIC TESTS
O MONITORING TESTS
150
>
100
90
80
70
60
50
40
30
20
10
LL
1963
r,
L
67 68
Y
oL
70
73
YEAR
Y
75
Y
Y
77 78
- 69 -
TABLE
4.3
AVERAGE NUMBER OF TESTS
FOR FIRST AND LAST FIVE DAYS OF HOSPITAL STAY PER PATIENT
Avg.# Tests
per day per
patient
Year
1963
67
68
70
73
75
77
5 days after
admission
3.8
4.7
6.2
6.9
9.7
11.0
14.9
5 days prior
to discharge
0.25
0.8
0.29
0.49
1.2
1.4
4.1
5.3
7.8
7.9
3.6
4.1
First 5 days
Last 5 days
15.1
5.8
21.3
14.1
78
21.8
-
70 -
five days of hospitalization increased progressively
in l 963 to 5.3
an average of 0.25 tests
value
in 1978
from
(the
is anomalous, result ing from a si n gle patient
967
i
whom
a large number of tests we re ordered ju s t before disc arae)
The significant
flects the
increase f rom
large increases
period (Figure 4.3).
The
1975-1977
predominantly re-
in LBCs obse r ved during the
rat ios of fir
a marked decline, indicati ng that late,
s
t/last days shows
sustained monitoring
is progressively increasin g a:s a propor t ion of the total.
A third approach to dem(onstrating
the changes
orderin g pattern from 1963 to the present
in test
involved the
compari son of the average number of tests ordered on each
day of the hospitalization.
periods are
illustrated
The data for 4 of the test
in Figure 4.8.
In 1963 and
1968,
the hig hest test frequencies occurred within the first 4-6
days of hospitalization, with averages falling below I test/
day soo n thereafter.
By
1973, the
initial
levels of testing
were su bstantially greater and, thereafter, persisted for a longer
duratio n, while averages
in the
range of 2 tests/day were
maintai ned throughout most of the period of hospitalization.
By
1978 , this pattern had become even more marked; very high
levels
were maintained for the first week, and moderate
levels
(in the
range of 6 to 8 tests/day) were sustained
through out most of the remainder of the hospitalization.
FIGURE 4.8
Averaae Number of Th.sts
3
flrdri.rdr
I
2
2
2
2
1963 &--
2
1968
1973
1978
[][
I
3 4
bO
IT
IU II II IO 1 1I
DAY OF HOSPITALIZATION
-
---
-
72 -
This also indicates that early progressively higher
of sustained monitoring activity is also evident
recent groups
(in 1973 and particularly
The time after admission required
level of testing was calculated
This result
in the more
in 1978).
to reach th e average
for each of the fou r years;
it decreased progressively from 6.9 days
in 1978.
levels
in 1963 to 6.0 days
is consistent with an
increas ed
level
of monitoring activity, giving rise to this shift o f the "day
of average testing"
to an earlier period after admi ssion.
- 73 -
Chapter 5. DISCUSSION
5.1
Laboratory Tes t Utilization
Our study of the total
led an exponential growth
1977,
chemistry test
in tests over the
in this medi cal center.
able to that obser ved
This growth
utilization
revea-
period of 1971p attern
is
compar-
in laboratory testing throughout the
(7, 8, 21,
United States, dur ing the same period
22).
Since
most of these stud ies provided aggregate da t a, summarizing the
behavior of many h ospitals or of all
a given hospital,
tors
they may be
labora t ory tests within
influenced by a variety of fac-
resulti ng from uncontrolled changes
lation.
For
in the patient
examp le, one might argue that
increases ar e the result of changes
in the
some of thes e
average seve rity
of illness, and th at patients now tend to b e relatively
ill than
in
prior
years.
What
is more like ly
patients who
had b een admitted in the past
work-up, for
exa mple) wou Id now obtain
popu-
more
is that s ome
(for a diag nostic
thes e tests as o ut-
patients due to increasin g pressure fr om PS RO organizat ions.
One might al so
s
imp rovement
peculate that with
in thera py(such
patients with chronic dis eases
as antibioti cs)
seriously
are kept ali ve
for longe r periods bef ore f inally succu mbing
to their ill
particularly
ness
These
facilitated
the
ill pat ien ts
terminally
intense trea tment and
might also s pecu late that
have
ill
1aborato ry mon
since both Medica re and
availability
of
ould
require
tori ng.
One
Medi caid
health care to two groups
- 74 (the elderly and the poor) who might be expected to have more
serious health problems, these have affected the case-mix of
the hospitals
in an adverse way.
In order to avoid these sources of variability, we have
also elected to study the
though
small,
lab utilization
group of patients.
uncomplicated myocardial
in a well-defined,
They were selected as having
infarction,
and are as close to a
homogeneous group of patients as we were able to identify.
If"
.
•
selected t his gr oup or patie nts since the diagnosis
ment
invol ved a substantial
utilization of clinical
and
We
treat-
laboratory
services, partic ularly chemi
ry.
some expan sion
d variety of laboratory tes
s
the period of this study.
or
useful
in the range
in these
patients ove
Furthermore, there has
example, s erum c reatinine ph
phokinase
came
ry
into
use in this
labora
in
isoenzyme determ ination of s
um la
and CPK wa s inst ituted
t 197
in ab
bout
serum
ic deh ydrogenase
(LDH)
The
also heavil
of the hea vy use
of serum el ectrolytes
and blood gases
different
Both showe
general
inc rease
on the se patients
in
beca
se
(par ticularly p otass
un
in their the rapeutic monito ring.
The data s uggest
tion seen
enzyme firs
1970, while
automation would
impac
(CPK)
been
that
his define
the
increased
laboratory utili
ho spital group wa s not subst ant ia ly
the behav or of the general hospital
remarkable
with exponential
increase
a-
gro wth
in the chem istry
in the range of
po pulat on.
tes ting,
14% per year from
- 75 1970 to 1977, a figu re which
is in general
experience throughou t the country.
crepancy was evident
at a greater ra
th an
the myocardial
inc remental
to about
the
From the
ported that
the
doubled every f
(The only significant dis-
in 1973 and 1974 when total
a consequence o
120% o
agreement with the
1972
increases
tests grew
infarction patients, as
in hospitals beds up
level).
d l 950s through the 1960s,
some hospitals
re-
umb er of laboratory tests ordered per patient
e y ears.
Typical was Yale-New Haven Hospital
if"
which in 1954 p rto med 48,000
98,000;
and
in
increase
in pat
hospital
testin
964
approximately 2 bil
in 1959,
200,000 associated with only a slight
ent census
ex
laboratory procedures;
anded
i
(23).
at
on
tests
a
In the 1970s
slightly
performed
the rate of
faster
in
pace.
From
the
1971
numb
r
climbed to about 3 b illion in 1974 and to roughly 5 billion
(24).
1977
Of these
tests, at a cost
5.2 Factors
5.2.1
approximately 30% were clinical chemistry
in 1977,of approximately 3 billion dollars.
Influenc ing Laboratory Utilization
Patient Factor s
The major pati ent
related fa ctor
is the nature of reim-
bursement for health care delivery ; namely
the increase
in
third party payment
for health car e with t he advent of Medi-
care and Medicaid.
In 1974, for e xample, 21%
bills
18%
in
of patients'
in the hospita 1 under study were rei mbursed by Medicare,
by Medicaid, 42% by Blue Cross and onl y 19%
themselves or through commercial
insurance.
by the patients
-
Technolog Iical
5.2.2
Factors
Laboratory factors
liability
76 -
include increased capabi lity
and
re-
of the results, and the development of automated
multichannel
analyzers.
Two majo )r technological
factors may
affect laboratory utili zation : th e introduction of new tests
without
retiring older
tests and
analyzers, particularly
those
the
impact of automated
whi ch perform a profile of tests
which the physician wou Id not
oth erwise have
study of both the total
ord ered
tests
ordered.
in the chemistry labora-
tory and tests ordered for a grou p of myocardial
patients has shown that 23 te sts
test volume.
The total
test
utilization
chemistry
account for over 90% of the
ou nted
Howe ver, most
o
10% represents a growing
f the growth
in laboratory
is d ue to greater us e of established
Two major
techno logical
labora tory d uring
ad v ances occurred
the
test volume for
Ilowing
LB
both
event
tests.
in the analyzed
p eriod studied:
duction of the f irst S MA6 in 197 2 and
GEMSAEC in 1976.
infarction
lume has, of course, been grow-
ing with time so that t he una
number of tests.
Our
the
he second
intro-
SMA 6 and
the rate of increase in
, SMA6-I I che mistr
es and enzymes was slightly
greater than wou Id
ave been expected
rom prior trend data.
It may be sugges ted
that suc h technolo
ical
a small,
albeit
shor t lived
of laboratory te sts.
LBCs
(1-2 year)
The te ndency for
advances do have
effect
in escalation
physicians to order
in groups of six was established before the
of the
SMA6
in
1972;
this
introduction
profile-ordering pattern certainly
- 77 became dominant
in subsequent years, which may be attributable
to the pre s ence of the SMA6.
b asic question which must be a nswered regarding the
The
impact of t echnology
is whether test uti lization expands to
the capaci t y of the laboratory or whethe r the laboratory
primarily k eeps pace with
other word
ical
s
,
is laboratory util ization s ubject
(inst r ument
Our
pull.
question.
induced) push
LBCs had apparently
to a technolog-
(physician demand)
is that
reached a plateau prior to
in 197 2 , but gr ew rapidly afterwards.
it may be argued
response t o an unmet
clusion
or a need
In
data do not provide an unambi guous answer to this
introducti o n of the SMA6
However,
the d emands of the physician.
that
the SMA6 was acquired
physician demand.
test
laboratory
in
Our subjective con-
volumes are largely determined
by physician need and that new automated technology has a
minimal causative effect.
5.2.3 Physician-Related Factors
Specialization:
play the predominant
oratory utilization.
The physician-related factors seem to
role
in the remarkable
The effect of a high degree of sub-
specialization has already been discussed.
hospital
used
increase in lab-
in the study,
In the particular
the degree of subspecialization
is rather extreme, with a high degree of subspecialization in
internal medicine, pediatrics and surgery and very few general
- 78 internists,
pediatricians or surgeons on
is essentially full-time
(and until
employees of the hospital); all
the medical
school,
the staff.
The staff
recently were salaried
have teaching appointments at
and many are actively engaged
in clinical
research.
There
is a very active house s taf f training program,
complemented by extensive fellowship
ities
tr aining
(particularly in internal medi cin e).
for ordering of laboratory tests
in subspecial-
Responsibility
res ts predominantly with
the house staff; daily rounds are he Id with the attending
physician on
service for the month, at which
status of e ach patient
is reviewed.
recognized as a n excellent hos pital
ical
12
traini ng;
Th e hospital
is generally
in which to obtain clin-
several hundred appl ica n ts apply each year for
internsh ip p ositions
tial
time the current
in med icine .
This
implies a substan-
degree of independence of the hou s e staff, in whom major
responsibil ity for day-to-day patient c are is vested.
medical
ser vice s are entirely
divisions
( e.g.
The
organize d into subspecialty
cardiology, he matology
nephrology, endocrin-
ology, gast roen terology, etc.) and all
patients are assigned
to one of t hese , as appropriat e.
resence of medical
The
student "ex tern s" on each serv ice may
oratory uti liza tion somewhat,
natural
cur iosi ty.
end to increase
lab-
in order to satisfy their
This degree of subspecialization undoubt-
edly contributes to the exceptionally high level of laboratory
- 79 testing.
lab-
Lack of effective interaction/consultation with
oratories:
is no formal
There
between clinical
laboratory directors and the clinicians/
house officers who utilize the
have been made
interaction/consultation
laboratory data.
to intervene or otherwise
No attempts
influence the order-
ing of laboratory tests by clinicians/house officers
than attempts to discourage unnecessary "stat"
(other
[urgent]
requests for exceptionally rapid reporting of results).
On
the contrary, the laboratory has attempted to respond to the
increased demands with
increased services, automation, and
more extensive coverage to almost 24 hour service, seven days
a week.
Malpractice:
physician related factor
increasing threat of malpractice suits,
is the
an
Another general
resulting in
increase in "defensive medicine", a term applied to the use
to avoid litigation.
of diagnostic procedures
felt compelled,
testing
in cases
in some instances,
in which
Physicians have
to carry out laboratory
the physical examination and history
have already yielded an unquestionable diagnosis. The major
increase
the
in rates began
in 1972 and may have contributed to
increased laboratory utilization evident at that time.
Influence of technology:
A much more significiant
factor relates to the attitide toward technology shared
physician and patient alike.
by
(For a comprehensive discussion
- 80 see Reiser's book "Medicine and the Reign of
of this point,
Technology"'' [4 ).Reiser notes that
the
increased
reliance
technologically-produced data at the expense of the physical
I
exam inati
cine
for
I D
and hi
number
tory h as be en a matter of concern in medif yea r s.
In 1921,
Tieken wrote, "For years,
phys icans and many of ou r well -known teachers have leaned so
much towards la borat ory m ethod s that physical
almost become a lost
art"
(2'5)
icantly increas ed si nce t hen.
accelerating by the
of medicine see med
of disease and
(4).
and clinical
and the dependency has signifBeginning
in the 1930s
and
the newer generations of professors
1950s
i ntere sted
principally in the biochemistry
"a nu mber shift ed almost directly from the lab-
oratory into pr ofess orial
schools"
diagnosis has
chai rs at hospitals and medical
The p hysic ian/s cientist,
steeped in laboratory
investi gatio n, wa s the model
academic physic ians
Furthermo re,
against which all
were measured.
a not he r factor
in th e increased attachment
to machine-prod uced evidence occurring during the 20th
century
"originated
in part
from contemporary faith
in scien ce
and technology, and a belief that a scientific spirit entered
clinical
practice through technology"
provided by means of a sophisticated
(4).
Laboratory data,
instrument,seemed to
satisfy the desire of the physician for precisely expressed
clinical evidence.
Progress
in medicine was measured by some
according to the degree to which quantitative measurements
- 81 replaced qualitative clinical
an
judgments.
At
the same time,
increasing proportion of health costs were borne by pri-
vate and government insurance, which also covered patients
for
laboratory and x-ray exami nations
creasing
to almost 75%
of the
(t he propor
populat ion
by
1974
is likely that, wi th litt le cos t constrain t,
and patient encour age the
technology in the
both
in-
26]).
physician
incre ased utiliz ation of laboratory
det ecti on and evaluation of
thermore, as a res ult
ion
illn ess.
of increa sed automat ion,
the
Fur-
labora-
tories tend to be increasingly cost e ffect ive, gen erating more
tests at a reduced real cost/test.
usually addressed,
A more crucial
is the question of the extent to which the
biochemical change s in blood
serum re flect the cri tical
alterations which a re occurring within
the diseased cells.
The ready availabi lity of serum shoul d not obscure
that
which
i
only
-eflects
indirectly
re more crucial
t
This h ah expectation of
to the older belief
the
the
the fact
intra cellular events
underst and ing of dis ease.
serum is, in some sense, analogous
(now discredited) that
position of the urine could prov ide major
nature of the disease processes
the che mical cominsights
While
in genera
of the characterization of serum elements
real
issue, not
into the
the value
is obvio us,
its
limitations should also be appreciated.
Behavioral
factor
Most marked changes seem to ref lect
behavorial characterist ics of the clinician - particular ly
-
evident
in the myocardial
the
Most noticeable was
levels, but the
increased
intensity of
there a marked increase
Not only was
istry testing.
-
82
infarct study.
chemin total
intervals b e tween consecutive tests p rogress-
ively shortened so that,
by 1978, most tests are orde red daily
(and the average intervals
thermore there are
greater
i s shortened
Fur-
to below I da y).
l evels of sustained monito ring
activity throughout the cou r se of hospitalization. Wh ile there
was some shortening of aver a ge
days),
this does not accoun t
for the markedly greater testing
levels throughout the hospi t al
(CCU
).
cardiac status also seems t o
in recent years. One facto r
The constant monitoring of
ead to an u rgency reg
t he current
other monitoring activities
daily orders for "LBCs"
stay
crisis orientation" evide nt within
undoubtedly relates to the
the Coronary Care Unit
length of stay (from 2 2 to 15
proced ures
en zymes for a 11
and
pat ient
ding
nclude
(orders
are written o n the day of admission).
Another
remarkable feature of the g rowth in cl nical
chemistry tes ts
is that the proportion of LBC/total
relatively un changed
(from about 52%
in spite of a n almost 3-fold
suggests an
in 1972 to 56%
emains
n 1978)
increase in testing. Th s also
i ncrease in intensity of test ing,
rather than a
change in q uality.
The increasing "packaging" of the LBC
remarkable. Although there
to combine BUN & crea tinine
is not a strong
requests
is also
rationale,a priori,
requests with e lectrolytes
(Na,
- 83 K,
Cl
and
CO2)
ing them.
which
ment.
Thi
this is now the preponderant manner of orderpattern antedated
the advent of the SMA-6
linked these 6 determinat ons
This
expec ed
linkage
is
more
in a multichannel
instru-
rema rkable when one considers the
rate of change in these
two groups of tests.
Serum
potas
ium levels might undergo qiuite rapid shifts, under cer-
tain
linical
circumstances, but
expec ed to show relatively
under
level
slow (and predictable) alterations
most clinical circumstance s.
might
require more
rapid
atini
changes.
requi
different freauencv of
different
practice of
5.3
BUN and creatinine would be
In
investigation
general,chai nge
freau
....
vI
Consequently, changing K
in these 2 groups should
s tudy and,
s
than BUN/cre-
consequently, the
universally grouping these tests seems to be questionable.
Determination of Appropriate Levels of Laboratory
Utilization
As
is evident from the foregoing discussion, there are
a great many causes of
increased laboratory utilization over
the past 30-40 years.
We have, thus far, avoided
whether this
remarkable growth
indicative of unneccessary and
be vigorously discouraged.
some of this growth
in utilization
is necessarily
inappropriate usage and should
While most people will
agree that
is unnecessary, there may be significant
differences of opinion on how much of it would fit
category.
the issue of
into that
Short of a case-by-case review of hospital
records,
how can the question of the appropriateness of laboratory
- 84 What sort ot evidence migh t be utilized
testing be resolved!
in establishing evidence of excess
unnecessa ry usage?
ve,
One simple approach
would be to determine
'
of abnormal
and normal
tests yielding normal
This simple view
values, wi th the
values were
is unfortunately
implication that the
likely to be unnecessary (15)
open to objection.
conversion of a test value from a bnormal
cate the efficacy of therapy, whi le
to normal
in other
normal
Fur thermore,
The
may indi-
instances,
required to ascertain that certai n undesirable
of therapy had not occurred.
the proportion
it is
side effects
in some instances,
values are very valuable i n resolving certain diagnos-
tic problems
(27).
For example, if alternative A has a higher
incidence of abnormal
values with
tion B, a normal value provides
the
given test than condi-
ind irect support favoring
condition B.
At
mal
the other
extreme, an
in
tests results does not necess
oratory utilization.
It might al
eased proportion of abnorily indicate effective
reflect unnecessary
lab-
lab-
oratory testing consequent to exc
sively frequent and
tious requests
were already demonstrably
abnormal,
for the tests whic
and could not
reasonabl
significantly altered
in the inte
dition of "necessity"
is introduc
the data (in terms of its
even
the
abnormality
initiati
repeti-
be expected to have been
al.
If an additional
con-
requiring "usefulness"
some change
of
in treatment),
of the test result would not guarantee
- 85 its "necessity".
Another approach to the establishment of appropriate
levels of laboratory utili zation
is the development of "group
norms" which would establi sh some level
a co rrect standard.
plic ity,
but
of current practice as
This has the advantage of relative sim-
it does not
ing that the "normal"
resolve the difficulty of establish-
behavior is actually appropriate.
This brief discussion of the problem of establishing an
uneq uivocal method of establishing the "proper"
level of
labo ratory utilization, for different categories of diseases
and
indicates
patients
moni toring schemes
test
5.4
the problems
inherent
in any sample
(such as determining percent abnormal, or
frequency).
General
Recommendations
The best alternative suggested
by a number of authors
~I
(3, 9,23,28)
involves
a controllin g role of the clinical
oratory directors and clinicians
ing of laboratory usage.
i n the education and monitor-
Educatio nal
programs might
the appropriate use of particular tests,
normal values and relation of test
as
involve
significance of ab-
results to disease, as well
the costs of laboratory tests; these programs would be
directed to medical
staff.
as
lab-
There
these can
programs
students, hous e officers and attending
is some evidence tha t educational
programs such
influence laboratory utilization, but these
must be repeatedly
reviewed to remain effective (29).
- 86 Key
individuals in the implementation are the directors of the
clinical
as
laboratories,
who would need to play an expanded role
laboratory consultant, advising on
tion of laboratory tests,
the use and interpreta-
design diagnostic test strategies
for different types of disease, and establish a test monitoring system to provide continuing review of current
laboratory
utilization.
The fundamental
problem of the consequences of increased
dependence of the clinician upon diagnostic technology still
remains unresolved.
Its use can enlarge the physicians'
knowledge of disease, but they may rely too much on such technology, and not enough on diagnostic capability based upon
their own ability and experience.
"risks becoming merely an
and
the medical
machines"
(4).
judgments
Otherwise the physician
intermediary between the patient
rendered by technical
experts and
- 87 -
REFERENCES
1.
Social
2.
4.
5.
Benson,
and What
Interest, S ummer 40 (1977).
Medicine and the Reign of Tech nology,
I T
Reiser, S.J.,
(1978).
H.,
Editors,
Logic
Laboratory Use,
Elsevier
N.Y.
E.S. and Rubin,
of Clinical
6.
"The High Cost of H ospitals
it", The Public
Cambridge
1974.
(1977).
Feldstein, M.S.,
do About
N ew York,
"The Twilight Zone", M edical Marketing and
R.L.,
Media, June 17
3.
Inc.,
Basic Books,
Choice.
Hughes,
Health , Economics and
Live?
Who Shall
Fuchs, V.R.,
Twine, E.H. and Po tchen,
and Economics
(1978).
E.J.; M.S. Thesis, Sloan School
of Management, M.I T. (1973).
7.
Zucker,
B.,
Editor
laboratories".
8.
Zucker, B.,
"National
La b oratory Management, March 33-4
chemistry laboratory
(1976).
survey of hospital clinical
Benson, Ellis, S., Strategies for
clinical
c linical
survey of hospital
La b oratory Management, May 17-39
Editor
laboratories".
9.
National
8(1979).
improved use of the
in patient care in
(5)
245-259, 1978.
10.
Freier, E.F.
chemistr
11.
Skeggs,
and R ausch, V.L.
Am. J.
Med. Techno
Quality contr ol
. 24195-207,
L. T. and Hochstrasser, J.,
sequential analysis.
Multiple
Clin. Che . 10:918-936,
in clinical
1958.
automatic
1964.
- 88 -
12.
Griner, P.F. and Liptzin, B.
teaching hospital:
implications for pa tient
education and hospital
163,
Use of la borat ory
costs.
Ann.
Int
in a
care,
Med
75:
157-
1971.
13. Martin,
S.P.,
Donald son, M.C.,
O.L. and Coulton, T.
30 years.
Jordon, C.D.
Peterson,
in corona ry ca re during
Outputs
A cost effectiveness study.
Ann.
Int.
Med.
t al.
Use
81:289-293, 1974.
14.
Schroeder, S.A.,
Kenders,
of laboratory tes ts and
K.,
Cooper, J
pharmaceuticals
riation
among physicians and eff ect of cost aud
use.
15.
JAMA 225:96 9-973,
Sheeley, D.R. and
resource use:
85:648-652,
16.
subsequent
1973.
Sherman, H.
Conservation in hospital
tr eatment of pneumonias.
Ann . Int.
Med.
1976.
Schroeder, S.A. a nd O'Leary,
D.S.
Differences
in
laboratory use an d length of stay between university and
community hospita Is.
17.
Skeggs,
L.T.,
analysis.
18. Theirs,
Jr.
J. Med.
Educ. 52:418-420, 1977.
An automated method for colorimetric
Am. J. Clin. Pathol.
R.E.,
in Clinical
Technics, edited by R.J.
28:311,
Chemistry,
Henry,
D.C.
1957.
Princi Iples and
Cannon and J.W.
Winkelman, Harpe r and Row, New York, 2nd ed ition,
Chapter
10,
1974
- 89 19. Moreland, F.B.
in Clinical
Methods, e dited by
Diagnosis by Laboratory
I. Davidsohn and B.B. Wells, Saun ders
Philadelph ia, 13th edition,
20.
Anderson, N.G. Analytical
XII.
A mul tiple-cuvet
system.
21.
A naly.
Chapter 8, 1962.
techniques
for cell
Biochem. 28:545, 1969.
Laboratory Management 21-34, March
22. Mohr, J.W. , Editor, National
Dis.
Clinical
survey of clinical
17-19, May 1971
laboratory automation.
J. Ch ronic
19:50 9, 1966.
24. Feinberg, H.V.,
tests.
The high cost of low-cost diagnostic
Su n Valley Forum on "Medical Technologist
The Culpri t Behind Health Care Cost?",
25.
laboratories.
1973.
Laboratory Management
23. Seligson, D.,
o
rotor for a new microanalytica
Mohr, J.W. , Editor, Survey of hospital
laboratori es.
fracti ons.
-
1977.
Tieken, G. , A plea for better understanding of physi cal
diagnosis.
JAMA 76:1736,
1921.
26. Mueller, M .S. and Piro, A.,
in 1974.
insurance
A review of coverage, enrollment and finan cial
experience .
27. Gorry, G.A .,
diagnostic
Private health
Soc. Secur.
Pauker,
Bull.
S.G.
and
39:9,
1976.
Schwartz, W.B.,
importance of the normal
finding.
The
New En g.
J. Med. 29 8:486-489, 1978.
28.
Sheinbach, J.,
The clinical
laboratories problems
cost containment as a challenge,
33-38, op.cit.
(5).
-
29.
Eisenberg, J.M.,
90
-
An educational
laboratory use by house staff.
578-581,
1977.
program to modify
J. Med.
Educ. 52: