Five Years' Experience with the Evaluation
of Diuretic Agents
By
CHiARf,EIS SWARTZ, M.D., ROBERT SELLER, M.D., MORTON Fucis, M.D.,
ALBERT N. BREST, M.D., AND JOHNI
H. MOYER, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
PRIOR TO the advent of potent oral diuretics, pareliteral mercurial drugs were
the mainstays of diuretic therapy. In the
search for orally effective diuretic agents, several events stand out. In 1949, Schwartz1
reported the diuretic effect of sulfanilamide
in patients with congestive heart failure. In
1953, acetazolamide, a sulfonamide derivative
and a potent carbonic anhydrase inhibitor,
became available for clinical use.2' 3 Then, in
1957, Novello and Sprague4 synthesized chlorothiazide, a benzothiadiazine compound. The
availability of the thiazide drugs represented
a major breakthrough in the search for potent
oral diuretics. Subsequently, in the continuing search for the ideal diuretic agent, the
pharmaceutical industry has made available
a large number of additional drugs. There
are at least nine analogues of the basic benzothiadiazine molecule now available or soon to
be released for clinical use. Also introduced
during recent years were chlorthalidone (a
phthalimidine diuretic), quinethazone (a
quinazolinone), and spironolactone (an antialdosterone agent). With such a plethora of
agents it is obviously important for the practicing physician to know which drugs are
assay method involved and its clinical application are also considered.
Methods and Materials
The investigation of each diuretic agent was
divided into three parts: inpatient bioassay, outpatient acute weight loss stud(ly, and ehronici outpatient administration.
Inpatient Bioassay
In a group of "normal" hospitalized patients,
the response of the urinary electrolyte (sodium,
potassium, chloride) excretion that occurs during
the acute administration of the drug was studied.
The subjects had a wide variety of primary diagnoses but were excluded from the study if any
clinical or laboratory evidences of renal disease,
cardiac decompensation, or edema were found.
The patients ranged in age from 20 to 70 years.
They were placed on a standard 50-mEq. sodium
diet for at least 5 days prior to drug administration. It has been our experience that this technic
produces a relatively constant sodium and potassium excretion. Daily 24-hour urinary excretion of
sodium, potassium-l, chloride, and creatinine were
measured for at least three control days. When
urinary sodium excretion reached 90 per cent of
dietary intake, the diuretic was administered in
different dosages to different subjects. The 24-hour
urinary electrolyte excretion was measured and
then compared with the control values. By employing varying dosages, a dose-response curve for
natriuresis was obtained and the maximum effective dose of a particular diuretic was determined.
outpatient Acute Weight-Loss Study
The studies were performed in an outpatient
diuretic study group that included 30 cardiac patients who were in chronic mild to moderate conrestive heart failure. All patients were on maintenance digitalis therapy and previously required
Dne or more weekly injections of niercurial diuretics in order to nmaintain an edema-free state.
Prior to the study of an individual drug, all diuretic therapy was discontinued for at least 2
weeks. In order to determine the weight-loss poteney of a diuretic, the patients were examined
and weighed, and a diuretic was administered for
18 hours. Weights were recorded before breakfast
superior.
During the past o years, we have investigated the clinical pharmacology of the newer
diuretic agents by a standardized methodology. The analysis of the data obtained in the
evaluation of 12 different diuretic agents
plus 2 combinations of drugs comprise the
body of this report. The sensitivity of the
Departmlent of Medicine, Hypertentsion- Rental Unit,
Hahnemann Medical College and Hospital, Philadlelphia, Pennsylvania.
Supported in part by grants from the Hahnemann
Cardiovascular Research Center (NIH H-6368) and
the Heart Association of Southeasterni Pennsylvania.
1042J
Circulation, Volume XXVIII, December 1903
EVALIUATION OF DIURETIC AESENTS
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
aiidI after voidinig. Up)on ompllies)ltionl of 2 (lays o-:
liuretic therapy, the patients wN-ere re-examined
and re-weighed. Cha nges in weights were considered to approxiniate fluid depletion. After 5
days of placebo therapy, a different dosage of the
diuretic being investigated was adnministered in
similar fashion. Thus a dose-response curve by
weight loss (fluid depletion) was obtained and a
maximum effective dose was determined.
Chronic Outpatient Administration Study
The clinic patients who participated in the acute
weight-loss study were also utilized in this investigation. The drug being studied was administered
for a 6-week period at its nmaximum effective dose
(i.e., the dose above which no greater natriuresis
or acute weight loss could be elicited). This study
was designed to provide information about clinical
toxicity and serum electrolyte changes that occur
during long-term drug administration.
Statistical Methods
In order to compare drugs with different milligram potencies, only data obtained at the maximum effective dose of each drug were utilized. The
maximum effective dose of an agent was defined
as that dose above which no increase in natriuresis
or acute weight loss could be obtained. The maxinum effective doses of the drugs under consideration are listed in table 1. The bioassay data
dealing with natriuresis, kaluresis, and acute
weight loss were subjected to an analysis of variance. Since the number of patients in each sample
was not the same, an average N was computed
according to the method of Snedecor.5 The standard error of a group mean between individuals
was computed by dividing the mean square by
the N of that group of data. The imieans were
then ranked and a Duncan multiple range test was
applied with the use of the 95 per cent confidence
limits.6
The data obtained from the study of chronic
administration of the drug to outpatients did not
lend themselves to interdrug statistical analysis;
therefore they will not be included in this report,
but some impressions that were gained from this
experience are described.
Results
The raw data are tabulated in tables 2, 3,
and 4. The analysis of variance is recorded
in tables 5, 6, and 7. In tables 8, 9, and 10,
the means are ranked and subjected to multiple F tests at the 95-per cent confidence
limits. For a detailed description of this
method, the reader is referred to Duncan's
original paper.6 In essence, the method sumCirculation, Volume XXVIII, December
1963
104;3
marizes imiultiple tests of signiificanlce between
group means. If table 8 (the sodium bioassay
data) is used as an example, 13 means must
be analyzed for significant differences. Although the familiar t test could be applied
to any two means in order to derive statistical
significance, this would require 156 separate
calculations. Instead, the pooled error (standard error) of a group mean was utilized, This
was 14 mEq. for sodium, 8 mEq. for potassium, and 0.47 pound for acute weight loss.
As demonstrated in table 8, the means of
24-hour sodium excretion above control were
ranked in ascending order and a line was then
drawn connecting those means that could not
be separated by chance alone. This resulted
in three overlapping lines. Any two means
not connected by the same black line were
significantly different at the 95-per cent level.
Again referring to table 8, it is seen that the
mean sodium excretion (above control) for
the combination of trichlormethiazide plus
meralluride was 227.3 mEq. per 24 hours. This
was not connected by a line to any other mean
and was therefore significantly greater than
any other regimens studied. At the opposite
end of the scale, benzthiazide has a mean of
88.8 mtq. per 24 hours and was not significantly less than any means underlined by line
c. However, the response of sodium excretion
to benzthiazide was significantly less than
the response to cyclothiazide and all the values
Table 1
Maximal Effective Dose of Diuretics Evaluated
Chlorothiazide
Chlorthalidonie
Quinethazone
Benzthiazide
Hydrochlorothiazide
Hydroflumethiazide
B3enzydroflumethiazide
Methyclothiazide
Cyclothiazide
Trichlormethiazide
Polythiazide
Meralluride
o.d.-once daily
b.i.d.-twice daily
i.m. intramuscularly
1,000 mg. b.i.d.
200 mg. o.d.
200 mg. o.d.
100 mg. b.i.d.
100 mg. b.i.d.
100 mg. b.i.d.
10 mg. o.d.
10 mg. o.d.
4 mg. b.i.d.
4 mg. b.i.d.
4 mg. b.i.d.
2 ml. i.m.
SWARTZ ET Al.
11044
Table 2
Acute Bioassay o( Sodiumn Excretioni (m.-Eq. per 24 Hours above Control)
aL)
a)
)
"0
C:5
a)
a
N
S._
N
a)
.0
X.
0
N-
N
.9
.
r.
N
0
10
IV
"0
0
0.
._
0
55
x11;
No. of Pts.
6
115
83
72
110
78
75
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
88.8
Meain
8
40
177
85
83
39
117
173
72
81
152
120
10
90
10
39
6Q)
11
13
1
9070
99.2
98.3
1177
104.0
103
112
813
64
51
605
101
7
5
109.3
113
120
129
157
164
170
127
9
111.6
9
12
108
195
55
189
154
148
188
194
180
71
117
168
174
145
153
126
193
208
189
145
163
99
127
134
140
57
146
51
185
152
134
116
198
22,6
2159
188
151.8
152.0
62
117
140
195
151.7
127.5
125.0
10
8
319
10
259
81
165
2-01
206
1.70
254
127
235
158
28
.292
208
153
153
139
259
1
69.7
227.3
Table 3
Acute Bioassay of Potassium Exeretion (nEq.
per
21
Hours abore Control)
"0~~~~~~~~~~~~~~~~~~~~~
x~~~~a
w
N0 of
C
No. of Pts.
2'8
53
-11
-10
8
7
12
5
5
-22
Mean
10
10
7.5
-13
9
63
8
-18
4
18
12
13
-
8.7
1
84
30
-21
6
36
49
-12
21.6
18
52
29
10
56
65
9
27
10
69
31
14
62
30.8
35.3
19
37
27
17
17
9
14
19
14
29
53
19
26
62
22.0
10
to the right of eyelothiazide, since no single
line conneeted any one of these to benzthia-
zide. Trichlormethiazide, with a illeani of 109.3
niEq. per 24 hours was connected bv line b to
$.0
11
21
56
16
25
36
45
13
54
58
53
34
37.4
8
9
22
18
77
105
0
43
28
37.8
9
37
2
53
20
29
79
47
46
60
41.4
-
6
12
7
51
56
6
66
54
17
73
62
69
27
20
60
77
81
31
21
41.7
46
0
76
59
78
24
15
6I
41
47.4
50.3
all drugs up to the coinbiination of hydrochlorothiazide plus chlorthalidone and the conmbination of nieralluride plus trichlornmethiazide.
Thiis indicated that triehliormethiazide proCirculation, Volume XXVIII, December 1963.
EVALUATION OF DIURETIC AGENTS
1 045"a
duced significantly less natriuresis than either
combination of drugs. On the other hand, trichlormethiazide was also connected by line c
to benzthiazide; and this indicates that it
cannot be considered to be significantly more
potent than any of the drugs evaluated. As
another example, polythiazide with a mean
sodiunm exeretion of 152 mEq. per 24 hours
was connected by line a down to methylelothiazide and by line b to trichlormethiazide.
It was also conliected by line a to the combination of hydrochlorothiazide plus chlorthalidolne. It can therefore be seen that polythiazide is a more potent natriuretic agent
than chlorothiazide, benzhydroflumethiazide,
quinethazone, or zenzthiazide and a less potent natriuretic than the combination of meralluride plus trichlormethiazide. This technic of summarizing tests of significance is
repeated for potassium in table 9 and acute
weight loss in table 10.
The maximum effective dose of each of the
various diuretics was administered daily to
outpatients for 6-week intervals. Althougb
the data obtained in these studies do not lend
themselves to statistical comparisons, certain
clinical observations can be made. All the
oral drugs tested were capable of maintaining
this group of patients clinically "dry" and
free from signs and symptoms of congestive
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Table 4
Acute Forty-Eight Hour TVeight Loss in Pounds
N
'a
N
0
a
Na
N
p:
4a
Eq
P._,
0
No. of Pts.
13
25
3.00
3.25
1.75
0.50
3.00
+1.25
0.25
3.50
1.50
2.75
1.00
30
1.50
2.00
2.00
3.00
2.75
3.00
2.50
1.50
1.00
4.50
2.50
0.50
2.75
±1.50
3.50
3.75
2.00
0.75
5.75
3.00
1.50
4.50
1.50
N0
a
.0
_~
4aa
'a
al
12
20
3.75
1.75
2.25
0.75
3.00
2.75
1.50
+0.25
5.00
4.00
0.75
0.50
3.75
1.00
1.50
4.50
7.00
1.75
2.00
0.50
3.75
1.00
7.50
1.50
5.25
3.00
4.75
+10.75
2.00
I
I
+1.50
5.00
2.00
0
0.75
2.25
1.25a
1.50
2.49
2.19
1.91
Mean
Circulation, Volume XXVIII, December 1963
2.69
.4
'
a
m
x
26
15
18
11
19
10
1.00
5.50
1.00
4.25
2.50
3.50
2.00
1.00
1.00
6.00
6.00
1.75
1.75
2.00
3.75
5.50
1.75
6.75
8.50
6.00
1.25
0.25
3.75
0.50
3.75
3.75
3.00
3.00
3.00
2.50
4.00
3.00
2.75
3.00
4.00
5.75
4.50
3.00
3.00
1.50
5.50
2.50
3.75
4.50
6.50
2.25
+0.25
3.00
6.50
3.00
5.00
11.50
7.50
3.50
3.55
4.45
2.50
0.25
3.25
6.00
0
2.50
0
8.00
5.25
3.50
4.00
3.50
±0.75
1.50
4.00
1.50
1.00
2.50
4.00
4.75
2.00
3.50
2.75
4.50
1.00
1.25
2.78
6.00
7.00
8.50
4.00
1.00
4.50
+2.00
2.00
+2.25
3.50
1.50
1.00
5.00
1.50
+2.50
3.03
3.17
---2.00
4.00
2.50
6.25
8.00
0.50
1.00
3.29
o
4.00
+1.25
1.00
4.00
7.00
±0.50
5.50
5.25
2.00
5.00
6.00
4.25
3.32
s'a
C
15
SWARTZ ET ALJ.
1046
Table 5
Analysis of Variance-Bioassay Data on Sodium
Degrees
Source of
variation
of
freedom
Mean
Sum of
squares
square
332,634
117
Total
11,481
12
137,774
Groups
Individuals
1,856
105
194,860
in groups
F = 5.85
P < 0.01
Standard error of group mean = J1850 = 14.4 mEq.
Table 7
Analysis of Variance-Acute TVeight Loss Data
Source of
variation
Total
Groups
Individuals
in groups
F = 3.07
Degrees of
freedom
Sum of
squares
183
10
749
173
636
113
Mean
square
11.3
P < 0.01
Standard error of group mean
=
3168 == 0.47
lb.
9.0
Table 6
Analysis of Variance-Bioassay Data, on Potassium
Source of Degrees of~~~~~
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Degrees of
freedom
Source of
variation
Total
Groups
Individuals
in groups
F = 3.08
P < 0.01
Standard
error
of
Sum of
Mean
squares
square
107
11
19,989
96
56,591
group mean
76,580
=
4-58 9
1,817
=
8.1 mEq.
heart failure. No significant differenees were
noted in the frequency of side effects among
the various drugs. Though electrolyte changes
were inconstant, there was some tendency
toward hypokalemic alkalosis. In addition,
there was a trend toward elevation of the
serum uric acid.
Discussion
Comparison of drug efficacy should be conducted preferably in a clinical setting that
will duplicate the ordinary use of such agents.
With regard to diuretics, this implies that
comparisons should be made in edematous
patients. Although this would seem to be an
ideal method, it involves certain practical difficulties. In particular, the severity of the
disease processes underlying the edema varies
so greatly from patient to patient that this
factor alone imposes sufficient spread, or variance, of data to obscure the effects caused by
the differences among drugs.
If the drugs under study have a physiologic
action on normal persons, as do diureties,
then the easiest way to minimize the variability of data is to perform the assay in persons in whom the factor of different degrees
of illness has been remioved, i.e., "normal"
subjects. In this sense, the spread of data
will be the minimum biologic variation present in any sample. A comnmonly employed
study group of "normal" or homogeneous
populations are healthy young medical students or laboratory technicians. Although
this group provides definitive data with which
to separate drug effects, they in no way represent the usual population for whom diuretics are intended. Therefore we have accepted
a compromise between the unacceptable wide
variance of data in patients with disease and
the less meaningful narrow variance of data
obtained in "normal" persons. In our studies we utilized a hospitalized population with
wide age spread and varying states of health
in whom evidence of renal disease or derangement in water or electrolvte metabolism was
absent.
There has been some discussion in the recent literature concerning the type of diet to
he employed during diuretic bioassay.7 A 50mEq. sodium (3-gram salt) diet was employed
in the present studies because it provides the
minimum sodium content consistent with palatability. hfigher salt intakes make it difficult
to achieve stable control excretions of sodium
and potassium. Although this low salt diet
may stimulate minimal endogenous hyperaldosteronism in subJects formerly accustomed
to higher sodium intake, it is consistent with
the elinical situiations in which diuretics are
CJirc4lation, Vohne XXVI!!, December
1963
EVALUATION OF D:IURETIC AGENTS
1047
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
employed, i.e., dietary salt restriction and
disease states frequently characterized by hyperaldosteronism. This technic has provided
bioassay data with a standard error of the
mean of 15 mEq. for sodium and 8 mEq. for
potassium (for samples of size nine). This
situation still, represents considerable variability and therefore this method of bioassay
can separate drugs into only a few large
categories.
To confirm diuretic efficacy under conditions of clinical usage, acute weight loss was
evaluated in a clinic population of 15 to 30
patients in mild to moderate congestive heart
failure. The average sample size is 17, and
the standard error of this sample was onehalf pound. Unfortunately this technic also
allows the drug to be categorized into only a
few groups.
It is not surprising that there is no correlation between weight loss and natriuretic
potency. This discrepancy occurs because
acute weight reduction reflects acute water
loss, and water loss does not necessarily parallel sodium loss. Thus the mercurial diuretics
cause the greatest acute weight loss but are
only moderate in natriuretic potenicy. Mercurial drugs tend to increase free water clearance and produce large amounts of dilute
urine.8' 9 Thus the mercurials are the diuretics of choice in the treatment of edema complicated by the presence of hyponatremia. On
the contrary, the thiazides decrease free water
clearance and produce a more concentrated
urine than do the mercurials.10 It is this attribute of the thiazides that has commended
Since the
their us.e in diabetes insipidus.1
mercurials cause a greater water (weight)
loss per mnilliequivalent of sodium excretion,
assays based on weight loss alone show inercurials to be most potent whereas those studies based on sodium excretion alone show the
thiazides and phthalimidine diureties to be
most potent.14-16
In addition to the difficulties involved in
choosing proper subjects for evaluation, the
problems imposed by the biologic variation in
response to a given drug, and the difficulties
inherent in using different measures to determine potency (i.e., weight loss versus sodium
excretion), there is still another limitation to
be considered. The methods of study employed are applicable only to the comparative
evaluation of diuretic agents that have a relatively rapid onset of action, i.e., marked increase in natriuresis within 24 hours of administration and decrease in body weight
within 48 hours. Thus these technics do not
lend themselves to the evaluation of such
drugs as the aldosterone antagonists. These
compounds do not elicit significant natriuresis in acute bioassay studies nor do they produce significant weight loss when administered
for only two days to patients with uncomplicated congestive heart failure.
It is our opinion that no single agent has
Table 8
Multiple F Tests-Ninety-Five Per Cent Confidence Limits; Bioassay Data on Sodium
a)
a)~~~~~~~~
a)
00
I.
~~~~~~~~~~
0
SIean 88.8
;
Drug
98.3
Ca
99.2~~~~~~
m 10 v4.0
o0.
11.
0
~~~~~~~~~0
12.
15.
12.
5.0
5.
16.7
22
(a)
(b)
*~
~
~ ~
~
~
~~~~~~~~____
(c)
Note: Any two means not underscored by the same line are significantly different.
Any two nieans anderscored by the same line are not significantly different.
Circulation, Volume XXVIII, December 1963
SWARTZ ET Al.
1048
Table 9
Multiple F 1Tests-Ninety-Five Per cent Confidence Limits; Bioassay Data
on Potassium
N~~~~~~~~~~~
C)~~~~~~~~~~~~~~~
4
0
Drug
Mean
8.7 21.6
22.0 30.8bytesm
35.3 37.4
37.8 41.4 41.7zee47.4
Noe
An1 7.5
two men
not W?,6'SO6
ieaesgcnl
50.3
Note: Any two means not underscored by the same line are significantly different.
N~~
i.Q line are not sigqificantly different.
Q
Any two means
tonderscored by .othe same
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Dru
,
.;
N
0
Table
10
m
Multiple F Tests-Ninety-Fire Per Cent Confidence Limits; Bioassay Data onl Acute
Forty-Eight Hour WVeight Loss
_E
V
0
X~ H
a
0~~~~~~~~
C;
It
IC
N
~~~~~~~t
0~~~~~~~~0
0
Drug.
Drug~~
.0
o
.Q
0
N
N
N=
7 40
~~~
0
.0~ ~ ~ ~~
Mean
1.91
2.19
2.49
2.69
2.78
3.03
3.17
3.2-9
3.32
3.55
4.45
Note: Any two means not underscored by the same line are significantly different.
Any two means underscored by the same line are not significantly different.
proved itself to be the ideal diuretic and that
insufficient evidenee exists to judge onie agent
superior to all others. In our studies all the
oral diuretics evaluated proved to be potent
agents capable of aehieving and maintaining
an edema-free state in patients with mild to
moderate congestive heart failure. The fact
that so many agents are comnmercially available reflects the continuing search for the
ideal compound, which hopefully will provide
maximal natriuresis and diuresis, minimal
kaluresis, and few or no side reactions. The
practicing physician nmay take comfort in the
knowledge that lie can choose fromri a larrge
group of potent agents with proved elinical
efficacy.
Summary
Twelve diuretics and 2 combinations of
diuretics were evaluated by a standard methodology. Electrolyte excretion patterns were
evaluated in "normnal" hospitalized patients,
and acute weight loss was evaluated in outpatients in congestive failure. The large
standard errors of the means for natriuresis,
kaluresis, and acute weight loss reflect the
inherent biologic variation of such assays and
permit separation of drugs into only a few
overlapping categories of potenicy. The combinatiol] of a merculrial and a thiazide caused
significanitly grreater sodiumll excretion than
any other drug used alone. No single diuretic
was significantly more potent than all others.
Circulation, Volume XXVIII, December
1963
1049
EVALUATION OF DIURETIC AGENTS
Acknowledgment
We are inidebted to Dr. S. Free, Chief statisticiani,
Researeh and Developmiient Sectioin, Smlith, Kliine and
French Laboratories for his assistance in preparing
the statistical analysis. We would also express our
appreciation for the technical assistance rendered
by Miss Ellen Lippuinuan and Miss Regina Burns.
References
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
1. SCHWARTZ, W. B.: Effect of sulfanilamide oIn
salt and water excretion in congestive heart
failure. New England J. Med. 240: 173, 1949.
2. BELSKY, H.: Use of new oral diuretic, Diamox,
ill coigestive heart failure. New England J.
Med. 249: 140, 1953.
3. FRIEDBERa, C. K., TAYMOR, R., MINOR, J. B.,
AND HALPERN, M.: Use of Diamox. a carbonic
anhydrase inhibitor, as an oral diuretic in
patients with congestive heart failure. Newv
England J. Med. 248: 883, 1953.
4. NOVELLO, F. C., AND SPRAGUE, J. M.: Benzothiadiazine dioxides as novel diuretics. J. Anm.
Chem. Soc. 79: 2028, 1957.
5. SNEDECOR, G. W.: Statistical Methods. Ames,
Iowa, The Iow^a State College Press, 1946.
6. DUNCAN, D. B.: Multiple ranige and mi-nultiple F
tests. 'Biomnetrics 11: 1, 1955.
7. M ARTZ, B. L., AND COKINos, G. C.: Diuretic
assay utilizing normiial subjects. Clin. Pharm.
Therap. 3: 340, 1962.
8. PORUSH, J. G., GOLDSTEIN, M. H., EISNER,
G. M., AND LEVITT, M. F.: Effects of organoinercurials oni the reiial concentratifng meclba-
nism in lhydropeniic nian. Coiiiniients on site of
action. J. Clin. Investigation 40: 1475, 1961.
9. LAMnIE, A. T., AND ROBSON, J. S.: Effect of
mersalyl on renal tubular reabsorption of
solute free water. Clini. Se. 20: 123, 1961.
1 0. HEINEMANN, H. O., DEMAARTINI, F. E., AND
LARAGGH, J. H.: Effect of chlorothiazide on
the renal excretion of electrolytes and free
water. Am. J. Med. 26: 853, 1959.
11. KENNEDY, G. C., AND CRAWFORD, J. D.: Treatment of diabetes insipidus with hydrochlorothiazide. Lancet. 1: 866, 1959.
12. HARVA-RD, C. W. H., AND WOOD, P. H. N.: Effect
of diuretics on the renal water exeretion of
diabetes iitsipidus. Cliln. Se. 21: 321, 1962.
13. DIES, F., AND RIVERA, A.: Possible mechanism
of the antidiuretic effect on the benzothiadiazinie derivatives. Clin. Pharmn. Therap. 3: 172,
1962.
14. GOLD, H., KUIRT, W. T., MESSELOFF, C. R.,
KRAMER, M., GOLFINOs, A. J., MEHTA, D.,
ZAHM, W., AND WARSHAW, L.: Comparison of
hydroflumethiazide and meralluride. New
method for quantitative bioassay of diuretics
in bedfast patients with edemiia. J.A.M.A. 177:
239, 1961.
15. FORD, R-. V.: Clinieal phlarim-acologie investigation of a newv beazothiadiazine diuretic,
CMR-807. Ani. J. Cardiol. 5: 407, 1960.
16. FuhcHis, M., MOYER, J. H., AND NEwAVAIN, P,. E.:
Human clinical pharmacology of the newer
diuretics: benzothiadiazine and phthalimidine.
Ann. New York Acad. Se. 88: 795, 1960.
William Withering
Withering testified hiimself to his excellent upbringing. As respects his education, it
was not remarkable. He received a good grounding apparently in the classical languages
from a neighboring clergyman, the Reverend Henry Wood of Ereall, and passed through
the usual course of study in mathematics, geography and history, necessary for entrance
into the University. He seems to have been a good student but showed little in the wa.y
of either precocity or brilliance of intellect. His father desired him to study medicine and
his own inclinations were toward his father's profession. In 1762, at the age of twentyone, he entered the University of Edinburg, then becom-ling celebrated for the excellence
of its medical school.-Louis H. RODDIS, M.D. William Withering: The Introduction of
Digitalis into Medical Practice. New York, Paul B. Hoeber, Inc., 1936, p. 4.
Circulation. Volume XXVIII, December 1956
Five Years' Experience with the Evaluation of Diuretic Agents
CHARLES SWARTZ, ROBERT SELLER, MORTON FUCHS, ALBERT N. BREST
and JOHN H. MOYER
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Circulation. 1963;28:1042-1049
doi: 10.1161/01.CIR.28.6.1042
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1963 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/28/6/1042
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Circulation can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of
the Web page under Services. Further information about this process is available in the Permissions
and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/