Bactericidal Activity and Toxicity

of Iodine-Containing Solutions in Wounds

George Rodeheaver, PhD; William Bellamy; Michael Kody; Grace Spatafora;
Leyden; Richard Edlich, MD, PhD
Lois Fitton; Keith

\s=b\ Complexing iodine with povidone (polyvinylpyrrolidone) or MATERIALS AND METHODS

surfactants significantly limits the quantity of free iodine. Reduc-
tion of the free iodine level eliminates the adverse properties of
staining, instability, and irritation and also alters bactericidal
activity. Addition of detergents to create surgical scrub solu-
tions further reduces the activity of iodine. In vitro testing
indicated that the bactericidal activity of iodophors was inferior
to that of uncomplexed aqueous iodine. In vivo tests proved that
aqueous iodine significantly potentiated the development of
infection. Although the povidone iodophor did not enhance the
rate of wound infection, it offered no therapeutic benefit when
compared with control wounds treated with saline solution.
Addition of detergents to the povidone iodophor was deleteri-
ous, with the wounds exposed to this combination displaying
significantly higher infection rates than untreated control
wounds. Based on these results, aqueous iodine solutions and
iodophor surgical scrub solutions should not be used on broken
skin. Aqueous iodophors can be used in wounds, but no thera-
peutic benefit from such use was found in this study.
(Arch Surg 1982;117:181-186)
Iodinegenerallong
Its
has been regarded as an excellent germicide.1
use has been restricted by its low solubility
in water (36 mg/L) and caustic reaction with skin and
mucosal membranes. Complexing the iodine with a suit¬
able polymer renders the iodine water soluble and stable.
These solutions of complexed iodine, called iodophors, have
very little free iodine, which makes them stable, nonstain-
ing, and nonirritating.2 Detergents have also been added to
iodophors to generate surgical scrub solutions.
Without complete documentation of safety and efficacy,
these iodophor solutions have been used to cleanse trau¬
matic wounds. Recent reports have questioned the validity
of using iodophors in wounds.3"5 The purpose of this study
was to evaluate critically the potential value of noncom-
plexed and complexed iodine solutions as antiseptic agents
for use in wounds.
Accepted for publication Aug 12, 1981.
From the Department of Plastic Surgery, University of Virginia Medical
Center, Charlottesville.
Read at the first annual meeting of the Surgical Infection Society,
Chicago, April 25, 1981.
Reprint requests to Plastic Surgery Research, University of Virginia
Medical Center, Box 332, Charlottesville, VA 22908 (Dr Rodeheaver).
Iodine Solutions
Three iodine solutions were evaluated: aqueous iodine, povi-
done-iodine antiseptic solution (Betadine), and povidone-iodine
surgical scrub solution (Betadine). Aqueous iodine was prepared
to contain 1% weight by volume (W/V) iodine and 2% W/V
sodium iodide. The iodophor solutions were obtained from com¬
mercial sources. Povidone-iodine antiseptic solution uses 10%
povidone (polyvinylpyrrolidone) to complex the 1% available
iodine. Adding a detergent to this antiseptic solution yields a
povidone-iodine surgical scrub solution.
Bacteria
The bacteria used during these experiments were Staph-
ylococcus aureus (CDC 2801, Centers for Disease Control, Atlanta)
and Escherichia coli (serotype 06; nonmotile). These bacteria were
maintained on tryptose blood agar base. Before each experiment, a
colony of the designated organism was transferred by a sterile
loop to 25 mL of trypticase soy broth and incubated at 37 °C with
agitation for 18 hours. The bacteria were harvested from the broth
by centrifugation. The bacterial pellet was washed twice with
0.9% saline solution and resuspended in a measured volume of
saline solution. This stock solution of bacteria was serially diluted
1:10 through eight tubes, and the number of bacteria in each tube
was determined by standard plating techniques.
Free Iodine Levels
In the formation of an iodophor, povidone or a surfactant is used
to complex and stabilize the iodine. A delicate equilibrium is then
established in which a small amount of free iodine exists that is
uncomplexed. This level of free iodine dictates the activity of the
iodophor. Free iodine levels were determined colorimetrically by
extracting the free iodine into an immiscible organic solvent.6
Spectroquality heptane (10.0 mL) was added to a test tube
containing 1.0 mL of iodine solution. This solution was agitated
and equilibrated at 25 °C. The amount of iodine extracted into the
heptane layer was determined by its absorption at 520 nm. Using a
standard curve, the concentration of free iodine in the iodophors
was determined in parts per million, taking into account that
iodine distributes itself between heptane and water in a ratio of
30:1.
Antimicrobial Activity of Iodine
Solutions (In Vitro)
Rate of Bacterial Kill.—Groups of six test tubes received 1.0-mL
aliquots of aqueous iodine, povidone-iodine antiseptic solution,

Downloaded From: by a UNIVERSITY OF ADELAIDE LIBRARY User on 01/05/2018

 povidone-iodine surgical scrub solution. In each test tube, 0.05 mL 11.00
of 0.9% saline solution containing IO10 E coli was added, and the
solutions were mixed. Using a calibrated loop, 0.01-mL samples -»· Aqueous Iodine
were removed from each tube at 10-s intervals and transferred to - Povidone-iodine Antiseptic Solution
1.0 mL of 0.9% saline solution containing sodium thiosulfate (1% )
9.00 -· Povidone-iodine Surgical
to neutralize the iodine in the samples. The number of viable
Scrub Solution
bacteria present in each test loop was determined by standard
microbiologie dilution and plating techniques. The sequential
monitoring of the viable bacteria that remained in each test
solution provided an indication of the rate at which each iodine 7.00
solution killed the bacteria.
Germicidal Capacity Test.—The germicidal capacity test7 mea¬
sures the capacity of an antiseptic to kill multiple challenges of
microorganisms. Escherichia coli were suspended in enriched
phenol coefficient broth (Association of Official Analytical Chem¬ 5.00-
ists procedure) to a concentration of 7.1 X 108 organisms per 0.05
mL. Tubes were prepared that contained 1.0-mL samples of each
of the iodine solutions. At designated time intervals, 0.05 mL of
bacterial suspension was added to each iodine solution. After one
minute of exposure, a calibrated loop (0.01 mL) was used to 3.00
subculture a sample of the iodine solution into 1.0 mL of nutrient
broth that contained sodium thiosulfate (1%). All subcultures
were incubated at 37 °C for three days and then checked for
turbidity. The inoculating and subculturing procedure was
repeated ten times for each test solution. The results were 1.00
-

reported as how many challenges the iodine solution could steril¬

ize within one minute before it failed.
Penetration of Agar.—Antibiotic Medium I agar was seeded with
E coli and dispensed into sterile Petri dishes (75 mL/150-mm
plate) to yield an agar thickness of 6 mm. After the agar had
solidified, 12 circular wells of a 4-mm diameter were created on
each test plate. Each well received 0.05 mL of either aqueous
iodine, povidone-iodine antiseptic solution, povidone-iodine surgi¬
cal scrub solution, or saline solution. All solutions were tested in
triplicate on each plate to eliminate any variability from plate to
plate. After incubation for 18 hours at 37 °C, the zone of bacterial
growth inhibition around each well was measured.
Penetration of Tissue.—The dorsal hair of anesthetized guinea
pigs was clipped with electric shears and then subjected to a
depilatory. After copious washing with water, their backs were
dried and swabbed with 70% alcohol. Using an aseptic technique,
a full-thickness piece of skin (thickness, 1.5 mm) was removed, and
test samples (10 X 12 mm) were obtained. Using a microliter
syringe, 0.05 mL of 0.9% saline solution containing IO6 E coli was
injected into each sample of skin. Each skin sample was placed in
1.0 mL of an iodine test solution for ten minutes before being
removed, rinsed with sterile water, and placed in 5 mL of 0.9%
saline solution containing sodium thiosulfate (1% ). After mechan¬
ical homogenization of the tissue, the number of viable bacteria
that remained in each skin sample was determined by standard
procedures.8
Activity of Iodine-Containing Solutions (In Vivo)
The activity of the iodine-containing solutions in vivo was
judged by their immediate effect on bacteria located at or near the
wound surface and by their therapeutic potential in reducing the
incidence of gross infection in contaminated wounds. Female,
Hartley guinea pigs (weight range, 300 to 350 g) were anesthetized
with an intraperitoneal injection of pentobarbital sodium (33
mg/kg). Their dorsal hair was clipped with electric shears and
removed with a depilatory. After copious washing with water,
their backs were swabbed with 70% alcohol and then draped for
surgery. Using aseptic technique, two paravertebral incisions
were made on each animal. Each incision was 3 cm in length and
extended through the panniculus carnosus. Any bleeding that
Test Interval, s
Fig 1. —Rate of bacterial kill in vitro was determined after single
challenge of Escherichia coli.
Antimicrobial Activity of Iodine Solutions in Agar Wells
Zone of Inhibition, mm
Povidone- Povidone-
iodine Iodine
Inoculum, Log/ Aqueous Antiseptic Surgical Scrub
mL of Agar Iodine Solution Solution
5.38 21 16 14
10
occurred was stopped by gentle pressure applied to gauze sponges
that contacted the wound surface. The wounds then received 0.05
mL of 0.9% saline solution containing a known number of S
aureus. After ten minutes, one wound on each animal received 0.05
mL of designated antiseptic solution. The contralateral wound on
each animal received 0.05 mL of 0.9% saline solution. After ten
minutes, the wounds were either closed with microporous tape or
assayed for the number of viable bacteria that remained.
To determine the immediate bactericidal effect of these iodine
solutions in contaminated wounds, the number of bacteria that
remained after exposure to the iodine was quantitated. Residual
iodine in the wounds was neutralized with 0.05 mL of sodium
thiosulfate solution (1% ). The wounds were swabbed with sterile,
cotton-tipped applicators, and the relative number of viable
bacteria that remained in the wound was determined by standard
techniques.9
To assess the therapeutic benefit of treating contaminated
wounds with an iodine solution, the wounds were closed with
microporous tape and bandaged and then reexamined after four
days. After killing the animals, each wound was opened and
examined for the presence of purulent exúdate. When pus was

Downloaded From: by a UNIVERSITY OF ADELAIDE LIBRARY User on 01/05/2018


4.27 ± 0.23
Fig 2.—Wounds exposed to aqueous iodine or povidone-iodine
antiseptic solution for ten minutes had significantly fewer Escheri¬
chia coli than contralateral control wounds. In contrast, treatment
with povidone-iodine surgical scrub solution was ineffective in
reducing level of contamination.
observed, the wound was judged to be infected. In addition, an
estimate of the number of the bacteria on the wound surface was
determined as described previously.
RESULTS
In the 1% aqueous solution of iodine, containing 2%
iodide, the free iodine level is 148 ppm. In an iodophor, the
iodine is highly complexed, and very little free iodine
exists. In povidone-iodine antiseptic solution and povi¬
done-iodine surgical scrub solution, the free iodine levels
were 1.67 and 0.42 ppm, respectively.
Solutions of iodine in a test tube were effective in rapidly
eliminating a single challenge of IO10 E coli (Fig 1). Because
of its large content of free iodine, aqueous iodine elimi¬
nated the bacteria more quickly than the complexed povi¬
done-iodine antiseptic solution or povidone-iodine sur¬
gical scrub solution. Aqueous iodine also displayed a
prolonged germicidal activity. A sample of aqueous iodine
was capable of sterilizing nine sequential challenges of
7.2 X IO8 E coli, while povidone-iodine antiseptic solution
only sterilized seven doses. In contrast, povidone-iodine
surgical scrub solution was incapable of sterilizing the
initial challenge.
Diffusion of the iodine solutions through contaminated
Downloaded From: by a UNIVERSITY OF ADELAIDE LIBRARY User on 01/05/2018
_ Aqueous Iodine
-o Control
Minimum Detectable Threshold
-//-
102 103 io4 io5 107
Inoculum
Fig 3.—Contaminated wounds treated with aqueous iodine had
significantly higher rates of infection and levels of bacteria than
wounds treated with saline solution.
agar and inhibition of bacterial growth was related to
their level of free iodine (Table). Thus, aqueous iodine had
zones of inhibition that were wider than those of povidone-
iodine antiseptic solution. Povidone-iodine surgical scrub
solution had the lowest degree of activity. The ability of
these iodine solutions to penetrate agar was not correlated
with their activity toward skin. The iodine solutions were
incapable of penetrating the skin and reducing the number
of bacteria injected. Despite a ten-minute exposure to the
iodine-containing solutions, all bacteria that were injected
into tissue were recovered.
The antibacterial activity of these iodine solutions was
evaluated in experimental animal wounds that contained
bacteria. Exposure of a contaminated wound to aqueous
iodine for ten minutes resulted in a 1.07 log,r> reduction in
the level of bacteria compared with untreated wounds (Fig
2). Similarly, povidone-iodine antiseptic solution signifi¬
cantly reduced the bacterial challenge by 1.31 log
(P < .001). However, the efficacy of povidone-iodine surgi¬
cal scrub solution was not statistically significant with the
reduction being only 0.3 log.
The bactericidal activity of aqueous iodine was insuffi¬
cient to compensate for the deleterious effects that
aqueous iodine had on the wound itself. Application of
 •-· Povidone-iodine Surgical Scrub Solution
100 a_ Povidone-iodine Antiseptic Solution o-o Control
o-o Control 100

80

60

40-

20-

7.5

6.0
^-*ß
to
m
4.5

3.0-

Minimum Detectable Threshold

1.5

-//-
102 103 10" 105 10b 10'
Inoculum Inoculum
Fig 4.—Treatment of wounds with povidone-iodine antiseptic solu¬ Fig 5. Use of povidone-iodine surgical scrub solution in wounds
tion offered no therapeutic benefit. Incidence of infection and was deleterious. When compared with untreated wounds, wounds
—

number of bacteria in this group of wounds were similar to those of exposed to solution had significantly higher incidence of infection
control wounds. and contained significantly more bacteria.

aqueous iodine to wounds containing 105 bacteria resulted COMMENT

in thedevelopment of infection in every wound (Fig 3). In
contrast, similarly contaminated wounds exposed only to
0.9% saline solution did not become infected. Even in a
group of wounds containing only 103 bacteria, application
of aqueous iodine produced an infection rate of 33%.
Complexing the iodine with povidone to form povidone-
iodine antiseptic solution eliminated the deleterious
effects of iodine. Wounds contaminated with 104 to 107
S aureus and treated with povidone-iodine antiseptic solu¬
tion had infection rates similar to those exposed to 0.9%
saline solution (Fig 4). Although povidone-iodine solution
was not deleterious to the wound, it was surprising that it
afforded no therapeutic benefit. Four days after treat¬
ment, wounds treated with povidone-iodine antiseptic
solution had the same level of viable bacteria as those that
received only saline solution.
When detergents were added to povidone-iodine antisep¬
tic solution to yield povidone-iodine surgical scrub solu¬
tion, an unacceptable solution resulted. Povidone-iodine
surgical scrub solution significantly potentiated (P < .002)
the development of wound infection when compared with
the incidence of infection in wounds treated with 0.9%
saline solution (Fig 5).
The results of this study demonstrate that aqueous
iodine and povidone-iodine surgical scrub solution signifi¬
cantly reduced (P < .002) the wound's ability to resist
infection. In the case of aqueous iodine, the inhibition of
wound defenses resulted from the presence of high levels
of reactive and caustic iodine. The damaging effects of
povidone-iodine surgical scrub solution on the wound's
defenses was probably related to its highly reactive deter¬
gents, which are cytotoxic. In the absence of these deter¬
gents, the resulting povidone-iodine antiseptic solution
was no longer debilitating to the wound, but the highly
complexed iodine offered no therapeutic benefit.
It is important for the physician to realize that the Food
and Drug Administration has not approved povidone-
iodine antiseptic solution or povidone-iodine surgical scrub
solution for use in wounds. In its review panel of Over-
the-Counter Drugs, the FDA concluded that insufficient
data were available to support the clinical safety and
efficacy of these complexed iodine solutions.10 Despite this
decision, the manufacturer continues to advocate the use of
povidone-iodine antiseptic solution in minor wounds. The
results of this experiment clearly indicate that povidone-
iodine antiseptic solution offers no therapeutic benefit in

Downloaded From: by a UNIVERSITY OF ADELAIDE LIBRARY User on 01/05/2018

 contaminated wounds.
In view of the extensive literature documenting the
antimicrobial activity of iodine, it may be difficult for the
physician to accept the fact that it is not effective in
treating wounds. However, a better understanding of the
pharmacokinetics of iodine in solution may be beneficial.11
In aqueous iodine solutions, the iodine is slightly com¬
plexed and thus yields a high level of free iodine. This high
level of free iodine in aqueous iodine accounts for its rapid
bactericidal action, as well as its irritating, corrosive, and
staining properties. The free iodine in aqueous solution can
be "tamed" by complexation with a compatible polymeric
compound. This complex is termed an iodophor, and solu¬
tions of these complexes have very little free iodine. In the
case of povidone-iodine surgical scrub solution, detergents
have been added that also complex the iodine and further
reduce the level of free iodine. Since the antimicrobial
activity of iodine solutions is directly related to their free
iodine level, the activity of iodophors must be significantly
different than that of aqueous iodine. An investigator's
ability to detect this difference is totally dependent on the
exact details of the testing procedure.
In most in vitro tests, a large bacterial challenge is
added to an aliquot of the iodine test solution and checked
for sterility after an extended time period. Because of large
dilutional effects and a long test interval, differences
between solutions cannot be detected. In our experiments,
the testing conditions were more critically controlled so
that the dynamics of each iodine solution could be better
determined. For example, in determining the rate of
bacterial kill, the bacterial challenge produced only a 5%
dilution, and the test solution was sampled every 10 s.
Under these conditions, the relative activity of each of the
iodine solutions was easily determined. Aqueous iodine,
with all of its iodine in the free state, eliminated the
bacteria within 10 s. When the level of free iodine was
reduced by complexation with povidone, as in povidone-
iodine antiseptic solution, the time required to kill the
bacteria was extended to 20 s. In povidone-iodine surgical
scrub solution, the added detergents further reduced the
References
1. Gershenfeld L: Iodine, in Reddish GF (ed): Antiseptics, Disinfectants,
Fungicides. Philadelphia, Lea & Febiger, 1954, pp 171-211.
2. Shelanski HA, Shelanski MV: PVP-iodine: History, toxicity and ther-
apeutic uses. J Int Coll Surg 1956;25:727-734.
3. Branemark PI, Albrektsson B, Lindstrom J, et al: Local tissue effects
of wound disinfectants. Acta Chir Scand 1966;357(suppl):166-176.
4. Bolton JS, Bornside GH, Cohn I Jr: Intraperitoneal povidone-iodine in
experimental canine and murine peritonitis. Am J Surg 1979;137:780-785.
5. Pietsch J, Meakins JL: Complications of povidone-iodine absorption in
topically treated burn patients. Lancet 1976;1:280-282.
6. Schmidt W, Winicov M: Detergent/iodine systems. Soap Chem Special-
ties 1967;43:61-64.
7. Cantor A, Shelanski HA: A 'capacity' test for germicidal action. Soap
Sanitary Chem 1951;27:133-135.
Discussion
John Heggers, PhD, Chicago: Through their endeavors, Dr
Rodeheaver and his colleagues have added another piece to this
most controversial puzzle. Like Lister's carbolic acid disinfectant,
iodine compounds are now experiencing the same notoriety.
Downloaded From: by a UNIVERSITY OF ADELAIDE LIBRARY User on 01/05/2018
level of free iodine with a commensurate extension in the
killing rate to 30 s. Although complexation of iodine
retarded its rate of bacterial kill, these complexes are
supposed to have prolonged germicidal activity. The
results of our germicidal capacity test did not support this
concept. Aqueous iodine was capable of sterilizing more
bacterial challenges than povidone-iodine antiseptic solu¬
tion. Povidone-iodine surgical scrub solution was unable to
sterilize the initial challenge.
For an antimicrobial agent to eliminate bacterial con¬
tamination, it must reach the bacteria in an active form.
Because of the insolubility of iodine in water and its rapid
complexation with tissue and body fluids, its ability to
reach and kill bacteria in a wound or tissue is highly
suspect. Bacteria present in tissue of only a 1.5-mm
thickness were unaffected by a 20-minute immersion in the
iodine solutions. Even the antimicrobial activity of iodine
solutions on contaminated wound surfaces was limited.
When contaminated wounds were exposed to aqueous
iodine solution for ten minutes, the bacterial level of the
wound was only reduced 1.07 log. Povidone-iodine antisep¬
tic solution was also capable of reducing the wound
contamination level by only 1.31 log. Even though these
reductions were statistically significant (P < .001) com¬
pared with untreated wounds, these treated wounds still
contained high levels of contamination. Povidone-iodine
surgical scrub solution was ineffective and did not reduce
the level of bacteria in the wound.
The limitations of iodophors have been further appre¬
ciated by recent reports of bacterial contaminants that
grow in commercially available iodophor solutions.12 This
discovery is a compelling argument for the sterile packag¬
ing of iodophors used in hospitals. Further investigation is
urgently needed to evaluate the safety and therapeutic
merit of all commercially available iodine-containing solu¬
tions. Until an iodophor has been proven to be effective
under simulated use conditions, its activity must be sus¬
pect. Until these studies are completed, it is inconceivable
to us how surgeons can treat wounds with these potentially
dangerous solutions.
8. Edlich RF, Rodeheaver GT, Spengler MS, et al: Practical bacteriologic
monitoring of the burn victim. Clin Plastic Surg 1977;4:561-569.
9. Edlich RF, Madden JE, Prusak M, et al: Studies in the management of
the contaminated wound: VI. The therapeutic value of gentle scrubbing in
prolonging the limited period of effectiveness of antibiotics in contaminated
wounds. Am J Surg 1971;212:668-672.
10. O-T-C topical antimicrobial products: Over-the-counter drugs gener-
ally recognized as safe, effective, and not misbranded. Federal Register
1978;43:1210-1249.
11. Rodeheaver GT, Trunbull V, Edgerton MT, et al: Pharmacokinetics of
a new skin wound cleanser. Am J Surg 1976;123:67-74.
12. Contaminated povidone-iodine solution: Northeastern United States.
Morbidity Mortality Weekly Rep 1980;29:553-555.
The authors' in vitro data concerning the microbicidal efficacy
of iodine-like compounds serves to confirm the data presented by
van den Broek and van Furth at the Second World Congress of
Antisepsis in 1980.
 I have several questions concerning the methods used by Dr
Rodeheaver and his associates. First, since Antibiotic Medium I
agar contains protein, is it not possible that more chemically
defined media would yield larger zones of inhibition?
It is well known that the free iodine irreversibly complexes with
tyrosine, and since this complex does form, then why is it
necessary to neutralize the tissues that are tested?
Another question related to tissue treatment revolves around
the period of time. In clinical trials involving severe peritonitis,
Groenewald irrigated the peritoneal cavity for 24 to 48 hours with
36 L of solution to include 1 L of povidone-iodine every three to
four hours. He reported a 6% mortality when compared with an
18.8% mortality for the control group.
Additional studies showed that with a povidone-iodine ointment
(radiolabeled with iodine 125) containing 1% iodine, the skin
retained 80% to 90% of the isotope after six hours. Therefore, is it
not possible that in the authors' experimental design, the elimina¬
tion of a neutralizer and a continuous irrigation may yield a
totally different result?
After an intensive review of the study by Dr Rodeheaver and his
colleagues, I found that the correlation of Fig 2 through 5, which
were graphs, seemed to be somewhat inconsistent. In Fig 2, a log
decrease to levels below those responsible for sepsis is shown.
However, in the subsequent Figures, the controls do not correlate
with each other. This, to me, indicates a possible variability in the
experimental design. Is it not possible that there may have been a
communication between the two paravertebral incisions, thus
causing this diversity of results? In essence, the controls should
not have shown so much variability.
Dr Rodeheaver and his associates have addressed the problem
of toxic effects and the deleterious effect of these products on the
wound. I believe, at this point, we have a problem in semantics.
What is their interpretation of the terms toxicity and deleteri¬
ous? Toxicity refers to toxins or poisons. The terms biocidal,
microbicidal, and its counterparts refer to death—death of the
microbes. Could not the decrease in infection rate be caused by
some other metabolic effect, other than that of the direct micro¬
bicidal activity of iodine-like compounds? Dr Rodeheaver and his
colleagues allude to the toxic effects of these compounds on the
host defense mechanism; however, no data were presented to
support this theory.
Earlier investigators have shown that varying concentrations of
iodine inhibit the host's initial inflammatory response. At the
American Burn Association meeting, Ninnemann showed that
periodate not only inhibits the primary antibody response, but the
cellular immune response as well.
In closing, I would like to say that Dr Rodeheaver and his
colleagues have admirably continued to sow the seeds of skepti¬
cism on the efficacy of iodine and iodine-like products as thera¬
peutic agents.
Ronald Lee Nichols, MD, New Orleans: I would like to report on
a series of studies that my associates and I have done in our
laboratory in New Orleans, trying to ascertain the efficacy of
various strengths of intraperitoneal povidone-iodine solution in
combating intra-abdominal infection.
In our experimental model, we evaluated the efficacy of both
full-strength and dilute povidone-iodine on peritoneal irrigation
in animals with fecal peritonitis. We also studied irrigations with
saline solution and kanamycin sulfate, as well as a control group.
We found the animals that were treated with peritoneal irrigation
with full-strength povidone-iodine experienced 100% mortality
within 24 hours. At autopsy, there was minimal evidence of
intraperitoneal infection, although the exact cause of death was
not ascertained.
Those animals that had irrigation with dilute povidone-iodine
Downloaded From: by a UNIVERSITY OF ADELAIDE LIBRARY User on 01/05/2018
had 100% mortality by 72 hours, and they had a septic death that
was also noted in the group treated with saline solution and the
control group. The only peritoneal irrigation that resulted in any
success was that of 0.1% kanamycin sulfate.
Subsequent study has shown, however, that effective parenteral
antibiotics work better than the antibiotic peritoneal irrigation in
our experimental model.
Richard L. Simmons, MD, Minneapolis: Povidone-iodine is toxic.
Dr Nichols' animals that received povidone-iodine in the perito¬
neal cavity died of hemorrhagic peritonitis. It is sterile and can be
cured with saline solution. With bugs, however, plus povidone-
iodine in the peritoneal cavity, they die, and they cannot be cured
with saline solution.
I wonder whether our clinical impression that povidone-iodine
does some good is not perfectly consistent with the observations
reported by Dr Rodeheaver and his associates, namely, that it is
tissue toxic, as well as microbicidal.
One of the things that impresses me clinically and experimen¬
tally is that the major lesion that we see, what we call pus, is
essentially autodigestion by neutrophil enzymes. Povidone-iodine
is toxic to neutrophils; a 1:1,000 concentration of povidone-iodine
will totally immobilize a neutrophil, and 1:100 concentration of
povidone-iodine will totally kill a neutrophil. Therefore, I wonder
when we pack this infected or contaminated wound with a sponge
soaked with povidone-iodine, whether what we are really doing is
turning off the autodigestive processes, as well as killing the
microbes.
Dr Rodeheaver: I hope that I can address the majority of
questions that Dr Heggers has raised.
The inability of iodine to migrate through agar correlates with
the disappointing results obtained in vivo. Since the purpose of the
test was only to show relative differences between the three iodine
solutions, there was little need to define the chemical composition
of the agar exactly.
The in vivo evaluation in wounds involved a brief exposure to
topical iodine. This procedure was meant to simulate the treat¬
ment of a simple laceration in the emergency room. No attempt
was made to evaluate the efficacy of iodine solution for peritoneal
irrigation.
Using 125I inan ointment on the skin and retaining 80% to 90%
of the label six hours later says nothing about the state of the
iodine nor its activity. It is most likely still bound to the povidone
or some skin protein, but, in either case, the iodine is inactive.
The results depicted in Fig 2 show the immediate kill of bacteria
after a ten-minute exposure to iodine. However, in Fig 3 through 5,
the results were obtained four days after closure. The results in
the latter Figures show the long-term effect of how iodine
solutions debilitate the normal tissue defenses and potentiate
otherwise subinfective inocula of bacteria into wound infection.
Every animal served as its own control since one wound was
treated with an iodine solution and the contralateral wound was
treated with saline solution. Regardless, the combined data from
all control groups were statistically similar and indicated no
variability in the experimental design.
The exact mechanism by which iodine solutions were detrimen¬
tal to wound defenses was not addressed in this study. Our
primary intent was to demonstrate that exposure of contaminated
wounds to iodine solutions results in a significantly higher
infection rate than similarly contaminated wounds treated only
with saline solution. These results indicated that iodine inhibits
wound defenses.
My colleagues and I continue to be impressed with the excellent
studies of Drs Nichols and Simmons that document the potential
consequences of using iodine solutions in the peritoneum.