- Email: [email protected]
Defatting catheter insertion sites in total parenteral nutrition is of no value as an infection control measure
Defatting Catheter Insertion Sites in Total Parenteral Nutrition Is of No Value as an Infection Control Measure Controlled Clinical Trial
DENNIS G. MAKI, M.D. KAREN N. McCORMACK, R.N., M.S.N. Madison,
Wisconsin
From the Section of Infectious Diseases, Department of Medicine, and the Infection Control Department, University of Wisconsin Hospitals and Clinics, University of Wisconsin Medical School, Madison, Wisconsin. Requests for reprints should be addressed to Dr. Dennis G. Maki, H4/ 574, University of Wisconsin Hospital and Clinics, Madison, Wisconsin 53792. Manuscript submitted February 19, 1987, and accepted June 29, 1987.
Defatting the skin with acetone or ether is widely used in the regimen for disinfection of insertion sites of central venous catheters in total parenteral nutrition. The fatty acids secreted by normal skin play an important role in regulation of the cutaneous microbial ecosystem, and it can be questioned whether application of a solvent might paradoxically promote colonization by pathogenic microorganisms. The clinical value of defatting catheter insertion sites was prospectively studied in a controlled, randomized trial: 100 subclavian catheters inserted for total parenteral nutrition were given identical site care except that with one half of the catheters, the site was defatted with acetone prior to catheter insertion and as part of the every-other-day site care regimen. Cutaneous colonization was found in only 130 (24.5 percent) of 531 site cultures in both groups, but was strongly predictive of concordant colonization of the catheter (relative risk, 22.1, p
November
1987
The
American
Journal
of Medicine
Volume
83
833
ACETONE
DEFATTING-MAKI
and
McCORMACK
the time it is inserted. A strong correlation has been shown between organisms present on skin of the catheter insertion site and organisms causing catheter-related bacteremia [3,11,12]. The importance of suppression of cutaneous floral populations with an effective antiseptic prior to insertion of an intravascular catheter cannot be overemphasized. In support of this thesis, hospitals that have implemented a stringent protocol approach to the management of hyperalimentation, usually through institution of nutritional support teams that assure consistency in aseptic technique during catheter insertion and follow-up care of the site, have reported very low rates of related sepsis, in the range of 2 percent or less [2-4,6,13171. Many programs routinely use an organic solvent, acetone or ether, to “defat” the skin as part of the regimen for disinfecting the site prior to insertion of the catheter and in follow-up care given to the site, usually every other day. t&fatting is widely practiced and included in many published protocols for care of hyperalimentation catheters [ 15-291 but appears to have no scientific basis that we can ascertain. Derrnatologic investigations suggest to the contrary that the natural skin lipids, especially the free fatty acids, contribute substantially to the intrinsic antimicrobial properties of normal skin, and application of organic solvents may actually enhance and prolong colonization by pathogenic microorganisms [30-321. We report the results of a prospective comparative clinical trial undertaken to determine whether the use of acetone in the regimen for disinfecting subclavian hyperalimentation catheter sites is beneficial in terms of recfucing cutaneous colonization and decreasing the incidence of catheter-related infection. The results show that use of acetone confers no discernible benefit in promoting microbial removal or in reducing catheter-related infection, whereas it increases local inflammation and discomfort to the patient.
of the nature of this comparative study before requesting their consent to participate. For consenting patients, at the time of insertion, the catheter was assigned by random allocation to have defatting with acetone used in the site preparation regimen or to receive all site care without acetone (control). During the study, 16-gauge catheters manufactured by Deseret Medical (Sandy, Utah) were used exclusively. Ten percent povidone-iodine on prepackaged applicators (Betadine Solution, Purdue-Frederick) was used for cutaneous disinfection prior to insertion of the catheter and for recleansing the site at subsequent every-otherday dressing changes. For catheters randomly assigned to the acetone group, USP-grade acetone (Pure Pat) was applied liberally over the site and left on for approximately one minute before application of the povidone-iodine. After vigorously scrubbing with the antiseptic for another minute, the catheter was inserted, 10 percent povidone-iodine ointment (Pharmadine, Sherwood Pharmaceutical) was applied to the catheter wound, and the site was dressed with fourinch by four-inch (11.2-cm by 11.2-cm), eight-ply, finemesh, sterile gauze (Johnson and Johnson) and elasticized tape (Elastoplast. Beiersdorf). Catheters in the control group received identical care except that defatting with acetone was omitted. Although every study catheter was inserted by a house officer, one of us (K.N.M.) randomly assigned each catheter to the appropriate treatment group and dressed the newly inserted catheter; thereafter, she evaluated each patient’s catheter daily, cultured sites and carried out follow-up dressing changes every other day, and personally received and cultured each catheter at removal. For each catheter, the patient was queried daily regarding pain or discomfort at the insertion site. On days the dressing was not scheduled to be changed, the site was gently palpated through the dressing. At scheduled dressing changes every other day and whenever dressings became wet, soiled, or nonadherent, the old dressing was removed, any remaining ointment was removed with sterile gauze, the site was scored by a protocol for cutaneous maceration, erythema, purulence, and tenderness, and skin surrounding the catheter was cultured quantitatively. After defatting with acetone for catheters randomly assigned to this group, the site was recleansed with povidone-iodine, and after reapplication of povidone-iodine ointment, a new sterile dressing was applied. Decisions to remove catheters were made independently by patients’ physicians. At the time of catheter removal, after the dressing and residual ointment were removed, the site was assessed, skin about the catheter site was again cultured, and the catheter was aseptically removed and cultured. Data obtained for each evaluable catheter included demographic characteristics of the patient, associated medical conditions, other invasive devices used for monitoring or medical care, all clinical and laboratory data pertaining to suspected or proven infections, and all antimicrobial therapy given. In addition, the condition of the insertion site and the number of days the catheter had remained in place were recorded. Each patient was followed for three days after
PAllENTS AND METHODS Sources of Data. Background: University of Wisconsin Hospital, a 450-bed referral institution, does not have a nurse intravenous therapy team. During this study, infusions received care in accordance with published guidelines [i3,6]. Central venous catheters were dressed with sterile gauze and tape; every 46 hours the dressing was removed, the site inspected and recleansed with an antiseptic solution, and a new dressing reapplied. Administration sets were changed routinely every 48 hours [33]; in-line microfilters were not used. All central venous catheters were inserted by house officers wearing sterile gloves and usually a sterile gown and mask, first draping the patient with sterile drapes. Catheter insertion sites were not routinely shaved. Study protocol: Adult patients over 18 years of age, without granulocytopenia, scheduled to receive a central venous catheter for total parenteral nutrition were informed
934
November
1987
The American
Journal
of Medicine
Volume
93
ACETONE
removal of the catheter to identify latently expressed local inflammation or systemic infection due to the catheter. Blood culture specimens drawn by separate percutaneous venipuncture were obtained from patients with fever or other signs of infection. Microbiologic Methods. Skin of the insertion site: Twenty square centimeters of skin about the catheter were cultured quantitatively by applying a sterile cardboard template and culturing the defined 20 cm2 of exposed skin with a sterile cotton-tip applicator (Culturette, American Scientific Products), with vigorous scrubbing of the entire area four times, twice in one direction and twice more perpendicularly. In the laboratory, the tip of the applicator was immersed in 1 ml of sterile saline and mixed vigorously (Vortex, Scientific Products), following which serial dilutions were inoculated onto the surface of 5 percent sheep blood agar that was incubated aerobically in a IO percent carbon dioxide atmosphere at 37% for three days before discarding. All colony types were enumerated and identified [34]. Catheters: Catheters were removed, with the externalized portion directed upward, to minimize contamination by organisms on the skin surface. Two 5-cm segments of the catheter were cultured: the proximal intracutaneous portion, extending from several millimeters inside the former skin-catheter interface to the vein, and the tip [35,36]; segments were amputated using sterile scissors and transported in a sterile humidified tube (Culturette). Details of the semiquantitative method for culturing vascular catheters have been reported [35,36]. Briefly, each catheter segment was transferred aseptically from its transport tube onto the surface of a IOO-mm 5 percent sheep blood agar plate, and with use of a sterile forceps, was rolled back and forth across the agar surface while gentle downward pressure was exerted to affirm adequate surface-to-surface contact. Plates were incubated aerobically in a IO percent carbon dioxide atmosphere at 37OC for at least 72 hours. All colony types appearing were enumerated and identified [34]. All cultures were processed in a monitored bticteriologic high-efficiency-filtration laminar-flow hood (Abbott Labora-
tories). Although it was not possible to mask the identity of the catheters’ treatment groups from the co-investigator assessing catheters, the research microbiologists processing all cultures were unaware of the identity of each catheter’s treatment group. Definitions. Local catheter-related infection: A positive result of a semiquantitative culture of the catheter (15 or more colony-forming units [cfu]) [35-371; a positive semiquantitative culture result represents colonization of the catheter. Catheter-related septicemia: (1) a semiquantitative
catheter culture and blood cultures demonstrating positive results for the same species;
(2) clinical
(or autopsy)
and
microbiologic data disclosed no other apparent source for the septicemia [33,35,36]. Statistics. The significance of differences between the two groups was determined with a chi-square test or Fisher’s exact test for categoric data and the Student t test for continuous data.
TABLE
DEFATTING-MAKI
and McCORMACK
Features of Patients and Catheters Two Site Care Groups
I
Feature Patients Age (years) Mean Range Male:female (percent) Service (percent) Surgical Medical Intensive care unit (percent) Neoplastic disease (percent) Type I diabetes mellitus (percent) Antimicrobial therapy (percent) Corticosteroid therapy (percent) Active infection unrelated to the catheter (percent) Catheters Days in place Mean f SD Range
Control (n = 51)
51
AcetoneTreated (n = 49)
54 20-80 58:42
15-80 61:39
13.5
in the
p Value
NS NS
84 16 45
75 2.5
NS
29
<0.05
33
35
NS
4
4
NS
84
80
NS
18
14
NS
53
69
NS
14.6 f 7.9 3-37
NS
f 10.0 4-57
RESULTS Characteristics of the Study Population. One hundred consecutively enrolled subclavian catheters were prospectively studied, 49 in which acetone was used in the site care regimen and 5 1 in which no acetone was used. Every catheter had been inserted percutaneously at a new
site into the subclavian vein by an infraclavicular approach; none of the catheters were inserted over a guidewire into an old site. Patients
in the two groups
were
very similar
demo-
graphically and, except for a higher percentage of control patients within an intensive care unit, were comparable in terms of risk factors predisposing to infection (Table I). The majority of catheters in each group were used in patients more susceptible to nosocomial infection [38], as reflected by the relatively high age of the study population, the preponderance of surgical patients, and the high frequency of neoplastic disease, recent surgery, and intensive care unit placement; over half of the patients in each group had an active infection unrelated to the catheter, and three fourths of the patients in each group received systemic antimicrobial therapy at some time while the study catheter was in place. The average length of the time study catheters remained in place was also very similar in the two groups, 14 and 15 days. Most catheters in each group were removed because
November 1987
The American Journal of Medicine
Volume 83
835
ACETONE
DEFATTING-MAKI
TABLE II
and McCORMACK
Colonization of Skin of the Insertion Site in the Two Site Care Groups
about 6 percent of the time with a potential pathogen other than coagulase-negative Staphylococcus, such as S. aureus, an enteric gram-n&gative bacillus, or Candida. Catheter-Related Infection. Four (8 percent) catheters in the control group and four (8 percent) in the acetone group were found to be colonized on removal, as defined by a positive semiquantitative catheter culture result (Table Ill). Two control catheters (4 percent) caused septicemia, one with S. aureus and the other with coagulasenegative Staphylococcus and Klebsiella pneumoniae. Two catheters in the acetone group caused septicemia (4 percent), one with Candida albicans and the other with Pseudomonas aeruginosa. Catheter-related infection was strongly associated with concordant cutaneous colonization of the insertion site prior to catheter removal: the site was shown to be colonized by the same species as the catheter in seven (85 percent) of the eight infected catheters, including all four tiausing bacteremia or candidemia. Colonization of the site was demonstrated with 24 catheters overall in both groups; as noted, in seven (29 percent), the catheter was found to be infected by the same species; in contrast, of 76 catheters without detectable colonization of the site, only one (1 percent) was infected on kernoval (relative risk of concordant catheter-related infection with a colonized site, 22.1, p
Acetone-
Feature
Control
Total
number of site cultures Number (percent) showing growth (cfu/cm*) None 1-9 10-99 2100 Number (percent) showing colonization by Coagulase-negative staphylococci Staphylococcus aureus Gram-negative bacilli Yeasts
TABLE III
256
187 17 17 35
Treated
p
Value
275
(73.1) (6.6) (6.6) (13.7)
214 22 10 29
(7T.8) (8.0) (3.6) (10.5)
NS NS NS NS
37 (14.4)
32 (11.6)
NS
8 (3.1) 14 (5.4) 7 (2.7)
6 (2.2) 12 (4.3) 7 (2.5)
NS NS NS
Catheter-Related Infection in the Two Site Care Groups
Parameter* Number of catheter-related infections (percent) Local+ Bacteremia Infecting organisms, causing local infection (bacteremia) Coagulase-negative staphylococci Staphylbcoccus aureus Gram-negative bacilli Candida Enterococcus * None of the differences significant at p <0.05. + Defined as 15 colony-forming [35-371.
between units
the
Control (n = 51)
AcetoneTreated (n = 49)
4 (6) 2 (4)
4 (6) 2 (4)
2 (1) 2 (1)
2
l(l) . . 1
i ;-i, l(l) . . .
groups
is statistically
on semiquantitative
culture
COMMENTS total parenteral nutrition had been terminated electively; : only IO catheters were removed because of suspected catheter-related sepsis. Cutaneous Colonization of Catheter Insertion Sites. There were no Significant differences between the two groups in skin colonization contiguous to the catheter (Table II). Approximately three fourths of the site cultures in each group showed no growth; about one fourth yielded organisms, with a similar distribution of quantitative colonization in the two groups (Table II). Most cutaneous colonization was with coagulase-negative staphylococci; colonization by Staphylococcus aureus, gram-negative bacilli, or yeasts was infrequent in both groups (Table II). Heavy colonization (more than 100 cfu/cm*) was found at approximately 10 percent of the sites in each group, 836
November
1987
The American
Journal
of Medicine
Volume
An abundance of clinical and microbiologic data suggests strongly that most intravenous catheter-related septicemias are caused by cutaneous organisms that invade the intracutaneous wound during catheter insertion or sometime thereafter: (1) Studies have shown a strong correlation between organisms present on skin surrounding the catheter wound and microorganisms recovered from catheters producing septicemia [3,11,12]. (2) Burn patients with massive populations of SUrfaCe microorganisms experience very high rates of catheterrelated sepsis [36,39-411. (3) Clusters of catheter-related septicemias have been traced to contaminated disinfectants used for cutaneous antisepsis [42-441 and to contaminated tape [45]. 83
ACETONE
(4) As shown in this study (Table III) and in numerous other recent studies of catheter-related infection [i -71, coagulase-negative Staphylococcus, the predom/nant aerobic species colonizing the human skin [46], is now the most common agent of catheter-related infection. (5) Hospitals with intravenous therapy teams that stress asepsis in catheter insertion and site care report substantially lower rates of catheter-related septicemia [i-7,13-17]. (6) Controlled trials have shown that application of a topical polyantimicrobial agent to peripheral venous catheter insertion sites provides some protection against catheter-related infection [47-521. (7) Heavy growth on semiquantitative culture of the external surface of a removed catheter has been shown to be strongly associated with bacteremia caused by the catheter [11,12,35-3753,541. (8) Microscopic studies of infected central venous catheters have shown organisms primarily on the external surface of the catheter [54]. Thus, the details of site care-the regimen for disinfecting the site prior to catheter insertion, the dressing, and the follow-up care-would seem likely to have a profound influence on the incidence of catheter-related infection. Floral populations can be effectively suppressed with an antiseptic solution prior to insertion of the catheter. However, extrinsic nosocomial contaminants or regrowth of the suppressed endogenous microflora can colonize the site and cause catheter-related infection. Dressings, intended to prevent extrinsic contamination of the insertion site, may, paradoxically, adversely affect cutaneous microbial populations in the vicinity of the catheter [55631. Although central venous catheters in most hospitals are redressed at periodic intervals-typically every other day-few studies [22,64], none of adequate size, have examined the effectiveness of periodic site care for central venous catheters, vis-a-vis, have determined whether periodically recleansing the skin about the catheter with an antiseptic and redressing the site reduces colonization sufficiently so as to significantly reduce the risk of systemic catheter-related infection. Many published protocols for disinfection of hyperalimentation catheter insertion sites and for follow-up care of the site advocate defatting the skin with acetone [ 15 291, presumably an the belief microbial removal is augmented. Use of acetone has become standard practice in many institutions without confirmation of its efficacy by controlled trial. We have been unable to find any clinical studies of the value of defatting as an adjunct to cutaneous disinfection. Healthy intact skin is one of the body’s most important barriers to deep infection. Pathogenic microorganisms such as Streptococcus pyogenes, S. aureus, and gramnegative bacilli placed on the surface of intact skin rapidly die and do not readily become established as part of the November
TABLE IV
DEFATTING-MAKI
and McCORMACK
Condition of Catheter Insertion Two Site Care Groups
Sites in the
Acetone-
Feature
Control
Number of catheters (total number of site evaluations) Catheters on one or more site evaluations showing (percent) Pain or tenderness Denuding Erythema Discomfort during site care One or more of the above Site evaluations showing (percent) Pain or tenderness Erythema Denuding
Treated pValue
51 (256)
49 (275)
10 2 29 12
76 18 65 82
35
80
5.6 18.7 0.8
59.1 42.2 10.2
normal cutaneous flora [30-321, which consists primarily of aerobic and anaerobic diphtheroids and micrococci, in concentrations ranging from 10 to lo6 microorganisms per cm*, depending on the method of sampling and the anatomic area sampled [46,65,66]; gram-negative bacilli and yeasts do not usually constitute a significant portion of the cutaneous microflora except on the perineum and feet. A variety of physical-chemical factors at the skin surface are thought to contribute to the intrinsic ability of normal skin to resist colonization by more pathogenic microorganisms, including a slightly acidic pH, surface desiccation, bacterial interference mediated by the less pathogenic normal flora, and the antimicrobial activity of sweat and sebum [66]. Sebum, the product of apocrine sweat glands, consists of a complex mixture of lipids including triglycerides (approximately 50 percent by weight), wax esters (25 percent), squalene (10 percent), cholesterol esters (2 percent), and cholesterol (2 percent), with considerable variation by anatomic location [66]. Triglycerides are hydrolyzed to free fatty acids, mainly by lipases produced by the normal skin flora [67]. In vitro, most of the skin’s fatty acids, ranging from C6 to GIG, have antibacterial activity and inhibit ‘growth of a variety of microorganisms, partricularly gram-positive cocci and yeasts, and to a much lesser extent, gramnegative bacilli [30,68]. Ricketts et al [30] first reported that S. pyogenes placed on normal skin rapidly died, even if desiccation was prevented; likewise, S. aureus survived only three days under the same conditions, and gram-negative bacilli died rapidly unless the skin surface was kept nonphysiologically humid. Topical application of albumin, which binds to free fatty acids, greatly prolonged the survival of both S. 1987
The American
Journal
of Medicine
Volume
93
837
ACETONE
DEFATTING-MAKI
and McCORMACK
pyogenes and S. aureus. Extraction of the skin lipids by application of acetone resulted in prolonged survival of S. pyogenes. Survival of different microorganisms after acetone treatment was found to be related to their in vitro susceptibility to the unsaturated free fatty acids, especially oleic acid, normally present on the skin surface. Escherichia coli and P. aeruginosa were found to be resistant to the cutaneous unsaturated fatty acids, and survival of these organisms was unaffected by defatting. Lacey [31] and Aly et a! [32] subsequently confirmed and extended these studies, showing that pretreatment of normal skin with acetone greatly prolonged the survival of S. aureus, S. pyogenes, and C. albicans, but not E. coli or P. aeruginosa. Aly and co-workers [32] further showed that reapplying the’lipid acetone extracts from normal skin to acetone-treated skin restored much of its intrinsic antimicrobial activity. These experimental studies led us to question whether defatting with acetone might actually be deleterious and promote colonization by staphylococci, Candida, and other potential pathogens and, paradoxically, increase the risk of catheter-related septicemia. The findings from our prospective randomized comparative trial suggest that the use of acetone in the regimen for disinfecting the insertion site does not measurably impair cutaneous antisepsis; however, it clearly does not augment it either. Using acetone did not measurably affect skin colonization (Table II); and, as might then be expected, also did not significantly influence the incidence of catheter-related infection (Table Ill). However, use of acetone greatly increased local inflammation of the catheter site and discomfort to the patient, typically pain during its application (Table IV). Type II error (beta) should be considered when designing clinical trials and analyzing the results [69]. Given the relatively low rates of catheter-related infection in both treatment groups in our study (Table Ill), to have been able to identify with statistical certainty a true difference of 25 percent in the incidence of related bacteremia would have required study of several thousand catheters. We point out, however, that even though there could well be a true (statistically significant) difference between the two treatment groups, the rate of bacteremic infection in hyperalimentation with reasonable aseptic care is now relatively low, and any differences, even if statistically significant, would probably be of marginal relevance clinically. We can conclude from our data with strong statistical power
(beta less than 10 percent) that use of acetone does not measurably improve the microbial removal of the disinfecting regimen (colonization of sites was unaffected, Table II), but greatly increases local inflammation and discomfort to the patient (Table IV). Thus, we believe that using acetone as part of the regimen of cutaneous antisepsis in vascular catheter care no longer has a rational basis. Our data are consonant with the findings of other studies reporting that most catheter-related infections in hyperalimentation derive from organisms colonizing the skin contiguous to the catheter [ 11,12]. In seven of the eight colonized catheters (more than 15 cfu on semiquantitative catheter culture), including all four catheters considered to have caused septicemia, the same microbial species isolated from the catheter was isolated one or more times in quantitative skin cultures performed at dressing changes prior to removal of the catheter. The importance of assiduous aseptic technique during catheter insertion and at follow-up dressing changes cannot be emphasized too strongly. These data further suggest that culturing the skin contiguous to the catheter at the time of dressing changes using a standardized technique, such as was employed in this study, may be of predictive value in monitoring for incipient catheter-related sepsis [ 11,121. Virtually all hyperalimentation programs with low rates of infection routinely provide periodic site care, usually daily or every other day. Our data suggest that heavy skin colonization contiguous to the catheter can be effectively suppressed by regular site care (Table II), In a small crossover study in 15 patients, Jarrard et al [22] found that recleansing the site daily with an iodophor solution was more effective than use of a topical iodophor ointment in preventing colonization of the site; but the best antiseptic to use and the cost-effectiveness of periodic site care, especially how frequently it should be performed, need to be prospectively studied. ACKNOWLEDGMENT We are grateful to Carol A. Hassemer and Carla A. Alvarado for microbiologic support, and the physicians and nurses of the University of Wisconsin Hospital and Clinics for permitting us to study their patients with Central venous catheters. We also express our appreciation to Ms. Peggy Barden for secretarial assistance in preparation of the manuscript.
REFERENCES 1. 2. 3.
838
Maki DG, Goldman DA, Rhame FS: Infection control in intravenous therapy. Ann Intern Med 1973; 79: 876-887. Goldmann DA, Maki DG: Infection control in total parenteral nutrition. JAMA 1973; 223: 1360-1364. Makf DG: Sepsis arising from extrinsic contamination of the infusion and measures for control. In: Phillips I, ed. Mfcrobiological hazards of infusion therapy. Lancaster, En-
November
1987
The American
Journal
of Medicine
4.
5.
Volume
83
gland: MTP Press, 1977; 99-141. Maki DG: Infections associated with intravascular lines. In: Swartz M, Remmington J, eds. Current topics in clinical infectious disease. New York: McGraw-Hill, 1982; 309363. Centers for Disease Control Working Group: Guidelines for prevention of intravenous therapy-related infections. In-
ACETONE
6.
7. 8.
9.
10.
11.
12.
13.
14. 15. 16. 17.
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
feet Control 1981; 3: 62-79. Allen JR: The incidence of nosocomial infection in patients receiving total parenteral nutrition. In: Johnston IDA, ed. Advances in parenteral nutrition. Lancaster, England: MTP Press, 1977; 339-377. Williams WW: Infection control during parenteral nutrition therapy. J Parenter Enter Nutr 1985; 9: 7352746. Curry CR, Quie PG: Fungal septicemia in patients receiving parenteral hyperalimentation. N Engl J Med 1971; 285: 1221-1225. Plouffe JF, Brown DG, Silva J Jr, Eck T, Stricof RL, Fekety FR Jr: Nosocomial outbreak of Candida parapsilosis fungemia related to intravenous infusions. Arch Intern Med 1977; 137: 1686-1689. Solomon SL, Khabbaz RF, Parker RH, et al: An outbreak of Candida parapsilosis bloodstream infections in patients receiving parenteral nutrition. J Infect Dis 1984; 149: 98102. Snydman DR, Pober ER, Murray SA, Gorbea JF, Majka JA, Perry LK: Predictive value of surveillance skin cultures in total-parenteral nutrition-related infection. Lancet 1982; II: 1385-1388. Bjornson HS, Colley R, Bower RH, Duty VP, Schwartz-Fulton JT, Fischer JE: Association between microorganism growth at the catheter site and colonization of the catheter in patients receiving total parenteral nutrition. Surgery 1982; 92: 720-727. Freeman JB, Lemire A, MacLean LD: Intravenous alimentation and septicemia. Surg Gynecol Obstet 1972; 135: 708-712. Nehme AE: Nutritional support of the hospitalized patient: the team conceot. JAMA 1980: 243: 1906-1908. Wilmore DW, Dudrick SJ: Safe long-term venous catheterization. Arch Surg 1969; 98: 258-258. Sanderson I, Dietel M: Intravenous hyperalimentation without sepsis. Surg Gynecol Obstet 1973; 136: 577-585. Copeland EM, MacFadyen BV, McGown C, Dudrick SJ: The use of hyperalimentation in patients with potential sepsis. Surg Gynecol Obstet 1974; 138: 377-381. Kaminski MV Jr, Stolar MH: Parenteral hyperalimentationa quality of care survey and review. Am J Hosp Pharm 1974; 31: 228-235. Ryan JA, Abel RM, Abbott WM, et al: Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. N Engl J Med 1974; 290: 757-761. Phillips KJ: Nursing care in parenteral nutrition. In: Fischer JE, ed. Total parenteral nutrition. Boston: Little, Brown, 1976; 101-l 10. Ota DM, lmbembo AL, Zuidema GD: Total parenteral nutrition. Surgery 1978; 83: 503-520. Jarrard MM, Olson CM, Freeman JB: Daily dressing change effects on skin flora beneath subclavian catheter dressings during total parenteral nutrition. J Parenter Enter Nub 1980; 4: 391-392. Palidar PJ, Simonowitz DA, Oreskovich MR, et al: Use of opsite as an occlusive dressing for total parenteral nutrition catheters. J Parenter Enter Nutr 1982; 6: 150-15 1. Schwartz-Fulton J, Colley R, Valanis 8, Fischer JE: Hyperalimentation dressings and skin flora. National Intravenous Therapy Association 1981; 4: 354-357. Padberg FT Jr, Ruggiero J, Blackburn GL, Bistrian BR: Central venous catheterization for parenteral nutrition. Ann Surg 1981; 193: 264-270. Bozzehi F, Terno G, Bonfanti G, et al: Prevention and treatment of central venous catheter sepsis by exchange via a guidewire. A prospective controlled trial. Ann Surg 1983; 198: 48-52. Nehme AE, Trigger JA: Catheter dressings in central parenteral nutrition: a prospective randomized comparative study. Nutritional Support Services 1984; 9: 42-43. Powell CR, Traetow MJ, Fabri PJ, Kudsk KA, Ruberg RL: Op-
November
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
1987
DEFAlTING-MAKI
and McCORMACK
site dressing study: a prospective randomized study evaluating povidone iodine ointment and extension set changes with 7-day Op-Site dressings applied to total parenteral nutrition subclavian sites. J Parenter Enter Nutr 1985; 9: 443-446. Sitzmann JV, Townsend TR, Siler MC, Bartlett JG: Septic and technical complications of central venous catheterization. A prospective study of 200 consecutive patients. Ann Surg 1985; 202: 766-770. Ricketts CR, Squire JR, Topley E: Human skin lipids with particular reference to the self-sterilizing power of the skin. Clin Sci 1951; 10: 89-l 11. Lacey RW: Antibacterial action of human skin. In vivo effect of acetone, alcohol, and soap on behavior of Staphylococcus aureus. Br J Exp Pathol 1968; 49: 209-215. Aly R, Maibach HI, Shinefield HR, Strauss WG: Survival of pathogenic microorganisms on human skin. J Invest Dermatol 1972; 58: 205-210. Band JD, Maki DG: Safety of changing intravenous delivery systems at longer than 24-hour intervals. Ann Intern Med 1979; 91: 713-718. Lennette E, Balows A, Hausler W, Shadomy H, eds. Manual of clinical microbiology, 3rd ed. Washington: American Society of Microbiology, 1980. Maki DG, Weise GE, Sarafin HW: A semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977; 296: 1305-1309. Maki DG, Jarrett F, Sarafin HW: A semiquantitative culture method for identification of catheter-related infection in the burn patient. J Surg Res 1977; 22: 513-520. Collignon PJ, Soni N, Pearson IY: Is semiquantitative culture of central vein catheter tips useful in the diagnosis of catheter-associated bacteremia? J Clin Microbial 1986; 24: 532-535. Haley RW, Hooton TM, Culver DH, et al: Nosocomial infections in U.S. hospitals, 1975-1976: estimated frequence by selected characteristics of patients. Am J Med 1981; 70: 947-959. Stein JM, Pruitt BA Jr: Suppurative thrombophlebitis. A lethal iatrogenic disease. N Engl J Med 1981; 70: 947-949. Popp MB, Law EJ, MacMillan BG: Parenteral nutrition in the burned child: a study of twenty-six patients. Ann Surg 1974; 179: 219-225. Pruitt BA Jr, McManus WF, Kim SH, Treat RC: Diagnosis and treatment of cannula-related intravenous sepsis in burn patients. Ann Surg 1980; 191: 546-554. Plotkin SA, Austrian R: Bacteremia caused by Pseudomonas sp. following the use of materials stored in solutions of a cationic surface-active agent. Am J Med Sci 1958; 235: 621-627. Dixon RE, Kaslow RA, Macket DC, Fulkerson CC, Mallison GF: Aqueous quaternary ammonium antiseptics and disinfectants. Use and misuse. JAMA 1976; 236: 2415-2417. Frank MJ, Schaffner W: Contaminated aqueous benzalkonium chloride. An unnecessary hospital infection hazard. JAMA 1976; 236: 2418-2419. Sheldon DL, Johnson WC: Cutaneous mucormycosis: two documental cases of suspected nosocomial cause. JAMA 1979; 241: 1032-1033. Kligman AM: The bacteriology of normal skin. In: Maibach HI, Hildick-Smith G, eds. Skin bacteria and their role in infection. New York: McGraw-Hill, 1965; 13-32.. Moran JM, Atwood RP, Rowe MI: A clinical bacteriologic study of infections associated wtih venous cut downs. N Engl J Med 1965; 272: 554-560. Norden CW: Application of antibiotic ointment to the site of venous catheterization-a controlled trial. J Infect Dis 1969; 120: 611-615. Zinner SH, Denny-Brown BC, Braun P, Burke JP, Kass EH: Risk of infection with intravenous indwelling catheters: effect of application of antibiotic ointment. J Infect Dis
The American
Journal
of Medicine
Volume
83
839
ACETONE
50.
51.
52.
53.
54.
55.
56. 57.
58.
840
DEFATTING-MAKI
and McCORMACK
1969; 120: 616-619. Levy RS, Goldstein J, Pressman RS: Value of a topical antibiotic ointment in reducing bacterial colonization of percutaneous venous catheters. J Albert Einstein Med Center 1970; 18: 67-70. Maki DG, Band JD: A comparative study of polyantibiotic and iodophor ointments in prevention of vascular catheterrelated infection. Am J Med 1981; 70: 739-744. Maki DG, Will L: Study of polyantibiotic and povidone-iodine ointments on central venous and arterial catheters dressed with gauze or polyurethane dressings. In: Program and Abstracts of the Twenty-Sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, September 30-October 2, 1986, New Orleans, Louisana. Washington: American Society for Microbiology, 1986. Snydman DR, Murray SA, Kornfeld SJ, Majka JA, Ellis CA: Total parenteral nutrition-related infections. Prospective epidemiologic study using semiquantitative methods. Am J Med 1982; 73: 695-699. Cooper GL, Hopkins CC: Rapid diagnosis of intravascular catheter-associated infection by direct gram staining of catheter segments. N Engl J Med 1985; 18: 1142-l 150. Hirman CD, Maibach HI: Effect of air exposure and occlusion on experimental human skin wounds. Nature 1963; 200: 377-378. Marples RR, Kligman AM: Growth of bacteremia under adhesive tapes. Arch Dermatol 1969; 99: 107-l 10. Aly R, Shirley C, Cunico B, et al: Effect of prolonged occlusion on the microbial flora, pH, carbon dioxide and transepidermal water loss on human skin. J Invest Dermatol 1978; 71: 378-381. Mertz PM, Eaglestein WH: The effect of a semiocclusive dressing on the microbial population in superficial wounds. Arch Surg 1984; 119: 287-289.
November
1987
The
American
Journal
of Medicine
59.
Powell C, Ellison EC, Blackson L, et al: Increased catheter sepsis with use of semipermeable membrane dressings system (Opsite) in total parenteral nutrition (abstr). J Parenter Enter Nutr 1979; 3: 515. Katich M, Band J: Local infection of the intravenous-cannulae wound associated with transparent dressings. J Infect Dis 1985; 151: 971-972. Mumford F, Grossman A, Hilton E: An outbreak of intravascular site infections (abstr). Am J Infect Control 1985; 13: 140. Powell C, Regan C, Fabri PJ, Ruberg RL: Evaluation of opsite catheter dressings for parenteral nutrition: a prospective, randomized study. J Parenter Enter Nutr 1982; 6: 43-46. Andersen PT, Herlevsen P, Schaumburg H: A comparative study of ‘Op-site’ and ‘Nobecutan gauze’ dressings for central venous line care. J Hosp Infect 1986; 7: 161-168. Jarrard MM, Freeman JB: The effects of antibiotic ointments and antiseptics on the skin flora beneath subclavian catheter dressings during intravenous hyperalimentation. J Surg Res 1977; 22: 521-526. Marples MJ: The ecology of the human skin. Springfield, Illinois: Charles C Thomas, 1965. Noble WC, Somerville DA: Microbiology of human skin. Philadelphia: WB Saunders, 1974; l-341. Scheimann LG, Knox G, Sher D, Rothman S: The role of bacteria in the formation of free fatty acids on the human skin surface, J Invest Dermatol 1960; 34: 171-174. Pillsbury DM, Rebel1 G: The bacterial flora of the skin. Factors influencing the growth of resident transient organisms. J Invest Dermatol 1952; 18: 173-186. Freiman JA, Chalmers TC, Smith H Jr, Kuebler RR: The importance of beta, and type II error and sample size in the design and interpretation of the randomized control trial. N Engl J Med 1978; 299: 690-694.
60.
6 1.
62.
63.
64.
65. 66. 67.
68.
69.
Volume
83
DENNIS G. MAKI, M.D. KAREN N. McCORMACK, R.N., M.S.N. Madison,
Wisconsin
From the Section of Infectious Diseases, Department of Medicine, and the Infection Control Department, University of Wisconsin Hospitals and Clinics, University of Wisconsin Medical School, Madison, Wisconsin. Requests for reprints should be addressed to Dr. Dennis G. Maki, H4/ 574, University of Wisconsin Hospital and Clinics, Madison, Wisconsin 53792. Manuscript submitted February 19, 1987, and accepted June 29, 1987.
Defatting the skin with acetone or ether is widely used in the regimen for disinfection of insertion sites of central venous catheters in total parenteral nutrition. The fatty acids secreted by normal skin play an important role in regulation of the cutaneous microbial ecosystem, and it can be questioned whether application of a solvent might paradoxically promote colonization by pathogenic microorganisms. The clinical value of defatting catheter insertion sites was prospectively studied in a controlled, randomized trial: 100 subclavian catheters inserted for total parenteral nutrition were given identical site care except that with one half of the catheters, the site was defatted with acetone prior to catheter insertion and as part of the every-other-day site care regimen. Cutaneous colonization was found in only 130 (24.5 percent) of 531 site cultures in both groups, but was strongly predictive of concordant colonization of the catheter (relative risk, 22.1, p
November
1987
The
American
Journal
of Medicine
Volume
83
833
ACETONE
DEFATTING-MAKI
and
McCORMACK
the time it is inserted. A strong correlation has been shown between organisms present on skin of the catheter insertion site and organisms causing catheter-related bacteremia [3,11,12]. The importance of suppression of cutaneous floral populations with an effective antiseptic prior to insertion of an intravascular catheter cannot be overemphasized. In support of this thesis, hospitals that have implemented a stringent protocol approach to the management of hyperalimentation, usually through institution of nutritional support teams that assure consistency in aseptic technique during catheter insertion and follow-up care of the site, have reported very low rates of related sepsis, in the range of 2 percent or less [2-4,6,13171. Many programs routinely use an organic solvent, acetone or ether, to “defat” the skin as part of the regimen for disinfecting the site prior to insertion of the catheter and in follow-up care given to the site, usually every other day. t&fatting is widely practiced and included in many published protocols for care of hyperalimentation catheters [ 15-291 but appears to have no scientific basis that we can ascertain. Derrnatologic investigations suggest to the contrary that the natural skin lipids, especially the free fatty acids, contribute substantially to the intrinsic antimicrobial properties of normal skin, and application of organic solvents may actually enhance and prolong colonization by pathogenic microorganisms [30-321. We report the results of a prospective comparative clinical trial undertaken to determine whether the use of acetone in the regimen for disinfecting subclavian hyperalimentation catheter sites is beneficial in terms of recfucing cutaneous colonization and decreasing the incidence of catheter-related infection. The results show that use of acetone confers no discernible benefit in promoting microbial removal or in reducing catheter-related infection, whereas it increases local inflammation and discomfort to the patient.
of the nature of this comparative study before requesting their consent to participate. For consenting patients, at the time of insertion, the catheter was assigned by random allocation to have defatting with acetone used in the site preparation regimen or to receive all site care without acetone (control). During the study, 16-gauge catheters manufactured by Deseret Medical (Sandy, Utah) were used exclusively. Ten percent povidone-iodine on prepackaged applicators (Betadine Solution, Purdue-Frederick) was used for cutaneous disinfection prior to insertion of the catheter and for recleansing the site at subsequent every-otherday dressing changes. For catheters randomly assigned to the acetone group, USP-grade acetone (Pure Pat) was applied liberally over the site and left on for approximately one minute before application of the povidone-iodine. After vigorously scrubbing with the antiseptic for another minute, the catheter was inserted, 10 percent povidone-iodine ointment (Pharmadine, Sherwood Pharmaceutical) was applied to the catheter wound, and the site was dressed with fourinch by four-inch (11.2-cm by 11.2-cm), eight-ply, finemesh, sterile gauze (Johnson and Johnson) and elasticized tape (Elastoplast. Beiersdorf). Catheters in the control group received identical care except that defatting with acetone was omitted. Although every study catheter was inserted by a house officer, one of us (K.N.M.) randomly assigned each catheter to the appropriate treatment group and dressed the newly inserted catheter; thereafter, she evaluated each patient’s catheter daily, cultured sites and carried out follow-up dressing changes every other day, and personally received and cultured each catheter at removal. For each catheter, the patient was queried daily regarding pain or discomfort at the insertion site. On days the dressing was not scheduled to be changed, the site was gently palpated through the dressing. At scheduled dressing changes every other day and whenever dressings became wet, soiled, or nonadherent, the old dressing was removed, any remaining ointment was removed with sterile gauze, the site was scored by a protocol for cutaneous maceration, erythema, purulence, and tenderness, and skin surrounding the catheter was cultured quantitatively. After defatting with acetone for catheters randomly assigned to this group, the site was recleansed with povidone-iodine, and after reapplication of povidone-iodine ointment, a new sterile dressing was applied. Decisions to remove catheters were made independently by patients’ physicians. At the time of catheter removal, after the dressing and residual ointment were removed, the site was assessed, skin about the catheter site was again cultured, and the catheter was aseptically removed and cultured. Data obtained for each evaluable catheter included demographic characteristics of the patient, associated medical conditions, other invasive devices used for monitoring or medical care, all clinical and laboratory data pertaining to suspected or proven infections, and all antimicrobial therapy given. In addition, the condition of the insertion site and the number of days the catheter had remained in place were recorded. Each patient was followed for three days after
PAllENTS AND METHODS Sources of Data. Background: University of Wisconsin Hospital, a 450-bed referral institution, does not have a nurse intravenous therapy team. During this study, infusions received care in accordance with published guidelines [i3,6]. Central venous catheters were dressed with sterile gauze and tape; every 46 hours the dressing was removed, the site inspected and recleansed with an antiseptic solution, and a new dressing reapplied. Administration sets were changed routinely every 48 hours [33]; in-line microfilters were not used. All central venous catheters were inserted by house officers wearing sterile gloves and usually a sterile gown and mask, first draping the patient with sterile drapes. Catheter insertion sites were not routinely shaved. Study protocol: Adult patients over 18 years of age, without granulocytopenia, scheduled to receive a central venous catheter for total parenteral nutrition were informed
934
November
1987
The American
Journal
of Medicine
Volume
93
ACETONE
removal of the catheter to identify latently expressed local inflammation or systemic infection due to the catheter. Blood culture specimens drawn by separate percutaneous venipuncture were obtained from patients with fever or other signs of infection. Microbiologic Methods. Skin of the insertion site: Twenty square centimeters of skin about the catheter were cultured quantitatively by applying a sterile cardboard template and culturing the defined 20 cm2 of exposed skin with a sterile cotton-tip applicator (Culturette, American Scientific Products), with vigorous scrubbing of the entire area four times, twice in one direction and twice more perpendicularly. In the laboratory, the tip of the applicator was immersed in 1 ml of sterile saline and mixed vigorously (Vortex, Scientific Products), following which serial dilutions were inoculated onto the surface of 5 percent sheep blood agar that was incubated aerobically in a IO percent carbon dioxide atmosphere at 37% for three days before discarding. All colony types were enumerated and identified [34]. Catheters: Catheters were removed, with the externalized portion directed upward, to minimize contamination by organisms on the skin surface. Two 5-cm segments of the catheter were cultured: the proximal intracutaneous portion, extending from several millimeters inside the former skin-catheter interface to the vein, and the tip [35,36]; segments were amputated using sterile scissors and transported in a sterile humidified tube (Culturette). Details of the semiquantitative method for culturing vascular catheters have been reported [35,36]. Briefly, each catheter segment was transferred aseptically from its transport tube onto the surface of a IOO-mm 5 percent sheep blood agar plate, and with use of a sterile forceps, was rolled back and forth across the agar surface while gentle downward pressure was exerted to affirm adequate surface-to-surface contact. Plates were incubated aerobically in a IO percent carbon dioxide atmosphere at 37OC for at least 72 hours. All colony types appearing were enumerated and identified [34]. All cultures were processed in a monitored bticteriologic high-efficiency-filtration laminar-flow hood (Abbott Labora-
tories). Although it was not possible to mask the identity of the catheters’ treatment groups from the co-investigator assessing catheters, the research microbiologists processing all cultures were unaware of the identity of each catheter’s treatment group. Definitions. Local catheter-related infection: A positive result of a semiquantitative culture of the catheter (15 or more colony-forming units [cfu]) [35-371; a positive semiquantitative culture result represents colonization of the catheter. Catheter-related septicemia: (1) a semiquantitative
catheter culture and blood cultures demonstrating positive results for the same species;
(2) clinical
(or autopsy)
and
microbiologic data disclosed no other apparent source for the septicemia [33,35,36]. Statistics. The significance of differences between the two groups was determined with a chi-square test or Fisher’s exact test for categoric data and the Student t test for continuous data.
TABLE
DEFATTING-MAKI
and McCORMACK
Features of Patients and Catheters Two Site Care Groups
I
Feature Patients Age (years) Mean Range Male:female (percent) Service (percent) Surgical Medical Intensive care unit (percent) Neoplastic disease (percent) Type I diabetes mellitus (percent) Antimicrobial therapy (percent) Corticosteroid therapy (percent) Active infection unrelated to the catheter (percent) Catheters Days in place Mean f SD Range
Control (n = 51)
51
AcetoneTreated (n = 49)
54 20-80 58:42
15-80 61:39
13.5
in the
p Value
NS NS
84 16 45
75 2.5
NS
29
<0.05
33
35
NS
4
4
NS
84
80
NS
18
14
NS
53
69
NS
14.6 f 7.9 3-37
NS
f 10.0 4-57
RESULTS Characteristics of the Study Population. One hundred consecutively enrolled subclavian catheters were prospectively studied, 49 in which acetone was used in the site care regimen and 5 1 in which no acetone was used. Every catheter had been inserted percutaneously at a new
site into the subclavian vein by an infraclavicular approach; none of the catheters were inserted over a guidewire into an old site. Patients
in the two groups
were
very similar
demo-
graphically and, except for a higher percentage of control patients within an intensive care unit, were comparable in terms of risk factors predisposing to infection (Table I). The majority of catheters in each group were used in patients more susceptible to nosocomial infection [38], as reflected by the relatively high age of the study population, the preponderance of surgical patients, and the high frequency of neoplastic disease, recent surgery, and intensive care unit placement; over half of the patients in each group had an active infection unrelated to the catheter, and three fourths of the patients in each group received systemic antimicrobial therapy at some time while the study catheter was in place. The average length of the time study catheters remained in place was also very similar in the two groups, 14 and 15 days. Most catheters in each group were removed because
November 1987
The American Journal of Medicine
Volume 83
835
ACETONE
DEFATTING-MAKI
TABLE II
and McCORMACK
Colonization of Skin of the Insertion Site in the Two Site Care Groups
about 6 percent of the time with a potential pathogen other than coagulase-negative Staphylococcus, such as S. aureus, an enteric gram-n&gative bacillus, or Candida. Catheter-Related Infection. Four (8 percent) catheters in the control group and four (8 percent) in the acetone group were found to be colonized on removal, as defined by a positive semiquantitative catheter culture result (Table Ill). Two control catheters (4 percent) caused septicemia, one with S. aureus and the other with coagulasenegative Staphylococcus and Klebsiella pneumoniae. Two catheters in the acetone group caused septicemia (4 percent), one with Candida albicans and the other with Pseudomonas aeruginosa. Catheter-related infection was strongly associated with concordant cutaneous colonization of the insertion site prior to catheter removal: the site was shown to be colonized by the same species as the catheter in seven (85 percent) of the eight infected catheters, including all four tiausing bacteremia or candidemia. Colonization of the site was demonstrated with 24 catheters overall in both groups; as noted, in seven (29 percent), the catheter was found to be infected by the same species; in contrast, of 76 catheters without detectable colonization of the site, only one (1 percent) was infected on kernoval (relative risk of concordant catheter-related infection with a colonized site, 22.1, p
Acetone-
Feature
Control
Total
number of site cultures Number (percent) showing growth (cfu/cm*) None 1-9 10-99 2100 Number (percent) showing colonization by Coagulase-negative staphylococci Staphylococcus aureus Gram-negative bacilli Yeasts
TABLE III
256
187 17 17 35
Treated
p
Value
275
(73.1) (6.6) (6.6) (13.7)
214 22 10 29
(7T.8) (8.0) (3.6) (10.5)
NS NS NS NS
37 (14.4)
32 (11.6)
NS
8 (3.1) 14 (5.4) 7 (2.7)
6 (2.2) 12 (4.3) 7 (2.5)
NS NS NS
Catheter-Related Infection in the Two Site Care Groups
Parameter* Number of catheter-related infections (percent) Local+ Bacteremia Infecting organisms, causing local infection (bacteremia) Coagulase-negative staphylococci Staphylbcoccus aureus Gram-negative bacilli Candida Enterococcus * None of the differences significant at p <0.05. + Defined as 15 colony-forming [35-371.
between units
the
Control (n = 51)
AcetoneTreated (n = 49)
4 (6) 2 (4)
4 (6) 2 (4)
2 (1) 2 (1)
2
l(l) . . 1
i ;-i, l(l) . . .
groups
is statistically
on semiquantitative
culture
COMMENTS total parenteral nutrition had been terminated electively; : only IO catheters were removed because of suspected catheter-related sepsis. Cutaneous Colonization of Catheter Insertion Sites. There were no Significant differences between the two groups in skin colonization contiguous to the catheter (Table II). Approximately three fourths of the site cultures in each group showed no growth; about one fourth yielded organisms, with a similar distribution of quantitative colonization in the two groups (Table II). Most cutaneous colonization was with coagulase-negative staphylococci; colonization by Staphylococcus aureus, gram-negative bacilli, or yeasts was infrequent in both groups (Table II). Heavy colonization (more than 100 cfu/cm*) was found at approximately 10 percent of the sites in each group, 836
November
1987
The American
Journal
of Medicine
Volume
An abundance of clinical and microbiologic data suggests strongly that most intravenous catheter-related septicemias are caused by cutaneous organisms that invade the intracutaneous wound during catheter insertion or sometime thereafter: (1) Studies have shown a strong correlation between organisms present on skin surrounding the catheter wound and microorganisms recovered from catheters producing septicemia [3,11,12]. (2) Burn patients with massive populations of SUrfaCe microorganisms experience very high rates of catheterrelated sepsis [36,39-411. (3) Clusters of catheter-related septicemias have been traced to contaminated disinfectants used for cutaneous antisepsis [42-441 and to contaminated tape [45]. 83
ACETONE
(4) As shown in this study (Table III) and in numerous other recent studies of catheter-related infection [i -71, coagulase-negative Staphylococcus, the predom/nant aerobic species colonizing the human skin [46], is now the most common agent of catheter-related infection. (5) Hospitals with intravenous therapy teams that stress asepsis in catheter insertion and site care report substantially lower rates of catheter-related septicemia [i-7,13-17]. (6) Controlled trials have shown that application of a topical polyantimicrobial agent to peripheral venous catheter insertion sites provides some protection against catheter-related infection [47-521. (7) Heavy growth on semiquantitative culture of the external surface of a removed catheter has been shown to be strongly associated with bacteremia caused by the catheter [11,12,35-3753,541. (8) Microscopic studies of infected central venous catheters have shown organisms primarily on the external surface of the catheter [54]. Thus, the details of site care-the regimen for disinfecting the site prior to catheter insertion, the dressing, and the follow-up care-would seem likely to have a profound influence on the incidence of catheter-related infection. Floral populations can be effectively suppressed with an antiseptic solution prior to insertion of the catheter. However, extrinsic nosocomial contaminants or regrowth of the suppressed endogenous microflora can colonize the site and cause catheter-related infection. Dressings, intended to prevent extrinsic contamination of the insertion site, may, paradoxically, adversely affect cutaneous microbial populations in the vicinity of the catheter [55631. Although central venous catheters in most hospitals are redressed at periodic intervals-typically every other day-few studies [22,64], none of adequate size, have examined the effectiveness of periodic site care for central venous catheters, vis-a-vis, have determined whether periodically recleansing the skin about the catheter with an antiseptic and redressing the site reduces colonization sufficiently so as to significantly reduce the risk of systemic catheter-related infection. Many published protocols for disinfection of hyperalimentation catheter insertion sites and for follow-up care of the site advocate defatting the skin with acetone [ 15 291, presumably an the belief microbial removal is augmented. Use of acetone has become standard practice in many institutions without confirmation of its efficacy by controlled trial. We have been unable to find any clinical studies of the value of defatting as an adjunct to cutaneous disinfection. Healthy intact skin is one of the body’s most important barriers to deep infection. Pathogenic microorganisms such as Streptococcus pyogenes, S. aureus, and gramnegative bacilli placed on the surface of intact skin rapidly die and do not readily become established as part of the November
TABLE IV
DEFATTING-MAKI
and McCORMACK
Condition of Catheter Insertion Two Site Care Groups
Sites in the
Acetone-
Feature
Control
Number of catheters (total number of site evaluations) Catheters on one or more site evaluations showing (percent) Pain or tenderness Denuding Erythema Discomfort during site care One or more of the above Site evaluations showing (percent) Pain or tenderness Erythema Denuding
Treated pValue
51 (256)
49 (275)
10 2 29 12
76 18 65 82
35
80
5.6 18.7 0.8
59.1 42.2 10.2
normal cutaneous flora [30-321, which consists primarily of aerobic and anaerobic diphtheroids and micrococci, in concentrations ranging from 10 to lo6 microorganisms per cm*, depending on the method of sampling and the anatomic area sampled [46,65,66]; gram-negative bacilli and yeasts do not usually constitute a significant portion of the cutaneous microflora except on the perineum and feet. A variety of physical-chemical factors at the skin surface are thought to contribute to the intrinsic ability of normal skin to resist colonization by more pathogenic microorganisms, including a slightly acidic pH, surface desiccation, bacterial interference mediated by the less pathogenic normal flora, and the antimicrobial activity of sweat and sebum [66]. Sebum, the product of apocrine sweat glands, consists of a complex mixture of lipids including triglycerides (approximately 50 percent by weight), wax esters (25 percent), squalene (10 percent), cholesterol esters (2 percent), and cholesterol (2 percent), with considerable variation by anatomic location [66]. Triglycerides are hydrolyzed to free fatty acids, mainly by lipases produced by the normal skin flora [67]. In vitro, most of the skin’s fatty acids, ranging from C6 to GIG, have antibacterial activity and inhibit ‘growth of a variety of microorganisms, partricularly gram-positive cocci and yeasts, and to a much lesser extent, gramnegative bacilli [30,68]. Ricketts et al [30] first reported that S. pyogenes placed on normal skin rapidly died, even if desiccation was prevented; likewise, S. aureus survived only three days under the same conditions, and gram-negative bacilli died rapidly unless the skin surface was kept nonphysiologically humid. Topical application of albumin, which binds to free fatty acids, greatly prolonged the survival of both S. 1987
The American
Journal
of Medicine
Volume
93
837
ACETONE
DEFATTING-MAKI
and McCORMACK
pyogenes and S. aureus. Extraction of the skin lipids by application of acetone resulted in prolonged survival of S. pyogenes. Survival of different microorganisms after acetone treatment was found to be related to their in vitro susceptibility to the unsaturated free fatty acids, especially oleic acid, normally present on the skin surface. Escherichia coli and P. aeruginosa were found to be resistant to the cutaneous unsaturated fatty acids, and survival of these organisms was unaffected by defatting. Lacey [31] and Aly et a! [32] subsequently confirmed and extended these studies, showing that pretreatment of normal skin with acetone greatly prolonged the survival of S. aureus, S. pyogenes, and C. albicans, but not E. coli or P. aeruginosa. Aly and co-workers [32] further showed that reapplying the’lipid acetone extracts from normal skin to acetone-treated skin restored much of its intrinsic antimicrobial activity. These experimental studies led us to question whether defatting with acetone might actually be deleterious and promote colonization by staphylococci, Candida, and other potential pathogens and, paradoxically, increase the risk of catheter-related septicemia. The findings from our prospective randomized comparative trial suggest that the use of acetone in the regimen for disinfecting the insertion site does not measurably impair cutaneous antisepsis; however, it clearly does not augment it either. Using acetone did not measurably affect skin colonization (Table II); and, as might then be expected, also did not significantly influence the incidence of catheter-related infection (Table Ill). However, use of acetone greatly increased local inflammation of the catheter site and discomfort to the patient, typically pain during its application (Table IV). Type II error (beta) should be considered when designing clinical trials and analyzing the results [69]. Given the relatively low rates of catheter-related infection in both treatment groups in our study (Table Ill), to have been able to identify with statistical certainty a true difference of 25 percent in the incidence of related bacteremia would have required study of several thousand catheters. We point out, however, that even though there could well be a true (statistically significant) difference between the two treatment groups, the rate of bacteremic infection in hyperalimentation with reasonable aseptic care is now relatively low, and any differences, even if statistically significant, would probably be of marginal relevance clinically. We can conclude from our data with strong statistical power
(beta less than 10 percent) that use of acetone does not measurably improve the microbial removal of the disinfecting regimen (colonization of sites was unaffected, Table II), but greatly increases local inflammation and discomfort to the patient (Table IV). Thus, we believe that using acetone as part of the regimen of cutaneous antisepsis in vascular catheter care no longer has a rational basis. Our data are consonant with the findings of other studies reporting that most catheter-related infections in hyperalimentation derive from organisms colonizing the skin contiguous to the catheter [ 11,12]. In seven of the eight colonized catheters (more than 15 cfu on semiquantitative catheter culture), including all four catheters considered to have caused septicemia, the same microbial species isolated from the catheter was isolated one or more times in quantitative skin cultures performed at dressing changes prior to removal of the catheter. The importance of assiduous aseptic technique during catheter insertion and at follow-up dressing changes cannot be emphasized too strongly. These data further suggest that culturing the skin contiguous to the catheter at the time of dressing changes using a standardized technique, such as was employed in this study, may be of predictive value in monitoring for incipient catheter-related sepsis [ 11,121. Virtually all hyperalimentation programs with low rates of infection routinely provide periodic site care, usually daily or every other day. Our data suggest that heavy skin colonization contiguous to the catheter can be effectively suppressed by regular site care (Table II), In a small crossover study in 15 patients, Jarrard et al [22] found that recleansing the site daily with an iodophor solution was more effective than use of a topical iodophor ointment in preventing colonization of the site; but the best antiseptic to use and the cost-effectiveness of periodic site care, especially how frequently it should be performed, need to be prospectively studied. ACKNOWLEDGMENT We are grateful to Carol A. Hassemer and Carla A. Alvarado for microbiologic support, and the physicians and nurses of the University of Wisconsin Hospital and Clinics for permitting us to study their patients with Central venous catheters. We also express our appreciation to Ms. Peggy Barden for secretarial assistance in preparation of the manuscript.
REFERENCES 1. 2. 3.
838
Maki DG, Goldman DA, Rhame FS: Infection control in intravenous therapy. Ann Intern Med 1973; 79: 876-887. Goldmann DA, Maki DG: Infection control in total parenteral nutrition. JAMA 1973; 223: 1360-1364. Makf DG: Sepsis arising from extrinsic contamination of the infusion and measures for control. In: Phillips I, ed. Mfcrobiological hazards of infusion therapy. Lancaster, En-
November
1987
The American
Journal
of Medicine
4.
5.
Volume
83
gland: MTP Press, 1977; 99-141. Maki DG: Infections associated with intravascular lines. In: Swartz M, Remmington J, eds. Current topics in clinical infectious disease. New York: McGraw-Hill, 1982; 309363. Centers for Disease Control Working Group: Guidelines for prevention of intravenous therapy-related infections. In-
ACETONE
6.
7. 8.
9.
10.
11.
12.
13.
14. 15. 16. 17.
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
feet Control 1981; 3: 62-79. Allen JR: The incidence of nosocomial infection in patients receiving total parenteral nutrition. In: Johnston IDA, ed. Advances in parenteral nutrition. Lancaster, England: MTP Press, 1977; 339-377. Williams WW: Infection control during parenteral nutrition therapy. J Parenter Enter Nutr 1985; 9: 7352746. Curry CR, Quie PG: Fungal septicemia in patients receiving parenteral hyperalimentation. N Engl J Med 1971; 285: 1221-1225. Plouffe JF, Brown DG, Silva J Jr, Eck T, Stricof RL, Fekety FR Jr: Nosocomial outbreak of Candida parapsilosis fungemia related to intravenous infusions. Arch Intern Med 1977; 137: 1686-1689. Solomon SL, Khabbaz RF, Parker RH, et al: An outbreak of Candida parapsilosis bloodstream infections in patients receiving parenteral nutrition. J Infect Dis 1984; 149: 98102. Snydman DR, Pober ER, Murray SA, Gorbea JF, Majka JA, Perry LK: Predictive value of surveillance skin cultures in total-parenteral nutrition-related infection. Lancet 1982; II: 1385-1388. Bjornson HS, Colley R, Bower RH, Duty VP, Schwartz-Fulton JT, Fischer JE: Association between microorganism growth at the catheter site and colonization of the catheter in patients receiving total parenteral nutrition. Surgery 1982; 92: 720-727. Freeman JB, Lemire A, MacLean LD: Intravenous alimentation and septicemia. Surg Gynecol Obstet 1972; 135: 708-712. Nehme AE: Nutritional support of the hospitalized patient: the team conceot. JAMA 1980: 243: 1906-1908. Wilmore DW, Dudrick SJ: Safe long-term venous catheterization. Arch Surg 1969; 98: 258-258. Sanderson I, Dietel M: Intravenous hyperalimentation without sepsis. Surg Gynecol Obstet 1973; 136: 577-585. Copeland EM, MacFadyen BV, McGown C, Dudrick SJ: The use of hyperalimentation in patients with potential sepsis. Surg Gynecol Obstet 1974; 138: 377-381. Kaminski MV Jr, Stolar MH: Parenteral hyperalimentationa quality of care survey and review. Am J Hosp Pharm 1974; 31: 228-235. Ryan JA, Abel RM, Abbott WM, et al: Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. N Engl J Med 1974; 290: 757-761. Phillips KJ: Nursing care in parenteral nutrition. In: Fischer JE, ed. Total parenteral nutrition. Boston: Little, Brown, 1976; 101-l 10. Ota DM, lmbembo AL, Zuidema GD: Total parenteral nutrition. Surgery 1978; 83: 503-520. Jarrard MM, Olson CM, Freeman JB: Daily dressing change effects on skin flora beneath subclavian catheter dressings during total parenteral nutrition. J Parenter Enter Nub 1980; 4: 391-392. Palidar PJ, Simonowitz DA, Oreskovich MR, et al: Use of opsite as an occlusive dressing for total parenteral nutrition catheters. J Parenter Enter Nutr 1982; 6: 150-15 1. Schwartz-Fulton J, Colley R, Valanis 8, Fischer JE: Hyperalimentation dressings and skin flora. National Intravenous Therapy Association 1981; 4: 354-357. Padberg FT Jr, Ruggiero J, Blackburn GL, Bistrian BR: Central venous catheterization for parenteral nutrition. Ann Surg 1981; 193: 264-270. Bozzehi F, Terno G, Bonfanti G, et al: Prevention and treatment of central venous catheter sepsis by exchange via a guidewire. A prospective controlled trial. Ann Surg 1983; 198: 48-52. Nehme AE, Trigger JA: Catheter dressings in central parenteral nutrition: a prospective randomized comparative study. Nutritional Support Services 1984; 9: 42-43. Powell CR, Traetow MJ, Fabri PJ, Kudsk KA, Ruberg RL: Op-
November
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
1987
DEFAlTING-MAKI
and McCORMACK
site dressing study: a prospective randomized study evaluating povidone iodine ointment and extension set changes with 7-day Op-Site dressings applied to total parenteral nutrition subclavian sites. J Parenter Enter Nutr 1985; 9: 443-446. Sitzmann JV, Townsend TR, Siler MC, Bartlett JG: Septic and technical complications of central venous catheterization. A prospective study of 200 consecutive patients. Ann Surg 1985; 202: 766-770. Ricketts CR, Squire JR, Topley E: Human skin lipids with particular reference to the self-sterilizing power of the skin. Clin Sci 1951; 10: 89-l 11. Lacey RW: Antibacterial action of human skin. In vivo effect of acetone, alcohol, and soap on behavior of Staphylococcus aureus. Br J Exp Pathol 1968; 49: 209-215. Aly R, Maibach HI, Shinefield HR, Strauss WG: Survival of pathogenic microorganisms on human skin. J Invest Dermatol 1972; 58: 205-210. Band JD, Maki DG: Safety of changing intravenous delivery systems at longer than 24-hour intervals. Ann Intern Med 1979; 91: 713-718. Lennette E, Balows A, Hausler W, Shadomy H, eds. Manual of clinical microbiology, 3rd ed. Washington: American Society of Microbiology, 1980. Maki DG, Weise GE, Sarafin HW: A semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977; 296: 1305-1309. Maki DG, Jarrett F, Sarafin HW: A semiquantitative culture method for identification of catheter-related infection in the burn patient. J Surg Res 1977; 22: 513-520. Collignon PJ, Soni N, Pearson IY: Is semiquantitative culture of central vein catheter tips useful in the diagnosis of catheter-associated bacteremia? J Clin Microbial 1986; 24: 532-535. Haley RW, Hooton TM, Culver DH, et al: Nosocomial infections in U.S. hospitals, 1975-1976: estimated frequence by selected characteristics of patients. Am J Med 1981; 70: 947-959. Stein JM, Pruitt BA Jr: Suppurative thrombophlebitis. A lethal iatrogenic disease. N Engl J Med 1981; 70: 947-949. Popp MB, Law EJ, MacMillan BG: Parenteral nutrition in the burned child: a study of twenty-six patients. Ann Surg 1974; 179: 219-225. Pruitt BA Jr, McManus WF, Kim SH, Treat RC: Diagnosis and treatment of cannula-related intravenous sepsis in burn patients. Ann Surg 1980; 191: 546-554. Plotkin SA, Austrian R: Bacteremia caused by Pseudomonas sp. following the use of materials stored in solutions of a cationic surface-active agent. Am J Med Sci 1958; 235: 621-627. Dixon RE, Kaslow RA, Macket DC, Fulkerson CC, Mallison GF: Aqueous quaternary ammonium antiseptics and disinfectants. Use and misuse. JAMA 1976; 236: 2415-2417. Frank MJ, Schaffner W: Contaminated aqueous benzalkonium chloride. An unnecessary hospital infection hazard. JAMA 1976; 236: 2418-2419. Sheldon DL, Johnson WC: Cutaneous mucormycosis: two documental cases of suspected nosocomial cause. JAMA 1979; 241: 1032-1033. Kligman AM: The bacteriology of normal skin. In: Maibach HI, Hildick-Smith G, eds. Skin bacteria and their role in infection. New York: McGraw-Hill, 1965; 13-32.. Moran JM, Atwood RP, Rowe MI: A clinical bacteriologic study of infections associated wtih venous cut downs. N Engl J Med 1965; 272: 554-560. Norden CW: Application of antibiotic ointment to the site of venous catheterization-a controlled trial. J Infect Dis 1969; 120: 611-615. Zinner SH, Denny-Brown BC, Braun P, Burke JP, Kass EH: Risk of infection with intravenous indwelling catheters: effect of application of antibiotic ointment. J Infect Dis
The American
Journal
of Medicine
Volume
83
839
ACETONE
50.
51.
52.
53.
54.
55.
56. 57.
58.
840
DEFATTING-MAKI
and McCORMACK
1969; 120: 616-619. Levy RS, Goldstein J, Pressman RS: Value of a topical antibiotic ointment in reducing bacterial colonization of percutaneous venous catheters. J Albert Einstein Med Center 1970; 18: 67-70. Maki DG, Band JD: A comparative study of polyantibiotic and iodophor ointments in prevention of vascular catheterrelated infection. Am J Med 1981; 70: 739-744. Maki DG, Will L: Study of polyantibiotic and povidone-iodine ointments on central venous and arterial catheters dressed with gauze or polyurethane dressings. In: Program and Abstracts of the Twenty-Sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, September 30-October 2, 1986, New Orleans, Louisana. Washington: American Society for Microbiology, 1986. Snydman DR, Murray SA, Kornfeld SJ, Majka JA, Ellis CA: Total parenteral nutrition-related infections. Prospective epidemiologic study using semiquantitative methods. Am J Med 1982; 73: 695-699. Cooper GL, Hopkins CC: Rapid diagnosis of intravascular catheter-associated infection by direct gram staining of catheter segments. N Engl J Med 1985; 18: 1142-l 150. Hirman CD, Maibach HI: Effect of air exposure and occlusion on experimental human skin wounds. Nature 1963; 200: 377-378. Marples RR, Kligman AM: Growth of bacteremia under adhesive tapes. Arch Dermatol 1969; 99: 107-l 10. Aly R, Shirley C, Cunico B, et al: Effect of prolonged occlusion on the microbial flora, pH, carbon dioxide and transepidermal water loss on human skin. J Invest Dermatol 1978; 71: 378-381. Mertz PM, Eaglestein WH: The effect of a semiocclusive dressing on the microbial population in superficial wounds. Arch Surg 1984; 119: 287-289.
November
1987
The
American
Journal
of Medicine
59.
Powell C, Ellison EC, Blackson L, et al: Increased catheter sepsis with use of semipermeable membrane dressings system (Opsite) in total parenteral nutrition (abstr). J Parenter Enter Nutr 1979; 3: 515. Katich M, Band J: Local infection of the intravenous-cannulae wound associated with transparent dressings. J Infect Dis 1985; 151: 971-972. Mumford F, Grossman A, Hilton E: An outbreak of intravascular site infections (abstr). Am J Infect Control 1985; 13: 140. Powell C, Regan C, Fabri PJ, Ruberg RL: Evaluation of opsite catheter dressings for parenteral nutrition: a prospective, randomized study. J Parenter Enter Nutr 1982; 6: 43-46. Andersen PT, Herlevsen P, Schaumburg H: A comparative study of ‘Op-site’ and ‘Nobecutan gauze’ dressings for central venous line care. J Hosp Infect 1986; 7: 161-168. Jarrard MM, Freeman JB: The effects of antibiotic ointments and antiseptics on the skin flora beneath subclavian catheter dressings during intravenous hyperalimentation. J Surg Res 1977; 22: 521-526. Marples MJ: The ecology of the human skin. Springfield, Illinois: Charles C Thomas, 1965. Noble WC, Somerville DA: Microbiology of human skin. Philadelphia: WB Saunders, 1974; l-341. Scheimann LG, Knox G, Sher D, Rothman S: The role of bacteria in the formation of free fatty acids on the human skin surface, J Invest Dermatol 1960; 34: 171-174. Pillsbury DM, Rebel1 G: The bacterial flora of the skin. Factors influencing the growth of resident transient organisms. J Invest Dermatol 1952; 18: 173-186. Freiman JA, Chalmers TC, Smith H Jr, Kuebler RR: The importance of beta, and type II error and sample size in the design and interpretation of the randomized control trial. N Engl J Med 1978; 299: 690-694.
60.
6 1.
62.
63.
64.
65. 66. 67.
68.
69.
Volume
83