The Effect of Prophylactic Ointment Therapy on Nosocomial Sepsis Rates and Skin Integrity in Infants With Birth Weights of 501 to 1000 g
Objective. Extremely low birth weight infants have a high risk of developing nosocomial bacterial sepsis (NBS). Immature fragile skin may represent an inadequate protective barrier to bacteria colonizing the skin. We conducted a randomized, multicenter trial to determine whether prophylactic application of an emollient ointment would result in a lower incidence of death and/or NBS in the first 28 days of life, compared with routine skin care.
Methods. Infants of birth weight 501 to 1000 g and gestational age ≤30 weeks were assigned randomly to receive generalized application of ointment twice a day through day 14 (prophylactic group [P]) or local application of ointment to the site of injury (routine skin care [R]). The study was conducted at 53 neonatal intensive care units that were members of the Vermont Oxford Network.
Results. Included in the analysis were 1191 infants (P: 602; R: 589). No difference was found in the combined primary outcome of NBS or death (33.6% P vs 30.3% R; relative risk [RR]: 1.10; 95% confidence interval [CI]: 0.89, 1.27). The incidence of death was no different between the groups (10.8% P vs 12.1% R; RR: 0.87; 95% CI: 0.59, 1.25). More infants in the prophylactic group had NBS (25.8% P vs 20.4% R; RR: 1.26; 95% CI: 1.02, 1.54), predominantly in the lower birth weight infants (501–750 g) and for infections caused by coagulase-negative staphylococci. Infants in the prophylactic group had better skin condition on days 1 to 14 of life and less skin injury on days 15 to 28 of life. There was no difference between groups in other complications of prematurity.
Conclusions. Prophylactic application of ointment did not lead to a difference in death and/or NBS in the first 28 days of life. There may be an increase in the risk of NBS associated with this practice.
- clinical trial
- randomized controlled trial
- multicenter trial
- skin care
- nosocomial sepsis
Nosocomial bacterial sepsis (NBS) is a frequent and serious complication of the care of premature infants. Extremely low birth weight (ELBW) infants with birth weights of <1000 g are particularly susceptible, with as many as 19% to 44% developing NBS.1 Infants who develop NBS have a higher mortality rate as well as prolonged and more expensive hospital stays.2–6
The increased susceptibility of ELBW infants to infection has been attributed to less-effective immune function, compared with more mature newborns,7 and the invasive nature of necessary supportive care. Risk factors for developing NBS include the presence and duration of indwelling intravascular catheters, exposure to parenteral alimentation (especially intravenous lipids), endotracheal intubation, and the presence and duration of mechanical ventilation.3,8–12 Breakdown of the barrier function of the skin may be an additional risk factor. Nopper et al13 demonstrated improved skin condition and a decreased rate of nosocomial infection associated with application of an emollient ointment. We performed a study to determine whether the prophylactic application of emollient ointment (Aquaphor Original Emollient Ointment, Beiersdorf Inc, Norwalk, CT) would reduce the risk of NBS and/or mortality in the first 28 days of life.
This multicenter, randomized, clinical trial was conducted from August 1998 through March 2000 in 53 neonatal intensive care units in the Vermont Oxford Network (see Appendixes 1–4). Inclusion criteria were: birth weight of 501 to 1000 g; gestational age of ≤30 weeks; blood culture(s) obtained and antibiotics initiated; and expected survival for >48 hours. Infants were excluded for life-threatening congenital anomalies, congenital skin anomalies, hydrops fetalis, congenital infection of the skin, or previous treatment with an ointment. Each participating center’s institutional review board approved the protocol. Written, informed consent was obtained from a parent of each infant.
Enrolled infants were randomized within 48 hours of birth to receive either prophylactic ointment or routine skin care. Randomization was stratified into 2 birth weight categories (501–750 and 751–1000 g) and performed at each study site using sealed, opaque envelopes.
To achieve consistency in conducting the study, each study site was provided a manual of operations and a videotape demonstrating application procedures and other study protocols. Use of ointment (Aquaphor) was prescribed by protocol. Aquaphor contains petrolatum, mineral oil, mineral wax, and wool wax alcohol. Infants in the prophylactic ointment group had a measured amount of ointment applied every 12 hours to the entire body surface except the scalp and face. Application began by 48 hours of age and continued through the 14th day of life. A 0.25-oz tube of ointment was issued daily for 2 applications and then discarded. Before application, the tube was placed in a clean basin of warm water for 2 to 3 minutes to enhance spreading. The dose for the prophylactic ointment group was a ribbon of ointment measured on a single-use sheet of ruled, nonabsorbent, translucent paper provided for the study. The dose for the 501- to 750-g infants was a 3- to 4-inch ribbon of ointment and for the 751- to 1000-g infants, a 4- to 5-inch ribbon. Areas of skin trauma, open wounds, or incisions did not receive application.
The routine skin-care group received no generalized application of ointments, oils, creams, or other products placed on the skin. The use of protective adhesive films was allowed only for securing catheters or treating localized skin breakdown. Ointment treatment limited to the area of involvement was allowed in either group for severe dermatitis or local skin injury for a maximum of 3 days. A second treatment was allowed for persistence or recurrence of the skin condition.
Recommendations were made for an initial bath using a pH-neutral soap within 48 hours of birth. Other skin-care practices were at the discretion of care providers and not regulated by study protocol.
Infants in both study groups received a minimum of 48 hours of broad-spectrum antibiotic therapy after birth. All additional decisions about evaluation for infection or duration and choice of antibiotic treatment were at the discretion of the care providers. Evaluation for bacterial or fungal sepsis required a minimum of a single blood culture from any site.
The primary outcome was the combination of death or NBS through day 28 of life. An episode of NBS was defined as a blood culture positive for a bacterial organism with intention to treat with antibiotics for ≥5 consecutive days. Mortality was defined as death occurring through day of life 28 from any cause. Secondary outcomes through day of life 28 included number of sepsis evaluations, number of episodes of NBS, incidence of nosocomial fungal sepsis, incidence of meningitis, number of endotracheal reintubations, weekly weights, and skin condition. The condition of the skin and the presence of local skin injury were assessed daily during the first 28 days. Skin condition was graded on a 5-point scale (0–4) modified from the one devised by Lane and Drost.14 Secondary outcomes through discharge included complications of prematurity and hospital length of stay. Complications relating to prematurity were prospectively defined using the Vermont Oxford Network database definitions.15 Complications potentially related to ointment administration were reported separately as adverse events.
A sample size of 1200 infants was needed to show an absolute reduction in mortality or NBS from 40% to 32% (α: .05; β: .80). Baseline group characteristics were compared by using t tests for continuous variables and χ2 analysis for categorical variables. Analysis of primary and secondary outcomes was based on assignment to group by intention to treat. Mantel-Haenszel techniques were used to determine relative risk (RR) and 95% confidence intervals (CIs) using a 2-tailed analysis. For the primary outcome of NBS and/or death and for both component outcomes of NBS and death, potential for confounding was assessed by logistic regression. Independent variables for this analysis included maternal race and all baseline comparison variables that either differed significantly between the treatment groups or themselves were significantly related to the primary outcome. Treatment group was forced into the model, and then a stepwise procedure using a .05 significance level was used to select the other variables for the model. The adjusted relative risk (ARR) then was calculated from the adjusted odds ratio for treatment group obtained from this logistic regression.16 Data were analyzed by using SAS 6.12 (SAS Institute Inc, Cary, NC).
An independent data and patient safety monitoring committee (see Appendix 3) reviewed 1 interim analysis after the first 6 months of the trial and a second, targeted report prepared to address concerns raised by participants about fungal infection rates.
Study enrollment, randomization, and exclusions from analysis are profiled in Fig 1. Of 2660 infants meeting gestational age and birth weight criteria admitted to participating centers during the study period, 1206 infants met all entry criteria and were enrolled and randomized. Similar numbers were excluded from analysis from each group because of inappropriate randomization, failure to document informed consent, or withdrawal by parent request. The final analysis was a comparison of 602 infants in the prophylactic ointment group with 589 infants in the routine skin-care group.
The baseline maternal and neonatal characteristics of the prophylactic ointment and routine skin-care groups were similar except for prenatal care, delivery by cesarean section, and gender (Table 1). More infants in the prophylactic ointment group had mothers receiving prenatal care (96.8% vs 94.4%; P = .04), were delivered by cesarean section (65.5% vs 60.2%; P = .06), and were male in gender (54.3% vs 47.3%; P = .02).
Care practices initiated in the first 24 hours of life were not significantly different between groups, including the use of radiant warmers, incubators, the placement of umbilical arterial or venous catheters, the administration of endotracheal surfactant therapy, or the initiation of conventional or high-frequency mechanical ventilation.
During the first 14 days of life, 98% of infants in the prophylactic group received ointment application for an average of 12.0 days. Infants in the routine skin-care group were allowed to receive ointment treatment limited to the area of dermatitis or injury. During the first 14 days of life, 34% of infants in the routine skin-care group received ointment for an average of 4.1 (±2.75 SD) days. During days 15 through 28 of life, 33% of infants in the prophylactic group received ointment for an average of 2.6 (±2.63 SD) days, compared with 15% in the routine skin-care group for 2.8 (±2.01 SD) days. This difference was felt largely to be caused by misinterpretation of the protocol, which specified application through day 14 of life rather than for 14 days from initiation of therapy.
The primary outcome of death or NBS by day 28 of life occurred in 33.6% of infants in the prophylactic ointment group and in 30.3% of infants in the routine skin-care group (Table 2). The ARR for death or sepsis for the prophylactic ointment group was 1.07 (95% CI: 0.89–1.27). The risk for the combined primary outcome of death or NBS was also not significantly different between treatment groups for either of the 2 birth weight subcategories. The independent variables used in calculating the ARR differed somewhat for analysis of the group as a whole or for the 2 birth weight subcategories but included some or all of the following: gestational age, male gender, birth weight, outborn place of birth, and surfactant in the first 24 hours of life. The other independent variables did not remain significant in the multivariate models.
A secondary analysis of the components of the primary outcome demonstrated that the risk for death was not significantly different between groups for the entire study population or for either of the 2 birth weight groups. The risk for NBS was significantly greater in the prophylactic ointment group (ARR: 1.27; 95% CI: 1.03–1.54). A subgroup analysis by birth weight demonstrated that this increased risk was only significant for the lower birth weight group of 501–750 g (ARR: 1.43; 95% CI: 1.05–1.86).
Organisms responsible for NBS were Gram positive in 82% of the bacterial blood culture isolates (Table 3). The distribution of organisms causing sepsis was similar between study groups. Coagulase-negative staphylococcus was the cause of NBS in >60% of the cases. The incidence of NBS was analyzed as episodes caused by coagulase-negative staphylococci and those caused by all other bacterial organisms (Table 4). A significant difference in the incidence of NBS caused by coagulase-negative staphylococci was found between groups. The RR of coagulase-negative staphylococcal sepsis for the prophylactic ointment group was 1.40 (95% CI: 1.08–1.83). A significant increase in RR was shown only for the lower birth weight group of 501 to 750 g (RR: 1.60; 95% CI: 1.07–2.39). The incidence of nosocomial fungal sepsis was not significantly different between groups (RR: 1.27; 95% CI: 0.75–2.17).
Infants in the prophylactic ointment group had more evaluations for possible sepsis, compared with the routine skin-care group (1.46 per infant vs 1.31; P = .02). The proportion of evaluations that were positive for a bacterial (20% vs 17%; P = .18) or a fungal (3.4% vs 3.4%; P = .96) organism was similar for each group. Cerebrospinal fluid (CSF) cultures were obtained in 7% of the sepsis evaluations. Only 8 CSF samples were positive for bacterial organisms, 4 in each study group. There were 2 CSF samples positive for fungal organisms, both in the routine skin-care group.
Infants in the prophylactic ointment group were less likely to have a high skin-condition score (>3) at any time during the first 28 days. They also had less-frequently documented local skin injury during days 15 to 28 (Table 5).
Care practices during the first 28 days were similar between groups for use of radiant warmers or incubators, the duration of umbilical arterial and venous lines, the presence and duration of other central venous lines, exposure to parenteral nutrition and intravenous lipids, postnatal steroids, and nasal continuous positive airway pressure (Table 6). Infants in the prophylactic ointment group had more days of assisted ventilation (17.2 vs 16.0, P = .04), had more days of high-frequency ventilation (5.0 vs 4.0; P = .02), and required reintubation more frequently (1.3 vs 1.1; P = .01).
There were no differences between the study groups for any of the complications of prematurity analyzed (Table 7). The incidence of chronic lung disease expressed as oxygen need at 28 days and 36 weeks post menstrual age was similar between groups, as was the percent of infants receiving steroids for treatment of chronic lung disease.
Infants in the prophylactic ointment group weighed less on days 14 and 28 of life and had a lower average weight gain over the 28 days (Table 8). Length of stay did not differ between the groups.
The skin of extremely premature newborn infants is a fragile barrier to the external environment. The stratum corneum is only a few cell layers thick.17 Although the skin matures rapidly after birth, there is evidence for compromised function for ≥4 weeks.18 During this time, the infant is susceptible to excessive transepidermal water loss, skin breakdown, abrasions, and epidermal stripping from removal of tapes and adhesives.13,14,19,20 Restoration of normative function by applying exogenous topical agents is a potential strategy for improving the care of these infants. The suggestion that ointment therapy may reduce the risk of NBS led to rapid adoption of this treatment as a care standard.13,21 Our trial was designed to study whether ointment therapy was safe and effective in reducing NBS in ELBW infants.
We found no difference between ELBW infants receiving prophylactic ointment through day 14, compared with those receiving routine skin care, including as-necessary application of ointment for the primary combined outcome of death or NBS by 28 days of age. For the component outcome of death by 28 days, there was no significant difference between groups. The incidence of NBS was significantly greater in the prophylactic ointment group. In the subgroup analysis, there were significant differences between groups for all causes of NBS and for sepsis caused by coagulase-negative staphylococcus in the lower weight stratum (501–750 g). There were no significant differences in infection rates between groups for other bacterial pathogens or for fungal organisms.
Despite randomization, the prophylactic ointment group had significantly more males and higher rates of prenatal care. Male infants have a higher mortality rate and have been reported to have higher rates of nosocomial sepsis.12,22 Adjusting for gender and prenatal care as well as other potential confounding variables by multivariate analysis did not change the results obtained from the crude RR analysis.
The difference in infection rates between study groups was mostly due to a commensal organism: coagulase-negative staphylococcus. It is difficult to know whether a positive blood culture with coagulase-negative staphylococcus is a skin contaminant or a true infection, particularly with a single blood culture. A number of findings make it less likely that study differences were caused by an increased rate of blood culture contaminants in the prophylaxis group. The infants in the prophylactic ointment group were evaluated for sepsis more frequently than those in the routine skin-care group, but the proportion of sepsis evaluations that were positive was the same. The prophylactic ointment group had more reintubations, more ventilator days, and less weight gain over 28 days, all findings compatible with morbidity from a higher incidence of NBS. The similarity between groups for birth weight, gestational age, and ventilator care in the first 24 hours make it less likely that the prophylactic ointment group had a higher acuity before study intervention. The study definition of an episode of NBS required both a positive culture and a clinical decision to treat. Even if a positive blood culture were a contaminant, the episode was only analyzed as a study outcome if it was a clinically diagnosed sepsis event.
Repeated application of an emollient ointment has been shown to improve skin condition without increasing bacterial or fungal skin colonization.13,14,23 Before this study, the strongest evidence for a benefit for prophylactic ointment treatment came from Nopper et al13, who found improved skin condition to be associated with a lower incidence of nosocomial infection. In their study, the primary outcome was transepidermal water loss. Infection was a secondary outcome. The study was small, with only 20 infants of birth weight <1000 g, and the infection rate in the infants of that weight in the control group was high (8 of 11) compared with the prophylactic ointment group (1 of 9). Five of the 9 cases of infection had positive CSF cultures, with no positive blood culture in 2 cases. In our study, CSF was infrequently obtained as part of the sepsis evaluation (7% of sepsis evaluations), with only 8 positive cultures in the entire study. In both studies, topical ointment therapy was used in the control group to treat poor skin condition. Nopper et al13 used Eucerin, an emulsion containing water, petrolatum, mineral oil, ceresin, lanolin alcohol, and preservatives, compared with our use of Aquaphor, an anhydrous mixture.
Two other reported studies have suggested a potential risk for infection with ointment therapy. Campbell et al,24 in a case-control study, demonstrated a potential relationship between the topical application of petrolatum ointment and systemic candidiasis in infants weighing <1500 g. In our study, there was no significant difference in the incidence of fungal sepsis between groups. Ramsey et al25 reported contamination of Aquaphor as a potential source of nosocomial infection. In our study, Aquaphor was issued in individual tubes for each infant daily and therefore unlikely to have served as a reservoir for infection or a source of cross-contamination.
Skin-condition scores, determined by the score adapted from Lane and Drost,14 were better in the infants with prophylactic ointment application. The incidence of local skin injury for infants in the prophylactic group was lower for days 15–28. Other potential benefits of prophylactic ointment application, such as reduction of insensible water loss, were not measured directly in this study. Weights at 7 days of age were not different between groups.
We demonstrated no reduction in the risk of the combined outcome of death or NBS with generalized prophylactic application of ointment over the first 2 weeks of life in ELBW infants. However, there may be an increase in the risk of NBS associated with that practice. Other strategies for supporting epidermal barrier function during maturation in ELBW infants need to be investigated.
This work was sponsored in part by a grant from Beiersdorf Inc and General Clinical Research Center grant MO1 RR109.
- Received March 25, 2002.
- Accepted September 24, 2003.
- Reprint requests to (W.H.E.) Department of Pediatrics, Children’s Hospital at Dartmouth, One Medical Center Drive, Lebanon, NH 03756. E-mail:
- ↵Horbar JD, Kenny M, Carpenter JH, eds. Vermont Oxford Network 1999 Database Summary Burlington, VT: Vermont Oxford Network; 2000
- ↵Gray JE, Richardson DK, McCormick MC, Goldmann DA. Coagulase-negative staphylococcal bacteremia among very low birth weight infants: relation to admission illness severity, resource use, and outcome. Pediatrics. 1995;95 :225– 230
- Stoll BJ, Holman RC, Schuchat A. Decline in sepsis-associated neonatal and infant deaths in the United States, 1979 through 1994. Pediatrics. 1998;102(2) . Available at: www.pediatrics.org/cgi/content/full/102/2/e18
- Makhoul IR, Sujov P, Smolkin T, Lusky A, Reichman B. Epidemiological, clinical, and microbiological characteristics of late-onset sepsis among very low birth weight infants in Israel: a national survey. Pediatrics. 2002;109 :34– 39
- Freeman J, Platt R, Epstein MF, Smith NE, Sidebottom DG, Goldmann DA. Birth weight and length of stay as determinants of nosocomial coagulase-negative staphylococcal bacteremia in neonatal intensive care unit populations: potential for confounding. Am J Epidemiol. 1990;132 :1130– 1140
- Gaynes RP, Edwards JR, Jarvis WR, Culver DH, Tolson JS, Martone WJ. Nosocomial infections among neonates in high-risk nurseries in the United States. National Nosocomial Infections Surveillance System. Pediatrics. 1996;98 :357– 361
- ↵Fanaroff AA, Korones SB, Wright LL, et al. Incidence, presenting features, risk factors and significance of late onset septicemia in very low birth weight infants. The National Institute of Child Health and Human Development Neonatal Research Network. Pediatr Infect Dis J. 1998;17 :593– 598
- ↵Lane AT, Drost SS. Effects of repeated application of emollient cream to premature neonates’ skin. Pediatrics. 1993;92 :415– 419
- ↵Vermont Oxford Network. Vermont Oxford Network Database Manual of Operations (Release 3.3) Burlington, VT: Vermont Oxford Network; 1997
- ↵Rutter N. The immature skin. Eur J Pediatr. 1996;155(suppl 2) :S18– S20
- ↵Okah FA, Wickett RR, Pickens WL, Hoath SB. Surface electrical capacitance as a noninvasive bedside measure of epidermal barrier maturation in the newborn infant. Pediatrics. 1995;96 :688– 692
- ↵National Association of Neonatal Nurses. Guidelines for Practice: Neonatal Skin Care Petaluma, CA: National Association of Neonatal Nurses; 1997
- ↵Horbar JD, Badger GJ, Lewit EM, Rogowski J, Shiono PH. Hospital and patient characteristics associated with variation in 28-day mortality rates for very low birth weight infants. Vermont Oxford Network. Pediatrics. 1997;99 :149– 156
- ↵Campbell JR, Zaccaria E, Baker CJ. Systemic candidiasis in extremely low birth weight infants receiving topical petrolatum ointment for skin care: a case-control study. Pediatrics. 2000;105 :1041– 1045
- ↵Ramsey K, Malone S, Fey P, et al. Aquaphor as a source of colonization and subsequent bloodstream infections among very low birthweight neonates [abstract 53]. Infect Control Hosp Epidemiol. 1998;19 :689
- Copyright © 2004 by the American Academy of Pediatrics