25

BIRTH 17:l March 1990

A Randomized Controlled Trial of a Nursery Ritual: Wearing Cover Gowns to Care for

Healthy Newborns Janet Rush, R.N., M .H.Sc., Rosalina Fiorino-Chiovitti, R.N., B.Sc.N., Karyn Kaufman, R.N., Dr.P.H., and Aha Mitchell, R. N., M.Sc. ABSTRACT: The routine wearing of individual cover gowns by nurses and visitors for direct care of healthy newborns was usual practice on the maternity ward of a regional referral center. We conducted a randomized trial in which cover gowns were not provided for care of infants in the experimental group (n = 222), but were maintained f o r control infants (n = 230). The principal outcome measured was Staphylococcus aureus colonization of the newborn nares or umbilicus on day 3 or day of discharge. Twenty percent (n = 51) of the experimental group (no gown) had a positive culture compared with 21 percent (n = 47) of the controls. Of the infants with positive cultures, two in each group exhibited symptoms of overt S. aureus infection. Experimental infants were similar to controls with respect to feeding method, route of delivery, amount of time spent rooming-in, and average number af visitors per day. I n the group of positively cultured infants, the mothers experienced longer labor, and more vaginal examinations in labor, and the number of males undergoing circumcision was higher. We concluded that routine use of cover gowns was unwarranted, and we have altered the ward policy accordingly. This also has had a positive economic effect. (BIRTH 17:1, March 1990) Several traditional practices in newborn nurseries have been directed toward reducing the prevalence of bacterial colonization of newborns, and thereby the likelihood of overt infection. At Chedoke-

Janet Rush is Director of Nursing, MaternallChild Health, S t . Joseph s Hospital, and Assistant Clinical Professor at McMasrer University School of Nursing, Hamilton, Ontario, Canada. Rosalina Fiorino-Chiovitti is a Public Health Nurse, HamiltonWentworth Public Health Unit, Hamilton, Ontario, Canada. Karyn Kaufman is Associate Professor at McMaster University School of Nursing and Department of Obstetrics-Gynecology, McMaster University, Hamilton, Ontario, Canada. Alba Mbchell is Associate Professor at McMaster University School of Nursing and Associate Member, Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada. This project was supported in p a r t b y a grant f r o m the McMaster University School of Nursing. Address correspondence to Janet Rush, Department of Nursing, St. Joseph's Hospital, 5 0 Charlton Avenue, East Hamilton, Ontario, LBN 4A6, Canada.

McMaster Hospitals in Hamilton, Ontario, we previously used hand washing, alcohol cord care, daily infant bathing, and long-sleeved gowns when providing direct care as the primary means of protecting infants from pathogens. A randomized, controlled trial (1) conducted in this setting demonstrated no significant difference in the prevalence of S . aureus colonization of newborns when routine daily bathing was compared with dry skin care. Based on the results of the study, the conventional practice of daily bathing was replaced with dry skin care and optional bathing. We then examined the usefulness of cover gowns in this setting. The literature we reviewed did not support the use of cover gowns. A 1958 controlled study showed no difference in colonization rates between two nurseries, only one of which used gowns and masks (2). Williams and Oliver demonstrated a constant rate of 4 percent of infant nares colonized over 6Y2 years during which many routines, including gowning, were discontinued (3). Two studies in neonatal intensive care units (NICUs) showed no difference in rates of infections and infection-related deaths during alternate months

BIRTH 17:1 March 1990

when gowns were not used when entering the units or providing incubator care (43). A third study in an NICU using the alternating-month method observed comparable bacterial colonization rates, but noted a greater incidence of necrotizing enterocolitis during gowning periods (6). A more recent study of healthy newborns that employed a beforeand-after design showed the gowned group (n = 74) had a higher colonization rate than the no-gown group ( n = 50) (80% vs 62%) and concluded that gowns were of no value (7). The presence of S . aureus was not specifically reported. Although these studies were similar in their conclusions, they frequently lacked adequate controls and were based on small sample sizes. These methodologic problems make it difficult to support changing from the present-day practice of wearing gowns. The use of gowns is not based on evidence of benefit, however, and the practice adds expense to nursery care. We therefore planned and carried out a controlled trial to provide a better basis for either continuing or stopping the use of gowns in our hospital. Our specific purpose was to determine whether the use of individual gowns by staff and visitors influenced newborn colonization with S. aureus. Methods

The study was conducted in the 34-bed postnatal unit of a teaching hospital. A central nursery is available, but babies room-in as much as mothers desire. The average length of stay is 4.2 days, but varies from a few hours to more than 10 days. A small proportion (5%) of women leave within 48 hours on an early-discharge program. There are no restrictions on the age or number of visitors. Six to eight nurses per shift provide combined mother-infant care. Colonization rates are monitored routinely; nurses obtain nasal and umbilical swabs on the third postnatal day or prior to discharge, whichever is sooner. They follow a defined procedure for obtaining the cultures, and specimens are plated on blood agar in the microbiology department. Identification of S . aureus from either swab is a positive result. A randomized, controlled trial of the discontinuation of gowns was carried out for 32 weeks between June 3, 1986, and January 9, 1987. Consecutively admitted, eligible infants whose parent(s) gave consent to the trial were randomly allocated to either the control group, in which nurses continued to use cover gowns, or the experimental group, in which nurses did not use cover gowns for direct in-

fant care. Staff members, parents, and visitors in both groups continued to be advised to wash their hands carefully before providing infant care. The outcome measure was the prevalence of S . aweus in either the nasal or umbilical cultures. All the eligible infants were delivered in the teaching hospital and were at least 37 weeks’ gestational age, with a birthweight of 2500 g or more and an Apgar score of greater than 7 at 5 minutes. Those who were initially admitted to the NICU and then transferred to the well-baby nursery were not eligible for the trial. All the mothers were English speaking, were cared for on the postpartum ward, and were well enough to participate in infant care. The sample size calculation was based on the following considerations. The mean weekly S . aureus colonization rate for well infants in the 12 months preceding the trial was 0.22 (SD 0.08). The infection-control department advised that they would consider an average prevalence of 0.33 or more as a clinically important level because of the increased risk of overt infection. We therefore wished to detect an increase of 0.10 or more, using a one-tailed alpha of 0.05 and beta of 0.2 as statistically significant; 230 infants per group were required to satisfy these conditions. We planned to enroll the sample over four to six months to have an adequate time period to assess any sustained rise in colonization. All infants admitted to the nursery were assessed for eligibility, and parent consent was sought when an investigator was available. Eligible infants whose parents provided consent were randomly allocated to a group by opening a consecutively numbered, sealed envelope containing a folded card with the group assignment. The order of assignment was based on a computer-generated random table of numbers with randomly ordered blocking factors of 4 and 6. A large sign was placed on the cribs to indicate each baby’s group allocation. Staff nurses obtained swabs of the nares and umbilical cord for culture following the defined protocol. Observations were made by student-nurse research assistants prior to and during the study to evaluate adherence to the protocol. Data were also collected about the infant’s sex, gestational age, weight, time between rupture of membranes and birth, length of labor, fetal distress in labor, use of an internal monitor, number of vaginal examinations in labor, method of delivery, length of stay, circumcision, rooming-in time, and number of visitors per day. Mothers maintained diaries of infants’ rooming-in time and the number of visitors. These were used to calculate the proportion of the total stay during which babies roomed-in. Mothers described and reinforced either the gown or no-gown

27

BIRTH 17:l March 1990

procedure for their visitors. Student research assistants made daily rounds to monitor mothers’ participation with and adherence to the assigned use of gowns. The results of cultures were known to the infection-control officer. She agreed to inform the principal investigator if the nursery colonization rate for any week approached 40 percent in order to implement special measures. Laboratory staff who read the cultures were blind to the group assignment of babies.

A11 consecutive newborn admissions: June 1986-January 1987 1,130 ~~

~

Eligible mother-baby pairs \

\ \

995

Results

During the study period, 995 (88.1%) of the 1130 infants admitted to the newborn nursery met the study criteria. The investigators were available to request participation from parents of 541 infants (54.4% of those eligible), and consent was secured for 489. Descriptive data were collected for 234 and 239 in the experimental and control groups, respectively, and colonization results were available for 222 and 230 infants, respectively. Missing data resulted when research assistants were unavailable to abstract chart information, or culture reports were missing (Fig. 1). The randomization procedure resulted in equivalent groups, as shown in Table 1 . No important differences were seen on any of the variables that were considered to be potential confounders of newborn colonization rates. A large proportion of infants in both groups were born vaginally and were breastfed. There was a similar distribution of females and males in each group, and nearly one-half of the males in each group was circumcised. The use of internal fetal monitors, time between membrane rupture and birth, number of vaginal examinations in labor, and overall length of labor were nearly identical in the groups. The infants had similar mean gestational ages, birthweights, and length of stay in hospital. Approximately 80 percent of mothers in each study group recorded information in their diaries about amount of time rooming-in and number of visitors. Based on the available information, the figures were similar for the two groups. The procedure for obtaining nasal and umbilical cultures was observed covertly on 20 occasions by trained students who used a written checklist to audit the nurses’ actions. In no instance were errors detected in the nurses’ performance of the procedure. The: mean colonization prevalence for the total newborn population over the study period was 0.26

16

R

Experimental

available

-f

r-l I I Cultures available 230

Figure 1 . Derivation of study sample.

(SD 0.07) compared with 0.22 in the prestudy year. Considerable week-to-week variation was noted in the prevalence of S.aweus in both groups, with the control group varying from 0.0 to 0.66 and the experimental group from 0.0 to 0.57. There was no significant difference in the means of the weekly rates: 0.21 (SD 0.16) for the control group versus 0.22 (SD 0.13) for the experimental group. A similar and large number of the positive cultures in each group was accounted for by a positive umbilical swab (Table 2). Four infants (two in each group) developed overt S . aureus infections, consisting of two infected circumcision sites, one conjunctivitis, and one skin lesion. Since gowning practice did not obviously exert an influence on colonization, we examined the data from infants with complete colonization reports to determine whether other study variables were significantly associated with newborn colonization by S. aureus. As shown in Table 3, being colonized

I

28

BIRTH 17:l March 1990

Table 1. Characteristics of Experimental and Control Groups Experimental

(no gown

n

Variables

Sex M F Breastfeeding Delivery method Vaginal Cesarean Internal fetal monitor Circumcision* Weight (8) Gestational age (wks) Length of stay (hrs) Length of labor (hrs) Time from membrane to birth (hrs) Rooming-in (% of stay)t Visitorsldayt Vaginal examinations in labor

f%)

=

Control

234)

Mean

(gown = 239)

(SD)

n

f%)

124 110 197

118 (49) 121 (51) 20 1 (85)

188 46 39 45

187 52 32 51 (434.0) (2.4) (33.0) (9.5) 6.3 48.7 6.9 3.0

(10.6) (20.5) (6.8) (2.1)

Mean

(SD)

P Value

0.49 0.95

(78) (22) (13) (43) 3490 (471.0) 39.7 (2.5) 103.3 (36.0) 9.8 (9.2) 6.4 46.4 6.8 3.2

(9.0) (20.2) (6.4) (2.1)

0.65 0.58 0.34 0.21 0.72 0.91 0.43 0.95 0.29 0.87 0.37

* Percentage of male infants. t Missing data for 48 experimental and 47 control mothers. was statistically significantly related to the number of vaginal examinations in labor, length of labor, vaginal delivery, and circumcision. No association was demonstrated with infant gender or feeding method, intrapartum use of internal monitoring, postpartum rooming-in time, visitors per day, or length of stay. Discussion

The results of this study do not support the routine use of gowns by staff or visitors as an infectioncontrol procedure when providing direct infant care. The prevalence rate of S. aureus was not importantly changed from the prestudy rate, and no statistically significant differences were found between the two groups. We have now changed the policy on the postnatal unit, with staff members using gowns only as desired to protect their clothing. Impervious gowns are worn for the initial newborn bath as per universal precautions.

Infants in this study who were colonized with S. aureus were delivered vaginally significantly more often than by cesarean section but did not differ from uncolonized infants in the proportion of time spent in the central nursery. These findings are inconsistent with those from an earlier trial of infant daily bathing, which showed that colonization was increased in infants having cesarean birth and those who spent more time in a central nursery (1,s). Perhaps greater compliance with hand-washing techniques has reduced the risks of becoming colonized while in the central nursery. The lack of association between colonization and rooming-in and the number of visitors is important because some professionals remain concerned about the hazards of visitors in postpartum rooming-in units. This study supports the view that ungowned, healthy visitors who are instructed about hand washing are not a hazard to newborns. The significant associations of a longer labor, number of vaginal examinations, and vaginal de-

Table 2. Colonization and Overt Infection of Newborns bs Studs GrouD

Positive nose culture Positive umbilical culture Positive nose or umbilical culture Overt S . aureus infection

Experimental

Control

(no gown = 222) n f%)

(gown = 230) n f%)

P Value

7 (3.2) 51 (23.0) 53 (23.9) 2 (0.9)

7 (3.0) 47 (20.4) 50 (21.7) 2 (0.9)

1.oo 0.59 0.58 0.85

29

BIRTH 17:l March 1990

Table 3. Comparison of Infants With and Without S. aureus ~

~~

Colonization

Variables

Vaginal examination in labor Length of labor (hrs) Internal monitor used Delivery method, vaginal Feeding method, breast Infant gender, male Circumcisiont Length of stay (days) 0-4 >4 Rooming-in (% of hospital stay) 0-50 >50

Visitors (meadday)$ 0-8 >8

n

Positive (N = 103) (%) Mean (SD)

3.5 11.1 17 90 90 57 31

n

Negative (N = 350) (%) Mean

3.0 8.9

(1.8) (9.6)

(16.8) (87.4) (87.4) (55.3) (54.4)

51 265 290 177 62

(SD)

P Value

(2.2) (8.8)

0.03* 0.03* 0.55 0.02* 0.50 0.50 0.01*

(14.7) (75.7) (84.1) (50.6) (34.8)

43 (41.7) 60 (58.2)

151 (43.1) 199 (56.8)

0.79

50 (63.3) 29 (36.7)

174 (62.8) 103 (37.2)

1 .oo

69 (85.2) 12 (14.8)

223 (80.2) 52 (19.8)

0.37

* Statistically significant difference. 't Percentage of male infants (57 in the colonized groups, 178 in the uncolonized). $ Missing data result in reduction of total number.

livery with increased colonization reflect a continuing need to evaluate intrapartum practices in relation to newborn outcomes. The significant association of circumcision and colonization is not surprising, because the infants receive more direct handling during the procedure itself and the subsequent dressing changes. This study was not designed to detect differences in overt infection rates between the experimental and control groups. We used S. aureus colonization as a helpful marker of environmental pathogens. The association between colonization and overt infection appears to be highly variable (9,101. The use and costs of gowns per month decreased during the study period. This downward trend has continued despite a stable census in the year after the study. The projected annual cost saving compared with the prestudy period of an official ward policy of discontinuing gowns is nearly $8000. The nursery nurse historically has been depicted as gowned or specially dressed. The 1987 Canadian guidelines for hospital maternal and newborn care had a cover illustration of a gowned family holding the baby (11). Hospital staff and the public are assumed to be sources of pathogenic bacteria to newborns: rules to control the access and dress of care providers, visitors, and family members persist despite other changes brought about by family-centered care. It is not surprising that, in a recent survey, the majority of physicians still support the

practice of using gowns (12). Our study challenges the assumptions behind this nursery ritual. We think it is important to continue to question and evaluate such practices to establish a scientific basis for the actions of health practitioners and thereby improve the care of healthy newborns. Acknowledgments

The authors acknowledge the assistance of the nursing staff, Ward 4A, McMaster Division, Chedoke-McMaster Hospitals, and the student-nurse research assistants, McMaster University School of Nursing. References 1 . Rush J. Does routine newborn bathing reduce staphylococcal colonization rates? A randomized controlled trial. Birth 1986;13(3):176-179. 2. Forfar JO, MacCabe AF. Masking and gowning in nurseries for the newborn infant. Br Med J 1958;1:76-79. 3. Williams CPS, Oliver TK. Nursery routines and staphylococcal colonization of the newborn. Part I. Pediatrics 1969;44(5):640-646. 4. Silverman WA, Sinclair JC. Evaluation of precautions before entering a neonatal unit. Pediatrics 1967;40:900-901. 5 . Evans HE, Akpata SO, Baki A. Bacteriologic and clinical evaluation of gowning in a premature nursery. J Pediatr 1971;8(5):883-886. 6. Agbaayani M, Rosenfeld W, Evans H , Salazar D, Jhaveri R, Braun J. Evaluation of modified gowning procedures in a neonatal intensive care unit. Am J Dis Child 1981;135: 650-652.

BIRTH 17:l March 1990 7. Renaud MT. Effects of discontinuing cover gowns in a postpartal ward upon cord colonization of the newborn. JOGN Nurs 1983;12:399-40 1. 8. Rush J, Hampel V, Dotzert L. Rooming-in and visiting in the maternity ward: Effects on colonization rates. Infecr Control Can 1987;8(1):30-33. 9. Gooch JJ, Britt EM. Staphylococcus aureus colonization and infection in normal newborn nursery patients. Am J Dis Child 1978;132:893-896.

10. Wang EL, Elder D, Mishkel N. Staphylococcal colonization and infection after discharge from a term newborn nursery. Infect Control 1987;8(suppl 1):30-33. 11. Ministry of National Health and Welfare. Family-Centered Maternity and Newborn Care: National Guidelines. Ottawa, Canada: Author, 1987. 12. Cloney DL, Donowitz LG. Overgown use for infection control in nurseries and neonatal intensive care units. Am J Dis Child 1986;140:680-683.