Analysis of the efficacy of urine culture as part of sepsis evaluation in the premature infant

Abstract

Background. 

Premature infants have a higher incidence of urinary tract infection (UTI) than full term infants. UTI in premature infants can present with signs of sepsis: poor weight gain; temperature instability; metabolic acidosis; poor feeding; and abdominal distention.

Objective. 

The purpose of this study was to determine the usefulness of routine urine culture as part of a sepsis evaluation in the preterm infants.

Methods. 

We conducted a retrospective review of all infants with birth weight <1500 g (very low birth weight) who underwent sepsis evaluation at MetroHealth Medical Center between January 1991 and February 1998. All infants from whom urine and blood specimens were collected concomitantly for culture as part of a sepsis evaluation were included.

Results. 

Included were 538 infants. Their mean gestational age was 28.5 ± 2.7 weeks, and mean birth weight was 1072 ± 276 g. Blood and urine specimens for culture were taken from 349 infants on admission or in the first 24 h of life (Group A), their mean birth weight was 1147 ± 244 g, and mean gestational age was 28.9 ± 2.6 weeks. None of these infants had positive urine cultures; 8 infants (2%) had positive blood cultures. Blood and urine specimens were obtained from 189 infants later between Days 6 and 150 of life (Group B); their mean birth weight was 933 ± 278 g, and mean gestational age was 27.5 ± 2.5 weeks. Forty-eight infants (25.3%) in Group B had positive urine cultures, and 79 infants (41.7%) had positive blood cultures. Eighteen infants (38%) with positive urine cultures had positive blood cultures, and 30 infants (62%) had negative blood cultures.

Conclusions. 

There is minimal benefit in obtaining urine cultures from very low birth weight infants as part of a sepsis evaluation in the first 24 h of life. It is important to obtain urine cultures from older infants with signs of sepsis to identify patients with UTI with or without bacteremia.

INTRODUCTION

The reported incidence of urinary tract infection (UTI) in premature infants is higher than that in full term infants. It varies from 0.1 to 1% in term infants but is as high as 10% in low birth weight infants. ¹ The incidence of UTI in very low birth weight (VLBW) infants has not been established. Poor weight gain, failure to thrive, vomiting, poor feeding, abdominal distension, jaundice, temperature instability, irritability, lethargy and metabolic acidosis are common signs of urinary tract infection in neonates. ^2–5

These signs are also seen in infants with sepsis; for this reason urine samples are usually obtained in infants being investigated for sepsis.

Procedures to obtain urine samples aseptically are invasive and carry the risk of trauma and infection. Hematuria has been reported in 0.6% of 654 infants with suprapubic aspiration. ⁶ Hematoma of the anterior wall of the bladder, ^{7, 8} peritonitis and anaerobic bacteremia have also been reported after suprapubic aspiration. ⁹ Bag collection has a high reported incidence of false positive results resulting from contamination. ^{10, 11}

The aim of our study was to determine the necessity of urine culture as part of a sepsis evaluation in VLBW infants at various times after admission to the neonatal intensive care unit (NICU).

METHODS

We reviewed the medical records of all infants with birth weight <1500 g admitted to the NICU at MetroHealth Medical Center between January 1991 and February 1998. All infants were included who had a sepsis investigation including a urine culture at admission or during their stay in the NICU. The decision to include a urine culture as a part of the investigation was made by the attending neonatologist on the basis of clinical evaluation of each case. Included patients were divided into two groups: infants with blood and urine cultures obtained at admission and on the first day of life (Group A); and infants with blood and urine cultures obtained between Days 6 and 150 during their stay in the NICU (Group B).

Urine specimens were obtained by one of three methods: Sterile urethral catheterization, suprapubic aspiration or bag collection. Bag collection was done by applying U-bag to the perineum after thorough cleansing with sterile or distilled water and antiseptic skin cleaner. After the patient urinated, the bag was removed and a small area of the bag’s surface was cleaned by alcohol, then the urine was aspirated by a sterile needle and syringe through the cleaned area. Bag collection was used to obtain urine specimens in all patients in group A and 56% of patients in group B. The remainder were obtained by sterile catheterization or suprapubic aspiration.

Specimen transport from the NICU to the laboratory was within 2 h of collection. Urine samples were placed in sterile containers and stored at 4°C until processed (not longer than 24 h).

For culture and quantification a platinum loop was calibrated to deliver 0.01 ml, or 0.001 ml was used to inoculate plates containing 5% sheep blood agar and MacConkey’s agar. All plates were incubated at 35–37°C and examined at 24 to 48 h for colony counts and bacteria identification. Urine specimens with bacterial growth of >10 000 colonies/ml when obtained by sterile catheterization or >50 000 colonies/ml when obtained by bag collection were considered positive. Any bacterial growth in the urine specimens obtained by suprapubic aspiration was considered significant. Cultures with mixed organism growth were considered contaminated.

Data were analyzed with Sigma Stat software Version 2.0. t, chi square and Mann-Whitney tests were used, and P < 0.05 was considered significant.

RESULTS

Sepsis was evaluated in 1129 preterm infants with birth weight <1500 g during the study period; 538 infants (48%) were evaluable, and 591 infants (52%) were excluded because urine cultures were not done as a part of sepsis evaluation.

The mean gestational age and birth weight of included infants were 28.5 ± 2.7 weeks and 1072 ± 276 g, respectively. Blood and urine specimens were obtained from 349 infants (65%) on admission or during the first day of life as part of a sepsis evaluation (Group A). Blood and urine specimens were obtained from 189 infants (35%) later between Days 6 and 150 of age (Group B) (Table 1). Concomitant blood and urine specimens were not evaluated on any infants between admission and Day of Life 6. Of the 538 infants included, 48 infants (8.9%) had positive urine cultures, and 87 infants (16.1%) had positive blood cultures.

TABLE 1: Study population demographics

The data from both groups were analyzed separately. The ages at which the infants were cultured spanned 6 to 150 days. The median age for urine culture was 19 days. Twenty-five percent of cultures were performed before Day 11 of life, and 75% were performed before Day 32. Forty-eight infants in Group B (25.3%) had positive urine cultures; 30 of these infants (62%) had negative blood cultures. Seventy-nine infants in Group B (41.7%) had positive blood cultures, 18 (22.8%) of whom had positive urine cultures (Table 2).

TABLE 2: Results of urine and blood cultures

The most common organism isolated from urine cultures was Candida (19 patients, 39.6%). Candidaalbicans was isolated from the urine cultures of 16 patients, 6 of them had positive blood cultures for the same organism and 1 had positive blood culture for coagulase-negative Staphylococcus. Candidatropicalis was isolated from the urine cultures of 2 patients who had positive blood cultures for the same organism. Candidaparapsilosis was isolated from the urine culture of 1 patient who had a positive culture of peritoneal fluids for the same organism.

The second most common organism was Staphylococcus spp. (10 patients, 20.8%); coagulase-negative Staphylococcus was isolated from the urine cultures of 9 patients, 7 of them had positive blood cultures for the same organism and 2 patients had negative blood cultures. Staphylococcus aureus was isolated from the urine culture of 1 patient with positive blood culture for the same organism.

Pseudomonas aeruginosa was isolated from the urine cultures of four patients (8.3%); one of them had positive blood culture for the same organism. Group B Streptococcus was isolated from the urine cultures of two patients (4.1%); one of them had positive blood culture for the same organism. Klebsiella spp. were isolated from the urine cultures of two patients (4.1%); one of them had positive growth of C. albicans in the urine and blood cultures.

Other organisms were isolated from urine cultures with negative blood cultures included:Enterococcus spp. (five patients,10.4%); Escherichia coli (two patients, 4.1%); Enterobacter (two patients, 4.1%) and Citrobacter (two patients, 4.1%).

The organisms isolated from the urine cultures of patients in Group B in relation to method of collection are shown in Table 3.

TABLE 3: Isolated organisms in relation to method of collection in Group B

DISCUSSION

The incidence of positive urine culture in VLBW infants undergoing sepsis evaluation on admission to the unit and thereafter was evaluated. No infant had positive urine culture at admission or during the first day of life, but approximately one-fourth of older infants had positive urine cultures.

The incidence of urinary tract infection in VLBW infants undergoing sepsis evaluation during their stay in the neonatal intensive care unit is unknown.

No studies have addressed the issue since 1978 when survival and management of such infants was very different. Visser and Hall ¹² reviewed 377 patients (term and preterm) in the neonatal intensive care unit in 1975 and 1976. During a 12-month period, 188 sets of concomitant blood and urine cultures were obtained in infants <72 h of age (early onset group) and 189 sets of cultures were obtained in infants >72 h of age (late onset group). Significantly more positive urine cultures were reported in infants >72 h of age (7.4%) than in infants in the first 72 h of life (1.6%). That study also attempted to identify clinical features of late onset infections discriminating infants with uncomplicated UTI from those with sepsis. The single infant in whom both blood and urine cultures were positive was excluded. No differences were found with the exception that four patients with bacteremia had one or more sites of bleeding.

DiGeronimo ¹³ studied 280 term and near term infants (35 weeks of gestation) with urine and blood cultures taken during an evaluation for suspected sepsis within 24 h of birth. There was one positive urine culture (<0.5%).

Maherzi et al. ⁵ studied the prevalence of neonatal urinary tract infection in 1762 high risk neonates in 1978. Preterm, term and postterm infants were included in the study; urine cultures were performed when symptoms were suggestive of UTI or at the end of the hospital stay in asymptomatic patients. Bacteriuria was found in 2% of symptomatic patients and 0.5% of asymptomatic patients.

These reports are consistent with our study. Term and preterm infants have an extremely low incidence of UTI at birth. Thereafter all infants evaluated for sepsis have a much higher incidence of UTI. Infants with both positive and negative blood cultures had positive urine cultures.

CONCLUSIONS

The incidence of positive urine cultures in VLBW infants undergoing sepsis evaluation during the first 24 h of life is extremely low (none in our study). Thereafter, however, symptomatic infants undergoing sepsis evaluation had a significant incidence of positive urine cultures both in the presence and absence of positive blood cultures.

We thank John J. Moore, M.D., Professor of Pediatrics and Reproductive Biology at Case Western Reserve University School of Medicine, Head, Division of Neonatology, MetroHealth Medical Center, for his invaluable input and guidance.