PEDIATRICS Vol. 107 No. 1 January 2001, pp. 61-66

Review of 49 Neonates With Acquired Fungal Sepsis: Further Characterization

Imad R. Makhoul, MD, DSc^*, Imad Kassis, MD, Tatiana Smolkin, MD^*, Ada Tamir, PhD^§, and Polo Sujov, MD^*

From the Departments of ^* Neonatology,  Infectious Diseases, and ^§ Community Medicine and Epidemiology, Rambam Medical Center, Rappaport Faculty of Medicine, Technion, Israel Institute of Technology, Haifa, Israel.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Background.  Neonatal acquired fungal sepsis (AFS) is a risky condition that warrants every effort for early diagnosis and management.

Methods.  We retrospectively reviewed the medical charts of all 4445 neonatal intensive care unit (NICU) admissions in the past 10 years and detected 49 neonates with AFS. We then compared their data with those of 49 matched control neonates who did not have AFS. The following details were collected: gestational, perinatal and neonatal courses; bacterial sepsis; antibacterial therapy; laboratory and imaging investigations; and antifungal therapy and its complications.

Results.  The incidence of AFS was .4 to 2 cases per 1000 live-births and 3.8% to 12.9% of very low birth weight (VLBW) infants. Compared with 1989 through 1992, between 1993 and 1995 the rate of AFS in VLBW neonates significantly increased (3.8%-5.6%  9.6%-12.9%), along with a significant increase of NICU admission rate (369-410  496-510 admissions/year). Compared with controls, AFS neonates had significantly longer hospitalizations, higher rates of mechanical ventilation, umbilical vein catheterization, and previous treatment with broad-spectrum antibacterial agents (amikacin, vancomycin, ceftazidime, or imipenem). At the onset of AFS, 42.8% of patients had hyperthermia and 40.9% had normal white blood cell count. Causative fungi were as follows: Candida albicans42.8% of cases, Candida parapsilosis26.5%, and Candida tropicalis20.4%. Fungal dissemination was rare, complications of antifungal therapy were infrequent, and no deaths occurred.

Conclusions.  First, non-albicans Candida have become more frequent in neonatal AFS. Second, mechanical ventilation and antibacterial agents are significant risk factors for AFS. Third, hyperthermia is a frequent presenting sign of AFS. Fourth, a normal white blood cell count does not rule out AFS. Fifth, meningeal involvement in neonatal AFS should be ruled out before initiation of antifungal therapy. Sixth, the policy of empiric antifungal therapy for AFS should be considered on an individual NICU basis.newborn infant, fungal sepsis, clinical signs, risk factors.

Asignificant increase in the survival of very low birth weight (VLBW) infants <1500 g has been achieved over the last 2 decades. However, there has also been a concomitant increase in the rate of acquired fungal sepsis (AFS) of up to 2.6% to 10% in these tiny premature neonates.^1-3 In the 1980s, Candida albicans was the leading causative organism in AFS,^4,5 but the rate of non-albicans Candida has recently increased.^5-7

AFS is a serious complication of modern NICU, which can cause severe morbidity, mandates extensive laboratory and imaging workup for ruling out fungal dissemination, and requires prolonged antifungal therapy and hospitalization, with close follow-up and management of complications of both the AFS as well as the antifungal therapy.

Known risk factors for AFS include prematurity, especially in VLBW neonates,^1,2 the use of broad-spectrum antibacterial medications,^8-10 prolonged mechanical ventilation, catheterization of central¹¹ and peripheral^10,12 vessels, prolonged parenteral nutrition,^13,14 prolonged hospitalization in a neonatal intensive care unit (NICU),^15,16 and glucocorticosteroid therapy.¹⁷

The aims of this retrospective study were: 1) to summarize the clinical and laboratory data of all neonates with AFS in our NICU in the last 10 years; 2) to compare the clinical characteristics of neonates with AFS with matched controls; and 3) to underline several meaningful clinical and laboratory trends as were observed during the study. According to the results of this study, we expected to discern characteristics that might assist in the elaboration of policies for early diagnosis and prevention of this troubling neonatal complication.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

Data Collection and Statistical Analysis

This retrospective survey includes all neonates with AFS who were detected during a 10-year period between January 1, 1989 and December 31, 1998, by reviewing the medical records of all NICU admissions. The study group consisted of neonates with AFS. AFS was defined as follows: 1) clinical signs of neonatal sepsis after 7 days of life and 2) simultaneous isolation of fungi (but no bacteria) in one or more of the following sites: bloodstream, urine or cerebrospinal fluid (CSF). Blood cultures were all drawn from femoral vein puncture. Urine cultures were all drawn via percutaneous suprapubic aspiration. Care was taken to avoid obtaining blood or urine cultures through an area with active mucoepithelial candidiasis. Each patient in the study group was matched to the subsequent neonate born at the same gestational age and hospitalized in the NICU but who did not have AFS (control neonate).

From every neonate admitted to the NICU in our institution, we routinely obtain cultures from external ear, eyes, nose, gastric aspirate, and bloodstream. However, this local protocol does not affect our decision regarding the initiation of antimicrobial therapy after birth. This is rather based on perinatal risk factors for early neonatal sepsis, gestational age, severity of disease, and the extent of endotracheal and vascular cannulation required.

The following data were collected from the medical records of each neonate, in both groups: demographic data, gestational and perinatal course (specifically those associated with maternal and fetal infection), initial neonatal course including septic episodes, interventions, complications of prematurity and hospitalization in the NICU, mode and duration of nutrition (enteral and parenteral), late-onset bacterial neonatal sepsis and complications, antibacterial therapy, laboratory examinations, and outcome (Tables 1 and 2).

In the study group, data related to AFS were also collected, including clinical presentation (temperature instability, abdominal distention, lethargy, gastric residuals, and poor peripheral perfusion), risk factors, laboratory and imaging investigations, identification of the causative organism, and antifungal therapy and its complications (Tables 3, 4, and 5).

For comparison of the results of the study and control groups, we used the Student's t test for continuous variables (ie, duration of antibacterial therapy), the Wilcoxon test for ordinal variables (ie, Apgar score), and the ² test for dichotomic variables. For testing the trend in the yearly distribution of AFS rate during the study, we used the ² for trend test. For comparing the monthly distribution of AFS rate during the study with the expected distribution, we used the ² for goodness-of-fit test. A logistic regression model was used to identify independent risk factors for AFS. A P value of <.05 was considered statistically significant.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

Epidemiology of AFS

During the 10-year study, there were 46 705 live-births, 4445 NICU admissions, and 515 VLBW neonates. Review of the medical records of all 4445 NICU admissions indicated a total of 52 patients with fungal sepsis. Three of these 52 patients had congenital fungal sepsis and were excluded from the study. The study group consisted of the 49 neonates with acquired fungal sepsis. Two of these 49 were outborn patients but were not VLBW infants. Between 2 and 10 cases of AFS were detected annually. Of the 49 AFS cases detected, 35 were VLBW infants (71.4%). The overall incidence of AFS was 1.05 cases per 1000 live-births (annual range: .4-2); or 1.1% of all NICU admissions (range: .53%-2%); or 6.8% of all VLBW infants (range: 3.8%-12.9%).

Compared with the period 1989 through 1992, the rate of NICU admissions increased significantly between 1993 and 1995 (369-410  496-510 admissions/year; P < .0001), specifically the non-VLBW infants. There was no significant change in the annual rate of VLBW infants born. We tested for annual and monthly trends in the rate of AFS during the entire 10-year study in 3 subpopulations (live-births, NICU admissions, or VLBW infants); there was no significant linear annual trend in the rate of AFS in any of the 3 subpopulations tested. However, the rate of AFS in VLBW infants was significantly higher between 1993 and 1995 (9.6%-12.9%), compared with 3.8% to 5.6% between 1989 and 1992 or with 3.8% to 6.8% between 1996 and 1998 (P < .045). No similar significant changes were observed in the 2 other subpopulations.

Comparison of the Study and Control Groups

Prenatal and Labor Characteristics The neonates in the study and control groups were not significantly different as to gender, rates of rupture of membranes, chorioamnionitis, maternal therapy with glucocorticosteroid and anti-bacterial agents, fetal heart rate abnormalities during labor, mode of delivery, need for resuscitation at birth, Apgar score, and birth weight and its appropriateness to gestational age. Thus, the matching process of control neonates to neonates with AFS was appropriate and allowed a reliable comparison of the study and control groups.

Neonatal Course Before the Onset of AFS The study and control groups were not significantly different as to the rates of respiratory distress syndrome, pneumonia, umbilical artery catheterization, peripheral venous cannulation, parenteral nutrition (peripheral or central), postnatal steroid use, orogastric enteral nutrition, and mortality (no death in either group). Compared with the control group, significantly more patients of the study were mechanically ventilated (P < .002) and more underwent umbilical vein catheterization (UVC; P < .06; Table 1). However, there was no significant difference between the groups regarding the duration of mechanical ventilation, umbilical artery catheterization, UVC, or central parenteral nutrition before onset of AFS. The duration of hospitalization (P < .007) and the period of peripheral parenteral nutrition (peripheral venous cannulation; P < .001) were significantly longer in the study group, compared with the control group (Table 1), probably because of the extra hospitalization time needed to complete the antifungal therapy. In a logistic regression model, only mechanical ventilation and bacterial growth in external ear at birth were significant independent risk factors for AFS (Table 1).

Bacteriologic and Hematologic Findings The bacterial growth rates at birth (eyes, nose, gastric aspirate, and bloodstream) were higher, albeit not statistically significant, in the study, compared with the control group: 8.5% versus 6.4%; 2.2% versus 0%; 17% versus 8.5%; and 6.3 versus 4.2%, respectively. However, a significantly higher bacterial colonization rate was observed in the ears of neonates of the study group (P < .03). Bacterial growth from endotracheal tubes, umbilical artery catheter tips, and bloodstream samples during the first month of life was not significantly different between the groups: 18.6% versus 22%; 5% versus 0%; and 25% versus 19.5%, respectively. However, there was a higher percentage of bacterial colonization of the UVC tips in the study group (P < .045; Table 1).

Antibacterial Treatment Compared with the control group, significantly more neonates in the study group received amikacin, vancomycin, ceftazidime, or imipenem. More patients in the study group received ampicillin, gentamycin, ceftriaxone, clindamycin, or cefotaxime, albeit not significantly higher than did those in the control group. However, the mean total number of days on each antibacterial agent (in those of who received antibacterial therapy) was not significantly different between the groups (Table 2).

Characteristics of Patients With Acquired Fungal Sepsis

Clinical and Laboratory Characteristics As shown in Table 3, the clinical onset of AFS was at 23.2 ± 19.5 days. Hyperthermia (>38°C) as a presenting sign of AFS was observed in 42.8% of patients, third most common after abdominal distention and poor peripheral perfusion. There were no deaths in the AFS group. As shown in Table 3, in 89.8% of the patients (44/49) fungi grew in the bloodstream. In the remaining 14.3% of the infants (7/49), the bloodstream cultures did not grow fungi or bacteria, but the patients were included in this study because their urine cultures grew fungi and they featured clinical signs of sepsis. In 2 patients, fungi were isolated from the bloodstream and from urine. C albicans was the leading causative organism of AFS and was isolated in 42.8% of the patients. Candida parapsilosis and Candida tropicalis were the causative organisms in 26.5% and in 20.4% of the AFS patients, respectively.

Retinal and Ultrasonographic (US) Findings All patients with AFS underwent ophthamological examinations, and none had retinal changes of AFS (Table 4). The majority of the patients had US examinations of the brain, heart, and abdomen. There was no US evidence of fungal dissemination to the brain. Fungal changes (ball or vegetation) in the kidneys, liver, and heart were found in 7.1%, 2.5%, and 6.6% of the patients, respectively (Table 4).

Antifungal Therapy All AFS patients were treated with amphotericin B monotherapy. Flucytosine and fluconazole were not given. The duration and dosage of amphotericin B are shown in Table 5. Complications that were noticed after >7 days of amphotericin B treatment included hypokalemia, impairment of renal function, leukopenia, thrombocytopenia, hypoalbuminemia, and anemia. No notable change was observed in the serum levels of bilirubin and liver enzymes.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

The results of this retrospective review show that significant risk factors for the development of neonatal AFS include mechanical ventilation before the onset of AFS and the use of broad-spectrum antibacterial agents (amikacin, vancomycin, ceftazidime, or imipenem). Prolonged broad-spectrum antibacterial therapy is a well-known risk factor for AFS in all age groups. Evidence is accumulating regarding increased rates of fungal colonization and of AFS in humans after treatment with imipenem,^19-23 vancomycin,^20,22 or ceftazidime,²¹ specifically in immunocompromised patients. Our results in sick premature infants with AFS conform with these data.

Compared with the controls, neonates in the AFS group were sicker before the onset of infection, as evidenced by a greater frequency of acute respiratory disease, parenteral nutrition, mechanical ventilation, and UVC placement. Furthermore, the higher rates of bacterial colonization in the AFS group of UVC tips (P < .045) and of bloodstream, both at birth and later on (albeit not significant), attest to the complicated neonatal course of these infants and their significantly greater need for broad-spectrum antibacterial agents. Although neonates of the AFS group required longer periods of peripheral vein cannulation, total parenteral nutrition, and hospitalization, these 2 parameters could be attributed to the extra time needed for antifungal therapy.

The relevance of the higher bacterial ear colonization at birth to AFS is not clear. This finding did not affect our decision to initiate antibacterial therapy after birth. It may be speculated that the higher ear bacterial colonization rate in the AFS group might indicate a greater vulnerability for colonization and late infections in this group.

The incidence of AFS in our neonatal subpopulations is comparable with that reported by others: 1.05 per 1000 live-births versus .74,²⁴ 1.1% of our NICU admissions versus 2.85%⁵ and 3.2%,¹⁰ and 3.8% to 12.9% of our VLBW infants versus 2.6% to 10%.^1-3 During 1993-1995, there was a significant increase in the number of NICU admissions and a concomitant significant rise in the annual rate of AFS in VLBW infants, without a significant change in the total number of VLBW births. The increase in the number of admissions was caused by a change in the admission policy to the NICU beginning in 1993, where outpatient neonates referred to the emergency department (age: 3-28 days, mostly non-VLBW infants) were admitted to the intermediate section of the NICU. This policy was instituted without a concomitant increase in the number of NICU beds or in the size of the nursing and medical staff. We speculate that the increase in the AFS rate that occurred during 1993-1995 might have been associated with overcrowding in the NICU, along with a decrease in the nurse/patient ratio.

Most of the clinical characteristics of the neonates with AFS in the current study conform to those already published. However, some observations and findings deserve further consideration, such as the incidence of hyperthermia, abnormal findings in CSF, WBC abnormalities, and the spectrum of causative Candida strains in neonates with AFS.

Hyperthermia as a presenting sign of neonatal bacterial sepsis is less frequent in premature than in term neonates,²⁵ ranging from 1% in premature neonates to 12% in term neonates.²⁶ In our work, hyperthermia was a major presenting sign of AFS, occurring in 42.8% of our patients, compared with 53% as reported by Wang et al²⁷ in neonates with AFS. This incidence seems higher than that observed with late-onset bacterial neonatal sepsis managed in the NICU.²⁶ Only 59.1% of AFS patients had abnormalities of WBC and differential suggestive of sepsis (leukopenia, leukocytosis, or increased I/T ratio), a lower rate than that observed in neonatal bacterial sepsis.^28,29 This implies that a normal WBC in neonates does not rule out evolving AFS.

Diagnosis of meningeal involvement by fungi is important because it mandates a modified antifungal therapeutic regimen. Seven of 21 AFS patients who underwent lumbar puncture had one or more cytologic or biochemical CSF abnormality. Those patients did not have any alternative explanation for these inflammatory CSF changes, such as IVH or previous bacterial meningitis. These findings might suggest that some of our AFS patients probably had intracranial fungal involvement, despite negative CSF cultures for fungi. A few of our AFS patients underwent lumbar puncture only after initiation of antifungal therapy; therefore, their high CSF protein concentration and pleocytosis but negative culture could have reflected partially treated fungal meningitis. In the others, failure to isolate a fungus from the CSF might have been attributable to the inherent difficulties of fungal isolation from a small CSF sample,³⁰ a relatively lower inoculum size of fungi in the CSF than in the bloodstream, or improper application of special isolation techniques for CSF. CSF should be obtained for fungal culture before initiating antifungal therapy, even if the decision to start such therapy is empirical.

The spectrum of the causative Candida strains observed in our study conforms with the recently reported trend of increasing rates of non-albicans Candida as the causative organism in AFS in all age groups. In the 1980s, C albicans was the leading causative fungi in neonatal AFS, accounting for 80% to 90% of cases.^4,5 In the late 1990s this rate has decreased, and 3 recently published studies have shown that C albicans was isolated in 40% to 70% of neonatal AFS cases.^5-7 A similar trend was seen in our survey, where C albicans was the causative organism in only 42.3% of the AFS episodes.

The dissemination rate of AFS in our patients, as evidenced by clinical signs, microbiologic findings, and US examinations, was very low (2.5%-7.1%), compared with previous reports, where meningitis and endophthalmitis occurred in 54.5% and 45.5%, respectively.^31,32 In addition, there were no deaths in our AFS patients, compared with significantly higher mortality rates of 17% to 63% reported by others in neonatal AFS.^{5,10,27,32,33} We think that the more favorable outcome in our patients might be attributed to the following factors: 1) a high index of clinical suspicion; 2) reduced time from obtaining culture until recovery noted; and 3) reduced time from sampling until antifungal therapy initiated (empiric antifungal therapy).

Our policy of empiric therapy in neonates with risk factors for AFS, ie, initiating antifungal therapy immediately after obtaining cultures, followed the outbreak of AFS in our NICU in the late 1980s.³⁴ The reduction in mortality that was achieved had its own price, ie, overtreatment with empiric amphotericin B in some neonates suspected of having AFS, who did not eventually prove to have AFS. When this remarkable reduction in mortality, achieved by our policy of empirical therapy, is weighed against the risk of increased complications of amphotericin B, our therapeutic approach proved to be worthy.

Complications of antifungal therapy that occurred after >7 days of amphotericin B treatment were lower than those reported in 2 previous studies^27,33 but conform with those recently reported by Kingo et al.³⁵ Hypokalemia and impairment of renal function were noted in 7% and 9.3% in our study versus 54% and 23% in the study by Wang et al,²⁷ respectively. In addition, 37% of our patients had hypoalbuminemia and 57% had anemia, which might be attributable to prematurity.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

We conclude that non-albicans Candida strains have recently become more frequent as the causative fungus in neonatal AFS. Mechanical ventilation, UVC, and treatment with broad-spectrum antibacterial agents are significant risk factors for neonatal AFS. In addition, hyperthermia is a frequent presenting clinical sign of AFS, and the WBC is often normal, which does not rule out AFS but might cause a delayed diagnosis and treatment. The incidence and extent of meningeal involvement in neonatal AFS might be underestimated, which mandates an examination of CSF for fungi before initiation of antifungal therapy.

	FOOTNOTES

Received for publication Nov 19, 1999; accepted Apr 21, 2000.

Reprint requests to (I.R.M.) Department of Neonatology, Rambam Medical Center, Bat-Galim, Haifa 31096, Israel. E-mail: makhoul{at}rambam.health.gov.il

	ABBREVIATIONS

VLBW, very low birth weight; AFS, acquired fungal sepsis; NICU, neonatal intensive care unit; CSF, cerebrospinal fluid; UVC, umbilical vein catheterization; WBC, white blood cell count; I/T, immature to total neutrophils; US, ultrasonographic.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Baley JE, Kliegman RM, Fanaroff AA Disseminated fungal infections in very low-birth-weight infants: clinical manifestations and epidemiology. Pediatrics 1984; 73:144 [Abstract/Free Full Text]
2. Faix RG, Kovarik SM, Shaw TR, Mucocutaneous and invasive candidiasis among very low birth weight (<1500 grams) infants in intensive care nurseries: a prospective study. Pediatrics 1984; 83:101 [Abstract/Free Full Text]
3. Huang YC, Li CC, Lin TY, Lien RI, Chou YH, Wu JL, Hsueh C Association of fungal colonization and invasive disease in very low birth weight infants. Pediatr Infect Dis J 1998; 17:819-822 [CrossRef][Medline]
4. Wees-Mayer DE, Fondriest DW, Risk factors associated with candidemia in the neonatal intensive care unit: a case control study. Pediatr Infect Dis 1987; 6:190 [Medline]
5. Kossoff EH, Buescher ES, Karlowicz MG Candidemia in a neonatal intensive care unit: trends during fifteen years and clinical features of 111 cases. Pediatr Infect Dis J 1998; 17:504-508 [CrossRef][Medline]
6. Narang A, Agrawal P, Chakraborti A, Kumar P Fluconazole in the management of neonatal systemic candidiasis. Ind Pediatr 1996; 33:823-826
7. Scarcella A, Pasquariello MB, Giugliano B, Vendemmia M, de Lucia A Liposomal amphotericin B treatment for neonatal fungal sepsis. Pediatr Infect Dis J 1998; 17:146-148 [CrossRef][Medline]
8. Seelig MS The role of antibiotics in the pathogenesis of Candida infections. Am J Med 1966; 40:887 [CrossRef][Medline]
9. Mullett MD, Cook EF, Gallagher R Nosocomial sepsis in the neonatal intensive care unit. J Perinatol 1998; 18:112-115 [Medline]
10. Narang A, Agrawal PB, Chakrabarti A, Kumar P Epidemiology of systemic candidiasis in a tertiary care neonatal unit. J Trop Pediatr 1998; 44:104-108 [Abstract/Free Full Text]
11. Dato M, Dahani A Candidemia in children with central venous catheters: role of catheter removal and amphotericin B therapy. Pediatr Infect Dis J 1990; 9:309 [Medline]
12. Leibovitz E, Iuster-Reicher A, Amitai M, Mogilner B Systemic candidal infections associated with use of peripheral venous catheters in neonates: 9-year experience. Clin Infect Dis 1992; 14:485-491 [Medline]
13. Noble HB, Lyne ED Candida osteomyelitis and arthritis from hyperalimentary therapy. J Bone Joint Surg Am 1974; 56:825 [Free Full Text]
14. Mercer HP, Gupta JM Candida meningitis causing aqueductal stenosis following parenteral nutrition in an infant with meconium peritonitis. Aust Paediatr J 1978; 14:286 [Medline]
15. Solomon SL, Alexander H, Eley JW, Nosocomial fungemia in neonates associated with intravascular pressure-monitoring devices. Pediatr Infect Dis J 1986; 5:680
16. Mazumdar PK, Marks MI Candida albicans infections in hospitalized children: a survey of predisposing factors. Clin Pediatr 1975; 14:123
17. Frencel JK Role of corticosteroids on the kinetics of mononuclear phagocytes. J Exp Med 1970; 131:429 [Abstract]
18. Starff LD, Platt LH, McCracken GH Jr Cerebrospinal fluid evaluation in neonates: comparison of high-risk infants with and without meningitis. J Pediatr 1976; 88:474
19. Park SY, Parker RH Review of imipenem. Infect Control 1986; 7:333-337 [Medline]
20. Jarvis WR Epidemiology of nosocomial fungal infections, with emphasis on Candida species. Clin Infect Dis 1995; 20:1526-1530 [Medline]
21. Samonis G, Gikas A, Anaissie EJ, Vrenzos G, Maraki S, Tselentis Y, Bodey GP Prospective evaluation of effects of broad-spectrum antibiotics on gastrointestinal yeast colonization of humans. Antimicrob Agents Chemother 1993; 37:51-53 [Abstract/Free Full Text]
22. Richet HM, Andremont A, Tancrede C, Pico JL, Jarvis WR Risk factors for candidemia in patients with acute lymphocytic leukemia. Rev Infect Dis 1991; 13:211-215 [Medline]
23. Welcon CJ, Long SS, Gilligan PH Effect of imipenem-cilastatin therapy on fecal flora. Antimicrob Agents Chemother. 1986; 29:741-743 [Abstract/Free Full Text]
24. Leibovitz E, Flidel-Rimon O, Juster-Reicher A, Sepsis at a neonatal intensive care unit: a four-year retrospective study (1989-1992). Isr J Med Sci 1997; 33:734-738 [Medline]
25. Mancilla-Ramirez J, Sanches-Saucedo LU Neonatal septicemia: differences in full-term and pre-term newborn infants. Bol Med Hosp Infant Mex 1990; 47:227-233 [Medline]
26. Klein JO, Marcy SM. Bacterial sepsis and meningitis. In: Remington JS, Klein JO, eds. Infectious Diseases of the Fetus and Newborn Infant. 4th ed. Philadelphia, PA: WB Saunders Company; 1995:858
27. Wang LW, Lin CH, Liu CC, Lin YJ Systemic fungal infection in very low birth weight infants. Acta Paediatr Sin 1996; 37:272-277
28. Zipursky A, Palko J, Milner R, Akenzua GI Hematology of bacterial infections in premature infants. Pediatrics 1976; 57:839 [Abstract/Free Full Text]
29. Manroe BL, Weinberg AG, Rosenfeld CR, Browne R Neonatal blood counts in health and disease: reference values for neutrophilic cells. J Pediatr 1979; 95:89 [CrossRef][Medline]
30. Woods GL, Washington JA. The clinician and the microbiology: central nervous system infections. In: Mandell G, Douglas J, Bennett J, eds. Principles and Practice of Infectious Diseases. 4th ed. New York, NY: Churchill Livingstone; 1995:178
31. Chen JY Neonatal candidiasis associated with meningitis and endophthalmitis. Acta Paediatr Jpn 1994; 36:261-265 [Medline]
32. Kirkpekar M, Abiri MM, Hilfer C, Enerson R Ultrasound in the diagnosis of systemic candidiasis (renal and cranial) in very-low-birth-weight premature infants. Pediatr Radiol 1986; 16:17-20 [CrossRef][Medline]
33. Faix RG Systemic Candida infections in infants in intensive care nurseries: high incidence of central nervous system involvement. J Pediatr 1984; 105:616-622 [CrossRef][Medline]
34. Finkelstein R, Reinhertz G, Hashman N, Merzbach D Outbreak of Candida tropicalis fungemia in a neonatal intensive care unit. Infect Control Hosp Epidemiol 1993; 14:587-590 [Medline]
35. Kingo ARM, Smyth JA, Waisman D Lack of evidence of amphotericin B toxicity in very low birth weight infants treated for systemic candidiasis. Pediatr Infect Dis J 1997; 16:1002-1003 [CrossRef][Medline]