Profile of neonatal candidiasis in tertiary neonatal intensive care unit: A report from a developing country

1 Introduction

Infections are a major cause of mortality and morbidity in newborn throughout the globe [1]. The risk for invasive fungal infections is higher in very low birth weight (VLBW) infants requiring prolonged hospital stay or multiple antibiotics for suspected sepsis or prolonged parenteral nutrition [2–4]. In contrast to only inborn infants, sepsis in outborn infants is related to different organisms and heterogeneous outcomes due to differences in source of infection, timing of referral, Anti-Microbial Resistance (AMR), and level of sickness at admission [5, 6].

Fungal sepsis was reported in low incidence series (2–3%) previously from developed countries [7–9], whereas some recent studies of the current decade have reported a higher incidence of fungal sepsis in NICU in the range of 7.1% to 9% [10, 11]. Although Candida albicans has historically been the most frequently isolated species, recently non-albicans candida (NAC) with higher intrinsic resistance to triazoles have emerged as important opportunistic pathogen, notably Candida tropicalis, C. parapsilosis, C. krusei, and C. Glabrata in Indian scenario [12, 13]. Candida blood stream infection (BSI) may be associated with very high crude mortality of 15–34% in some studies [14, 15], but this can also be influenced by several factors including the population at risk, level of care, Candida spp involved and antifungal resistance.

However, there is a paucity of high-quality epidemiological data on neonatal candidiasis in the tertiary care referral NICUs in developing countries. This study has been undertaken to evaluate the epidemiology of candidial sepsis in a tertiary referral NICU from eastern India with special emphasis to its risk factors, species pattern, outcome, and antifungal resistance pattern.

2 Methods

This study was conducted by retrospective review of case records from a tertiary referral NICU of a medical college hospital in eastern India. The NICU has 16 NICU beds along with 9 steps down beds and 8 intensive care beds for surgical infants. It also serves as a referral center from district level II neonatal units for babies with congenital surgical and cardiac problems. All new-born admitted in the NICU between 2014 and 2019 in whom fungal organisms were isolated in blood culture were recruited in the study as a part of convenience sampling. Infants with lethal congenital malformations and inborn error of metabolism and co infection with bacterial isolates were excluded. Early onset and late onset sepsis were defined as infection within first 72 h of life and >72 hours of life respectively. We retrospectively collected data from case records and reviewed demographic data, risk factors, blood culture results isolated species, antifungal resistance pattern and outcomes of neonatal candidemia.

Initial blood culture was sent if infection was suspected at any time after admission to NICU. Additionally for outborn infants, blood culture was taken at admission. Candidemia was defined as the presence of at least one positive blood culture containing pure growth of Candida spp with supportive clinical features.

Blood culture was done with the help of an automated Blood culture system (BACTEC 9050). Nearly, 1–3 ml of blood was collected from neonates suspected of septicemia from peripheral vein under aseptic precaution. It was inoculated into pediatric blood culture bottle (BACTEC Peds Plus/F) and introduced into BACTEC 9050 system. After the system signaled for positive cultures, samples were subjected to gram stain to confirm the presence of yeast cells. Subcultures were done on Sabouraud dextrose agar and blood agar. Species identification was done with the help of CHROM agar, germ tube test, sugar fermentation and assimilation tests. Antifungal susceptibility was carried out by disk diffusion method as per Clinical Laboratory Standard Institute (CLSI) guidelines for Amphotericin-B, and Fluconazole, Itraconazole and Voriconazole [16].

3 Results

During the study period, 11486 neonates were admitted of which 103 had positive fungal (candida) cultures (0.89%) and 536 had any culture proven sepsis (bacterial and fungal). Among the admitted infants the number of preterm infants was 5882 (51.2%) and term infants was 5602 (48.8%). The outborn admission rate was 25.2% (2894/11486).

Basic characteristics of the study population have been depicted in Table 1. There was significant higher incidence of candida sepsis among outborn (p < 0.001, RR of outborn 18.84, 95% CI 10.74–33.05). Among fungal sepsis cases, 91 infants (88.3%) weighed ≥1500 g, 95 babies (92%) were born at or after 32 weeks’ gestation and 66 infants (64%) were term. Alternatively, the incidence of fungal sepsis among admitted VLBW infants was 0.64% (12/1875) while the incidence among infants >1500 g was 0.94% (91/9681) in our study. Similarly, incidence of fungal sepsis among admitted term infants was 1.1% (66/6000) whereas the similar incidence among admitted preterm was 0.62% (37/5968). Majority of infants with fungal sepsis (96/103, 93.8%) presented as late onset sepsis.

Outborn infants with candida septicemia had significantly higher birth weight (2500 g vs. 1935 g, p = 0.008), were more mature (35 vs. 30 weeks, p = 0.009) than inborn infants with candida septicemia. Prolonged antibiotic usage (>14 days), meropenem usage (>5 days), central catheterization (>5 days), invasive mechanical ventilation (>5 days), surgical intervention was found in 64 (62.1%), 46 (44.6%), 31(30.0%), 40 (38.8%) and 39 (37.8%) infants respectively (Table 2). Among the infants receiving prolonged antibiotics, 40/64 infants (62.5%) had received such intervention in the referral hospital before admission to study unit.

Among total 103 candida positive cases, Candida albicans was isolated in 21 (20.8%) cases where as Non albicans candida (NAC) species were isolated in 82 (79.2%) infants. Incidence of various candida species are given in Table 3. Resistance to fluconazole was found in 53 (51.4%) infants and resistance to amphotericin B was found in 36 (34.9%) infants whereas resistance to both of fluconazole and amphotericin B was found in only 19 (18.4%) infants.

Out of the 103 infants with candida septicemia, 35 (33.9%) expired despite best of treatment effort. Median birth weight and gestation of expired infants were significantly higher in case of outborn infants. In univariate analysis significant association of mortality due to fungal infection was seen for the risk factors like invasive ventilation >5 days (OR 2.65, 95% CI 1.14–6.16), central line >5 days (OR 2.43, 95% CI 1.01–5.83), Meropenem >5 days (OR 3.10, 95% CI 1.33–7.23), resistance to both fluconazole and amphotericin B (OR 3.43, 95% CI 1.23–9.59) and NAC infection (OR 6.39, 95% CI 1.23–9.59. Other risk factors like out born, male gender, late onset sepsis, cesarean section, surgical infants and prolonged broad-spectrum antibiotics did not show any significant association with mortality. However, in multiple logistic regression analysis, mortality in candida positive babies was found independently associated with presence of NAC infection (adjusted OR 7.62, 95% CI 1.21–47.95) and resistance to both amphotericin B and fluconazole (adjusted OR 3.74 95% CI 1.15–12.07)) (Table 4).

In this study, fungal sepsis was found out to be positive in 0.89% (103/11486) cases among admitted infants in NICU during the study period, which is comparable to incidence reported from developed and developing countries in many studies [4, 7–9], but less than that of reported by Niranjan et al., and Benjamin et al., (7.1% and 9% respectively) [2, 10].

Out of all culture positive sepsis cases, the fungal sepsis cases were found as high as 19.3% (103/536). In this study 88% of fungal infection was among babies more than 1500 g birth weight, and 64% was found in babies with term gestation. The low incidence of fungus sepsis in VLBW infants here was corroborating with the reports from the developed world [10, 18]. This could be attributed by some quality initiative measures in the unit like early enteral feeding policy, use of noninvasive ventilation, robust hand hygiene, antimicrobial stewardship, and rational use of antimicrobials for long years. This resulted in a low incidence of sepsis in the unit.

In the Indian perspective, Femitha et al., [11] also reported only 3.1% incidence of fungal sepsis in VLBW infants from India. The reason for low incidence of candidemia in above study was routine use of fluconazole prophylaxis in VLBW infants in NICU but we did not use fluconazole prophylaxis in our NICU. Moreover, all the VLBW infants were fed human milk from the human milk bank in the unit. This along with strict infection control policy could be attributed to the lower incidence of fungal sepsis amongst the inborn infants in the unit.

One of the interesting findings of our study was significantly higher incidence of candida sepsis amongst infants with higher birth weight in contrast to reports from India where higher incidence of fungal septicemia was found among VLBW infants [15, 20]. Here fungal sepsis was seen amongst the outborn infants of higher birth weight referred for surgical or cardiac morbidities. In those infants, fungal culture was positive within 24 hours of admission in most cases and most of the infants had history of previous exposure to broad-spectrum antibiotics. Prior hospitalization was one of the strong risk factors of fungal sepsis in the referred neonates, most of whom came from level II neonatal units with sub-optimal infection control practices and irrational antibiotic policies. There is a definite need to optimize infection control practices and facilitate implementation of quality control measures in these units. This has also posed emergent threat of risk for large tertiary neonatal care unit where large number of infants are admitted after being referred from peripheral units.

In literature [11, 21], prolonged antibiotics usage, H₂ blockers usage, mechanical ventilation, total parenteral nutrition/central line were found as risk factors for fungal sepsis. Similar scenario was observed in the present study, except for few differences like we did not use H2 blockers here and additionally surgical babies were found as susceptible for fungal sepsis.

Previously, most cases of neonatal candidemia were caused by Candida albicans. However, there are few reports that there is an emergence of non-albicans Candida (NAC) species as an important cause of fungus septicemia [22, 23]. Here in this study the important observation was the predominance of NAC infection in both inborn and outborn infants. Among NAC species various organisms were reported in earlier reports [15, 24], but we found C. tropicalis as most predominant species (34.9%) which was also very virulent and quite uncommon in neonatal population. Resistance to fluconazole and amphotericin B was found in higher proportion here as in previous studies [15, 26]. Resistance to fluconazole, amphotericin B and both of fluconazole and amphotericin B were found in 51.4%, 34.9% and 18.4% of fungal sepsis cases respectively in this study.

There is a heterogeneous report of mortality (15–34%) of candida sepsis among newborn [14, 15] but the mortality can be even higher in case of NAC infection [11]. However, in the present study, 33.9% (35/103) of fungal sepsis cases expired and the presence of NAC infection and resistance to both amphotericin B and fluconazole were found as independent risk factors for mortality. Hence, the local epidemiology of fungal isolates with their resistance pattern is critical in terms of better management and prevention of fungal sepsis in NICU.

5 Conclusion

Higher incidence of fatal fungal infection and its threat amongst infants with higher birth weight was one of the striking findings of our study. NAC species are present day’s predominant organism with resistance to common antifungal drugs. In view of case fatality rate of fungal sepsis and fungal resistance, clinicians should refrain from administration of antibiotics for babies who do not require it. This study highlights the importance of implementation of strict antimicrobial policy and adherence to standard treatment protocol especially early enteral feed with human milk for management of sick neonates in the sick newborn care units. This will not only reduce sepsis related newborn mortality but also reduce of antifungal as well as antimicrobial resistance.