A Case–Control Study of Candidaemia in Very Low Birthweight Infants in a Tertiary Hospital in Johannesburg

Abstract

Background: Candidaemia is a significant cause of morbidity and mortality in infants. The mortality rate ranges between 21% and 76%. Non-albicans Candida (NAC) species are increasing in incidence as well as resistance to azoles. Very low birthweight (VLBW) infants have numerous risk factors which predispose them as a group to invasive candidaemia.

Methods: A retrospective case–control study of candidaemia in VLBW infants admitted to the neonatal unit at Charlotte Maxeke Johannesburg Hospital between 1 January 2015 and 31 December 2017 was undertaken. Clinical and demographic characteristics of VLBW infants who developed candidaemia, commonest Candida species, antifungal susceptibility profiles and outcomes (death) were identified. A total of 71 infants with confirmed positive blood cultures for candidaemia from the National Health Laboratory Services database were selected, and each case was allocated 3 controls; the final sample comprised 284 infants.

Results: Bacterial sepsis, chronic lung disease, necrotising enterocolitis (NEC) and NEC surgery, other surgery, anaemia and ventilation, all showed a strong association with development of candidaemia in the infants. The most common isolate was Candida parapsilosis (59.1%), followed by Candida albicans (30.9%). The cases of candidaemia overall and NAC isolates increased over the study years. Resistance to azoles by NAC was demonstrated. Mortality was 31.2% and 28.2% in controls and cases, respectively. The difference in death between the two groups was not statistically significant.

Conclusions: The study demonstrated an increasing rate of candidaemia, predominance of NAC isolates and increased resistance to azoles. Stricter infection control measures and medical intervention strategies should be implemented.

Main article text

INTRODUCTION

The commonest cause of invasive fungal infections in infants is Candida species. Candidaemia is associated with significant morbidity and mortality in infants.(1) The mortality rate in infants with candidaemia ranges between 21% and 76%.(2,3) Ballot et al. in their review of neonatal fungal bloodstream infections (BSI) at Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) for the period January 2007 to December 2011 reported a mortality rate of 45.8%.(4) Candidaemia is responsible for 9–13% of BSI in infants.(4) A high index of suspicion for candidaemia needs to be maintained in infants of very low birthweight (VLBW), i.e. less than 1500 g at birth, low gestation, chronic lung disease (CLD), anaemia, surgery, necrotising enterocolitis (NEC), bacterial sepsis, hypothermia and low blood glucose in whom blood cultures remain negative.(4) Symptoms and signs of candidaemia are non-specific and blood cultures have a suboptimal sensitivity for detection of candidaemia. The burden of disease, poor outcomes and cost of medical treatment in infants with candidaemia is significant and necessitates the implementation of prevention strategies.

Infants at most risk for fungal BSI are VLBW and extremely low birthweight (ELBW) infants, i.e. less than 1000 g at birth, infants born at less than 30 weeks gestation age and those with surgical problems.(5) The risk of sepsis is inversely proportional to birthweight and gestation age.(6) VLBW infants are prone to numerous medical complications that require invasive treatment, and a significant number of the treatments can increase the risk for candidaemia.

Infants with candidaemia present with non-specific clinical signs. Sudden clinical deterioration, such as elevated or low blood glucose levels, unexplained abdominal distention, intolerance to feeds (large gastric aspirates or aspirates with abnormal discoloration), hypothermia, suddenly increased ventilation or oxygen requirements, lethargy, platelet levels dropping and circulatory insufficiency require urgent treatment.(4,5)

Appropriate choice of antifungal agent and timely implementation of therapy are significant with respect to patient outcomes. A delay of more than 24 h in administration of appropriate therapy has been shown to independently increase the patient's probability of demising by two fold.(7) The selection of appropriate empiric therapy must take into account the local epidemiology. In settings with a high prevalence of non-albicans and azole-resistant Candida species, the use of fluconazole is not suitable for empiric therapy.

Management guidelines must be guided by local current epidemiological data. This study therefore aimed to review candidaemia in VLBW infants at CMJAH, South Africa, with respect to associated risk factors, fungal organisms isolated, antifungal susceptibility patterns and to determine outcomes of neonates with candidaemia.

METHODS

This was a retrospective case–control study of VLBW infants admitted to the neonatal unit at CMJAH between 1 January 2015 and 31 December 2017. The study was approved by the Human Research Ethics Committee of the University of the Witwatersrand.

Data used were obtained from Research Electronic Data Capture hosted by the University of the Witwatersrand.(8) The VLBW infants identified with candidaemia were classified as cases. The next three VLBW infants, without candidaemia, admitted to the unit were classified as controls.

Culture results (organism identity and antifungal sensitivity) were obtained from the National Health Laboratory Services database (Corporate Data Warehouse). Those infants with suspected candidaemia who were found to have negative cultures were excluded from the study.

During the study period, Amphotericin B was used as empiric therapy for treatment of suspected candidaemia. Treatment was revised based on microbiology results.

The BacT/AlerT^® (BioMerieux, France) blood culture system was used. Yeasts were identified using Vitek^® MS, Vitek^® 2 or API^® 20C AUX (BioMerieux). Antifungal susceptibility testing was performed using ETESTs^® or Vitek^® 2 (BioMerieux) and interpreted using Clinical and Laboratory Standards Institute guidelines (CLSI M27-A3).(9)

The participants were categorised according to their weight, those <1000 and >1000 g and gestational ages <30 and >30 weeks. The cut-off for weight was 1500 g and gestational age <36 weeks.

Definitions

CLD was defined as supplemental oxygen at 28 days of age.

NEC was defined as modified Bell's stage 2 and 3.

Bacterial sepsis was defined as culture-proven sepsis at any age and included both early-onset and late-onset sepsis.

Statistical analysis

IBM SPSS statistics version 25 was used for statistical analysis. Categorical data were described using frequencies and percentages. Cases (with candidaemia) were compared with controls (without candidaemia). Chi-square analysis was used to compare categorical data. A P-value <0.05 was considered statistically significant.

RESULTS

Flow diagram showing a selection of study participants.

The majority of the patients were females (54.6%). Most infants were VLBW (65.1%) and below 30 weeks gestation (64.0%). The overall mortality rate was 30.6%. There was no association with candidaemia for gender, birthweight, gestational age or mortality in the current study. Anaemia, CLD, NEC, the need for ventilator support, surgery and bacterial sepsis were all significantly associated with candidaemia (Table 1). The microbiology of the organisms isolated and their antibiotic susceptibility is shown in Table 2.

Table 1:

Patient characteristics in VLBW infants with and without candidemia

Characteristics	Total(n%) 284	No FBSI(n%) 213	FBSI(n%) 71	P-value
Gender
Male	129 (45.4)	99 (46.5)	30 (42.3)	0.583
Gestational age <30 weeks	181 (64.0)	132 (62.0)	49 (69.0)	0.322
Birth weight <1000g	97 (34.1)	75 (35.2)	22 (31.0)	0.536
HMD	206 (72.5)	153 (71.8)	53 (75.0)	0.643
CLD	32 (11.3)	16 (7.5)	16 (22.5)	0.002
Anaemia	85 (29.9)	54 (25.4)	31 (43.7)	0.005
Maternal HIV	93 (32.7)	65 (30.5)	28 (39.4)	0.205
Antenatal steroids	118 (41.5)	84 (39.4)	34 (47.9)	0.404
Antenatal care	202 (71.1)	151 (70.8)	51 (71.8)	0.761
NEC	41 (14.4)	24 (11.3)	17 (23.9)	0.012
NEC surgery	32 (11.3)	19 (8.9)	13 (18.3)	0.049
Other surgery	29 (10.2)	10 (4.7)	19 (26.8)	<0.001
Ventilated
NCPAP	206 (72.5)	142 (66.7)	64 (90.1)	0.001
Conventional	99 (34.9)	55 (25.8)	44 (62.0)	<0.001
Bacterial sepsis	118 (41.5)	62 (29.1)	56 (78.9)	<0.001
Demised	87 (30.6)	67 (31.5)	20 (28.2)	0.324

HMD = Hyaline membrane disease, CLD = chronic lung disease, NEC = necrotising enterocolitis, NCPAP = nasal continuous positive pressure ventilation. Other surgery (arthrotomy, arthroplasty, laparotomy for various indications), FBSI = fungal blood stream infection.

Table 2:

Organisms isolated and antimicrobial sensitivity

Candida species isolated	Total	Frequency	Azole susceptible	Total percentage of resistant
C. parapsilosis	*42	59.2%	20/34	41.2%
C. albicans	22	31.0	22	0
Other Candida species	4	5.6	-	-
C. glabrata	3	4.2	2	-

*

Data missing for C. parapsilosis microbiology results, only 34 out of 42 reported.

No Clinical and Laboratory Standards Institute breakpoint for Candida glabrata and voriconazole.

The Candida isolates during the study period is shown in Figure 1. The incidence of candidaemia in this VLBW study population was calculated as the total number of candidaemia cases divided by the number of total admissions of VLBW infants (18/475, 23/487, 30/515 for 2015, 2016 and 2017, respectively) to get percentage total per annum. Cases with C. parapsilosis also showed a progressive but significant (P-value <0.001) increase during the study period. The increase in Candida isolates from 2015 to 2017 was not statistically significant (P = 0.835).

Fig 1:

Number of Candida species (isolates per year). The percentages indicate the percentage of VLBW infants with candidemia per year of the study.

DISCUSSION

The study found an increase in the cases of candidaemia in VLBW infants at CMJAH. There was an increase in the cases of candidaemia during the study period from 3.79% in 2015 to 5.83% in 2017, an increase of almost 50% over a 3-year period. This is in keeping with a previous report from the same study unit which demonstrated an increase in candidaemia over a 5-year period.(5) The current study also demonstrated a significant increase in the cases of C. parapsilosis from 2015 to 2017. This is in concert with findings by Lovero et al. who also showed an increase in the incidence of non-albicans Candida (NAC) with particular reference to C. parapsilosis.(10)

The risk factors significantly associated with candidaemia in VLBW infants in the present study were similar to those reported in western China and India.(11–13) The association between candidaemia and NEC, mechanical ventilation and surgery reported in the current study are comparable with other studies evaluating candidaemia in infants.(5,14–16) The association between NCPAP and candidaemia described in the current study has not been previously reported. The plausible explanation would be poor adherence to infection control measures. When the tubing and humidifiers are contaminated by the hands of the healthcare personnel possibly colonised with fungal organisms, it can lead to contamination and spread of fungal infections.(17) Therefore, adhering to strict infection control measure is essential. CLD has been found to be a risk factor for candidaemia in VLBW in the present study. Won et al. demonstrated a similar association.(18) Low gestation (<30 weeks) and ELBW were not statistically significant risk factors in this study. However, risk factors that are found to be common to VLBW infants are seen in this study and have shown a strong association with candidaemia.(12,17,19)

Prior bacterial sepsis was a common risk factor in patients presenting with candidaemia. Broad spectrum antibiotics are used for the treatment of bacterial sepsis. This is associated with a selective pressure for Candida species.(1,20) The study showed bacterial sepsis to have a significant association with candidaemia in VLBW infants and is consistent with results in similar studies.(1,5,15) At the time of the study, the empiric antibiotics for late onset bacterial sepsis were meropenem and vancomycin. Fu et al. demonstrated a strong association between carbapenem use and development of candidaemia.(13)

Candida species is a third leading cause of late onset sepsis in blood culture isolates and leading cause of fungal infection in infants, in particular VLBW.(7,10,14,21) The present study has shown an increase in the incidence of candidaemia. A decline in the number of C. albicans blood culture isolates was also seen in this study, with C. parapsilosis blood culture isolates increasing annually and surpassing C. albicans. Studies from Brazil, Iran, Uganda and the Netherlands have demonstrated an increase in NAC; however, C. parapsilosis was the second isolate to C. albicans.(12,15,21,22) Fu et al. showed C. glabrata as the second common NAC isolate in a recent study done at a tertiary hospital in western China.(14)

Hospitals in Italy and in the USA demonstrated a strong association between C. parapsilosis and the use of total parenteral nutrition (TPN) and adherence to central catheters.(23,24) However, due to the lack of information in our database, these two risk factors were not included in the present study.(1,6,13,16,25,26)

The increase in resistance to azoles by NAC is well demonstrated in this study with 41.2% of isolates of C. parapsilosis showing resistance to both fluconazole and voriconazole. Sarvikivi et al. in their study also demonstrated an increase in resistance to fluconazole by C. parapsilosis isolates.(27) C. glabrata that is susceptible to fluconazole requires high doses to treat it. Both Talarmin et al. and Ota and McGowan demonstrated dose-dependent susceptibility to fluconazole by C. parapsilosis.(28,29) The unit where the study was conducted, Amphotericin B was used to treat C. glabrata candidaemia.

The successful use of fluconazole as prophylaxis in preventing invasive fungal infections with Candida species has been reported; however, a statistically significant decrease in mortality has not been demonstrated.(30–33) The CMJAH neonatal unit does not give fluconazole prophylaxis to VLBW neonates. The high number of fluconazole-resistant C. parapsilosis isolates made the use of fluconazole prophylaxis questionable.

Death as an outcome in infants with candidaemia was 28.2%, which was not statistically significantly higher than in controls. The high number of deaths in the control group is attributed to the high number of ELBW infants. Sriram demonstrated higher mortality rate in infants with candidaemia 19.6% versus 10.7% for controls.(34)

STUDY LIMITATIONS

The study duration was short and the statistical power was reduced due to exclusion of a significant number of infants due to missing information in the clinical records. Information on central venous catheters and TPN was also not available. The study cohort was limited to VLBW premature infants. Thus, the findings of the study cannot be generalised to all infants.

CONCLUSION

CLD, anaemia, ventilatory support, bacterial sepsis and surgery were associated with candidaemia. The increase in incidence of NAC isolates was demonstrated with increasing resistance to azoles. Stricter infection control measures and medical intervention strategies to reduce the incidence of candidaemia in VLBW infants should be implemented. The use of Amphotericin B in empiric treatment of candidaemia should be continued in the unit and treatment only changed once results on the isolates and susceptibility patterns are obtained. A study over a longer period and a larger cohort is needed.

REFERENCES

1. MirandaL, RodriguesEC, CostaSF, et al. Candida parapsilosis candidaemia in a neonatal unit over 7 years: a case series study. BMJ Open. 2012; 2(4): 2.e0009992.

2. RodriguezD, AlmiranteB, ParkBJ, et al. Candidemia in neonatal intensive care units: Barcelona, Spain. Pediatr Infect Dis J. 2006; 25(3): 224–229.

3. XavierPC, ChangMR, NunesMO, et al. [Neonatal candidemia in a public hospital in Mato Grosso do Sul]. Rev Soc Bras Med Trop. 2008; 41(5): 459–463.

4. BallotDE, BosmanN, NanaT, RamdinT, CooperPA. Background changing patterns of neonatal fungal sepsis in a developing country. J Trop Pediatr. 2013; 59(6): 460–464.

5. JuyalD, SharmaM, PalS, RathaurVK, SharmaN. Emergence of non-albicans Candida species in neonatal candidemia. N Am J Med Sci. 2013; 5(9): 541.

6. TrofaD, GacserA, NosanchukJD. Candida parapsilosis, an emerging fungal pathogen. Clin Microbiol Rev. 2008; 21(4): 606–625.

7. ZilberbergMD, KollefMH, ArnoldH, et al. Inappropriate empiric antifungal therapy for candidemia in the ICU and hospital resource utilization: a retrospective cohort study. BMC Infect Dis. 2010; 10(1): 150.

8. HarrisPA, TaylorR, ThielkeR, PayneJ, GonzalezN, CondeJG. Research electronic data capture (REDCap) – a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42(2): 377–381.

9. WayneP. Clinical and Laboratory Standards Institute: reference method for broth dilution antifungal susceptibility testing of yeasts; approved standard. In: CLSI document M27-A3 and Supplement S; 2008; 3.

10. HammoudMS, Al-TaiarA, FouadM, RainaA, KhanZ. Persistent candidemia in neonatal care units: risk factors and clinical significance. Int J Infect Dis. 2013; 17(8): e624–e628.

11. WadileRG, BhateVM. Study of clinical spectrum and risk factors of neonatal candidemia. Indian J Pathol Microbiol. 2015; 58(4): 472–474.

12. FuJ, WangX, WeiB, JiangY, ChenJ. Risk factors and clinical analysis of candidemia in very-low-birth-weight neonates. Am J Infect Control. 2016; 44(11): 1321–1325.

13. FuJ, DingY, WeiB, et al. Epidemiology of Candida albicans and non-C. albicans of neonatal candidemia at a tertiary care hospital in western China. BMC Infect Dis. 2017; 17(1): 329.

14. SpiliopoulouA, DimitriouG, JelastopuluE, GiannakopoulosI, AnastassiouED, ChristofidouM. Neonatal intensive care unit candidemia: epidemiology, risk factors, outcome, and critical review of published case series. Mycopathologia. 2012; 173(4): 219–228.

15. AriffS, SaleemAF, SoofiSB, SajjadR. Clinical spectrum and outcomes of neonatal candidiasis in a tertiary care hospital in Karachi, Pakistan. J Infect Dev Countries. 2011; 5(3): 216–223.

16. GarzilloC, BagattiniM, BogdanovicL, et al. Risk factors for Candida parapsilosis bloodstream infection in a neonatal intensive care unit: a case-control study. Ital J Pediat. 2017; 43(1): 10.

17. WonS-P, ChenC-Y, ChouH-C, TsaoP-N, HsiehW-S. Clinical outcomes of very low birth weight infants with candidemia. Survival. 2007; 22(57.9): 0.567.

18. LoveroG, De GiglioO, MontagnaO, et al. Epidemiology of candidemia in neonatal intensive care units: a persistent public health problem. Ann Ig. 2016; 28(4): 282–287.

19. KotwalA, BiswasD, SharmaJP, GuptaA, JindalP. An observational study on the epidemiological and mycological profile of Candidemia in ICU patients. Med Sci Monit. 2011; 17(11): Cr663–Cr668.

20. MaXL, SunW, LiuT. [Clinical characteristics of Candida septicemia seen in a neonatal intensive care unit: analysis of 9 cases]. Zhonghua Er Ke Za Zhi. 2006; 44(9): 694–697.

21. CharsizadehA, MirhendiH, NikmaneshB, EshaghiH, MakimuraK. Microbial epidemiology of candidaemia in neonatal and paediatric intensive care units at the Children's Medical Center, Tehran. Mycoses. 2018; 61(1): 22–29.

22. AbdallahY, Kaddu-MulindwaD, NankundaJ, MusokePM. Prevalence and immediate outcome of candida colonized preterm neonates admitted to Special Care Unit of Mulago Hospital, Kampala Uganda. Afr Health Sci. 2015; 15(1): 197–205.

23. CrivaroV, BogdanovicL, BagattiniM, et al. Surveillance of healthcare-associated infections in a neonatal intensive care unit in Italy during 2006–2010. BMC Infect Dis. 2015; 15: 152.

24. BenjaminDK, StollBJ, GantzMG, et al. Neonatal candidiasis: epidemiology, risk factors, and clinical judgment. Pediatrics. 2010; 126(4): e865–e873.

25. MilstoneAM, ReichNG, AdvaniS, et al. Catheter dwell time and CLABSIs in neonates with PICCs: a multicenter cohort study. Pediatrics. 2013; 132(6): e1609–e1615.

26. BasuS, KumarR, TilakR, KumarA. Candida blood stream infection in neonates: experience from a tertiary care teaching hospital of Central India. Indian Pediatr. 2017; 54(7): 556–559.

27. SarvikiviE, LyytikainenO, SollDR, et al. Emergence of fluconazole resistance in a Candida parapsilosis strain that caused infections in a neonatal intensive care unit. J Clin Microbiol. 2005; 43(6): 2729–2735.

28. TalarminJP, BoutoilleD, TattevinP, et al. [Epidemiology of candidemia: a one-year prospective observational study in the west of France]. Med Mal Infect. 2009; 39(12): 877–885.

29. OtaKV, McGowanKL. Declining incidence of candidemia in a tertiary inpatient pediatric population. J Clin Microbiol. 2012; 50(3): 1048–1050.

30. KaufmanD, FairchildKD. Clinical microbiology of bacterial and fungal sepsis in very-low-birth-weight infants. Clin Microbiol Rev. 2004; 17(3): 638–680, table of contents.

31. PfallerMA, DiekemaDJ, GibbsDL, et al. Results from the ARTEMIS DISK Global Antifungal Surveillance Study, 1997 to 2007: a 10.5-year analysis of susceptibilities of Candida Species to fluconazole and voriconazole as determined by CLSI standardized disk diffusion. J Clin Microbiol. 2010; 48(4): 1366–1377.

32. AustinN, McGuireW. Prophylactic systemic antifungal agents to prevent mortality and morbidity in very low birth weight infants. Cochrane Database Syst Rev. 2013; Apr 30 (4): Cd003850.

33. ClerihewL, AustinN, McGuireW. Prophylactic systemic antifungal agents to prevent mortality and morbidity in very low birth weight infants. Cochrane Database Syst Rev. 2007; Oct 17 (4): Cd003850.

34. SriramB, AgarwalPK, TeeNW, RajaduraiVS. Systemic candidiasis in extremely low birthweight (ELBW) neonates despite the routine use of topical miconazole prophylaxis: trends, risk factors and outcomes over an 11-year period. Ann Acad Med Singapore. 2014; 43(5): 255–262.