Incidence, Etiology and Risk Factors Associated with Neonatal Healthcare-Associated Conjunctivitis: A Prospective Study from a Tertiary Care Hospital in India

Abstract

Introduction: Healthcare-associated conjunctivitis (HAC) can lead to serious sequelae including blindness. We conducted a one-year prospective study to determine the epidemiology of neonatal HAC at a tertiary-care hospital in India.

Methods: From the neonates fulfilling a set of predefined inclusion criteria, cases of HAC were diagnosed based on CDC guidelines. Conjunctival swabs, obtained from neonates with suggestive clinical signs, were processed using standard protocols. Twenty-eight potential risk factors were analyzed.

Results: We detected 24 cases of HAC among 591 enrolled neonates, with Escherichia coli being the most frequently isolated microorganism. On multivariate analysis, intubation at birth (p = 0.046) and orogastric feeding (p = 0.029) had a statistically significant association with neonatal HAC. Average hospitalization increased from 9.6 to 20.8 days for neonates diagnosed with HAC.

Conclusion: A standardized case-definition and physician awareness of potential serious sequelae would help improve detection rates and timely institution of therapy. Hand hygiene could help control the menace of neonatal HAC.

INTRODUCTION

Conjunctivitis is one of the commonest ocular diseases of neonates. It is associated with potentially serious complications and long-term sequelae, and is therefore considered to be a major public health problem [1]. A relatively weak immune system of the neonates, developmental immaturity of their lacrimal ducts and frequent colonization of conjunctivae that occurs during the newborn care, together result in an increased susceptibility to develop conjunctivitis [2]. Mechanical trauma from therapeutic interventions, such as eye patching for phototherapy and routine neonatal care, further increases the likelihood of hospitalized neonates to develop conjunctivitis [3]. Moreover, the proximity of newborns in many neonatal units may facilitate the horizontal transfer of pathogens, in this highly susceptible population [4]. As a consequence, many outbreaks of healthcare-associated conjunctivitis (HAC) have been reported to develop in the neonatal units/nurseries across the world [5–7]. It is, therefore, not surprising that HAC is one of the commonest healthcare-associated infections (HAIs) among neonates [8–12].

Untreated neonatal HAC, especially the cases associated with highly virulent pathogens such as Pseudomonas aeruginosa, can lead to several serious complications. These include both, local complications such as corneal perforation and visual impairment leading to blindness [13, 14], as well as a multitude of systemic complications including sepsis, meningitis and pneumonia [5, 15]. Despite being a fairly common HAI with a high risk of developing serious health sequelae, only a few epidemiological studies that evaluate the risk factors and microbiological profile of neonatal HAC have been reported in the literature [2, 16, 17]. The absence of neonate-specific diagnostic criteria, administration of empiric antibiotics without undertaking diagnostic cultures and a general ignorance of the potentially serious complications have together resulted in a lack of adequate surveillance for this disease. With this background, we conducted an epidemiological study to determine the incidence, the etiological profile and the risk factors associated with neonatal HAC.

METHODS

This one-year prospective study was conducted in the 23-bedded neonatal unit of Smt. Sucheta Kriplani Hospital, which is the major teaching hospital and academic center affiliated with the Lady Hardinge Medical College in New Delhi, India. The nursery was staffed by three neonatologists, along with fellows-in-training, registered nurses and ancillary personnel. Laboratory support was provided by the microbiology, pathology and biochemistry departments along with consult support from the departments of nutrition, radiology and pediatric surgery. Universal precautions including standard hand washing procedures were mandatory for all staff members and visitors entering the nursery. The study protocol was approved by the ethics committee and the institutional review board (IRB) of the Lady Hardinge Medical College, University of Delhi. Due to the observational nature of the study, consent was waived as per the IRB policy. This study was conducted in collaboration with the National Centre for Disease Control, New Delhi.

All neonates admitted to the nursery between 17th February 2010 and 16th February 2011 were screened for the presence of four predefined inclusion criteria: (i) age between 0 to 28 days at the time of admission, (ii) stay in the nursery for >48 h, (iii) absence of any signs of sepsis at the time of admission, and (iv) absence of prior hospitalization in the neonatal intensive care unit (NICU). Neonates that fulfilled all the above inclusion criteria were included in the study and further evaluated for identification of HAC cases. A detailed birth history was collected from the parent/guardian of all the enrolled neonates, and daily examinations were performed to identify any clinical evidence of conjunctivitis (e.g. presence of conjunctival redness, chemosis and purulent exudates) until discharge from the nursery or death. The cases of neonatal HAC were detected based on the diagnostic criteria established by the Centers for Disease Control and Prevention (CDC)/National Healthcare Safety Network (Table 1) [18].

Table 1.

CDC Criteria for HAC [18]

1. Patient has pathogens cultured from purulent exudate obtained from the conjunctiva or contiguous tissues, such as eyelid, cornea, meibomian glands or lacrimal glands. 2. Patient has pain or redness of conjunctiva or around eye                AND At least 1 of the following: White blood cells and organisms seen on Gram’s stain of exudate Purulent exudate Positive laboratory test (e.g., antigen tests such as ELISA or IF for Chlamydia trachomatis, herpes simplex virus, adenovirus) on exudate or conjunctival scraping Multinucleated giant cells seen on microscopic examination of conjunctival exudate or scrapings Positive viral culture Diagnostic single antibody titer (IgM) or 4-fold increase in paired sera (IgG) for pathogen.

The following are not included in HAC:

Conjunctivitis caused by silver nitrate (AgNO3). Conjunctivitis occurring as a part of a more widely disseminated viral illness (such as measles, chickenpox or an upper respiratory infection).

Conjunctival scrapings were taken from both eyes in all clinically suspicious cases of conjunctivitis. Scrapings were obtained using (i) sterile cotton swabs pre-moistened with sterile normal saline, and (ii) sterile polyester swabs that were transported to the laboratory in viral transport medium (VTM) at 4°C. One cotton swab from each eye was used to prepare gram-stain and Giemsa-stained smears while the second cotton swab was processed for the isolation and identification of bacterial and fungal pathogens using standard procedure [19]. The antimicrobial susceptibility testing was performed by the Clinical and Laboratory Standards Institute disc diffusion method [20, 21]. Swabs in VTM were inoculated into the HEp-2 cells for isolation of viruses such as adenovirus and enterovirus [22].

A total of 28 potential risk factors related to the demographic profile, birth-related events, maternal comorbidities, prior exposure to antimicrobials, blood transfusions and the presence of invasive catheters/devices in the neonate were evaluated to determine any association with the occurrence of HAC. The data were compiled and analyzed using the Epi Info™ version 7.1.4 (CDC, Atlanta, USA) statistical software. Univariate logistic regression analysis was performed. Variables that were statistically significant in univariate analysis were entered into a backward stepwise regression model in multivariate analysis. All p-values were two sided and values <0.05 were considered significant. Odds ratios (OR) and 95% confidence intervals (CIs) were calculated for all significant variables. The impact of HAC on outcomes such as the length of hospital stay and neonatal mortality was also evaluated.

RESULTS

A total of 899 intramural neonates were admitted to the neonatal nursery during the defined one-year study period, out of which 591 neonates fulfilled the inclusion criteria and were enrolled into the study. Of these, 24 neonates were eventually diagnosed with HAC. The calculated incidence of neonatal HAC was 4.06 infections per 100 eligible admissions in this single institution study. Majority (71%; 17/24) of the HAC cases were diagnosed during day 3–day 7 of the hospital stay. Cases were detected throughout the year with maximum in the month of May (30%; 7/24). All neonates with HAC were found to have purulent discharge from the eyes (either one or both), while conjunctival erythema and chemosis was observed in only 75% (18/24) of the cases. Most of the HAC cases (83%; 20/24) were culture-positive. The four cases without culture positivity were diagnosed based on the clinical presentation and fulfillment of other CDC diagnostic criteria (Table 1) [18].

Among the 20 culture-positive neonatal HAC cases in our study, gram-negative bacteria (GNB) accounted for 12 (60%), while gram-positive bacteria (GPB) were associated with 8 (40%) cases (Table 2). Escherichia coli (35%), Staphylococcus aureus (25%), Klebsiella sp. (15%) and Pseudomonas aeruginosa (10%) were the most frequently isolated bacterial pathogens. Interestingly, GNBs were the more frequently (71%; 10/14) implicated microorganisms in the cases diagnosed during the first week of hospital stay, with a shift in trend towards GPBs (67%; 4/6) for neonates that developed HAC after day 7. Viral or fungal etiology was not established in any of the conjunctivitis cases.

Table 2.

Micro-organisms Isolated from Neonates with HAC

Group of Pathogens	Micro-organism	Number of Cases (%)
Gram-negative bacteria (GNB)	Klebsiella sp.	3 (15)
	Escherichia coli	7 (35)
	Pseudomonas aeruginosa	2 (10)
Gram-positive bacteria (GPB)	Staphylococcus aureus	5 (25)
	Coagulase negative staphylococci	1 (5)
	Enterococcus sp.	2 (10)
Fungi	Not isolated	0 (0)
Viruses	Not isolated	0 (0)
Total		20 (100)

In our study, all the GPB strains were found to be multidrug resistant, and exhibited resistance most notably to penicillin (100%), erythromycin (87.5%), ampicillin (75%), gentamicin (75%), tetracycline (62.5%) and amoxicillin-clavulanate (50%). However, good in vitro sensitivity to vancomycin (100%) was seen among all GPBs (Table 3). Among the GNBs, two strains of Pseudomonas aeruginosa were isolated, and exhibited a similar sensitivity pattern (both resistant to gentamicin, amikacin and ciprofloxacin). Strains of Enterobacteriaceae family (E.coli and Klebsiella sp.) showed a variable sensitivity pattern, with many strains being resistant to ampicillin (70%), amoxicillin-clavulanate (60%), ciprofloxacin (60%) and gentamicin (60%). Most Enterobacteriaceae strains were, however, susceptible to imipenem (100%), aztreonam (90%) and third-generation cephalosporins (70%).

Table 3.

Antimicrobial susceptibility pattern of bacterial isolates

Antimicrobial tested	Number (%) resistant
	Staphylococcus aureus (n = 5)	CoNSa (n = 1)	Enterococcus sp. (n = 2)	Escherichia coli (n = 7)	Klebsiella sp. (n = 3)	Pseudomonas aeruginosa (n = 2)
Ampicillin	4 (80)	1 (100)	1 (50)	5 (71.4)	2 (66.7)
Amoxycillin-clavulanate	3 (60)	0 (0)	1 (50)	5 (71.4)	1 (33.3)
Penicillin	5 (100)	1 (100)	2 (100)	7 (100)	3 (100)
Cefotaxime				3 (42.9)	0 (0)
Ceftazidime						0 (0)
Aztreonam				1 (14.3)	0 (0)	0 (0)
Imipenem				0 (0)	0 (0)	0 (0)
Gentamicin	3 (60)	1 (100)	2 (100)	4 (57.1)	2 (66.7)	2 (100)
Amikacin				1 (14.3)	1 (33.3)	2 (100)
Tetracycline	4 (80)	1 (100)	0 (0)
Erythromycin	4 (80)	1 (100)	2 (100)
Vancomycin	0 (0)	0 (0)	0(0)
Azithromycin	3 (60)	0 (0)
SMX/TMPb	4 (80)	1(100)
Ciprofloxacin				5 (71.4)	1 (33.3)	2 (100)

^aCoNS: Coagulase negative staphylococci.

^bSMX/TMP: Sulphamethoxazole-Trimethoprim.

Out of the 28 potential risk factors that were analyzed, intubation at birth (p = 0.021; OR = 3.10; CI, 1.18–8.14), prior blood transfusion (p = 0.037; OR = 3.34; CI, 1.08–10.36) and orogastric feeding (p = 0.018; OR = 2.95; CI, 1.20–7.22) were the only factors found to be statistically significant on univariate analyses (Table 4). On subsequent multivariate analysis, only intubation at birth (p = 0.046; OR = 2.74; CI, 1.02–7.37) and orogastric feeding (p = 0.029; OR = 2.74; CI, 1.11–6.78) retained significant association with the development of neonatal HAC (Table 5).

Table 4.

Risk factors for neonatal HAC

Characteristic	Neonates without HAC	Neonates with HAC	p-value	Odds ratio (95% CI)
	n = 567	n = 24
	Number (%)	Number (%)
Neonatal characteristics
Male Sex	313 (55.2)	14 (58.3)	0.763	1.14 (0.49–2.60)
Prematurity	356 (62.8)	17 (70.8)	0.426	1.44 (0.59–3.53)
Multiple births	82 (14.5)	7 (29.2)	0.055	2.44 (0.98–6.05)
Birth weight
ELBWa	14 (2.5)	1 (4.2)	0.609	1.72 (0.22–13.63)
VLBWb	120 (21.2)	9 (37.5)	0.064	2.24 (0.95–5.23)
LBWc	285 (50.3)	9 (37.5)	0.225	0.59 (0.26–1.38)
Meconium stained liquor	86 (15.2)	6 (25)	0.199	1.86 (0.72–4.83)
APGAR <4
1 min	64 (11.3)	1 (4.2)	0.297	0.34 (0.05–2.57)
5 min	8 (1.4)	1 (4.2)	0.304	3.04 (0.36–25.32)
Respiratory distress at birth	331 (58.4)	18 (75)	0.112	2.14 (0.84–5.47)
Maternal characteristics
Multiparity	340 (59.9)	15 (62.5)	0.804	1.11 (0.48–2.59)
Antepartum hemorrhage	45 (7.9)	3 (12.5)	0.427	1.66 (0.48–5.77)
Gestational hypertension	87 (15.3)	5 (20.8)	0.469	1.45 (0.53–3.99)
Hypothyroidism	14 (2.5)	0 (0)	0.971	0.00 (0.00–>1.0E12)
Premature rupture of membranes	40 (7.1)	1 (4.2)	0.590	0.57 (0.07–4.35)
Antenatal steroids	155 (27.3)	11 (45.8)	0.054	2.25 (0.99–5.13)
Interventional characteristics
Type of delivery
Assisted vaginal	15 (2.6)	1 (4.2)	0.656	1.60 (0.20–12.64)
Caesarian	187 (32.9)	7 (29.2)	0.697	0.84 (0.34–2.05)
Resuscitation at birth
Bag and mask ventilation	69 (12.2)	1 (4.2)	0.259	0.31 (0.04–2.36)
Intubation	55 (9.7)	6 (25)	0.021	3.10 (1.18–8.14)
Vaccination	538 (94.9)	22 (91.7)	0.493	0.59 (0.13–2.64)
Kangaroo mother care	86 (15.2)	5 (20.8)	0.454	1.47 (0.54–4.05)
Prior blood transfusion	32 (5.6)	4 (16.7)	0.037	3.34 (1.08–10.36)
Prior antibiotics	143 (25.2)	6 (25.0)	0.981	0.99 (0.38–2.54)
Prior H2 blockers	3 (0.5)	0 (0.0)	0.969	0.00 (0.00–>1.0E12)
Prior steroids	0	0
Prior surgery	0	0
Total parenteral nutrition	242 (42.7)	10 (41.7)	0.922	0.96 (0.42–2.20)
Orogastric feed	256 (45.1)	17 (70.8)	0.018	2.95 (1.20–7.22)
Peripheral venous catheter insertion	301 (53.1)	13 (54.2)	0.917	1.04 (0.46–2.37)
Central catheter insertion	17 (3.0)	2 (8.3)	0.166	2.94 (0.64–13.53)
Mechanical ventilation	22 (3.9)	2 (8.3)	0.291	2.25 (0.49–10.18)
Oxygen hood/prongs	187 (32.9)	11 (45.8)	0.196	1.72 (0.76–3.91)

^aELBW: extremely low birth weight (<1000 g).

^bVLBW: very low birth weight (1000–1500 g).

^cLBW: low birth weight (1500–2500 g).

p-value < 0.05 was considered as statistically significant.

Table 5.

Multivariate analysis of risk factors

Characteristic	p-value	Odds ratio (95% CI)
Intubation at birth	0.046	2.74 (1.02–7.37)
Orogastric feed	0.029	2.74 (1.11–6.78)
Prior blood transfusion	0.128	2.48 (0.77–7.96)

p-value < 0.05 was considered as statistically significant.

The average length of hospital stay for neonates with HAC was 20.8 days (range 3–52), which was significantly longer compared with those without HAC (mean 9.6 days, range 2–77; p < 0.001). There was no significant difference in the mortality rate between the two groups (p = 0.689). In our study, one neonate with HAC died during the hospital stay, which was attributed to sepsis from underlying acute respiratory distress syndrome and bacteremia.

DISCUSSION

As described previously, HAC is one of the commonest HAIs among the neonates that can lead to serious complications including blindness and death [14]. However, the currently published literature investigating the epidemiology of this disease entity is relatively scarce. This study was conducted to evaluate the incidence, etiological profile and the risk factors associated with the development of neonatal HAC.

The incidence of neonatal HAC in our study was estimated at 4.06 infections per 100 eligible admissions. A similar incidence rate was reported in two previously published studies, one from Portugal (4.1 per 100 admissions) and the other from New York (5.2 per 100 admissions) [2, 17]. As in our study, both of these studies also utilized the CDC criteria for the establishment of HAC diagnosis. Other studies from Spain, South Korea and Israel have reported a higher incidence rate of HAC, ranging from 10–15% [8, 11, 16]. In contrast, a 5-year study from Iran and a 3-year study from Italy documented a much lower HAC incidence rate of 0.2% and 1.4%, respectively [9, 23]. Such a wide variation in the reported incidence of HAC across studies is partly attributed to differences in case definition employed by different studies. As an example, the study from Iran required a positive culture for establishing the diagnosis of conjunctivitis, thereby resulting in an exceedingly low incidence rate of 0.2% [9]. In the absence of neonate-specific diagnostic guidelines, we utilized the CDC criteria for the diagnosis of neonatal HAC in our study. Although the CDC definitions are helpful and commonly used, there exist a few concerns regarding their applicability to neonates. As an example, the presence of clinical features such as pain or redness of conjunctiva/ periorbital area is one of the criteria for the diagnosis of HAC [18]. It is known that in neonatal bacterial conjunctivitis, conjunctiva typically remains white and rarely gets red or pink. Moreover, eye pain cannot be elicited in the neonates [2, 24]. The CDC criteria are unable to differentiate between infections that result from the passage through the maternal birth canal versus those truly related to neonatal hospitalization, and consider both as HAIs. As a result, none of the existing definitions can be perfectly applied to the neonates, and therefore standardized neonate-specific guidelines for HAC diagnosis are much needed [17].

Majority (70.8%) of the HAC episodes in our study occurred during the first week (day 3–day 7) of hospitalization with the median time from admission to the onset of conjunctivitis being 4.5 days (range 3–28). In contrast, a study from Israel reported only 20% of HAC episodes during the first 10 days of hospital stay, with the majority (43%) cases occurring during days 11–23 [16]. In another study, the median time from admission to onset of HAC was documented as 16 days (range, 3–118) [11]. The clustering of HAC episodes in our study during the early days of hospitalization could probably find its source in the maternal birth canal.

Of the 20 culture-positive HAC cases in our study, majority of the infections (60%) were caused by GNBs. This is consistent with the previously published data that also reports the predominance of GNBs in neonatal HAC [16, 17]. Furthermore, GNBs were the more frequently isolated microorganisms from HAC cases diagnosed during the first 7 days of hospitalization (71.4%), with a shift in trend towards GPBs for cases that were diagnosed after day 7 (66.7%). This trend is similar to that observed in a study from Israel where the distribution of neonatal HAC causing microorganisms shifted from gram-negative to gram-positive bacteria during the NICU stay [16]. Moreover, we observed that 8 of the 17 HAC cases (47.1%) diagnosed in the first week of hospital stay were caused by either E. coli or Klebsiella sp. microorganisms. The high frequency of these pathogens in our study is also similar to the Israel study [16]. The isolation of these microorganisms suggests that the early HAC infections are likely acquired during passage through the maternal birth canal. In fact, the frequency of neonatal conjunctivitis has been suggested to be a reflection of the sexually transmitted infections prevalent in the community [14].

P. aeruginosa is often associated with serious sequelae including corneal scarring and visual impairment [13]. Previous studies have found this organism to be a frequent isolate, accounting for 8–18% of infections, with a likely transmission from the hands of health care workers in NICU [2, 16, 17]. We found that only two cases of HAC were associated with P. aeruginosa among our neonates. Neisseria gonorrheae can be acquired from maternal birth canal [17], and can lead to neonatal conjunctivitis that presents with continuous, profuse purulent discharge within days after birth [24]. In the present era of effective broad-spectrum antibiotics and neonatal ophthalmic prophylaxis, the incidence of N. gonorrheae has gone down tremendously. It is, therefore, not surprising that this organism was not isolated from any of our conjunctivitis patients. A few studies have reported coagulase-negative staphylococci (CoNS) as a frequent pathogen accounting for up to 20–60% of the conjunctivitis cases [2, 8, 11, 16, 23]. This high percentage of CoNS suggests repeated handling of neonates by either health care professionals or their families without adequate hand washing. Only a single case of CoNS-related neonatal HAC was diagnosed in our study, likely due to strict implementation of hand washing protocols for the staff as well as the visitors in our neonatal unit. Other bacteria that have been previously reported to be associated with neonatal HAC but not found in our study are Serratia marscecens, Enterobacter cloacae, Enterobacter aerogenes, Citrobacter koseri, Haemophilus influenzae, Proteus mirabilis, Streptococcus pneumoniae, Streptococcus pyogenes and diphtheroids [2, 8, 17, 24]. A previously published study has reported 5.8% neonatal HAC cases to be associated with Candida sp. [16]. We, however, did not encounter any case of fungal conjunctivitis in our study. Although isolated outbreaks of viral HAC caused by adenoviruses have been previously reported [6, 7], viral etiology was also not established in any of our cases.

Both strains of P. aeruginosa showed similar resistance pattern, which suggests a common origin in the neonatal unit. The variable resistance pattern of other GNBs, i.e. E.coli and Klebsiella sp., supports the possible maternal origin of the strains. A previous study has reported a highly resistant pattern of the gram-positive cocci to erythromycin [16]. We also observed a similar pattern in our study with most gram-positive strains being resistant to erythromycin, which is a commonly used antibiotic in the prophylaxis and treatment of ophthalmia neonatorum.

Prematurity and low birth weight (LBW) increase the general vulnerability of the neonate to infections including HAC [2, 17]. We, however, did not find gestational age or birth weight to be significant risk factors associated with the occurrence of neonatal HAC. No significant association was observed between the gender of the neonate and HAC, unlike some previous reports [2, 17]. We found intubation at birth to have statistically significant association with neonatal HAC. Flow of respiratory secretions to the eyes through the nasolacrimal duct, especially during suctioning is the likely explanation [2]. Orogastric feeding was the only other statistically significant risk factor for the development of HAC in our study, probably by a similar mechanism. Other previously reported risk factors, which were not found to be significant in our study, include noninvasive ventilation and parenteral nutrition [2, 17]. Potential risk factors such as mechanical trauma to the conjunctiva due to eye patch during phototherapy and prior eye exam for retinopathy of prematurity were not analyzed in our study.

We found a significant increase in the mean hospital length of stay among neonates diagnosed with HAC (20.8 versus 9.6 days; p < 0.001). No significant increase in the mortality rate was noted among these patients. Eight out of the 20 neonates with HAC were found to have one or more additional episodes of sepsis during their stay in the hospital. The implicating microorganisms for conjunctivitis and sepsis were, however, different in all eight cases. The occurrence of sepsis and conjunctivitis was, therefore, not considered a part of the same pathologic process.

We believe that a strict adherence to universal precautions and hand sanitation guidelines in the neonatal unit was the major contributory factor resulting in a relatively low incidence rate of HAC in our study. Among the HAC cases, we found a very few isolates of P. aeruginosa and CoNS, which are microorganisms known to transmit through the hands of the handlers. Hand hygiene has been long identified as the single most effective preventive factor in the transmission of infections. In fact, World Health Organization (WHO) recognizes inappropriate hand hygienic measures as the leading cause of HAIs and the spread of multidrug-resistant organisms [25]. Strict hand hygiene protocols are, therefore, pivotal in the control of neonatal HAC. It is unfortunate that low hand hygiene compliance rates have been reported from across the world due a variety of causes including ignorance among healthcare providers and lack of institutional priority towards hand hygiene.

Our study is also associated with a few limitations. The neonates in our study were not followed post-discharge from the neonatal unit. As a result, some cases of conjunctivitis that might have been incubating at the time of discharge were probably missed. Also, the long-term sequelae of neonatal HAC could not be determined in our study for the same reason. It is often difficult to accurately distinguish between the HAC infections acquired during the stay in the neonatal unit and those acquired from the maternal birth canal. Consequently, some of the HAC cases in our study, especially the ones diagnosed early during hospital stay, might have been acquired in the maternal birth canal leading to an overestimation of the real HAC incidence. Lastly, conjunctivae of hospitalized neonates are frequently colonized by microorganisms [3], and it is sometimes difficult to differentiate between colonization and true infection [16]. Therefore, some of the HAC cases in our study might in fact represent a mere conjunctival colonization rather than a true infection.

CONCLUSION

In this one-year prospective cohort study, the incidence of neonatal HAC was estimated at 4.06 infections per 100 eligible admissions. The microbiological profile and the anti-microbial sensitivity patterns were characterized in detail for the established cases of HAC. A significant association was observed between conjunctivitis and some common interventions in the neonatal units, such as intubation and orogastric feeding, which suggest that extra eye care is warranted especially in the presence of these local insults. A strict adherence to hand sanitation guidelines is likely effective in preventing neonatal HAC. Further studies to specifically test this hypothesis are needed and should be strongly encouraged. Lastly, standardized case definition and awareness of the possible serious outcomes of neonatal conjunctivitis will aid in better surveillance.

FUNDING

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.