*Corresponding author
Hospital del Niño DIF Hidalgo.
Nosocomial neonatal sepsis (NNS) is one of the principal causes of childhood morbidity and mortality and has a large impact on public health. Therefore, our objective was to determine the prevalence and risk factors associated with NNS in patients at Hospital del Niño DIF Hidalgo (Children’s Hospital, Hidalgo DIF). A retrospective case-control study was carried out by reviewing the medical records of 29 cases and 62 controls. The relative risk and odds ratio were calculated, and the Chi-squared test was used. The prevalence of NNS was found to be 13.6%. The statistically significant risk factors for NNS in premature infants were the male gender, the use of histamine H2 receptor blockers, a birth weight <1500 g, and an Apgar score <6. In full-term patients, the risk factors were respiratory insufficiency and anemia. The common risk factors in both groups were ventilatory assistance, presence of a central line, total parenteral nutrition, >7 days of antibiotic use, and presence of a bladder catheter. We found that the prevalence and risk factors in this study are similar to those reported in medical literature.
Keywords: Newborn; Nosocomial Sepsis; Risk Factors.
Nosocomial neonatal sepsis (NNS) is an infectious disorder that is characterized in newborns (NBs) by a systemic inflammatory response syndrome that occurs 3 days following admission to a hospital unit.1-5 It should be noted that admission was due to a different cause, and the germs present on personnel or furniture within the hospital infected the patient during the treatment of the patient by the personnel or performance of invasive procedures.1-7 NNS is a severe disease, and its natural evolution leads to multi-organ failure and death.5-9 It is the principal cause of mortality in neonatal intensive care units,1-8 and its presence reflects the quality of attention paid by the hospital unit in which the infant is admitted. A diagnosis of NNS is suspected based on epidemiologic data, known risk factors, and clinical data for systemic inflammatory response; these markers are corroborated through laboratory studies.4-13 Thus, the presence of a local focus of infection is identified and/or pathogenic germs of nosocomial origin are isolated from blood, urine, or body fluids.1-12 The control of NNS includes the treatment of the primary pathology, general measures, infection control, and intensive supportive therapy. Neonatal risk factors exist that are inherent to the disease, as do perinatal factors, which are difficult to control and require inter-institutional changes in conduct for their abatement. The following risk factors have been identified: immunologic state, low birth weight, perinatal hypoxia, use of antibiotics, use of corticosteroids, parenteral feeding, use of permanent venous catheters, prolonged mechanical ventilation, male gender, invasive and/or surgical procedures, and others.1-15 Thus, nosocomial risk factors are the most statistically likely to lead to the development of NNS, and close care and control of preventive measures already in use must be maintained. This evidence suggests that understanding the impact and risk factors for NNS is useful. The presence of risk factors should lead to an early diagnosis of NNS to avoid the worsening of the disease or the development of complications and to decrease, to a certain degree, of its impact on public health, operating costs, and quality of life.11-15
An observational, descriptive, cross-sectional, analytical, retrospective case-control study was carried out on male and female neonatal patients between 0 and 28 days old in the neonatology unit during the year 2008. Patients were included if they developed NNS that fit the criteria published in 2005 and of Manroe et al. 16-18 3 days following admission. Patients hospitalized during the same period who did not develop NNS and who remained hospitalized more than 3 days were included as controls. Patients who did not meet the criteria for NNS diagnosis or who were admitted with this pathology acquired at another institution were excluded.
Data for each patient were collected, such as age at admission, gender, length of hospital stay (HS), age at discharge, diagnosis at admission and discharge, date of admission and discharge, and total length of stay. The following clinical and para-clinical data were used to discriminate the diagnosis of NNS: age at diagnosis, length of stay to diagnosis, presence of tachycardia or bradycardia, tachypnea, hypotension, body temperature, leukocyte and platelet counts, presence of bandemia, total bands and ratio of immature neutrophils, glycemic level, C-reactive protein level, erythrocyte sedimentation rate, presence of infectious focus, and the results from cultures (blood, urine, intravascular device tips, body fluids, stool, and cerebrospinal fluid).
The NNS risk factors studied were divided into 4 categories:
Neonatal: birth weight, presence of congenital anomalies, anemia, respiratory insufficiency, gastrointestinal alterations, gestational age at birth in weeks, hemodynamic disorders, and jaundice.
Maternal: maternal age, infections (vaginal, chorioamnionitis, urinary tract infection), previous delivery with infection, pre-eclampsia, gestation number, premature rupture of membranes (>12 hours), prolonged rupture of membranes (>18 hours), maternal group B Streptococcus, diabetic mother, and type of birth.
Peripartum: perinatal asphyxia, meconium, Apgar score <6, O2 requirement, neonatal hypotension, medication use, blood loss, fetal tachycardia, and obstetric trauma.
Nosocomial: H2 blocker use, prophylactic antibiotic use, methylxanthine use, parenteral nutrition and lipid levels, intubation and ventilation, pleural drainage, bladder catheterization, endoscopy, surgery, peripheral venous catheter, fasting, corticosteroid use, >7 days of antibiotic use, orogastric tube, ventriculo-peritoneal shunt, and artificial formula.
A database was made in Excel using the obtained data. A descriptive statistical analysis was carried out using distribution tables for frequencies and percentages. Similarly, the means and standard deviations were determined. To analyze the risk factors for NNS, contingency tables were made with the control group (patients exposed to the different risk factors who did not develop NNS) and the case group (patients exposed to the different risk factors who had NNS). Each risk factor was analyzed in both groups with the statistical program Sigma Stat version 2.0. Odds-ratios and relative risks were obtained, and the Pearson Chi-squared test was used. In all analyses, p <0.05 was considered to be statistically significant.
Cases
A total of 213 hospitalized patients were identified in the neonatology clinic, 29 of which developed NNS, corresponding to a prevalence of 13.6 per 100 hospital discharges. Of these, 10 (34.5%) were premature and 19 (65.5%) were full-term; 17 (58.6%) were male and 12 (41.4%) female. The ages of the patients at admission ranged from 0 to 27 days of life, with a mean ± standard deviation of 4.9 ± 1.4 days, and from 29 to 40 weeks gestational age (WGA), with a mean of 36.6 ± 0.6. Patient weight ranged from 800 g to 3850 g, with an average of 2399.7 ± 158.8 g. More than one pathology per patient was diagnosed at admission (Table 1). Prior to diagnosis, the average HS was 9.6 days (range 4-23) for premature patients and 22 days (range 11-82) for full-term patients.
Controls
There were 62 controls, of which 21 (33.8%) were premature and 41 (66.1%) full-term; 35 (56.5%) were male and 27 (43.5%) female. Their ages at admission ranged from 0 to 27 days of life (mean 7.1) and from 29 to 42 WGA (mean 37.1). Patient weight ranged from 1700 g to 4100 g (mean 2745.2 g). More than one pathology per patient was diagnosed at admission (Table 2). The average HS was 12.2 days (4-33 days) and 8.9 days (4-29 days) for premature and full-term NBs, respectively.
Diagnosis of NNS
In this study, the average age at the diagnosis of NNS was 14.6 ± 7.9 days, and the average HS was 9.7 days (4-25 days) at the time of the first manifestations of NNS. Of these patients, 7 (24.1%) presented with NNS at HS day 4, 5 (17.2%) presented with NNS between HS days 4 and 7, and 17 (58.6%) presented with NNS after HS day 7. Infectious foci were found in 22 (75.9%) of the cases; the other 7 (24.1%) did not have any. The most common focus location was the airway, (12 cases; 41.4%), followed by the intestine (6 cases; 20.7%), urinary tract (1 case; 3.4%), a surgical wound (1 case; 3.4%), a central line insertion site (1 case; 3.4%), and an ophthalmic focus (1 case; 3.4%).
Table 1. Diagnosis at admission/NB (cases).
Table 2. Diagnosis at admission/NB premature and full-term (controls).
Patients diagnosed with NNS most frequently presented with tachycardia (22 cases; 75.9%); none presented with bradycardia. Tachypnea and dysthermia were the second most common (21 cases; 72.4% each). Ten (47.6%) patients were found to present with hyperthermia above 38.5º C, and 11 (52.4%) presented with hypothermia below 36º C. Eight patients did not meet the criteria for dysthermia (<36 and >38.5º C), 6 patients had temperature ranges from 36.4 to 38º C with an average of 37.7º C, and the other 2 cases each had variations during the day between 36 and 38º C.
In all cases, systematic and continuous routines were conducted to culture urine, peripheral blood, and/or central line blood. Urine culture was positive for infection in 1 patient (3.4%). Peripheral blood culture was positive in 12 cases (38.7%), and central line blood was positive in 16 (51.6%) cases. Stool and bronchial secretion cultures were each positive in 1 case (3.4%). Staphylococcus was isolated in 16 cases (55.2%); S. epidermidis was the most frequent species isolated (12 cases; 41.1%). Other isolated species were S. aureus (2 cases; 6.9%), S. hominis, and S. haemoliticus (1 case each; 3.4%). Enterobacter cloacae was isolated in 3 cases (10.3%). In addition, the genus Klebsiella was isolated from 3 cases (10.3%), 2 of which (6.9%) were the species pneumoniae and 1 (3.4%) oxitoca. E. coli was isolated in 2 cases (6.9%). Citrobacter freundii, Pseudomonas aeruginosa, and Enterococcus faecalis were each isolated in 1 case (3.4%). Candida was isolated in only 1 case (3.4%). Interestingly, in 8 cases (27.6%), no microorganism was isolated, despite the severity of the sicknesses. Finally, 23 cases (79.3% with hypotension) presented with septic shock that required treatment with vasopressors.
A. Premature patients
The following data regarding neonatal risk factors for NNS were found: in the group of premature patients, the first risk factor with statistical significance (SS) was male gender and low birth weight (<1500 g) (Table 3). Gestational age at birth, congenital anomalies, hemodynamic disorders, anemia, gastrointestinal alterations, jaundice, and respiratory insufficiency did not reach SS (p >0.05).
B. Full-term patients
In these patients, respiratory insufficiency and anemia were the principal risk factors with SS (p <0.05) (Table 4). In contrast, none of the other recorded factors in this category reached SS (p >0.05).
A. Premature patients
No SS relationship was found between NNS and the following factors: premature rupture of membranes, labor >15 hours, Caesarean section, maternal genitourinary infections, pre-eclampsia, gestation number, and maternal diabetes (p >0.05).
B. Full-term patients
None of the maternal risk factors reached SS for the development of NNS in these patients (p >0.05).
A. Premature patients
In these patients, an Apgar score <6 was the only factor that reached SS (p = 0.027). Other factors, such as perinatal asphyxia, hypotension, blood loss, fetal tachycardia, use of medications, meconium, and O2 requirement, did not show an association with the development of NNS (p >0.05).
B. Full-term patients
None of the peripartum factors were risk factors for NNS (p >0.05).
A. Premature patients
The nosocomial factors for the development of NNS with SS were ventilatory assistance, use of histamine H2 receptor blockers, central venous line, use of lipids and parenteral nutrition, use of antibiotics >7 days, and bladder catheterization (Table 3; p <0.05). Associated factors without SS were use of prophylactic antibiotics, use of corticosteroids, use of methylxanthines, pleural drainage, use of peripheral catheter, use of orogastric tube, fasting, use of artificial formula, and surgery (p >0.05).
B. Full-term patients
The nosocomial factors for the development of NNS with SS were use of central line, ventilatory assistance, use of lipids and parenteral nutrition, placement of bladder catheter, and use of antibiotics >7 days (Table 4, p <0.05). The other nosocomial factors did not reach SS (p >0.05).
Table 3. Risk factors for NNS that reached SS. Premature patients: 10 patients with sepsis and 21 patients without sepsis.
Table 4. Risk factors for NNS that reached SS. Full-term patients: 19 patients with sepsis and 41 patients without sepsis.
As a result of studies performed in 1996 in secondary and tertiary care hospitals, it is assumed that the percentage of nosocomial infection in Mexico is 10-15%. We found the prevalence of NNS to be 13.6 per 100 hospital discharges, which is in agreement with the expected global and national range (Colombia, 8.4%; Mexico, 19 to 40%; Durango, 30.54%; “November 20 CMN,” 2.1%; Culiacan, 7.5%).19-22
Of the statistically significant neonatal risk factors for the development of NNS in our study, male gender was the principal risk factor in premature infants. As such, we found a prevalence of NNS of 90% in male infants, which is much higher than the level of 54% reported in another secondary level hospital similar to ours (Zone 1A, Los Venados General Hospital of the Mexican Social Security Institute). 8 Those authors proposed that there is a susceptibility factor involved in thymus function and synthesis of immunoglobulins encoded by a gene on the X chromosome; because females have two X chromosomes, they have a greater resistance to infection.8 Compared with controls, for each single healthy male patient, there are 12.4 patients with NNS, so there is a 6.4-fold greater risk to develop NNS as a result of being male. The second most important neonatal risk factor for NNS in the premature patient group was low birth weight ( < 1500 g, described in various publications as the principal risk factor for the development of nosocomial neonatal sepsis). 2,10, 19, 23, 24 The influence of low birth weight is due to immunologic immaturity, the greater use of invasive technology, and the greater length of hospital stay during which enough weight should be gained for discharge.10 In the present study, we found 6 (60.6%) premature patients weighing < 1500 g. This percentage is similar to that reported in literature: “71.5% of nosocomial infections were present in neonates weighing less than 1500 grams.25” The survival of premature newborns with a very low birth weight complicated by late sepsis continues to be a challenge because it increases the duration of hospitalization, cost, and mortality rate. 26
In the full-term newborns, the principal neonatal risk factor for NNS was respiratory insufficiency, which necessitates ventilatory support. In general, respiratory insufficiency complicates up to 17% of neonatal sepsis.27 Respiratory insufficiency is the final pathway of many pathologies, and it increases energy and oxygen requirements in the tissues; this increase in metabolism leads to an increased production of deleterious substances.28 This same process causes a depression of the bone marrow, thereby compromising cellular immunity. The second and final neonatal risk factor for NNS in full-term newborns is anemia. Independently of its dependence on the socioeconomic and cultural conditions of the pregnant mother, it has been demonstrated that gestational anemia can predispose the child for anemia until the fourth year of life. Iron deficiency decreases cellular immunity and consequently increases susceptibility to infections, particularly in the respiratory system. Five of the 10 cases (50%) with anemia (94.7% of the full-term group) presented with respiratory infectious foci.29
In our study, the maternal factors for NNS were not statistically significant risk factors and were only found to be associated with prematurity. Prolonged rupture of membranes and labor longer than 15 hours both conditioned for an increased risk of colonization and hypoxia with repercussions on an immunological level, which added to the immaturity of the immune system.1,2,8,9 Undoubtedly, these products would be obtained through Caesarean section, and it has been demonstrated that newborns delivered by Caesarean section have fewer intestinal lactobacilli than vaginally delivered newborns, thus affecting bacterial colonization.30
The only peripartum risk factor for NNS was an Apgar score of less than 6, and this only presented in the premature patient group. There are studies that have reported values below 5.31 An Apgar score below 6 is indicative of moderate depression; although this score does not normally translate to criteria for neonatal asphyxia, minimal hypoxia is capable of producing mitochondrial changes and liberation of deleterious substances, such as nitric oxide. An oxygen supply is given to these patients, which then increases the production of free oxygen radicals. 27
Various studies have demonstrated the great importance of nosocomial risk factors for NNS.2,8,10,14 In our study, ventilatory assistance was required for 9 (90%) preterm patients and 12 (41.1%) from the full-term group. This factor was the principal nosocomial factor for the premature group, producing 8.4 fold more risk for NNS, and a secondary factor in the full-term group, producing 3.4 fold more risk for NNS. The lung has been described as a primary site of infection related to use of mechanical ventilation during prolonged endotracheal intubation.8 Of the cases that required mechanical ventilation, 2 (20%) of the premature and 6 (50%) of the full-term cases presented with a respiratory infectious foci. In 7 patients (24.1%), no infectious foci were reported. However, a primary respiratory focus was not dismissed because bronchial secretion cultures were not systematically performed, with the exception of a case in which P. aeruginosa was isolated. These cases are similar to the report of ventilator-associated pneumonia, which is a common and severe complication in premature infants of low weight; during with the development of P. aeruginosa and Acinetobacter predominate. Similarly, the duration of mechanical ventilation is important because it is related to the severity of the underlying illness and prolonged invasive procedures in the NICU.32 Ranitidine, a histamine H2 receptor blocker used in 100% of the premature infants in our study, was the second highest nosocomial risk factor for NNS in this group. Previously, an association of ranitidine with the development of necrotizing enterocolitis was reported, as it provokes bacterial translocation by alkalinizing the intestinal pH and favoring the growth of pathogenic microbes. 33,34 Although only 3 (30%) premature infants presented with intestinal infectious foci, it cannot be assumed that they were not present in others or that the intestine was a route of entry for pathogens. Moreover, H2 blockers are agonistic to mechanical ventilation and other factors in the development of NNS. 32 In premature infants, the use of a central line as a nosocomial factor for NNS in our study was the third most common risk factor, increasing the occurrence of NNS up to 4.9 fold in exposed patients and 5.7 fold in full-term newborns; in the latter, it is the principal nosocomial factor. Total parenteral nutrition (TPN) and lipids is the fourth most common nosocomial risk factor for NNS in premature infants and the second most common in full-term infants, with an increased risk for acquiring NNS of 3.7 and 3.4 fold, respectively. Sepsis as a complication of TPN is principally associated with the use of central lines for administration. Thus, 100% of the premature infants (8 of the 10 cases) and 75% of full-term newborns (12 of 16 cases) required TPN with central lines. It is also known that a birth weight below 1500 g, use of a central line, and TPN are significantly associated with the development of NNS.35,36 In 5 cases in the premature group, we found the joint presence of these 3 factors in the development of NNS. On the subject of pharmacotherapy, our study found that the prolonged use of antibiotics (>7 days) was associated with an increased risk of NNS of 3.5 and 3.3 fold in premature and full-term newborns, respectively. Of the 11 (37.9%) cases treated with antibiotics for more than 7 days, 8 had central lines, corresponding to an association of 11.3/1000 catheter-days. Antibiotics compromise the immune response of the host, and NNS in more common in the patients with central lines.37, 38
Finally, bladder catheterization was shown to be an important risk factor for NNS, with an increased risk of 3.5 and 2.3 fold in premature and full-term newborns, respectively. The placement of urinary catheters used to collect urine samples and/or quantify the urinary flow is not always beneficial for patients and can injure the epithelium of the urinary tract and allow the further passage of germs into the bloodstream.4
In the Hospital del Niño DIF Hidalgo, the prevalence of NNS was found to be 13.6%, which is in the expected range for a secondary-level hospital. The statistically significant risk factors for NNS included 4 factors exclusive to the premature group (male gender, use of H2 blockers, birth weight <1500 g, and Apgar score <6); 2 factors exclusive to the group of full-term newborns (respiratory insufficiency and anemia); and 5 risk factors common to both groups (ventilatory assistance, use of central line, total parenteral nutrition, antibiotic use >7 days, and bladder catheterization). The most important neonatal risk factor for premature newborns is male gender, which was higher than that reported in the literature, and NSS had a direct relationship with a birth weight <1500 g. The principal neonatal risk factors for NNS in full-term patients were respiratory insufficiency and anemia. The principal perinatal risk factors of an Apgar score <6 (premature) and respiratory insufficiency (full-term) are in accordance with the literature.
Nosocomial risk factors had greater weight than the rest of the factors. A conjunction of risk factors exists for the development of NNS. Neonatal risk factors are inherent to patients and are not modifiable. To decrease the perinatal factors, conventions are recommended for private and governmental health care institutions to provide better neonatal resuscitation. For nosocomial factors, these measures include avoiding the indiscriminant use of H2 blockers except in cases in which their use is justified, maintaining a constant follow-up in surveillance programs, and maintaining prevention measures. These measures all support the goal of further decreasing the incidence of the hospital complication of NNS.