IM Antibiotics for Secondary Prevention of Neonatal Sepsis
IM Antibiotics for Secondary Prevention of Neonatal Sepsis
Evaluation and treatment approaches for the secondary prevention of EOS have been controversial. Current clinical guidelines do not provide clear-cut recommendations regarding the optimal method of antibiotic administration in the treatment of an asymptomatic infant at risk for, or with, suspected sepsis. The issue of the need for a complete versus limited sepsis evaluation before initiating antibiotic therapy has been a topic of debate over many years. In 2010, the CDC published an updated algorithm for the secondary prevention of group B streptococcal disease in neonates recommending a limited evaluation for the asymptomatic infant at risk for EOS. Current recommendations continue to advise clinicians to initiate antibiotic therapy while awaiting laboratory data confirming infection. However, this too remains an area of controversy.
EOS often presents within the first hours of life. Most infants with EOS exhibit symptoms within the first 24 to 48 h, with 90% of infants symptomatic by 24 h of age. In addition, clinical signs of sepsis have been found to be a more sensitive indicator of neonatal sepsis than hematological tests. Although cultures are known to be the gold standard for detection of bacterial infection, there continues to be concern about the validity of body fluid cultures after maternal pretreatment with antibiotics. Consequently, clinicians must rely on other clinical parameters for evidence of infection in infants with suspected sepsis. Although hematological indices such as elevated CRP and elevated I:T ratios have been shown to have good negative predictive value, their specificity is poor, and therefore false positives are high. Although maternal GBS colonization might increase clinical suspicion for early onset GBS disease in a symptomatic infant, in the era of universal screening, more than 60% of early onset infection caused by GBS occurred among infants born to women who had a negative prenatal GBS culture screen. Thus, clinical signs of sepsis in any newborn can be an indication of early onset GBS disease, regardless of maternal colonization status. An additional confounding factor relates to the potential over-diagnosis of chorioamnionitis, when a pregnant woman develops intrapartum fever. Approximately 1.6% of all pregnancies and 10 to 30% of nulliparous women who have received epidural analgesia will develop intrapartum fever. This may lead to unnecessary diagnostic evaluations and unwanted exposure to empirical treatment with antibiotics in the neonate.
Thus, it is important to identify effective strategies for the management of well-appearing, asymptomatic neonates undergoing evaluation and treatment for suspected EOS. In our cohort study, the majority of infants (91.7%) who were considered at risk for sepsis underwent a limited evaluation and were observed clinically and not treated with antibiotics. For the past two decades, our approach to the secondary prevention of EOS in well-appearing, asymptomatic infants has included a limited sepsis evaluation in conjunction with IM administration of antibiotics. Retrospective analyses of data in our large cohort of 5045 infants who underwent EOS evaluation revealed that only seven neonates became symptomatic, primarily exhibiting respiratory symptomatology that was readily identified by the mother–baby nurses caring for the babies. There were no delays in treatment of infants who developed symptomatology. Although six neonates required transfer to the NICU for IV antibiotic therapy because of positive cultures, they were clinically asymptomatic; the organisms identified were considered to be possible contaminants. None of the organisms cultured were the typical agents associated with neonatal EOS, nor were the clinical courses of the infants consistent with neonatal septicemia or meningitis. Significantly, no infant exhibited hemodynamic instability or required respiratory support. We also examined maternal risk factors among the neonates who were treated with IM antibiotics for the secondary prevention of EOS and those infants who were deemed treatment protocol failures. There were no significant differences in risk factors (Table 2) among infants considered protocol successes in comparison with the protocol failures. The small sample size of infants who were considered protocol failures may preclude more complete analysis of differences between these patient groups.
Concern exists regarding the safety of this treatment approach in infants considered partially treated as their mothers received intrapartum antibiotic therapy for presumed chorioamnionitis. In all, 9 of the 14 infants in the 'inadequate IM antibiotic protocol' (protocol failures) group were born to mothers diagnosed with chorioamnionitis and treated with intrapartum antibiotics. None of these infants exhibited signs of sepsis or meningitis; all remained healthy on follow-up contact. Furthermore, of the 407 infants who remained in the NBN (at risk, asymptomatic and culture-negative), 71% (n = 288) of their mothers received intrapartum antibiotics for a diagnosis of chorioamnionitis with or without GBS colonization.
Data on post-discharge rehospitalization within the first 14 days of life did not reveal any long-term morbidity/mortality associated with this treatment approach. There was only one infant rehospitalized within 2 weeks after NBN discharge among the 5045 infants in the study cohort. It was determined that the bacterial growth observed in the CSF culture of this infant was a contaminant. Thus, this infant was not an actual protocol failure. This adds further support to the efficacy and safety of this treatment approach. None of the infants treated with IM antibiotics developed complications such as cellulitis or abscess at the injection sites. At the 2-week interval follow-up visit none of the infants in the study cohort had manifestations of bacterial infection. Escobar et al. reported on a large population-based study of 18 299 newborns with birth weights >2000 g who were without major congenital anomalies. The rehospitalization rate within 7 days post-nursery discharge among neonates who underwent sepsis evaluation was 2.4%. Noting the low rate of positive blood cultures among their study population, Escobar et al. have suggested that treatment of infants at risk for EOS may not be justified.
There is limited information available on the drug pharmacokinetics and dosage requirements for infants who are treated with IM gentamicin, as most of the available data relate to neonates given IV. It has been shown that infants, especially neonates, have a higher volume of distribution per kg body weight and a lower clearance compared with older children and adults. Theoretically, there may also be delayed absorption after IM antibiotic administration caused by poor muscle perfusion and decreased bioavailability during the first week of life. This can lead to variability in peak drug concentrations. However, Hayani et al. found no differences in serum concentrations or pharmacokinetic parameters between two groups of neonates treated twice daily with IM or IV gentamicin therapy for presumed sepsis. Paisley et al. found that peak gentamicin concentrations occurred 15 to 30 min after IM injection in newborn infants. Furthermore, this study determined that the mean half-life of an IM dose of gentamicin was 4.0 ± 2.1 h. No significant difference was observed in the 1-h post-treatment kinetics of IM or IV gentamicin administered to neonates treated for suspected sepsis. Ampicillin has been shown to have effective systemic absorption following IM administration.
A potential criticism of this approach relates to the pain associated with IM injections. However, IV catheter insertion is also associated with pain. The small, fragile veins of newborn infants often lead to multiple attempts to establish IV access. Comfort measures such as sucking on a pacifier, gentle containment and breastfeeding can be utilized to lessen infant discomfort. In addition, topical local anesthetics can be applied before IM antibiotic injection. However, the depth of penetration is variable and increases with increasing application time up to ~2 to 3 mm after 1 to 2 h, and 6 mm after 3 to 4 h. Furthermore, the presence of an IV catheter may be perceived by staff and the parents as a barrier to breastfeeding and imposes a separation between the mother and infant, as most nurseries don't permit newborns with an IV line or a medication lock to room-in. Even among family-centered neonatal units that provide parents with continuous access to their babies, maternal–infant bonding may potentially be impeded. Another factor to consider is the medicalization of neonates treated in a special or NICU. Babies in such units are frequently viewed as at greater risk and often subject to unnecessary lab tests, resulting in unnecessary interventions; there is a common tendency (everywhere) to 'do more' and 'get more tests'.
Current management practices for the treatment of well-appearing, asymptomatic infants with suspected sepsis may lead to prolonged antibiotic use and extended hospital stays and excess hospital costs. Even among physicians who treat these infants in a NICU or special care nursery for only 48 to 72 h while awaiting blood culture and other laboratory test results, the costs of care are higher than costs associated with treatment of infants in the regular NBN. In comparison with routine NBN care, NICU charges are 7- to 10-fold higher (http://marchofdimes.com/peristats/pdfdocs/cts/ThomsonAnalysis2008_SummaryDocument_final121208.pdf). Practice variations are additional important contributors to the costs of care. This includes variations in the application of guideline criteria that identify infants requiring evaluation for suspected infection, as well as wide variation in laboratory testing recommendations and treatment protocols.
The Centers for Medicare and Medicaid Services reports that 2 of the top 20 patient diagnosis-related groups that account for hospitalizations in the US are neonates with major problems and neonates with minor problems. In 2009, 29% of newborn infants discharged from the hospital had a clinical problem or complicating diagnosis at a cost of nearly $7.5 billion. Total charges during this period for complicated newborn stays averaged $13 600 compared with $1700 for an uncomplicated newborn stay. In 2000, 0.3% of neonates were treated for infections in the perinatal period. In the ensuing decade health care costs have risen steadily. In contrast, since 2003 despite relative stability in the case mix index (the weight given to a medical condition to indicate its severity or complexity) applied to these diagnosis-related groups, and non-significant changes in the mean and median lengths of hospital stay for neonates with major problems or other significant problems, there has been a 1.4-fold increase in both the mean and median hospital charges. In our own institution, current hospital bed charges for the NICU for intensive/intermediate care are $10 850.00 per day. In contrast, the daily bed charge for a bassinet in the NBN is $1735.00. Although physician services for neonatal care are bundled, hospital charges are currently still reimbursed on a per diem basis. On an annual basis, we have ~60 asymptomatic, well-appearing infants at risk for EOS who are treated in the NBN with IM antibiotics. If all these infants were admitted to the NICU for evaluation and therapy, costs for care would be in the range of $652 550 annually. In contrast, treating these infants in the NBN is associated with a projected cost of $104 343 per year, a 6.25-fold reduction in hospital costs. As hospital payments shift more toward reimbursement for an episode of care, the high cost of providing certain types of care in the NICU, such as parenteral antibiotics to asymptomatic infants at risk for, or with suspected, sepsis will likely be scrutinized more closely. At the present time, many insurers determine hospital payments on the basis of the level of care required by the infant using acuity-based criteria rather than site where the care is provided. Consequently, hospitals may not receive payment for the higher care costs associated with admission of otherwise healthy infants with minimal symptomatology to the NICU for parenteral antibiotic therapy.
In summary, in a large retrospective cohort study of 5045 newborn infants evaluated for suspected EOS, 96.7% of the 421 well-appearing, asymptomatic neonates with suspected EOS treated with IM antibiotics in the regular NBN were discharged home without complication. Among the 3.33% of babies treated with IM antibiotics, seven infants required transfer to the NICU for intensive care because of the development of clinical symptomatology. The other infants required IV therapy because of positive cultures. There were no deaths or complications related to IM antibiotic treatment. IM antibiotic therapy administered in the regular NBN is a means to achieving significant cost savings to hospitals, insurers and families.
Discussion
Evaluation and treatment approaches for the secondary prevention of EOS have been controversial. Current clinical guidelines do not provide clear-cut recommendations regarding the optimal method of antibiotic administration in the treatment of an asymptomatic infant at risk for, or with, suspected sepsis. The issue of the need for a complete versus limited sepsis evaluation before initiating antibiotic therapy has been a topic of debate over many years. In 2010, the CDC published an updated algorithm for the secondary prevention of group B streptococcal disease in neonates recommending a limited evaluation for the asymptomatic infant at risk for EOS. Current recommendations continue to advise clinicians to initiate antibiotic therapy while awaiting laboratory data confirming infection. However, this too remains an area of controversy.
EOS often presents within the first hours of life. Most infants with EOS exhibit symptoms within the first 24 to 48 h, with 90% of infants symptomatic by 24 h of age. In addition, clinical signs of sepsis have been found to be a more sensitive indicator of neonatal sepsis than hematological tests. Although cultures are known to be the gold standard for detection of bacterial infection, there continues to be concern about the validity of body fluid cultures after maternal pretreatment with antibiotics. Consequently, clinicians must rely on other clinical parameters for evidence of infection in infants with suspected sepsis. Although hematological indices such as elevated CRP and elevated I:T ratios have been shown to have good negative predictive value, their specificity is poor, and therefore false positives are high. Although maternal GBS colonization might increase clinical suspicion for early onset GBS disease in a symptomatic infant, in the era of universal screening, more than 60% of early onset infection caused by GBS occurred among infants born to women who had a negative prenatal GBS culture screen. Thus, clinical signs of sepsis in any newborn can be an indication of early onset GBS disease, regardless of maternal colonization status. An additional confounding factor relates to the potential over-diagnosis of chorioamnionitis, when a pregnant woman develops intrapartum fever. Approximately 1.6% of all pregnancies and 10 to 30% of nulliparous women who have received epidural analgesia will develop intrapartum fever. This may lead to unnecessary diagnostic evaluations and unwanted exposure to empirical treatment with antibiotics in the neonate.
Thus, it is important to identify effective strategies for the management of well-appearing, asymptomatic neonates undergoing evaluation and treatment for suspected EOS. In our cohort study, the majority of infants (91.7%) who were considered at risk for sepsis underwent a limited evaluation and were observed clinically and not treated with antibiotics. For the past two decades, our approach to the secondary prevention of EOS in well-appearing, asymptomatic infants has included a limited sepsis evaluation in conjunction with IM administration of antibiotics. Retrospective analyses of data in our large cohort of 5045 infants who underwent EOS evaluation revealed that only seven neonates became symptomatic, primarily exhibiting respiratory symptomatology that was readily identified by the mother–baby nurses caring for the babies. There were no delays in treatment of infants who developed symptomatology. Although six neonates required transfer to the NICU for IV antibiotic therapy because of positive cultures, they were clinically asymptomatic; the organisms identified were considered to be possible contaminants. None of the organisms cultured were the typical agents associated with neonatal EOS, nor were the clinical courses of the infants consistent with neonatal septicemia or meningitis. Significantly, no infant exhibited hemodynamic instability or required respiratory support. We also examined maternal risk factors among the neonates who were treated with IM antibiotics for the secondary prevention of EOS and those infants who were deemed treatment protocol failures. There were no significant differences in risk factors (Table 2) among infants considered protocol successes in comparison with the protocol failures. The small sample size of infants who were considered protocol failures may preclude more complete analysis of differences between these patient groups.
Concern exists regarding the safety of this treatment approach in infants considered partially treated as their mothers received intrapartum antibiotic therapy for presumed chorioamnionitis. In all, 9 of the 14 infants in the 'inadequate IM antibiotic protocol' (protocol failures) group were born to mothers diagnosed with chorioamnionitis and treated with intrapartum antibiotics. None of these infants exhibited signs of sepsis or meningitis; all remained healthy on follow-up contact. Furthermore, of the 407 infants who remained in the NBN (at risk, asymptomatic and culture-negative), 71% (n = 288) of their mothers received intrapartum antibiotics for a diagnosis of chorioamnionitis with or without GBS colonization.
Data on post-discharge rehospitalization within the first 14 days of life did not reveal any long-term morbidity/mortality associated with this treatment approach. There was only one infant rehospitalized within 2 weeks after NBN discharge among the 5045 infants in the study cohort. It was determined that the bacterial growth observed in the CSF culture of this infant was a contaminant. Thus, this infant was not an actual protocol failure. This adds further support to the efficacy and safety of this treatment approach. None of the infants treated with IM antibiotics developed complications such as cellulitis or abscess at the injection sites. At the 2-week interval follow-up visit none of the infants in the study cohort had manifestations of bacterial infection. Escobar et al. reported on a large population-based study of 18 299 newborns with birth weights >2000 g who were without major congenital anomalies. The rehospitalization rate within 7 days post-nursery discharge among neonates who underwent sepsis evaluation was 2.4%. Noting the low rate of positive blood cultures among their study population, Escobar et al. have suggested that treatment of infants at risk for EOS may not be justified.
There is limited information available on the drug pharmacokinetics and dosage requirements for infants who are treated with IM gentamicin, as most of the available data relate to neonates given IV. It has been shown that infants, especially neonates, have a higher volume of distribution per kg body weight and a lower clearance compared with older children and adults. Theoretically, there may also be delayed absorption after IM antibiotic administration caused by poor muscle perfusion and decreased bioavailability during the first week of life. This can lead to variability in peak drug concentrations. However, Hayani et al. found no differences in serum concentrations or pharmacokinetic parameters between two groups of neonates treated twice daily with IM or IV gentamicin therapy for presumed sepsis. Paisley et al. found that peak gentamicin concentrations occurred 15 to 30 min after IM injection in newborn infants. Furthermore, this study determined that the mean half-life of an IM dose of gentamicin was 4.0 ± 2.1 h. No significant difference was observed in the 1-h post-treatment kinetics of IM or IV gentamicin administered to neonates treated for suspected sepsis. Ampicillin has been shown to have effective systemic absorption following IM administration.
A potential criticism of this approach relates to the pain associated with IM injections. However, IV catheter insertion is also associated with pain. The small, fragile veins of newborn infants often lead to multiple attempts to establish IV access. Comfort measures such as sucking on a pacifier, gentle containment and breastfeeding can be utilized to lessen infant discomfort. In addition, topical local anesthetics can be applied before IM antibiotic injection. However, the depth of penetration is variable and increases with increasing application time up to ~2 to 3 mm after 1 to 2 h, and 6 mm after 3 to 4 h. Furthermore, the presence of an IV catheter may be perceived by staff and the parents as a barrier to breastfeeding and imposes a separation between the mother and infant, as most nurseries don't permit newborns with an IV line or a medication lock to room-in. Even among family-centered neonatal units that provide parents with continuous access to their babies, maternal–infant bonding may potentially be impeded. Another factor to consider is the medicalization of neonates treated in a special or NICU. Babies in such units are frequently viewed as at greater risk and often subject to unnecessary lab tests, resulting in unnecessary interventions; there is a common tendency (everywhere) to 'do more' and 'get more tests'.
Current management practices for the treatment of well-appearing, asymptomatic infants with suspected sepsis may lead to prolonged antibiotic use and extended hospital stays and excess hospital costs. Even among physicians who treat these infants in a NICU or special care nursery for only 48 to 72 h while awaiting blood culture and other laboratory test results, the costs of care are higher than costs associated with treatment of infants in the regular NBN. In comparison with routine NBN care, NICU charges are 7- to 10-fold higher (http://marchofdimes.com/peristats/pdfdocs/cts/ThomsonAnalysis2008_SummaryDocument_final121208.pdf). Practice variations are additional important contributors to the costs of care. This includes variations in the application of guideline criteria that identify infants requiring evaluation for suspected infection, as well as wide variation in laboratory testing recommendations and treatment protocols.
The Centers for Medicare and Medicaid Services reports that 2 of the top 20 patient diagnosis-related groups that account for hospitalizations in the US are neonates with major problems and neonates with minor problems. In 2009, 29% of newborn infants discharged from the hospital had a clinical problem or complicating diagnosis at a cost of nearly $7.5 billion. Total charges during this period for complicated newborn stays averaged $13 600 compared with $1700 for an uncomplicated newborn stay. In 2000, 0.3% of neonates were treated for infections in the perinatal period. In the ensuing decade health care costs have risen steadily. In contrast, since 2003 despite relative stability in the case mix index (the weight given to a medical condition to indicate its severity or complexity) applied to these diagnosis-related groups, and non-significant changes in the mean and median lengths of hospital stay for neonates with major problems or other significant problems, there has been a 1.4-fold increase in both the mean and median hospital charges. In our own institution, current hospital bed charges for the NICU for intensive/intermediate care are $10 850.00 per day. In contrast, the daily bed charge for a bassinet in the NBN is $1735.00. Although physician services for neonatal care are bundled, hospital charges are currently still reimbursed on a per diem basis. On an annual basis, we have ~60 asymptomatic, well-appearing infants at risk for EOS who are treated in the NBN with IM antibiotics. If all these infants were admitted to the NICU for evaluation and therapy, costs for care would be in the range of $652 550 annually. In contrast, treating these infants in the NBN is associated with a projected cost of $104 343 per year, a 6.25-fold reduction in hospital costs. As hospital payments shift more toward reimbursement for an episode of care, the high cost of providing certain types of care in the NICU, such as parenteral antibiotics to asymptomatic infants at risk for, or with suspected, sepsis will likely be scrutinized more closely. At the present time, many insurers determine hospital payments on the basis of the level of care required by the infant using acuity-based criteria rather than site where the care is provided. Consequently, hospitals may not receive payment for the higher care costs associated with admission of otherwise healthy infants with minimal symptomatology to the NICU for parenteral antibiotic therapy.
In summary, in a large retrospective cohort study of 5045 newborn infants evaluated for suspected EOS, 96.7% of the 421 well-appearing, asymptomatic neonates with suspected EOS treated with IM antibiotics in the regular NBN were discharged home without complication. Among the 3.33% of babies treated with IM antibiotics, seven infants required transfer to the NICU for intensive care because of the development of clinical symptomatology. The other infants required IV therapy because of positive cultures. There were no deaths or complications related to IM antibiotic treatment. IM antibiotic therapy administered in the regular NBN is a means to achieving significant cost savings to hospitals, insurers and families.
Source...