Original Paper Pediatr Neurosurg 2013;49:75–80 DOI: 10.1159/000357384
Received: May 31, 2013 Accepted after revision: October 10, 2013 Published online: January 9, 2014
Determinants and Outcomes of Ventriculoperitoneal Shunt Infections in Enugu, Nigeria E.O. Uche E. Onyia U.C. Mezue E. Okorie I.I. Ozor M.C. Chikani Neurological Surgery Unit, Department of Surgery, College of Medicine, University of Nigeria Teaching Hospital, Enugu, Nigeria
Key Words Ventriculoperitoneal shunt · Shunt infection · Risk factors · Outcome
Abstract Background: To evaluate the determinants and outcomes of shunt infection (SI). Methods: One hundred ninety-eight pediatric hydrocephalic patients treated with a ventriculoperitoneal (VP) shunt between January 2008 and August 2012 were retrospectively studied. Patients with SI were compared to those without SI in terms of the occurrence of risk factors and outcomes. Data was analyzed using Statistical Package for the Social Sciences software (version 15). Results: The age range was 2 weeks to 13 years, with a mean age of 3.1 ± 0.19 years for the SI group versus 2.7 ± 0.2 years for those without SI. One hundred and twelve patients were female and 86 were male. SI was recorded in 17 (8.6%) patients. Postinfective hydrocephalus (n = 6) was the most common cause of hydrocephalus in the SI group. Individuals in the SI group, compared to those without infection, were more likely to be underweight (χ2 = 23.4, p < 0.01). The mean interval between VP shunt placement and SI was 1.83 ± 1.25 months. Coagulase-negative Staphylococcus (29.4%) was
© 2014 S. Karger AG, Basel 1016–2291/14/0492–0075$39.50/0 E-Mail [email protected] www.karger.com/pne
the most common pathogen. The mortality rate in our series was 21.4% in patients with SI compared to 2.7% in those without SI. Conclusion: Coagulase-negative Staphylococcus is currently the most common cause of SI and underweight children appear have a higher risk. © 2014 S. Karger AG, Basel
Introduction
Although ventriculoperitoneal (VP) shunt placement remains the mainstay of surgical treatment for hydrocephalus, infection has become a formidable cause of morbidity and mortality in hydrocephalic children treated with a VP shunt [1]. Although the source of infection is quite often the patient’s skin, other foci may also contribute significantly. Early detection coupled with prompt institution of appropriate intervention has proven to be the determinant of successful treatment outcomes in the experience of many a neurosurgeon. This study aims to evaluate the determinants of shunt infection (SI), the microbial and treatment patterns, and the outcome of treatment in a reemerging neurological surgery unit in Southeastern Nigeria. Dr. Enoch Ogbonnaya Uche Lecturer/Consultant Neurosurgeon Department of Surgery, College of Medicine University of Nigeria, Enugu Campus, Enugu (Nigeria) E-Mail kechyenny @ yahoo.com
Materials and Methods Two hundred and eleven children with hydrocephalus were treated with a VP shunt in our center over a period of 4 years and 8 months (January 2008 to August 2012), and 198 had complete records. Patients with complete records were retrospectively studied using data from case notes, cranial CT/MRI reports, and laboratory and surgical records, as well as follow-up notes. Data on demography, the etiology of the hydrocephalus, the interval between presentation and surgery, CT/MRI findings/diagnosis, the caliber of the surgeon, the theater population during the surgery, the duration of antibiotic prophylaxis, the interval between the surgery and the infection, and the clinical features of the infection were recorded. Cerebrospinal fluid (CSF) and blood culture profiles including fungal studies, treatment strategy, number of shunt revisions, duration of hospital admission, and outcome of treatment were acquired and analyzed using Statistical Package for the Social Sciences (SPSS) software (version 15). Chabbra Slit n Spring differential pressure VP shunt hardware was used for all patients. The shunt hardware was irrigated with antibiotic-impregnated saline prior to insertion and a no-touch technique was observed. All of the patients received prophylactic antibiotics at induction of anesthesia and for a period of 24 h post-op. SI was defined as a positive CSF and shunt component culture or other supportive CSF or hematological findings in culture or Gram stain-negative cases associated with clinical features of infection. Patients with SI were managed with a protocol of antibiotics/external ventricular drainage with shunt removal. The duration of the antibiotic treatment varied from 10 to 50 days. Cure was defined as 2 negative CSF cultures in a 48-hour interval, in conjunction with a supportive clinical correlation. Shunt revision was performed following cure through a different site. The weight (kg) of the patients was assessed. The SI rates of patients with a normal weight for their age and underweight patients were compared. Patients were followed up for a period of 6 months to 4 years (mean 2.7 years). Patients older than 18 years and those with infection from other CSF diversion procedures such as endoscopic third ventriculostomy were excluded from this study.
Age group, n 0–2 years 3–5 years 6–8 years 9–11 years 12–14 years Total Mean age (95% Cl), years Sex, n Male Female Cause of hydrocephalus, n Aqueductal stenosis Dandy-Walker malformation Associated with MM Chiari 1 malformation Postinfectious Tumor associated Posthemorrhagic Posttraumatic Vein of Galen malformation Clinical features, n Fever Vomiting Headache Altered sensorium Exposed shunt hardware Shunt track erythema Poor feeding Seizures Abdominal distension Association with underweight, n (%) No infection (n = 181) SI (n = 17)
Patients
SI cases
130 41 15 7 5 198 2.7±0.2
11 3 1 1 1 17 3.1±0.19
86 112
12 5
49 17 43 5 35 24 13 11 1
2 1 5 0 6 1 1 1 0 14 8 6 7 5 5 5 3 3 19 (10.5) 9 (52.9)
χ2 = 23.4, p < 0.01, d.f. = 1.
Results
One hundred ninety-eight patients were studied. The age range of the patients was 2 weeks to 13 years (table 1), with a mean age of 2.7 years (95% CI 2.5–2.9 years, SD = 1.5, SEM = 0.1) for those without infection compared to a mean age of 3.1 years (95% CI 2.91–3.29 years, SD = 1.5, SEM = 0.1) for those with SI. One hundred twelve patients were female and 86 were male (female-to-male ratio = 1.3). Hydrocephalus resulted from congenital causes in 115 patients, with aqueductal stenosis in 49 patients, Dandy-Walker malformation in 17 patients, Chiari 1 malformation in 5 patients, and vein of Galen malformation in 1 patient. Hydrocephalus occurred in association with myelomeningocele in 43 patients. Among 83 patients with acquired causes, postinfectious hydrocephalus occurred 76
Table 1. Summary of demographic profiles, etiologies of hydrocephalus, clinical presentations, CSF culture profiles, treatments, and outcomes
Pediatr Neurosurg 2013;49:75–80 DOI: 10.1159/000357384
in 35 cases. Other causes were intracranial tumors in 24 patients, posthemorrhagic hydrocephalus in 13 patients, and posttraumatic hydrocephalus in 11 patients. SI was recorded in 17 (8.6%) patients, and 12 of them were male while 5 were female (female-to-male ratio = 1:2.4). The causes of hydrocephalus were as follows: postinfective hydrocephalus (n = 6), myelomeningocele-associated hydrocephalus (n = 5), aqueductal stenosis (n = 2), DandyWalker malformation (n = 1), tumor (n = 1), posttraumatic hydrocephalus (n = 1), and posthemorrhagic hydrocephalus (n = 1). The most common clinical presentations of SI were: fever (n = 14; 82.4%), headache (n = 6; Uche/Onyia/Mezue/Okorie/Ozor/Chikani
35.2%), vomiting (n = 8; 47.1%), an altered sensorium (n = 7; 41.2%), exposed shunt hardware (n = 5; 29.4%), shunt tract erythema (n = 5; 29.4%), poor feeding (n = 5; 29.4%), seizures (n = 3; 17.6%), and abdominal distension (n = 3; 17.6%). Nine (52.9%) patients in the SI group had associated malnutrition, while 19 (10.5%) patients in the no- infection group had malnutrition as well and this relationship was highly significant (χ2 = 23.4, p < 0.01, d.f. = 1). The interval between VP shunt placement and SI was 1 month or less in 9 patients, 1–3 months in 5 patients, and more than 3 months in 3 patients, with a range of 3 days to 9 months and a mean time between surgery and infection of 1.83 ± 1.25 months (95% CI). The most frequently isolated pathogen was staphylococci in 52.9% of patients [9 cases, i.e. coagulase-negative Staphylococcus (n = 5) and Staphylococcus aureus (n = 4)], Escherichia coli in 2 patients, Klebsiella in 2 patients, and streptococci in 2 patients. A negative culture was found in 2 cases. The 2 culture-negative cases had supportive clinical features (fever, an altered sensorium, and vomiting) as well as CSF evidence of infection [hypoglycorrhachia and white blood cell (WBC) pleocytosis]. An elevated WBC count (>10,000/ mm3) was recorded in 7 patients, while in 4 patients the WBC count was normal (between 3,000 and 10,000/mm3). In 6 patients, the WBC count was
Received: May 31, 2013 Accepted after revision: October 10, 2013 Published online: January 9, 2014
Determinants and Outcomes of Ventriculoperitoneal Shunt Infections in Enugu, Nigeria E.O. Uche E. Onyia U.C. Mezue E. Okorie I.I. Ozor M.C. Chikani Neurological Surgery Unit, Department of Surgery, College of Medicine, University of Nigeria Teaching Hospital, Enugu, Nigeria
Key Words Ventriculoperitoneal shunt · Shunt infection · Risk factors · Outcome
Abstract Background: To evaluate the determinants and outcomes of shunt infection (SI). Methods: One hundred ninety-eight pediatric hydrocephalic patients treated with a ventriculoperitoneal (VP) shunt between January 2008 and August 2012 were retrospectively studied. Patients with SI were compared to those without SI in terms of the occurrence of risk factors and outcomes. Data was analyzed using Statistical Package for the Social Sciences software (version 15). Results: The age range was 2 weeks to 13 years, with a mean age of 3.1 ± 0.19 years for the SI group versus 2.7 ± 0.2 years for those without SI. One hundred and twelve patients were female and 86 were male. SI was recorded in 17 (8.6%) patients. Postinfective hydrocephalus (n = 6) was the most common cause of hydrocephalus in the SI group. Individuals in the SI group, compared to those without infection, were more likely to be underweight (χ2 = 23.4, p < 0.01). The mean interval between VP shunt placement and SI was 1.83 ± 1.25 months. Coagulase-negative Staphylococcus (29.4%) was
© 2014 S. Karger AG, Basel 1016–2291/14/0492–0075$39.50/0 E-Mail [email protected] www.karger.com/pne
the most common pathogen. The mortality rate in our series was 21.4% in patients with SI compared to 2.7% in those without SI. Conclusion: Coagulase-negative Staphylococcus is currently the most common cause of SI and underweight children appear have a higher risk. © 2014 S. Karger AG, Basel
Introduction
Although ventriculoperitoneal (VP) shunt placement remains the mainstay of surgical treatment for hydrocephalus, infection has become a formidable cause of morbidity and mortality in hydrocephalic children treated with a VP shunt [1]. Although the source of infection is quite often the patient’s skin, other foci may also contribute significantly. Early detection coupled with prompt institution of appropriate intervention has proven to be the determinant of successful treatment outcomes in the experience of many a neurosurgeon. This study aims to evaluate the determinants of shunt infection (SI), the microbial and treatment patterns, and the outcome of treatment in a reemerging neurological surgery unit in Southeastern Nigeria. Dr. Enoch Ogbonnaya Uche Lecturer/Consultant Neurosurgeon Department of Surgery, College of Medicine University of Nigeria, Enugu Campus, Enugu (Nigeria) E-Mail kechyenny @ yahoo.com
Materials and Methods Two hundred and eleven children with hydrocephalus were treated with a VP shunt in our center over a period of 4 years and 8 months (January 2008 to August 2012), and 198 had complete records. Patients with complete records were retrospectively studied using data from case notes, cranial CT/MRI reports, and laboratory and surgical records, as well as follow-up notes. Data on demography, the etiology of the hydrocephalus, the interval between presentation and surgery, CT/MRI findings/diagnosis, the caliber of the surgeon, the theater population during the surgery, the duration of antibiotic prophylaxis, the interval between the surgery and the infection, and the clinical features of the infection were recorded. Cerebrospinal fluid (CSF) and blood culture profiles including fungal studies, treatment strategy, number of shunt revisions, duration of hospital admission, and outcome of treatment were acquired and analyzed using Statistical Package for the Social Sciences (SPSS) software (version 15). Chabbra Slit n Spring differential pressure VP shunt hardware was used for all patients. The shunt hardware was irrigated with antibiotic-impregnated saline prior to insertion and a no-touch technique was observed. All of the patients received prophylactic antibiotics at induction of anesthesia and for a period of 24 h post-op. SI was defined as a positive CSF and shunt component culture or other supportive CSF or hematological findings in culture or Gram stain-negative cases associated with clinical features of infection. Patients with SI were managed with a protocol of antibiotics/external ventricular drainage with shunt removal. The duration of the antibiotic treatment varied from 10 to 50 days. Cure was defined as 2 negative CSF cultures in a 48-hour interval, in conjunction with a supportive clinical correlation. Shunt revision was performed following cure through a different site. The weight (kg) of the patients was assessed. The SI rates of patients with a normal weight for their age and underweight patients were compared. Patients were followed up for a period of 6 months to 4 years (mean 2.7 years). Patients older than 18 years and those with infection from other CSF diversion procedures such as endoscopic third ventriculostomy were excluded from this study.
Age group, n 0–2 years 3–5 years 6–8 years 9–11 years 12–14 years Total Mean age (95% Cl), years Sex, n Male Female Cause of hydrocephalus, n Aqueductal stenosis Dandy-Walker malformation Associated with MM Chiari 1 malformation Postinfectious Tumor associated Posthemorrhagic Posttraumatic Vein of Galen malformation Clinical features, n Fever Vomiting Headache Altered sensorium Exposed shunt hardware Shunt track erythema Poor feeding Seizures Abdominal distension Association with underweight, n (%) No infection (n = 181) SI (n = 17)
Patients
SI cases
130 41 15 7 5 198 2.7±0.2
11 3 1 1 1 17 3.1±0.19
86 112
12 5
49 17 43 5 35 24 13 11 1
2 1 5 0 6 1 1 1 0 14 8 6 7 5 5 5 3 3 19 (10.5) 9 (52.9)
χ2 = 23.4, p < 0.01, d.f. = 1.
Results
One hundred ninety-eight patients were studied. The age range of the patients was 2 weeks to 13 years (table 1), with a mean age of 2.7 years (95% CI 2.5–2.9 years, SD = 1.5, SEM = 0.1) for those without infection compared to a mean age of 3.1 years (95% CI 2.91–3.29 years, SD = 1.5, SEM = 0.1) for those with SI. One hundred twelve patients were female and 86 were male (female-to-male ratio = 1.3). Hydrocephalus resulted from congenital causes in 115 patients, with aqueductal stenosis in 49 patients, Dandy-Walker malformation in 17 patients, Chiari 1 malformation in 5 patients, and vein of Galen malformation in 1 patient. Hydrocephalus occurred in association with myelomeningocele in 43 patients. Among 83 patients with acquired causes, postinfectious hydrocephalus occurred 76
Table 1. Summary of demographic profiles, etiologies of hydrocephalus, clinical presentations, CSF culture profiles, treatments, and outcomes
Pediatr Neurosurg 2013;49:75–80 DOI: 10.1159/000357384
in 35 cases. Other causes were intracranial tumors in 24 patients, posthemorrhagic hydrocephalus in 13 patients, and posttraumatic hydrocephalus in 11 patients. SI was recorded in 17 (8.6%) patients, and 12 of them were male while 5 were female (female-to-male ratio = 1:2.4). The causes of hydrocephalus were as follows: postinfective hydrocephalus (n = 6), myelomeningocele-associated hydrocephalus (n = 5), aqueductal stenosis (n = 2), DandyWalker malformation (n = 1), tumor (n = 1), posttraumatic hydrocephalus (n = 1), and posthemorrhagic hydrocephalus (n = 1). The most common clinical presentations of SI were: fever (n = 14; 82.4%), headache (n = 6; Uche/Onyia/Mezue/Okorie/Ozor/Chikani
35.2%), vomiting (n = 8; 47.1%), an altered sensorium (n = 7; 41.2%), exposed shunt hardware (n = 5; 29.4%), shunt tract erythema (n = 5; 29.4%), poor feeding (n = 5; 29.4%), seizures (n = 3; 17.6%), and abdominal distension (n = 3; 17.6%). Nine (52.9%) patients in the SI group had associated malnutrition, while 19 (10.5%) patients in the no- infection group had malnutrition as well and this relationship was highly significant (χ2 = 23.4, p < 0.01, d.f. = 1). The interval between VP shunt placement and SI was 1 month or less in 9 patients, 1–3 months in 5 patients, and more than 3 months in 3 patients, with a range of 3 days to 9 months and a mean time between surgery and infection of 1.83 ± 1.25 months (95% CI). The most frequently isolated pathogen was staphylococci in 52.9% of patients [9 cases, i.e. coagulase-negative Staphylococcus (n = 5) and Staphylococcus aureus (n = 4)], Escherichia coli in 2 patients, Klebsiella in 2 patients, and streptococci in 2 patients. A negative culture was found in 2 cases. The 2 culture-negative cases had supportive clinical features (fever, an altered sensorium, and vomiting) as well as CSF evidence of infection [hypoglycorrhachia and white blood cell (WBC) pleocytosis]. An elevated WBC count (>10,000/ mm3) was recorded in 7 patients, while in 4 patients the WBC count was normal (between 3,000 and 10,000/mm3). In 6 patients, the WBC count was
