Infections in neurosurgical patients incidence, microbiology and risk factors for postcraniotomy meningitis

Abstract

Nosocomial infections related to the central nervous system are a relatively small but important category of hospital- acquired infections. These infections span a spectrum from superficial wound infections to ventricular shunt infections, meningitis and deep-seated abscesses of the brain parenchyma. They are serious infections, if not life threatening and can be associated with a poor functional outcome. These infections may present many challenges in diagnosis. Nevertheless, a heightened awareness may result to declining rates of infections. The association of neurosurgical infections with certain risk factors may further help in the prevention or the early diagnosis of the neurosurgical infections. The determination of the most commonly isolated offending pathogens and their sensitivities may lead to the most effective treatment. Neurosurgical infections have not been extensively investigated in Greece. With this research we attempted to further determine the bacteriology, the rates and the factors associated with neurosurgical infections. Since there are not detailed reports on post-craniotomy meningitis (PCM) in the international literature, we performed three studies on this subject (two retrospective and one prospective). One of these retrospective studies was performed in NYU Medical Center where the candidate was an Infectious Disease fellow. The retrospective studies were performed in different time periods in NYU Medical Center and in the Medical Center of the University of Crete and we had the opportunity to compare the trends in the offending pathogens and the associated risk factors. In the University of Crete there was a report on the infections after craniotomy in a Thesis for doctoral degree presented in 2008 by M. Roumbelaki, RN. In this Thesis, the rate of SSIs after craniotomy was reported to be 11.1% [287], which was higher when compared with countries such as USA and Spain. Nevertheless, there was no specific mention on the rates of meningitis/ventriculitis after craniotomy, and no mention on the SSIs in the general neurosurgical population. In the retrospective study we performed in a 3-year period and in 1112 general neurosurgical patients, the rate of SSIs was 12.5%. This is a rate that is considerable higher than the ones reported in the literature [16, 24, 51, 194, 195]. In one of the largest studies on postoperative infections in neurosurgery, the infection rates was <1% (0.8%) [195]. This study underwent extensive critique since all the procedures were elective, although not all of them were ―clean?. The same group of authors has reported that 142 the postoperative wound infection after intracranial neurosurgery is nearly three times more likely in European vs. North American studies [198]. Our population mainly consisted of Trauma patients (56.3%) and this could account for the high rate of infections. Meningitis rate was higher than reported in the literature (4.2%). Malignancy was for the first time reported as an independent risk factor in patients undergoing neurosurgery. The placement of drains achieved a statistical significance in the development of infections and the prevalence of SSIs was significantly higher in patients who also developed VAP, UTI and BSI/CAB. The most important result of this study was the absence in mortality increase among patients who developed SSIs and any infection in general but there was an increased duration of ICU hospitalization and an increased length of stay (LOS) in general, therefore increasing the hospital costs. This is a topic that should be further investigated in the future. The first retrospective study on post-craniotomy meningitis (PCM) was performed in NYU Medical Center and it included patients that underwent elective or emergency craniotomy between January 1996 and March 2000. The second retrospective study on PCM was performed in the UOC Medical Center between January 1999 and December 2005. The cohorts in comparison consisted of similar rates of male patients but the UOC population included more trauma patients and consequently more emergent procedures. The PCM rate was higher in the UOC cohort but the difference was not statistically significant. The offending pathogens in the NYU cohort were mainly gram-positives, a fact that reflected the general re-emergence of gram-positive as nosocomial pathogens in the mid-1990s. In the UOC cohort there was a slight predominance of the gram-negatives with Acinetobacter spp. being the most prevalent with a percentage of 16%. These data are consistent with most recent reports. A great limitation of the UOC study is the lack of ASA score inclusion in the risk factor analysis. ASA score is included in the Risk Index Score (RIS) for the SSIs as proposed by the NNIS system [202]. Nevertheless the RIS does not effectively stratify the risk development after craniotomy and CSF shunt placement [202, 203]. Malignancy was determined as an independent risk factor for PCM in the UOC cohort, which has not been reported before. Entering a sinus was a major risk factor for PCM in the NYU cohort but not in the UOC cohort. This probably reflects the trend for avoiding generally this technique in the most recent years. The presence of ventricular drainage was confirmed as independent risk factor in both cohorts but the duration of EVD was an independent risk factor in the NYU 143 cohort. The role of lumbar drains in the development of meningitis was extensively underscored in the UOC retrospective study although it had not been as well recognized in the past. Based on our results and those of recent reports [237], we should emphasize on the role of lumbar drains and the need for removal by the clinicians when they are no longer needed. ICP monitoring and its duration were confirmed as independent risk factors in the NYU cohort but not in the UOC cohort. In the UOC cohort the significance of other infections was underscored for the development of PCM, a fact that should alert the clinicians for prophylaxis against the development of any infection in order to avoid PCM. The percentage of craniotomy patients that developed any kind of infection reached 26%, a percentage that underscores the significance of the very good nosocomial care of these individuals. A prospective study on PCM was performed in the UOC Medical Center for the time period between 2006 and 2008. In this time period traumatic brain injury (TBI) was the most common reason for craniotomy. The percentage of 4.8% for PCM was consistent with the literature. The results of the cultures seem to reflect the increasing role of gram-negative pathogens (especially Acinetobacter spp. which represented 45% of the isolates) in the neurosurgical infections. In this analysis, the use of ventricular drains was once more confirmed as an independent risk factor for the development of PCM, but the duration of EVD drainage was not. CSF leak was demonstrated as an independent risk factor for PCM as repeatedly in previous reports [16, 20, 51, 204]. The use of perioperative steroids achieved statistical significance, an association that has not been described before. ...