Clinical Features and Prognostic Factors in Adults with Bacterial Meningitis
http://www.100md.com
《新英格兰医药杂志》
ABSTRACT
Background We conducted a nationwide study in the Netherlands to determine clinical features and prognostic factors in adults with community-acquired acute bacterial meningitis.
Methods From October 1998 to April 2002, all Dutch patients with community-acquired acute bacterial meningitis, confirmed by cerebrospinal fluid cultures, were prospectively evaluated. All patients underwent a neurologic examination on admission and at discharge, and outcomes were classified as unfavorable (defined by a Glasgow Outcome Scale score of 1 to 4 points at discharge) or favorable (a score of 5). Predictors of an unfavorable outcome were identified through logistic-regression analysis.
Results We evaluated 696 episodes of community-acquired acute bacterial meningitis. The most common pathogens were Streptococcus pneumoniae (51 percent of episodes) and Neisseria meningitidis (37 percent). The classic triad of fever, neck stiffness, and a change in mental status was present in only 44 percent of episodes; however, 95 percent had at least two of the four symptoms of headache, fever, neck stiffness, and altered mental status. On admission, 14 percent of patients were comatose and 33 percent had focal neurologic abnormalities. The overall mortality rate was 21 percent. The mortality rate was higher among patients with pneumococcal meningitis than among those with meningococcal meningitis (30 percent vs. 7 percent, P<0.001). The outcome was unfavorable in 34 percent of episodes. Risk factors for an unfavorable outcome were advanced age, presence of otitis or sinusitis, absence of rash, a low score on the Glasgow Coma Scale on admission, tachycardia, a positive blood culture, an elevated erythrocyte sedimentation rate, thrombocytopenia, and a low cerebrospinal fluid white-cell count.
Conclusions In adults presenting with community-acquired acute bacterial meningitis, the sensitivity of the classic triad of fever, neck stiffness, and altered mental status is low, but almost all present with at least two of the four symptoms of headache, fever, neck stiffness, and altered mental status. The mortality associated with bacterial meningitis remains high, and the strongest risk factors for an unfavorable outcome are those that are indicative of systemic compromise, a low level of consciousness, and infection with S. pneumoniae.
The epidemiology of bacterial meningitis has changed. Meningitis due to Haemophilus influenzae type b has been nearly eliminated in the Western world since vaccination against H. influenzae type b was initiated,1 and the introduction of conjugate vaccines against Streptococcus pneumoniae is expected to reduce the burden of childhood pneumococcal meningitis significantly.2 Although vaccination with a pneumococcal conjugate vaccine is producing herd immunity among adults, the age distribution of meningitis has now shifted to older age groups.2,3 Several studies of clinical features and prognostic factors in adults with bacterial meningitis have been performed; however, all were retrospective and relatively small in size.4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23 We performed a nationwide prospective study of clinical features and prognostic factors in adults with community-acquired bacterial meningitis in the Netherlands.
Methods
We identified adults (defined as patients older than 16 years of age) who had bacterial meningitis and were listed in the database of the Netherlands Reference Laboratory for Bacterial Meningitis from October 1998 to April 2002. This laboratory receives cerebrospinal fluid and blood isolates from approximately 85 percent of all patients with bacterial meningitis in the Netherlands (population, 16.2 million).3,24 The laboratory provides daily updates of the names of hospitals where patients with bacterial meningitis have been admitted in the preceding two to six days and the names of physicians, usually neurologists. Physicians were informed about the study by telephone. Subsequently, patients or their legal representatives received written information concerning the study and were asked to give written informed consent for participation; only patients for whom consent was obtained were enrolled. Case-record forms were used to collect data on patients' history, symptoms and signs on admission, laboratory findings at admission, clinical course, outcome and neurologic findings at discharge, and treatment.
Patients were categorized as having either community-acquired or hospital-acquired meningitis (the latter was defined as meningitis that occurred during hospitalization or within one week after discharge). Patients with an altered immune status owing to the use of immunosuppressive drugs or splenectomy, diabetes mellitus, or alcoholism were considered immunocompromised, as were patients infected with the human immunodeficiency virus. At discharge, all patients underwent a neurologic examination performed by a neurologist, and the outcome was graded according to the Glasgow Outcome Scale. A score of 1 on this scale indicates death; a score of 2, a vegetative state (the patient is unable to interact with the environment); a score of 3, severe disability (the patient is unable to live independently but can follow commands); a score of 4, moderate disability (the patient is capable of living independently but unable to return to work or school); and a score of 5, mild or no disability (the patient is able to return to work or school). A favorable outcome was defined as a score of 5, and an unfavorable outcome as a score of 1 to 4. The Glasgow Outcome Scale is a well-validated instrument with good interobserver agreement.25
The susceptibility of meningococci to penicillin was determined by inoculating strains onto chocolate agar containing 0.1 mg of penicillin per liter. A 1-μg oxacillin disk was used to identify penicillin-resistant strains of pneumococci. Whenever a strain showed antibiotic resistance, the E rosette assay was used to determine the minimal inhibitory concentration of the antibiotic. The inoculation procedure and susceptibility testing were performed as described elsewhere.26 Parts of this cohort study have been reported previously.7,26,27,28
This observational study used patient data that had been rendered anonymous and was carried out in accordance with Dutch privacy legislation. The study was announced in the journal of the Dutch Neurologic Society, followed by periodic reminders. Before the study, all Dutch neurologists received information about the study, including a case-record form.
Parametric and nonparametric tests were used to identify differences between groups in continuous outcomes, and chi-square tests were used to compare categorical outcomes. We used logistic regression to examine the association between potential predictors and the likelihood of an unfavorable outcome. Odds ratios and 95 percent confidence intervals were used to quantify the strength of these associations. On the basis of previous research and pathophysiologic interest, 20 potentially relevant predictors were chosen.
Despite the low median percentage of missing values for individual variables (2 percent), only 320 of the 696 patients had complete data on all predictors — which presented a considerable limitation for multivariate models. Therefore, we used multiple imputation techniques to reduce this loss.29 All predictors together were used to impute missing values on the basis of multivariate normal distributions. The coefficients of five rounds of imputation were combined to obtain the final estimates for the multivariate model. The low rate of missing values per variable, the large number of predictors, and the large sample size maximized the benefits of the multiple imputation methods.30 All statistical tests were two-tailed, and P values of less than 0.05 were considered to indicate statistical significance.
Statistical analyses were performed with use of SAS software, version 8.02 (SAS Institute). The study was designed, conducted, and analyzed independently of the sponsor.
Results
A total of 1108 episodes of bacterial meningitis were identified by the reference laboratory. A total of 994 case-record forms were sent out, and the response rate was 76 percent (754 of 994) (Figure 1). The demographic characteristics of patients with meningitis identified by the laboratory and those included in the analysis were similar for each causative organism. Fifty-eight patients were excluded — 50 patients with hospital-acquired meningitis, 3 patients with a recent history of neurosurgery, and 5 patients with a neurosurgical device — leaving a total of 696 episodes of community-acquired meningitis in 671 patients. The annual incidence of community-acquired bacterial meningitis was 2.6 cases per 100,000 adults.
Figure 1. Selection of Patients.
Among the 19 strains of Haemophilus influenzae, 16 were nontypeable, 1 was type b, 1 was type e, and 1 was type f. Neurosurgery and neurosurgical devices included ventriculostomy, ventriculoperitoneal or ventriculoatrial shunt, lumbar epidural catheter, and a dorsal-column stimulator. Among the 57 of 696 episodes of meningitis due to bacteria other than Streptococcus pneumoniae, Neisseria meningitidis, and Listeria monocytogenes, 14 were due to H. influenzae, 9 to Staphylococcus aureus, 6 to group A streptococcus, 5 to group B streptococcus, 4 to Streptococcus suis, 4 to Escherichia coli, 2 to Streptococcus salivarius, 1 to group C streptococcus, 1 to group G streptococcus, 1 to H. parainfluenzae, 1 to Klebsiella pneumoniae, 1 to Streptococcus bovis, 1 to Streptococcus mitis, 1 to Streptococcus oralis, 1 to Streptococcus constellatus, 1 to Staphylococcus epidermidis, and 1 merely to "streptococci"; 3 specimens were not viable.
Characteristics of the study population are shown in Table 1. Seizures occurred before admission in 32 of 666 episodes (5 percent). Predisposing conditions were present in 48 percent of episodes, the most common of which were otitis or sinusitis in 25 percent, pneumonia in 12 percent, and an immunocompromised state in 16 percent. Patients with pneumococcal meningitis were more likely to have distant foci of infection than were patients with meningococcal meningitis (62 percent vs. 9 percent, P<0.001).
Table 1. Characteristics of the Study Population.
Classic symptoms and signs of bacterial meningitis were present in a large proportion of patients (Table 1). Headache occurred in 87 percent of episodes, neck stiffness in 83 percent, fever in 77 percent, and a change in mental status (defined by a Glasgow Coma Scale score below 14) in 69 percent. Only 44 percent of episodes were characterized by the classic triad of fever, neck stiffness, and a change in mental status. At least two of four signs (the classic triad plus headache) were present in 664 episodes (95 percent), only one in 28 episodes (4 percent), and none in 4 episodes (1 percent). The classic triad was significantly more likely to be present in patients with pneumococcal meningitis than in those with meningococcal meningitis (58 percent vs. 27 percent, P<0.001).
Rash was present in 176 of 683 episodes (26 percent); the causative species was Neisseria meningitidis in 162 episodes, S. pneumoniae in 8, Staphylococcus aureus in 2, group B streptococcus in 2, H. influenzae in 1, and Listeria monocytogenes in 1. The rash was petechial in 157 of these 176 episodes (89 percent). The rash was petechial in 13 of 14 episodes that were characterized by a rash and a nonmeningococcal cause of meningitis.
In 14 percent of episodes, the patients were comatose on admission, and in 33 percent of episodes, focal neurologic deficits were present on admission. Patients with pneumococcal meningitis had more severe disease than did patients with meningococcal meningitis, as reflected by a higher frequency of seizures (P=0.001) and focal neurologic deficits (P<0.001) and a lower level of consciousness (P<0.001). Results of funduscopic examination were recorded for 386 episodes; 13 episodes were characterized by papilledema (3 percent).
Lumbar puncture was performed in all patients. The time between admission and lumbar puncture was not recorded. Cerebrospinal fluid pressure was evaluated in 216 episodes. Opening pressures were normal (less than 200 mm of water) in 38 episodes (18 percent) and exceeded 400 mm of water in 85 (39 percent). Patients with opening pressures of more than 400 mm of water were more likely to be admitted in a coma than were those with lower opening pressures (14 of 131 vs. 20 of 85 , P=0.01). However, the percentage of patients with an unfavorable outcome was similar in the two groups (30 percent and 35 percent, respectively). The mean opening pressures were similar among patients with papilledema and those without papilledema. At least one individual cerebrospinal fluid finding predictive of bacterial meningitis (a glucose level of less than 34 mg per deciliter , a ratio of cerebrospinal fluid glucose to blood glucose of less than 0.23, a protein level of more than 220 mg per deciliter, or a white-cell count of more than 2000 per cubic millimeter)31 was present in 567 of 645 episodes (88 percent).
Cranial computed tomography (CT) was performed on admission in 496 episodes (71 percent); the results were normal in 325. Abnormalities were recorded in 171 episodes (34 percent): cerebral edema was identified in 48 of 496 episodes (10 percent), sinusitis or otitis in 48 (10 percent), evidence of recent infarction in 30 (6 percent), and hydrocephalus in 15 (3 percent). Thirty-eight episodes involved other abnormalities on cranial CT: old vascular lesions in 12, cerebral atrophy in 6, skull fracture in 5, pneumatocephalus in 5, arachnoid cyst in 3, vascular aneurysm in 2, subarachnoid hemorrhage in 1, brain-parenchyma hemorrhage in 1, subdural empyema in 1, meningioma in 1, and Dandy–Walker malformation in 1; in 8 episodes, two abnormalities were identified.
Cranial CT was performed before lumbar puncture in 337 of 696 episodes (48 percent). Focal neurologic deficits (not including cranial-nerve abnormalities), a score of less than 10 on the Glasgow Coma Scale, or both were present in 313 of 696 episodes (45 percent), and cranial CT was performed before lumbar puncture in 197 of these 313 episodes (63 percent). Therapy was initiated before CT in 35 percent of episodes in which CT was performed before lumbar puncture.
The most common microorganism was S. pneumoniae, accounting for 51 percent of isolates cultured from cerebrospinal fluid (Figure 1). The distribution of serotypes among 352 pneumococci is shown in Table 2. N. meningitidis was responsible for 37 percent of the episodes, with group B identified in 173 episodes, group C in 79, group Y in 3, group H in 1, and group W135 in 1. Gram's staining of cerebrospinal fluid revealed the microorganism in 524 of 652 episodes (sensitivity, 80 percent; specificity, 97 percent). The yield of Gram's staining was similar in patients who had previously received antimicrobial therapy and those who had not. Antibiotic susceptibility was tested in 351 pneumococcal and 256 meningococcal strains; 2 pneumococcal and 4 meningococcal strains showed intermediate resistance to penicillin; all other strains were sensitive to penicillin. Initial antibiotic treatment consisted of penicillin or amoxicillin in 43 percent of episodes, third-generation cephalosporins in 16 percent, and a combination of penicillin or amoxicillin with a third-generation cephalosporin in 25 percent; other regimens were used in 16 percent of episodes.
Table 2. Capsular Pneumococcal Serotypes from 352 Adults with Meningitis.
During the clinical course, focal neurologic abnormalities were found in half the episodes (Table 3); most were present on admission. Seizures occurred in 15 percent of episodes, and cardiorespiratory failure in 29 percent. Complications were significantly more likely to develop among patients with pneumococcal meningitis than among patients with meningococcal meningitis (P<0.001).
Table 3. Clinical Course, Outcome, and Neurologic Findings at Discharge.
The mortality rate was 21 percent (Table 3) and varied depending on the causative organism; it was 30 percent for pneumococcal meningitis, as compared with 7 percent for meningococcal meningitis (P<0.001) (Table 3) and 20 percent for meningitis due to other pathogens (P=0.05). Thirty-four percent of episodes had an unfavorable outcome. A neurologic examination was performed at discharge in 550 of 553 surviving patients (99 percent); the most common abnormalities identified were hearing loss (14 percent) and hemiparesis (4 percent).
Corticosteroids were administered in 118 episodes (17 percent), and episodes involving corticosteroid therapy were more likely to have an unfavorable outcome than were episodes that did not involve corticosteroid therapy (51 of 118 vs. 186 of 578 , P=0.03). Episodes in which corticosteroids were administered before antibiotics were less likely to have an unfavorable outcome than episodes in which corticosteroids were administered after antibiotics (3 of 24 vs. 48 of 94 , P=0.001). Opening pressures of more than 400 mm of water were recorded in 46 percent of patients who received corticosteroids and 38 percent of those not receiving corticosteroids.
In the multivariate model several characteristics were significantly associated with an unfavorable outcome (Table 4): advanced age, the presence of otitis or sinusitis, the absence of rash, a heart rate of more than 120 beats per minute, a low score on the Glasgow Coma Scale, a cerebrospinal fluid white-cell count of fewer than 1000 per cubic millimeter, a positive blood culture, an elevated erythrocyte sedimentation rate, and a reduced platelet count. Six other characteristics tended toward statistical significance: the presence of symptoms for less than 24 hours before admission, seizures, pneumonia, an immunocompromised state, a heart rate below 60 beats per minute, and hypotension (defined as a diastolic blood pressure of less than 60 mm Hg).
Table 4. Multivariate Analysis of Factors Associated with an Unfavorable Outcome.
The causative organism had an independent effect when it was added to the multivariate model. The odds of an unfavorable outcome was six times as high (95 percent confidence interval, 2.61 to 13.91; P<0.001) among patients infected with S. pneumoniae as among patients infected with N. meningitidis, even after adjustment for other clinical predictors.
During episodes of meningococcal meningitis, hypotension was more common among patients with a cerebrospinal fluid white-cell count of fewer than 100 cells per cubic millimeter than among patients with a higher cerebrospinal fluid white-cell count (8 of 18 vs. 24 of 211 , P<0.001). This association was absent during episodes of pneumococcal meningitis.
Discussion
We found that the prevalence of the classic triad of fever, neck stiffness, and an altered mental status is low among adults with community-acquired bacterial meningitis. However, almost all patients (95 percent) presented with at least two of the four symptoms of headache, fever, neck stiffness, and an altered mental status. In addition, a high percentage of patients (33 percent) were admitted with focal neurologic deficits. These percentages are similar to those found in retrospective studies; however, we found an exceptionally high prevalence of aphasia.4,5,6,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23
The majority of patients who presented with moderate or severe impairment of consciousness, neurologic deficits (not including cranial-nerve abnormalities), or both, which are contraindications to performing lumbar puncture, underwent CT before lumbar puncture (63 percent). However, a large number of patients without these "red flags" also underwent CT first, and many patients with such red flags did not undergo CT before lumbar puncture. In most patients who underwent CT before lumbar puncture, therapy was not initiated before CT was performed. Cranial CT has been recommended as a precaution before lumbar puncture to predict the likelihood and avoid the possibility of brain herniation.9,32,33 The withdrawal of cerebrospinal fluid removes counterpressure from below, thus increasing the effect of compression from above and exacerbating the brain shift already present.33
In patients with red flags indicating space-occupying lesions and in those with moderate or severe impairment of consciousness (as indicated by a score of less than 10 on the Glasgow Coma Scale), lumbar puncture should be preceded by cranial CT. However, a delay in the initiation of antimicrobial therapy may lead to a poor outcome, especially in patients with clinical deterioration.8 Therefore, in patients with suspected meningitis in whom cranial CT is performed before lumbar puncture, CT should be preceded by blood cultures and the initiation of antibiotic therapy and corticosteroids.33
Community-acquired bacterial meningitis has a high rate of an unfavorable outcome in adults (34 percent). Our multivariate model identified several unfavorable prognostic factors, most of which pointed to systemic compromise. A low level of consciousness on admission was predictive of an unfavorable outcome, as was a low cerebrospinal fluid white-cell count. Finally, factors predictive of pneumococcal infection were associated with an unfavorable outcome (advanced age; presence of otitis or sinusitis, pneumonia, or immunocompromised status; and absence of rash).
In our cohort, one of every six patients received corticosteroids. The European Dexamethasone Study showed adjunctive treatment with dexamethasone to be beneficial in adults with bacterial meningitis.27 Dexamethasone reduced the rate of an unfavorable outcome from 25 percent to 15 percent (relative risk, 0.59; 95 percent confidence interval, 0.37 to 0.94; P=0.03). A post hoc analysis showed that the beneficial effect of dexamethasone on pneumococcal meningitis was attributable to a reduction in systemic rather than neurologic complications.34 A systemic review showed that adjunctive corticosteroid therapy also reduced the frequency of neurologic sequelae among survivors.35 The increased likelihood of an unfavorable outcome among those receiving corticosteroids in our cohort was probably due to the fact that corticosteroids were used more frequently in patients with clinical deterioration or those with factors associated with a poorer prognosis.
A low cerebrospinal fluid white-cell count was associated with an adverse outcome. This association has been described before.22,36 In patients with meningococcal meningitis, a low cerebrospinal fluid white-cell count may be found in those with sepsis and early meningitis. Studies of animals with pneumococcal meningitis showed a relation among a large bacterial cerebrospinal fluid load, lack of response of cerebrospinal fluid leukocytes, and intracranial complications.37 It probably indicates excessive bacterial growth and lack of a leukocyte response in the cerebrospinal fluid.
In our study, patients with pneumococcal meningitis were at risk for an unfavorable outcome, even after correction for other clinical predictors. Thus, for clinicians, knowledge of the causative organism of meningitis is important in predicting the risk of an unfavorable outcome. Gram's staining of cerebrospinal fluid permits rapid and accurate identification of the causative bacteria in patients with bacterial meningitis and should therefore be routine.4,5,6,7,9
Several studies of prognostic factors in bacterial meningitis have been performed; however, all were retrospective and relatively small in size.4,5,6,10,11,12,13,14,15,16,17,18,19,20,21,22 Our study was nationwide and, therefore, we were able to study a representative sample of adults with acute bacterial meningitis. Furthermore, our prospective approach allowed us to collect comprehensive data on signs and symptoms, clinical course, and outcome. In addition, our large sample gave us the statistical power to perform multivariate analysis.
Our study has one important limitation: only patients who underwent lumbar puncture and who had a positive cerebrospinal fluid culture were included. Negative cerebrospinal fluid cultures occur in 11 to 30 percent of patients with bacterial meningitis.4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23 No significant differences in clinical presentation have been reported between patients with culture-positive bacterial meningitis and those with culture-negative bacterial meningitis.4,5,9 Therefore, it is unlikely that this factor confounded our results. Patients with space-occupying lesions on cranial CT do not undergo lumbar puncture,32,33 and patients with meningitis and a florid rash or septic shock also may not undergo lumbar puncture initially. Thus, an unknown number of such patients were excluded from the cohort, which may have resulted in an underestimation of the mortality rate, especially among patients with meningococcal meningitis.
Rates of antibiotic resistance among meningococcal and pneumococcal isolates were very low. Similar rates have been found in other studies in the Netherlands.26,38 In the United States, France, Spain, and other countries, antibiotic-resistant S. pneumoniae strains are highly prevalent and have emerged as a major problem in the treatment of patients with bacterial meningitis.39 In response to this epidemiologic trend, treatment recommendations for suspected and confirmed cases of bacterial meningitis continue to evolve.40 In areas with high rates of pneumococci that are resistant to broad-spectrum cephalosporins, a combination of a third-generation cephalosporin and vancomycin should be the initial choice for patients with acute bacterial meningitis.40 Treatment failures in patients with meningitis due to multidrug-resistant bacterial isolates have been described.9,41,42 The outcome was not significantly influenced by the presence of antibiotic-resistant pneumococcal isolates in several studies20,23,41,42; however, the relationship of antibiotic resistance to the outcome of meningitis remains to be elucidated.
Despite the fact that the case fatality rate is decreased by the use of adjunctive dexamethasone, there is still room for improvement. Prompt use of dexamethasone and appropriate antibiotics, together with optimal supportive care, can further reduce the mortality and morbidity associated with bacterial meningitis. Most risk factors for an unfavorable outcome that we identified indicated the presence of systemic compromise, a low level of consciousness, and infection with S. pneumoniae. Therefore, aggressive supportive care of patients with bacterial meningitis and systemic complications, preferably in specialized care units, is needed.
Supported in part by a grant from Roche Pharmaceuticals.
We are indebted to Professor P. McIntyre for his comments on the manuscript and to many physicians in the Netherlands for their cooperation.
Source Information
From the Departments of Neurology (D.B., J.G., M.W., M.V.), Medical Microbiology (L.S.), and Clinical Epidemiology and Biostatistics (J.B.R.), Academic Medical Center; and the Netherlands Reference Laboratory for Bacterial Meningitis (L.S.) — both in Amsterdam.
Address reprint requests to Dr. van de Beek at the Academic Medical Center, University of Amsterdam, Department of Neurology H2, P.O. Box 22660, 1100 DD Amsterdam, the Netherlands, or at d.vandebeek@amc.uva.nl.
References
Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995. N Engl J Med 1997;337:970-976.
Whitney CG, Farley MM, Hadler J, et al. Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N Engl J Med 2003;348:1737-1746.
Netherlands Reference Laboratory for Bacterial Meningitis. Bacterial meningitis in The Netherlands: annual report 2002. Amsterdam: University of Amsterdam, 2003.
Durand ML, Calderwood SB, Weber DJ, et al. Acute bacterial meningitis in adults: a review of 493 episodes. N Engl J Med 1993;328:21-28.
Sigurdardottir B, Bjornsson OM, Jonsdottir KE, Erlendsdottir H, Gudmundsson S. Acute bacterial meningitis in adults: a 20-year overview. Arch Intern Med 1997;157:425-430.
Hussein AS, Shafran SD. Acute bacterial meningitis in adults: a 12-year review. Medicine (Baltimore) 2000;79:360-368.
van de Beek D, Schmand B, de Gans J, et al. Cognitive impairment in adults with good recovery after bacterial meningitis. J Infect Dis 2002;186:1047-1052.
Aronin SI, Peduzzi P, Quagliarello VJ. Community-acquired bacterial meningitis: risk stratification for adverse clinical outcome and effect of antibiotic timing. Ann Intern Med 1998;129:862-869.
Tunkel AR. Bacterial meningitis. Philadelphia: Lippincott Williams & Wilkins, 2001.
Olsson RA, Kirby JC, Romansky MJ. Pneumococcal meningitis in the adult: clinical, therapeutic, and prognostic aspects in forty-three patients. Ann Intern Med 1961;55:545-549.
Dodge PR, Swartz MN. Bacterial meningitis -- a review of selected aspects. II. Special neurologic problems, postmeningitis complications, and clinicopathological correlations. N Engl J Med 1965;272:954-960.
Weiss W, Figueroa W, Shapiro WH, Flippin HF. Prognostic factors in pneumococcal meningitis. Arch Intern Med 1967;120:517-524.
Magnussen CR. Meningitis in adults; ten-year retrospective analysis at community hospital. N Y State J Med 1980;80:901-906.
Bohr V, Rasmussen N, Hansen B, et al. Pneumococcal meningitis: an evaluation of prognostic factors in 164 cases based on mortality and on a study of lasting sequelae. J Infect 1985;10:143-157.
Bruyn GAW, Kremer HPH, de Marie S, Padberg GW, Hermans J, van Furth R. Clinical evaluation of pneumococcal meningitis in adults over a twelve-year period. Eur J Clin Microbiol Infect Dis 1989;8:695-700.
Hoen B, Viel JF, Gerard A, Dureux JB, Canton P. Mortality in pneumococcal meningitis: a multivariate analysis of prognostic factors. Eur J Med 1993;2:28-32.
Kirkpatrick B, Reeves DS, MacGowan AP. A review of the clinical presentation, laboratory features, antimicrobial therapy and outcome of 77 episodes of pneumococcal meningitis occurring in children and adults. J Infect 1994;29:171-182.
Attia J, Hatala R, Cook DJ, Wong JG. The rational clinical examination: does this adult patient have acute meningitis? JAMA 1999;282:175-181.
Kragsbjerg P, Kallman J, Olcen P. Pneumococcal meningitis in adults. Scand J Infect Dis 1994;26:659-666.
Auburtin M, Porcher R, Bruneel F, et al. Pneumococcal meningitis in the intensive care unit: prognostic factors of clinical outcome in a series of 80 cases. Am J Respir Crit Care Med 2002;165:713-717.
Lu CH, Huang CR, Chang WN, et al. Community-acquired bacterial meningitis in adults: the epidemiology, timing of appropriate antimicrobial therapy, and prognostic factors. Clin Neurol Neurosurg 2002;104:352-358.
Kastenbauer S, Pfister HW. Pneumococcal meningitis in adults: spectrum of complications and prognostic factors in a series of 87 cases. Brain 2003;126:1015-1025.
Flores-Cordero JM, Amaya-Villar R, Rincon-Ferrari MD, et al. Acute community-acquired bacterial meningitis in adults admitted to the intensive care unit: clinical manifestations, management and prognostic factors. Intensive Care Med 2003;29:1967-1973.
Statistics Netherlands. StatLine, Voorburg/Heerlen, 2001. (Accessed October 4, 2004, at http://www.cbs.nl.)
Jennett B, Teasdale G. Management of head injuries. 2nd ed. Philadelphia: F.A. Davis, 1981.
van de Beek D, de Gans J, Spanjaard L, Vermeulen M, Dankert J. Antibiotic guidelines and antibiotic use in adult bacterial meningitis in The Netherlands. J Antimicrob Chemother 2002;49:661-666.
de Gans J, van de Beek D. Dexamethasone in adults with bacterial meningitis. N Engl J Med 2002;347:1549-1556.
de Gans J, van de Beek D. Dexamethasone in adults with bacterial meningitis. N Engl J Med 2003;348:955-956.
Little RJA, Rubin DB. Statistical analysis with missing data. New York: Wiley, 1987.
Arnold AM, Kronmal RA. Multiple imputation of baseline data in the Cardiovascular Health Study. Am J Epidemiol 2003;157:74-84.
Spanos A, Harrell FE Jr, Durack DT. Differential diagnosis of acute meningitis: an analysis of the predictive value of initial observations. JAMA 1989;262:2700-2707.
Hasbun R, Abrahams J, Jekel J, Quagliarello VJ. Computed tomography of the head before lumbar puncture in adults with suspected meningitis. N Engl J Med 2001;345:1727-1733.
van Crevel H, Hijdra A, de Gans J. Lumbar puncture and the risk of herniation: when should we first perform CT? J Neurol 2002;249:129-137.
van de Beek D, de Gans J. Dexamethasone and pneumococcal meningitis. Ann Intern Med 2004;141:327-327.
van de Beek D, de Gans J, McIntyre P, Prasad K. Steroids in adults with bacterial meningitis: a systematic review. Lancet Infect Dis 2004;3:139-143.
McMillan DA, Lin CY, Aronin SI, Quagliarello VJ. Community-acquired bacterial meningitis in adults: categorization of causes and timing of death. Clin Infect Dis 2001;33:969-975.
Tauber MG, Kennedy SL, Tureen JH, Lowenstein DH. Experimental pneumococcal meningitis causes central nervous system pathology without inducing the 72-kd heat shock protein. Am J Pathol 1992;141:53-60.
van de Beek D, Hensen EF, Spanjaard L, de Gans J, Enting RH, Dankert J. Meropenem susceptibility of Neisseria meningitidis and Streptococcus pneumoniae from meningitis patients in The Netherlands. J Antimicrob Chemother 1997;40:895-897.
Whitney CG, Farley MM, Hadler J, et al. Increasing prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. N Engl J Med 2000;343:1917-1924.
Aronin SI. Current pharmacotherapy of pneumococcal meningitis. Expert Opin Pharmacother 2002;3:121-129.
Arditi M, Mason EO Jr, Bradley JS, et al. Three-year multicenter surveillance of pneumococcal meningitis in children: clinical characteristics, and outcome related to penicillin susceptibility and dexamethasone use. Pediatrics 1998;102:1087-1097.
Kellner JD, Scheifele DW, Halperin SA, et al. Outcome of penicillin-nonsusceptible Streptococcus pneumoniae meningitis: a nested case-control study. Pediatr Infect Dis J 2002;21:903-910.
Related Letters:
Prognostic Factors in Adults with Bacterial Meningitis
Joffe A. R., ?stergaard C., Klussmann J. P., Guntinas-Lichius O., Altschuler E. L., van de Beek D., de Gans J., Swartz M. N.(Diederik van de Beek, M.D)
Background We conducted a nationwide study in the Netherlands to determine clinical features and prognostic factors in adults with community-acquired acute bacterial meningitis.
Methods From October 1998 to April 2002, all Dutch patients with community-acquired acute bacterial meningitis, confirmed by cerebrospinal fluid cultures, were prospectively evaluated. All patients underwent a neurologic examination on admission and at discharge, and outcomes were classified as unfavorable (defined by a Glasgow Outcome Scale score of 1 to 4 points at discharge) or favorable (a score of 5). Predictors of an unfavorable outcome were identified through logistic-regression analysis.
Results We evaluated 696 episodes of community-acquired acute bacterial meningitis. The most common pathogens were Streptococcus pneumoniae (51 percent of episodes) and Neisseria meningitidis (37 percent). The classic triad of fever, neck stiffness, and a change in mental status was present in only 44 percent of episodes; however, 95 percent had at least two of the four symptoms of headache, fever, neck stiffness, and altered mental status. On admission, 14 percent of patients were comatose and 33 percent had focal neurologic abnormalities. The overall mortality rate was 21 percent. The mortality rate was higher among patients with pneumococcal meningitis than among those with meningococcal meningitis (30 percent vs. 7 percent, P<0.001). The outcome was unfavorable in 34 percent of episodes. Risk factors for an unfavorable outcome were advanced age, presence of otitis or sinusitis, absence of rash, a low score on the Glasgow Coma Scale on admission, tachycardia, a positive blood culture, an elevated erythrocyte sedimentation rate, thrombocytopenia, and a low cerebrospinal fluid white-cell count.
Conclusions In adults presenting with community-acquired acute bacterial meningitis, the sensitivity of the classic triad of fever, neck stiffness, and altered mental status is low, but almost all present with at least two of the four symptoms of headache, fever, neck stiffness, and altered mental status. The mortality associated with bacterial meningitis remains high, and the strongest risk factors for an unfavorable outcome are those that are indicative of systemic compromise, a low level of consciousness, and infection with S. pneumoniae.
The epidemiology of bacterial meningitis has changed. Meningitis due to Haemophilus influenzae type b has been nearly eliminated in the Western world since vaccination against H. influenzae type b was initiated,1 and the introduction of conjugate vaccines against Streptococcus pneumoniae is expected to reduce the burden of childhood pneumococcal meningitis significantly.2 Although vaccination with a pneumococcal conjugate vaccine is producing herd immunity among adults, the age distribution of meningitis has now shifted to older age groups.2,3 Several studies of clinical features and prognostic factors in adults with bacterial meningitis have been performed; however, all were retrospective and relatively small in size.4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23 We performed a nationwide prospective study of clinical features and prognostic factors in adults with community-acquired bacterial meningitis in the Netherlands.
Methods
We identified adults (defined as patients older than 16 years of age) who had bacterial meningitis and were listed in the database of the Netherlands Reference Laboratory for Bacterial Meningitis from October 1998 to April 2002. This laboratory receives cerebrospinal fluid and blood isolates from approximately 85 percent of all patients with bacterial meningitis in the Netherlands (population, 16.2 million).3,24 The laboratory provides daily updates of the names of hospitals where patients with bacterial meningitis have been admitted in the preceding two to six days and the names of physicians, usually neurologists. Physicians were informed about the study by telephone. Subsequently, patients or their legal representatives received written information concerning the study and were asked to give written informed consent for participation; only patients for whom consent was obtained were enrolled. Case-record forms were used to collect data on patients' history, symptoms and signs on admission, laboratory findings at admission, clinical course, outcome and neurologic findings at discharge, and treatment.
Patients were categorized as having either community-acquired or hospital-acquired meningitis (the latter was defined as meningitis that occurred during hospitalization or within one week after discharge). Patients with an altered immune status owing to the use of immunosuppressive drugs or splenectomy, diabetes mellitus, or alcoholism were considered immunocompromised, as were patients infected with the human immunodeficiency virus. At discharge, all patients underwent a neurologic examination performed by a neurologist, and the outcome was graded according to the Glasgow Outcome Scale. A score of 1 on this scale indicates death; a score of 2, a vegetative state (the patient is unable to interact with the environment); a score of 3, severe disability (the patient is unable to live independently but can follow commands); a score of 4, moderate disability (the patient is capable of living independently but unable to return to work or school); and a score of 5, mild or no disability (the patient is able to return to work or school). A favorable outcome was defined as a score of 5, and an unfavorable outcome as a score of 1 to 4. The Glasgow Outcome Scale is a well-validated instrument with good interobserver agreement.25
The susceptibility of meningococci to penicillin was determined by inoculating strains onto chocolate agar containing 0.1 mg of penicillin per liter. A 1-μg oxacillin disk was used to identify penicillin-resistant strains of pneumococci. Whenever a strain showed antibiotic resistance, the E rosette assay was used to determine the minimal inhibitory concentration of the antibiotic. The inoculation procedure and susceptibility testing were performed as described elsewhere.26 Parts of this cohort study have been reported previously.7,26,27,28
This observational study used patient data that had been rendered anonymous and was carried out in accordance with Dutch privacy legislation. The study was announced in the journal of the Dutch Neurologic Society, followed by periodic reminders. Before the study, all Dutch neurologists received information about the study, including a case-record form.
Parametric and nonparametric tests were used to identify differences between groups in continuous outcomes, and chi-square tests were used to compare categorical outcomes. We used logistic regression to examine the association between potential predictors and the likelihood of an unfavorable outcome. Odds ratios and 95 percent confidence intervals were used to quantify the strength of these associations. On the basis of previous research and pathophysiologic interest, 20 potentially relevant predictors were chosen.
Despite the low median percentage of missing values for individual variables (2 percent), only 320 of the 696 patients had complete data on all predictors — which presented a considerable limitation for multivariate models. Therefore, we used multiple imputation techniques to reduce this loss.29 All predictors together were used to impute missing values on the basis of multivariate normal distributions. The coefficients of five rounds of imputation were combined to obtain the final estimates for the multivariate model. The low rate of missing values per variable, the large number of predictors, and the large sample size maximized the benefits of the multiple imputation methods.30 All statistical tests were two-tailed, and P values of less than 0.05 were considered to indicate statistical significance.
Statistical analyses were performed with use of SAS software, version 8.02 (SAS Institute). The study was designed, conducted, and analyzed independently of the sponsor.
Results
A total of 1108 episodes of bacterial meningitis were identified by the reference laboratory. A total of 994 case-record forms were sent out, and the response rate was 76 percent (754 of 994) (Figure 1). The demographic characteristics of patients with meningitis identified by the laboratory and those included in the analysis were similar for each causative organism. Fifty-eight patients were excluded — 50 patients with hospital-acquired meningitis, 3 patients with a recent history of neurosurgery, and 5 patients with a neurosurgical device — leaving a total of 696 episodes of community-acquired meningitis in 671 patients. The annual incidence of community-acquired bacterial meningitis was 2.6 cases per 100,000 adults.
Figure 1. Selection of Patients.
Among the 19 strains of Haemophilus influenzae, 16 were nontypeable, 1 was type b, 1 was type e, and 1 was type f. Neurosurgery and neurosurgical devices included ventriculostomy, ventriculoperitoneal or ventriculoatrial shunt, lumbar epidural catheter, and a dorsal-column stimulator. Among the 57 of 696 episodes of meningitis due to bacteria other than Streptococcus pneumoniae, Neisseria meningitidis, and Listeria monocytogenes, 14 were due to H. influenzae, 9 to Staphylococcus aureus, 6 to group A streptococcus, 5 to group B streptococcus, 4 to Streptococcus suis, 4 to Escherichia coli, 2 to Streptococcus salivarius, 1 to group C streptococcus, 1 to group G streptococcus, 1 to H. parainfluenzae, 1 to Klebsiella pneumoniae, 1 to Streptococcus bovis, 1 to Streptococcus mitis, 1 to Streptococcus oralis, 1 to Streptococcus constellatus, 1 to Staphylococcus epidermidis, and 1 merely to "streptococci"; 3 specimens were not viable.
Characteristics of the study population are shown in Table 1. Seizures occurred before admission in 32 of 666 episodes (5 percent). Predisposing conditions were present in 48 percent of episodes, the most common of which were otitis or sinusitis in 25 percent, pneumonia in 12 percent, and an immunocompromised state in 16 percent. Patients with pneumococcal meningitis were more likely to have distant foci of infection than were patients with meningococcal meningitis (62 percent vs. 9 percent, P<0.001).
Table 1. Characteristics of the Study Population.
Classic symptoms and signs of bacterial meningitis were present in a large proportion of patients (Table 1). Headache occurred in 87 percent of episodes, neck stiffness in 83 percent, fever in 77 percent, and a change in mental status (defined by a Glasgow Coma Scale score below 14) in 69 percent. Only 44 percent of episodes were characterized by the classic triad of fever, neck stiffness, and a change in mental status. At least two of four signs (the classic triad plus headache) were present in 664 episodes (95 percent), only one in 28 episodes (4 percent), and none in 4 episodes (1 percent). The classic triad was significantly more likely to be present in patients with pneumococcal meningitis than in those with meningococcal meningitis (58 percent vs. 27 percent, P<0.001).
Rash was present in 176 of 683 episodes (26 percent); the causative species was Neisseria meningitidis in 162 episodes, S. pneumoniae in 8, Staphylococcus aureus in 2, group B streptococcus in 2, H. influenzae in 1, and Listeria monocytogenes in 1. The rash was petechial in 157 of these 176 episodes (89 percent). The rash was petechial in 13 of 14 episodes that were characterized by a rash and a nonmeningococcal cause of meningitis.
In 14 percent of episodes, the patients were comatose on admission, and in 33 percent of episodes, focal neurologic deficits were present on admission. Patients with pneumococcal meningitis had more severe disease than did patients with meningococcal meningitis, as reflected by a higher frequency of seizures (P=0.001) and focal neurologic deficits (P<0.001) and a lower level of consciousness (P<0.001). Results of funduscopic examination were recorded for 386 episodes; 13 episodes were characterized by papilledema (3 percent).
Lumbar puncture was performed in all patients. The time between admission and lumbar puncture was not recorded. Cerebrospinal fluid pressure was evaluated in 216 episodes. Opening pressures were normal (less than 200 mm of water) in 38 episodes (18 percent) and exceeded 400 mm of water in 85 (39 percent). Patients with opening pressures of more than 400 mm of water were more likely to be admitted in a coma than were those with lower opening pressures (14 of 131 vs. 20 of 85 , P=0.01). However, the percentage of patients with an unfavorable outcome was similar in the two groups (30 percent and 35 percent, respectively). The mean opening pressures were similar among patients with papilledema and those without papilledema. At least one individual cerebrospinal fluid finding predictive of bacterial meningitis (a glucose level of less than 34 mg per deciliter , a ratio of cerebrospinal fluid glucose to blood glucose of less than 0.23, a protein level of more than 220 mg per deciliter, or a white-cell count of more than 2000 per cubic millimeter)31 was present in 567 of 645 episodes (88 percent).
Cranial computed tomography (CT) was performed on admission in 496 episodes (71 percent); the results were normal in 325. Abnormalities were recorded in 171 episodes (34 percent): cerebral edema was identified in 48 of 496 episodes (10 percent), sinusitis or otitis in 48 (10 percent), evidence of recent infarction in 30 (6 percent), and hydrocephalus in 15 (3 percent). Thirty-eight episodes involved other abnormalities on cranial CT: old vascular lesions in 12, cerebral atrophy in 6, skull fracture in 5, pneumatocephalus in 5, arachnoid cyst in 3, vascular aneurysm in 2, subarachnoid hemorrhage in 1, brain-parenchyma hemorrhage in 1, subdural empyema in 1, meningioma in 1, and Dandy–Walker malformation in 1; in 8 episodes, two abnormalities were identified.
Cranial CT was performed before lumbar puncture in 337 of 696 episodes (48 percent). Focal neurologic deficits (not including cranial-nerve abnormalities), a score of less than 10 on the Glasgow Coma Scale, or both were present in 313 of 696 episodes (45 percent), and cranial CT was performed before lumbar puncture in 197 of these 313 episodes (63 percent). Therapy was initiated before CT in 35 percent of episodes in which CT was performed before lumbar puncture.
The most common microorganism was S. pneumoniae, accounting for 51 percent of isolates cultured from cerebrospinal fluid (Figure 1). The distribution of serotypes among 352 pneumococci is shown in Table 2. N. meningitidis was responsible for 37 percent of the episodes, with group B identified in 173 episodes, group C in 79, group Y in 3, group H in 1, and group W135 in 1. Gram's staining of cerebrospinal fluid revealed the microorganism in 524 of 652 episodes (sensitivity, 80 percent; specificity, 97 percent). The yield of Gram's staining was similar in patients who had previously received antimicrobial therapy and those who had not. Antibiotic susceptibility was tested in 351 pneumococcal and 256 meningococcal strains; 2 pneumococcal and 4 meningococcal strains showed intermediate resistance to penicillin; all other strains were sensitive to penicillin. Initial antibiotic treatment consisted of penicillin or amoxicillin in 43 percent of episodes, third-generation cephalosporins in 16 percent, and a combination of penicillin or amoxicillin with a third-generation cephalosporin in 25 percent; other regimens were used in 16 percent of episodes.
Table 2. Capsular Pneumococcal Serotypes from 352 Adults with Meningitis.
During the clinical course, focal neurologic abnormalities were found in half the episodes (Table 3); most were present on admission. Seizures occurred in 15 percent of episodes, and cardiorespiratory failure in 29 percent. Complications were significantly more likely to develop among patients with pneumococcal meningitis than among patients with meningococcal meningitis (P<0.001).
Table 3. Clinical Course, Outcome, and Neurologic Findings at Discharge.
The mortality rate was 21 percent (Table 3) and varied depending on the causative organism; it was 30 percent for pneumococcal meningitis, as compared with 7 percent for meningococcal meningitis (P<0.001) (Table 3) and 20 percent for meningitis due to other pathogens (P=0.05). Thirty-four percent of episodes had an unfavorable outcome. A neurologic examination was performed at discharge in 550 of 553 surviving patients (99 percent); the most common abnormalities identified were hearing loss (14 percent) and hemiparesis (4 percent).
Corticosteroids were administered in 118 episodes (17 percent), and episodes involving corticosteroid therapy were more likely to have an unfavorable outcome than were episodes that did not involve corticosteroid therapy (51 of 118 vs. 186 of 578 , P=0.03). Episodes in which corticosteroids were administered before antibiotics were less likely to have an unfavorable outcome than episodes in which corticosteroids were administered after antibiotics (3 of 24 vs. 48 of 94 , P=0.001). Opening pressures of more than 400 mm of water were recorded in 46 percent of patients who received corticosteroids and 38 percent of those not receiving corticosteroids.
In the multivariate model several characteristics were significantly associated with an unfavorable outcome (Table 4): advanced age, the presence of otitis or sinusitis, the absence of rash, a heart rate of more than 120 beats per minute, a low score on the Glasgow Coma Scale, a cerebrospinal fluid white-cell count of fewer than 1000 per cubic millimeter, a positive blood culture, an elevated erythrocyte sedimentation rate, and a reduced platelet count. Six other characteristics tended toward statistical significance: the presence of symptoms for less than 24 hours before admission, seizures, pneumonia, an immunocompromised state, a heart rate below 60 beats per minute, and hypotension (defined as a diastolic blood pressure of less than 60 mm Hg).
Table 4. Multivariate Analysis of Factors Associated with an Unfavorable Outcome.
The causative organism had an independent effect when it was added to the multivariate model. The odds of an unfavorable outcome was six times as high (95 percent confidence interval, 2.61 to 13.91; P<0.001) among patients infected with S. pneumoniae as among patients infected with N. meningitidis, even after adjustment for other clinical predictors.
During episodes of meningococcal meningitis, hypotension was more common among patients with a cerebrospinal fluid white-cell count of fewer than 100 cells per cubic millimeter than among patients with a higher cerebrospinal fluid white-cell count (8 of 18 vs. 24 of 211 , P<0.001). This association was absent during episodes of pneumococcal meningitis.
Discussion
We found that the prevalence of the classic triad of fever, neck stiffness, and an altered mental status is low among adults with community-acquired bacterial meningitis. However, almost all patients (95 percent) presented with at least two of the four symptoms of headache, fever, neck stiffness, and an altered mental status. In addition, a high percentage of patients (33 percent) were admitted with focal neurologic deficits. These percentages are similar to those found in retrospective studies; however, we found an exceptionally high prevalence of aphasia.4,5,6,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23
The majority of patients who presented with moderate or severe impairment of consciousness, neurologic deficits (not including cranial-nerve abnormalities), or both, which are contraindications to performing lumbar puncture, underwent CT before lumbar puncture (63 percent). However, a large number of patients without these "red flags" also underwent CT first, and many patients with such red flags did not undergo CT before lumbar puncture. In most patients who underwent CT before lumbar puncture, therapy was not initiated before CT was performed. Cranial CT has been recommended as a precaution before lumbar puncture to predict the likelihood and avoid the possibility of brain herniation.9,32,33 The withdrawal of cerebrospinal fluid removes counterpressure from below, thus increasing the effect of compression from above and exacerbating the brain shift already present.33
In patients with red flags indicating space-occupying lesions and in those with moderate or severe impairment of consciousness (as indicated by a score of less than 10 on the Glasgow Coma Scale), lumbar puncture should be preceded by cranial CT. However, a delay in the initiation of antimicrobial therapy may lead to a poor outcome, especially in patients with clinical deterioration.8 Therefore, in patients with suspected meningitis in whom cranial CT is performed before lumbar puncture, CT should be preceded by blood cultures and the initiation of antibiotic therapy and corticosteroids.33
Community-acquired bacterial meningitis has a high rate of an unfavorable outcome in adults (34 percent). Our multivariate model identified several unfavorable prognostic factors, most of which pointed to systemic compromise. A low level of consciousness on admission was predictive of an unfavorable outcome, as was a low cerebrospinal fluid white-cell count. Finally, factors predictive of pneumococcal infection were associated with an unfavorable outcome (advanced age; presence of otitis or sinusitis, pneumonia, or immunocompromised status; and absence of rash).
In our cohort, one of every six patients received corticosteroids. The European Dexamethasone Study showed adjunctive treatment with dexamethasone to be beneficial in adults with bacterial meningitis.27 Dexamethasone reduced the rate of an unfavorable outcome from 25 percent to 15 percent (relative risk, 0.59; 95 percent confidence interval, 0.37 to 0.94; P=0.03). A post hoc analysis showed that the beneficial effect of dexamethasone on pneumococcal meningitis was attributable to a reduction in systemic rather than neurologic complications.34 A systemic review showed that adjunctive corticosteroid therapy also reduced the frequency of neurologic sequelae among survivors.35 The increased likelihood of an unfavorable outcome among those receiving corticosteroids in our cohort was probably due to the fact that corticosteroids were used more frequently in patients with clinical deterioration or those with factors associated with a poorer prognosis.
A low cerebrospinal fluid white-cell count was associated with an adverse outcome. This association has been described before.22,36 In patients with meningococcal meningitis, a low cerebrospinal fluid white-cell count may be found in those with sepsis and early meningitis. Studies of animals with pneumococcal meningitis showed a relation among a large bacterial cerebrospinal fluid load, lack of response of cerebrospinal fluid leukocytes, and intracranial complications.37 It probably indicates excessive bacterial growth and lack of a leukocyte response in the cerebrospinal fluid.
In our study, patients with pneumococcal meningitis were at risk for an unfavorable outcome, even after correction for other clinical predictors. Thus, for clinicians, knowledge of the causative organism of meningitis is important in predicting the risk of an unfavorable outcome. Gram's staining of cerebrospinal fluid permits rapid and accurate identification of the causative bacteria in patients with bacterial meningitis and should therefore be routine.4,5,6,7,9
Several studies of prognostic factors in bacterial meningitis have been performed; however, all were retrospective and relatively small in size.4,5,6,10,11,12,13,14,15,16,17,18,19,20,21,22 Our study was nationwide and, therefore, we were able to study a representative sample of adults with acute bacterial meningitis. Furthermore, our prospective approach allowed us to collect comprehensive data on signs and symptoms, clinical course, and outcome. In addition, our large sample gave us the statistical power to perform multivariate analysis.
Our study has one important limitation: only patients who underwent lumbar puncture and who had a positive cerebrospinal fluid culture were included. Negative cerebrospinal fluid cultures occur in 11 to 30 percent of patients with bacterial meningitis.4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23 No significant differences in clinical presentation have been reported between patients with culture-positive bacterial meningitis and those with culture-negative bacterial meningitis.4,5,9 Therefore, it is unlikely that this factor confounded our results. Patients with space-occupying lesions on cranial CT do not undergo lumbar puncture,32,33 and patients with meningitis and a florid rash or septic shock also may not undergo lumbar puncture initially. Thus, an unknown number of such patients were excluded from the cohort, which may have resulted in an underestimation of the mortality rate, especially among patients with meningococcal meningitis.
Rates of antibiotic resistance among meningococcal and pneumococcal isolates were very low. Similar rates have been found in other studies in the Netherlands.26,38 In the United States, France, Spain, and other countries, antibiotic-resistant S. pneumoniae strains are highly prevalent and have emerged as a major problem in the treatment of patients with bacterial meningitis.39 In response to this epidemiologic trend, treatment recommendations for suspected and confirmed cases of bacterial meningitis continue to evolve.40 In areas with high rates of pneumococci that are resistant to broad-spectrum cephalosporins, a combination of a third-generation cephalosporin and vancomycin should be the initial choice for patients with acute bacterial meningitis.40 Treatment failures in patients with meningitis due to multidrug-resistant bacterial isolates have been described.9,41,42 The outcome was not significantly influenced by the presence of antibiotic-resistant pneumococcal isolates in several studies20,23,41,42; however, the relationship of antibiotic resistance to the outcome of meningitis remains to be elucidated.
Despite the fact that the case fatality rate is decreased by the use of adjunctive dexamethasone, there is still room for improvement. Prompt use of dexamethasone and appropriate antibiotics, together with optimal supportive care, can further reduce the mortality and morbidity associated with bacterial meningitis. Most risk factors for an unfavorable outcome that we identified indicated the presence of systemic compromise, a low level of consciousness, and infection with S. pneumoniae. Therefore, aggressive supportive care of patients with bacterial meningitis and systemic complications, preferably in specialized care units, is needed.
Supported in part by a grant from Roche Pharmaceuticals.
We are indebted to Professor P. McIntyre for his comments on the manuscript and to many physicians in the Netherlands for their cooperation.
Source Information
From the Departments of Neurology (D.B., J.G., M.W., M.V.), Medical Microbiology (L.S.), and Clinical Epidemiology and Biostatistics (J.B.R.), Academic Medical Center; and the Netherlands Reference Laboratory for Bacterial Meningitis (L.S.) — both in Amsterdam.
Address reprint requests to Dr. van de Beek at the Academic Medical Center, University of Amsterdam, Department of Neurology H2, P.O. Box 22660, 1100 DD Amsterdam, the Netherlands, or at d.vandebeek@amc.uva.nl.
References
Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995. N Engl J Med 1997;337:970-976.
Whitney CG, Farley MM, Hadler J, et al. Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N Engl J Med 2003;348:1737-1746.
Netherlands Reference Laboratory for Bacterial Meningitis. Bacterial meningitis in The Netherlands: annual report 2002. Amsterdam: University of Amsterdam, 2003.
Durand ML, Calderwood SB, Weber DJ, et al. Acute bacterial meningitis in adults: a review of 493 episodes. N Engl J Med 1993;328:21-28.
Sigurdardottir B, Bjornsson OM, Jonsdottir KE, Erlendsdottir H, Gudmundsson S. Acute bacterial meningitis in adults: a 20-year overview. Arch Intern Med 1997;157:425-430.
Hussein AS, Shafran SD. Acute bacterial meningitis in adults: a 12-year review. Medicine (Baltimore) 2000;79:360-368.
van de Beek D, Schmand B, de Gans J, et al. Cognitive impairment in adults with good recovery after bacterial meningitis. J Infect Dis 2002;186:1047-1052.
Aronin SI, Peduzzi P, Quagliarello VJ. Community-acquired bacterial meningitis: risk stratification for adverse clinical outcome and effect of antibiotic timing. Ann Intern Med 1998;129:862-869.
Tunkel AR. Bacterial meningitis. Philadelphia: Lippincott Williams & Wilkins, 2001.
Olsson RA, Kirby JC, Romansky MJ. Pneumococcal meningitis in the adult: clinical, therapeutic, and prognostic aspects in forty-three patients. Ann Intern Med 1961;55:545-549.
Dodge PR, Swartz MN. Bacterial meningitis -- a review of selected aspects. II. Special neurologic problems, postmeningitis complications, and clinicopathological correlations. N Engl J Med 1965;272:954-960.
Weiss W, Figueroa W, Shapiro WH, Flippin HF. Prognostic factors in pneumococcal meningitis. Arch Intern Med 1967;120:517-524.
Magnussen CR. Meningitis in adults; ten-year retrospective analysis at community hospital. N Y State J Med 1980;80:901-906.
Bohr V, Rasmussen N, Hansen B, et al. Pneumococcal meningitis: an evaluation of prognostic factors in 164 cases based on mortality and on a study of lasting sequelae. J Infect 1985;10:143-157.
Bruyn GAW, Kremer HPH, de Marie S, Padberg GW, Hermans J, van Furth R. Clinical evaluation of pneumococcal meningitis in adults over a twelve-year period. Eur J Clin Microbiol Infect Dis 1989;8:695-700.
Hoen B, Viel JF, Gerard A, Dureux JB, Canton P. Mortality in pneumococcal meningitis: a multivariate analysis of prognostic factors. Eur J Med 1993;2:28-32.
Kirkpatrick B, Reeves DS, MacGowan AP. A review of the clinical presentation, laboratory features, antimicrobial therapy and outcome of 77 episodes of pneumococcal meningitis occurring in children and adults. J Infect 1994;29:171-182.
Attia J, Hatala R, Cook DJ, Wong JG. The rational clinical examination: does this adult patient have acute meningitis? JAMA 1999;282:175-181.
Kragsbjerg P, Kallman J, Olcen P. Pneumococcal meningitis in adults. Scand J Infect Dis 1994;26:659-666.
Auburtin M, Porcher R, Bruneel F, et al. Pneumococcal meningitis in the intensive care unit: prognostic factors of clinical outcome in a series of 80 cases. Am J Respir Crit Care Med 2002;165:713-717.
Lu CH, Huang CR, Chang WN, et al. Community-acquired bacterial meningitis in adults: the epidemiology, timing of appropriate antimicrobial therapy, and prognostic factors. Clin Neurol Neurosurg 2002;104:352-358.
Kastenbauer S, Pfister HW. Pneumococcal meningitis in adults: spectrum of complications and prognostic factors in a series of 87 cases. Brain 2003;126:1015-1025.
Flores-Cordero JM, Amaya-Villar R, Rincon-Ferrari MD, et al. Acute community-acquired bacterial meningitis in adults admitted to the intensive care unit: clinical manifestations, management and prognostic factors. Intensive Care Med 2003;29:1967-1973.
Statistics Netherlands. StatLine, Voorburg/Heerlen, 2001. (Accessed October 4, 2004, at http://www.cbs.nl.)
Jennett B, Teasdale G. Management of head injuries. 2nd ed. Philadelphia: F.A. Davis, 1981.
van de Beek D, de Gans J, Spanjaard L, Vermeulen M, Dankert J. Antibiotic guidelines and antibiotic use in adult bacterial meningitis in The Netherlands. J Antimicrob Chemother 2002;49:661-666.
de Gans J, van de Beek D. Dexamethasone in adults with bacterial meningitis. N Engl J Med 2002;347:1549-1556.
de Gans J, van de Beek D. Dexamethasone in adults with bacterial meningitis. N Engl J Med 2003;348:955-956.
Little RJA, Rubin DB. Statistical analysis with missing data. New York: Wiley, 1987.
Arnold AM, Kronmal RA. Multiple imputation of baseline data in the Cardiovascular Health Study. Am J Epidemiol 2003;157:74-84.
Spanos A, Harrell FE Jr, Durack DT. Differential diagnosis of acute meningitis: an analysis of the predictive value of initial observations. JAMA 1989;262:2700-2707.
Hasbun R, Abrahams J, Jekel J, Quagliarello VJ. Computed tomography of the head before lumbar puncture in adults with suspected meningitis. N Engl J Med 2001;345:1727-1733.
van Crevel H, Hijdra A, de Gans J. Lumbar puncture and the risk of herniation: when should we first perform CT? J Neurol 2002;249:129-137.
van de Beek D, de Gans J. Dexamethasone and pneumococcal meningitis. Ann Intern Med 2004;141:327-327.
van de Beek D, de Gans J, McIntyre P, Prasad K. Steroids in adults with bacterial meningitis: a systematic review. Lancet Infect Dis 2004;3:139-143.
McMillan DA, Lin CY, Aronin SI, Quagliarello VJ. Community-acquired bacterial meningitis in adults: categorization of causes and timing of death. Clin Infect Dis 2001;33:969-975.
Tauber MG, Kennedy SL, Tureen JH, Lowenstein DH. Experimental pneumococcal meningitis causes central nervous system pathology without inducing the 72-kd heat shock protein. Am J Pathol 1992;141:53-60.
van de Beek D, Hensen EF, Spanjaard L, de Gans J, Enting RH, Dankert J. Meropenem susceptibility of Neisseria meningitidis and Streptococcus pneumoniae from meningitis patients in The Netherlands. J Antimicrob Chemother 1997;40:895-897.
Whitney CG, Farley MM, Hadler J, et al. Increasing prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. N Engl J Med 2000;343:1917-1924.
Aronin SI. Current pharmacotherapy of pneumococcal meningitis. Expert Opin Pharmacother 2002;3:121-129.
Arditi M, Mason EO Jr, Bradley JS, et al. Three-year multicenter surveillance of pneumococcal meningitis in children: clinical characteristics, and outcome related to penicillin susceptibility and dexamethasone use. Pediatrics 1998;102:1087-1097.
Kellner JD, Scheifele DW, Halperin SA, et al. Outcome of penicillin-nonsusceptible Streptococcus pneumoniae meningitis: a nested case-control study. Pediatr Infect Dis J 2002;21:903-910.
Related Letters:
Prognostic Factors in Adults with Bacterial Meningitis
Joffe A. R., ?stergaard C., Klussmann J. P., Guntinas-Lichius O., Altschuler E. L., van de Beek D., de Gans J., Swartz M. N.(Diederik van de Beek, M.D)