Etiology, Clinical Presentation, and Outcomes of Bacterial Meningitis in Adult Patients: A Retrospective Study in Lithuania (2018–2021)

Introduction

Acute bacterial meningitis remains a devastating infectious disease with substantial mortality rates globally [1]. According to the World Health Organization, around 1 in 6 people who get bacterial meningitis die and 1 in 5 have severe complications [2]. For example, bacterial meningitis was responsible for approximately 500 deaths annually between 2003 and 2007 in the United States [3]. Developing countries face an even higher burden of disease [4].

Despite progress in the prevention of community-acquired bacterial meningitis by implementing intensive vaccination programs against the 3 most common causative pathogens – Streptococcus pneumoniae, Neisseria meningitidis, and Haemophilus influenza – diagnostic and treatment dilemmas still occur [5]. For example, typical clinical characteristics of bacterial meningitis – headache, neck stiffness, fever – are only present in 40–50% of patients [6]. Further, such symptoms are not specific [7]. Due to absence of typical clinical features, treatment of bacterial meningitis may be delayed [8]. Moreover, efficacy of vaccines is also limited due to emergence of non-vaccine and non-typeable serotypes [9]. Lastly, treatment of bacterial meningitis may also be intricate. For example, data shows that dexamethasone therapy can improve outcomes of patients with bacterial meningitis, especially caused by S. pneumoniae [10,11]. However, a prospective cohort study found a reduced survival rate of patients with neurolisteriosis treated with adjunctive dexamethasone [12].

Studies also indicate various clinical and laboratory variables that may determine the outcome of patients with acute bacterial meningitis [6,13–16]. For example, 1 study highlighted that factors indicative of systemic compromise, a low level of consciousness, and infection with S. pneumoniae are among the strongest risk factors for an unfavorable outcome [6].

According to the data from the National Information System on Communicable Diseases and Infectious Agents, the incidence of bacterial meningitis in Lithuania in 2019 was 0.9/100 000 [17]. From 2010 to 2019, new yearly cases of bacterial meningitis in Lithuania ranged from 26 to 78 [17], and almost all these patients were hospitalized [17]. Similar prevalence rates have been estimated in other European countries, such as The Netherlands and Finland [18,19]. A few studies of causes, socio-demographic factors, clinical presentation, and outcomes of infectious diseases have been conducted in Lithuania [15,20]. For example, Matulyte et al investigated data of patients with a diagnosis of bacterial meningitis treated between 2009 and 2016 [15]. However, since then, no known studies have been conducted to investigate community-acquired bacterial meningitis cases in Lithuania.

As current data on causes, diagnostics, treatment, and prediction of outcomes of bacterial meningitis are heterogeneous, and morbidity and mortality rates remain high, this retrospective study aimed to identify the causes, presentation, and outcomes of community-acquired bacterial meningitis in 86 adults in Vilnius, Lithuania between 2018 and 2021.

Material and Methods

ETHICS STATEMENT:

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Regional Biomedical Research Ethics Committee of Vilnius (no. 158200-18-984-490). All medical records analyzed in the study were completely de-personalized; thus, an informed consent statement was not required.

STUDY SETTINGS:

A retrospective study was performed by analyzing data extracted from the Vilnius University Hospital Santaros Clinics database. The current population of Lithuania is around 2.62 million people and Vilnius is the capital and the largest city in Lithuania. Vilnius University Hospital Santaros Clinics is the main reference center for adult neuroinfectious diseases in Vilnius.

STUDY SUBJECTS:

A retrospective analysis was performed of adult patients (18 years or older) diagnosed with acute bacterial meningitis from January 1, 2018, through December 31, 2021, at the Infectious Diseases Center or Center for Neurology of Vilnius University Hospital Santaros Clinics. Inclusion criteria were: patients with a diagnosis of bacterial meningitis (ICD-10: A32.1, A39.0, G00.0-3, G00.8-9, G03.8, G04.2). Exclusion criteria were: 1) patients with meningitis caused by Borrelia burgdorferi and Mycobacterium tuberculosis due to significant differences in the course of the disease; 2) incomplete medical records.

DATA COLLECTION:

Demographic factors (age, sex), clinical characteristics (symptoms, onset of symptoms, chronic diseases, meningeal signs during first neurological examination after admission, site, and duration of hospitalization), laboratory results of CSF sampling (white blood cell count,% of granulocytes, protein concentration), and a specific aspect of treatment – administration of glucocorticosteroids – were analyzed.

To assess details of differences in demographic, clinical and laboratory variables based on etiological agents, rare cases (S. aureus, H. influenzae, and G+ cocci) in the study were considered “others.” Bacterial meningitis of unknown origin was determined by a negative CSF culture and neutrophilic pleocytosis.

STATISTICAL ANALYSIS:

Statistical analysis of the data was conducted in Microsoft Excel v16 and IBM SPSS v26. Categorical variables were expressed as counts and percentages, and frequency distributions were assessed with Fisher’s exact test. Continuous variables were expressed as the median and minimal–maximal values. Nonparametric tests (Mann-Whitney U or Kruskal-Wallis) were used to identify differences between groups in continuous variables. Unfavorable outcome was defined as death. Univariate analysis was used to explore unadjusted associations between variables and outcome. P values less than 0.1 were included for further multivariate analysis. The likelihood ratio test was used to build a multivariate model. Odds ratios and 95% confidence intervals were used to quantify the strength of these associations. Forest plots were created to visualize odds ratios for an unfavorable outcome. All statistical tests were two-tailed, and P values less than 0.05 were considered to indicate statistical significance.

Results

GENERAL CHARACTERISTICS:

We extracted data on 114 patients diagnosed with bacterial meningitis. Of them, 86 (75.44%) were included in further analysis. Fifty-four (62.79%) patients were men. The median age of patients was 58 years (range, 18–83). Median time from onset of symptoms to hospitalization was 2 days (range, 1–14). The most common clinical features at presentation were headache (66.28%), fever (56.98%), fatigue (53.49%), confusion/sleepiness (52.33%), and other symptoms (Figure 1). Meningeal signs (neck stiffness and/or positive Kernig’s sign) were identified in 73.8% of all cases.

Causative agents of meningitis were identified in 57 (66.28%) cases. Of confirmed cases, meningitis was most often caused by Listeria monocytogenes (17, 19.8%), followed by Streptococcus pneumoniae (16, 18.6%), Neisseria meningitidis (16, 18.6%), and other pathogens (Table 1).

Median time of hospitalization was 20 (range, 3–92) days. Fifty-one (59.3%) patients were first hospitalized in the ICU. During hospitalization, 64 patients (74.4%) were treated in the ICU. The median duration of hospitalization in the ICU was 4 (range, 1–37) days. The median number of white blood cells (cytosis) in CSF samples was 1181 (range, 25–32 426) cells per μl. The median percentage of polymorphonuclear cells in CSF samples was 88%. The median neutrophil-to-lymphocyte ratio was 6.82, and the median concentration of protein in CSF samples was 2.97 (range, 0.29–27.88). Treatment with glucocorticoids was administered to 54 patients (62.79%).

ETIOLOGICAL FACTOR-BASED COMPARISON:

Patients with meningitis caused by L. monocytogenes were the oldest (median of 70 years, compared to 67 years in the group “other”, 61 years in the S. pneumoniae group, and 43 years in the N. meningitidis group, P=0.003) and had the longest hospitalization (median of 46 days, compared to 28 days in the S. pneumoniae group, 17 days in the “other” group, and 13 days in the N. meningitidis group, P<0.001). Further, patients with meningitis caused by L. monocytogenes had the lowest levels of white blood cells in the CSF samples (median of 466 cells, compared to 2991 cells in the N. meningitidis group, 3988 cells in the “other” group, and 4738 cells in the S. pneumoniae group, P=0.002). Monomorphonuclear cells dominated (84.2%, whereas the percentage of monomorphonuclear cells in “other” groups ranged from 2.9% to 7%, P=0.004). Among patients with meningococcal meningitis, 81.2% indicated they felt fatigue, compared to 62.5% in the “other” group, 41.2% in the L. monocytogenes group, and 25% in the S. pneumoniae group (P=0.01) and were most often also treated with glucocorticosteroids (87.5%, compared to 62.5% in the “other” and S. pneumoniae group and 35.3% in the L. monocytogenes group, P=0.02). Patients with meningitis caused by S. pneumoniae often did not have meningeal signs (46.7%, compared to 29.4%, 14.3%, and 6.2% in the L. monocytogenes, “other,” and N. meningitidis groups, respectively, P=0.06), and most often were confused and sleepy (75%, compared to 25% in the N. meningitidis group, 62.5% in the “other”, group and 64.7% in the L. monocytogenes group, P=0.03).

Differences in demographic, clinical, laboratory, and hospitalization-related characteristics based on all etiological factors of meningitis and unknown origin of meningitis are summarized in Tables 2–4.

OUTCOMES:

Twelve (13.95%) patients died. Based on univariate regression analysis, older patients (65 years or more) were more likely to die from bacterial meningitis in our sample (OR 6.65 [1.65–26.88]). Further, <100 CSF WBC count was associated with higher risk of death (OR 5.17 [1.42–18.76]). In contrast, the presence of headache was related with a favorable outcome (OR 0.12 [0.03–0.50]). Factors and their impact on the outcome are presented in Table 5 and Figure 2. However, multivariate analysis did not reveal factors that may be significant in predicting unfavorable outcome (Table 6).

Discussion

Our study provides additional data regarding causes, presentation, and outcomes of community-acquired bacterial meningitis in adults. Unexpectedly, Listeria monocytogenes was the most common etiological factor of bacterial meningitis cases in the current sample. The literature suggests that 4 out of 5 cases of bacterial meningitis in adults are caused by either Neisseria meningitidis or Streptococcus pneumoniae [21,22]. Listeria monocytogenes is the third most frequent cause [23,24]. The latter pathogen mainly causes meningitis in the elderly and immunocompromised patients [16,25]. Further, the incidence of listeria meningitis tends to be low in countries where food hygiene requirements are well regulated [26]. Such discrepancy between global and current-sample prevalence of listeria meningitis may be partially explained by the relatively low sample of patients analyzed in the study. Further, meningitis caused by L. monocytogenes accounted for only 3.1% of all cases in the previous study from Lithuania, in which similar methodological procedures were used. Subjects in the current sample were significantly older than in the study by Matulyte et al (58 vs 36 years) [15]. However, the incongruent results shed a light on the need to be attentive to possible immunodeficiency conditions among patients with listeria meningitis, as none of the patients in our sample had documented immunodeficiency status before or during hospitalization.

Further, our study revealed that the most common clinical manifestations of bacterial meningitis in adults were headache, fever (≥38.5°C), fatigue, and confusion/sleepiness. These findings are quite similar compared to those found in the earlier Lithuanian study, which showed that the most common symptoms of bacterial meningitis were headache, fever (≥38°C), and nausea [15]. However, headache was reported in 84.3% of patients, more frequently than in our study (66.28%). In addition, we compared our findings to a Dutch study in which headache (87%) and fever (77%) occurred more often as well [6]. However, another retrospective study from western Canada noted more commensurate results as 66% of patients had headache and 56% had confusion [24]. On one hand, such differences again shed light on methodological inconsistencies between studies that prevent comparison of the research. On the other hand, the results of the present study highlight that only a complex of characteristic symptoms allows diagnosing bacterial meningitis accurately, as solitary symptoms are nonspecific.

Our study indicates that the highest CSF protein level was in the S. aureus and S. pneumoniae meningitis groups, whereas the lowest CSF protein level was in the N. meningitidis meningitis group. The highest CSF white cell count was in the G+ cocci (n=1), S. pneumoniae (n=16), and H. influenzae (n=2) meningitis groups. The previously mentioned study from Lithuania observed the highest CSF white cell count in the H. influenzae, S. aureus, and S. pyogenes meningitis groups [15]. The lowest CSF white cell count was found in the L. monocytogenes meningitis group both in our study (n=17) and the latter Lithuanian study. The literature suggests that listeria meningitis often presents with relatively low CSF white cell counts rather than with characteristic CSF findings of bacterial meningitis [4]. However, a prospective nationwide study in the Netherlands revealed that there was no relation between low CSF white cell count in L. monocytogenes cases and age or immunocompetence [25].

Acute changes in the level of consciousness, severe sepsis, or septic shock with multiple-organ failure are the main reasons for admission of patients with meningitis to the ICU [16]. In our study, 74.4% of patients were treated in the ICU, and 59.3% of patients were admitted directly to the ICU from the hospital emergency department. The need for hospitalization in the ICU was more common in the S. pneumoniae group, but the duration of hospitalization in the ICU was significantly longer in the S. aureus group. We consider that one of the reasons patients with bacterial meningitis tend to be hospitalized in the ICU is the prevalence of non-susceptible strains, which can complicate the choice of treatment [22,27]. In addition, bacterial meningitis can cause severe central nervous system (CNS) complications that require intensive care [28].

In this study, the mortality rate was 13.9% and did not differ significantly by etiological factor. In contrast to the latter finding, a previously mentioned Dutch study revealed that patients with S. pneumoniae meningitis were at risk for an unfavorable outcome [6]. In addition, another study indicated non-meningococcal etiology to be an independent predictor for unfavorable outcome [29]. We found that older patients (65 years or more) and patients with <100 CSF white cell count were at risk for an unfavorable outcome. The latter variables as risk factors were also noted in other studies [15,18]. In contrast to a previous study, we did not observe that protein concentration in the CSF were related with the outcome of bacterial meningitis cases [30]. Additionally, the literature suggests that other factors such as the absence of fever and low level of consciousness at admission, a delay in antibiotic therapy initiation, and comorbidities are also associated with an adverse outcome [15,16]. However, we determined that headache was correlated with a favorable prognosis. This finding can be explained as headache is a more typical symptom of bacterial meningitis, which makes it easier to suspect, diagnose the disease, and to start proper treatment earlier [6].

Our study has a few limitations to mention. Firstly, the retrospective nature of the study did not allow us to minutely examine some variables in the diagnosis and management of bacterial meningitis. For example, due to lack of information in the medical records we were not able to measure the exact period from the time of hospitalization to administration of antibiotics. The latter aspect may be of great significance when predicting outcomes of bacterial meningitis cases [29–31]. Moreover, the number of patients in the current sample was relatively small. Further, our study was limited to 1 tertiary care center hospital; thus, the patients may not accurately represent the population of bacterial meningitis cases within Lithuania.

Conclusions

Bacterial meningitis remains a dangerous and fatal condition, whether left untreated or, in some cases, even when treated properly. By analyzing the data of the patients hospitalized with bacterial meningitis, this study established that clinical characteristics and laboratory results differ based on etiological factors of the disease. The study indicates that Listeria monocytogenes is the most common causative agent of bacterial meningitis cases treated at Vilnius University Hospital Santaros Clinics. Furthermore, age of >65 and low CSF WBC count (<100 cells per μL) were associated with poor prognosis, whereas headache was related with a favorable prognosis of bacterial meningitis. Future research on community-acquired bacterial meningitis should focus on establishing more specific prognostic factors for outcomes of the disease.