Clinical data and factors associated with poor outcome in pneumococcal meningitis

Clinical Picture and Risk Factors for Poor Outcome in Streptococcus pneumoniae Meningitis of Childhood on Three Continents

BACKGROUND

Streptococcus pneumoniae meningitis (SpM) remains a major health burden worldwide, particularly in low- and middle-income countries. Identifying the patients at highest risk for mortality and disabling sequelae may reveal potentially avoidable predisposing factors and identify patients most in need of intensive care. We searched for factors that do not require laboratory facilities.

METHODS

This study was a secondary analysis of prospectively collected data from 5 clinical trials of childhood bacterial meningitis on 3 continents between 1984 and 2017. SpM cases were analyzed by study site and predictors for poor outcome (death or severe sequelae) were identified from the whole series, Latin America and Angola.

RESULTS

Among a total of 1575 children (age range: 2 months to 15 years), 505 cases were due to pneumococci. Compared to other etiologies, SpM doubled the death rate (33% vs. 17%) and tripled poor outcome (15% vs. 6%). In SpM, Glasgow Coma Score <13 [odds ratio (OR): 4.73] and previous antibiotics in Angola (OR: 1.70) were independent predictors for death. Predictors for poor outcome were age <1 year (OR: 2.41) and Glasgow Coma Score <13 (OR: 6.39) in the whole series, seizures in Latin America (OR: 3.98) and previous antibiotics in Angola (OR: 1.91). Angolan children had a 17-fold increased risk for poor outcome when compared with Finnish children ( P = 0.011).

CONCLUSIONS

Our study proved the severity of SpM when compared with other etiologies. The outcome was especially poor in Angola. Most patients at risk for poor outcome are easily identified by clinical factors on admission.

Risk factors and pathogen characteristics associated with unfavorable outcomes among adults with pneumococcal meningitis in Japan, 2006 to 2016

PURPOSE

In this study, we aimed to clarify the risk factors associated with unfavorable outcomes in adults with pneumococcal meningitis (PnM).

METHODS

Surveillance was conducted between 2006 and 2016. Adults with PnM (n = 268) were followed up for outcomes within 28 days after admission using the Glasgow Outcome Scale (GOS). After classifying the patients into the unfavorable (GOS1-4) and favorable (GOS5) outcome groups, i) the underlying diseases, ii) biomarkers at admission, and iii) serotype, genotype, and antimicrobial susceptibility for all isolates were compared between both groups.

RESULTS

Overall, 58.6% of patients with PnM survived,15.3% died, and 26.1% had sequelae. The number of living days in the GOS1 group was highly heterogeneous. Motor dysfunction, disturbance of consciousness, and hearing loss were the commonest sequelae. Of the underlying diseases identified in 68.9% of the PnM patients, liver and kidney diseases were significantly associated with unfavorable outcomes. Of the biomarkers, creatinine and blood urea nitrogen, followed by platelet and C-reactive protein had the most significant associations with unfavorable outcomes. There was a significant difference in the high protein concentrations in the cerebrospinal fluid between the groups. Serotypes 23F, 6C, 4, 23A, 22F, 10A, and 12F were associated with unfavorable outcomes. These serotypes were not penicillin-resistant isolates possessing three abnormal pbp genes (pbp1a, 2x, and 2b), except for 23F. The expected coverage rate of the pneumococcal conjugate vaccine (PCV) was 50.7% for PCV15 and 72.4% for PCV20.

CONCLUSIONS

In the introduction of PCV for adults, the risk factors for underlying diseases should be prioritized over age, and serotypes with unfavorable outcomes should be considered.

Otitis Media-associated Bacterial Meningitis in Children in a Low-income Country

BACKGROUND

Otitis media (OM) is a common childhood infection that may result in bacterial meningitis (BM). However, OM-associated BM remains poorly characterized. We aimed to study the occurrence, clinical presentation and outcome of this type of childhood BM in Luanda, Angola.

METHODS

Five hundred twelve children from our previous clinical BM trial, with the ear meticulously examined, were analyzed whether they had or not OM, and according to their age, ≤12 month old and >12 month old. Prospectively collected clinical data, laboratory test results and outcome for these groups were assessed.

RESULTS

Sixty-two children (12%) had OM-associated BM, of whom 39 had otorrhea. Ear discharge was more common in older children (median age 45 months old vs. 12 months old; P < 0.001). Children with OM often showed an additional infectious focus (n = 20, 32% vs. n = 82, 18%; P = 0.016), were dehydrated (n = 16, 26% vs. n = 66, 15%; P = 0.04), and showed higher odds of complicated clinical course or death (odds ratios 2.27, 95% CI: 1.004-5.15, P = 0.049) compared with children without OM. The >12-month-old children with OM often arrived in poor clinical condition with coma and/or ptosis. Otorrhea was associated with HIV positivity. Infants with otorrhea frequently lived under poor socioeconomic conditions.

CONCLUSIONS

Children with OM-associated BM were prone to many problems, such as being especially ill at presentation, undergoing a difficult clinical course and showing a higher risk of complicated or fatal outcome. HIV infection and malnutrition were common in children with otorrhea, which was also associated with low socioeconomic status.

Adult Patients with Pneumococcal Meningitis at a Neurosurgical Neurologic Center: Different Predisposing Conditions?

BACKGROUND

In previous studies of pneumococcal meningitis in adults within general hospitals or national cohorts, the most common predisposing conditions were otitis media, sinusitis, pneumonia, immunosuppression, alcoholism, and diabetes. The epidemiology of pneumococcal meningitis is changing because of the use of vaccines in childhood, and antibiotic resistance has increased.

METHODS

We retrospectively reviewed the cases of patients with diagnoses of pneumococcal meningitis proved by an inflammatory cerebrospinal fluid (CSF) with a positive culture, treated during a period of 14 years at an adult neurosurgical neurologic referral center. Our aim was to define their clinical course, predisposing conditions, antimicrobial susceptibilities, and outcome.

RESULTS

We reviewed the cases of 30 patients, 17 men and 13 women, with a mean age of 36.7 years. Fourteen patients (46.6%) had previous neurosurgery, 12 patients (40%) had CSF fistula, 8 had remote head trauma, and 8 also presented recurrent meningitis. Resistance to ceftriaxone or vancomycin was less than 5%, and penicillin resistance was 53%. Eight patients (26.7%) had died. An increased risk of death was associated with coma at admission, septic shock, upper gastrointestinal bleeding, mechanical ventilation, thrombocytopenia, and a low CSF opening pressure.

CONCLUSIONS

We conclude that patients with pneumococcal meningitis treated at neurosurgical neurologic centers have different predisposing conditions with severe disease and high mortality, thus prompting us to recommend aggressive pneumococcal vaccination in patients with CSF leaks and severe head trauma. Prospective studies to identify which neurosurgical patients may benefit in the long term from a pneumococcal vaccine are urgently needed.

Systematic Review: Incidence and Course of Hearing Loss Caused by Bacterial Meningitis: In Search of an Optimal Timed Audiological Follow-up

OBJECTIVE

The objective of this study was to clarify the incidence and course of hearing loss after bacterial meningitis to optimize the audiological follow-up.

DATA SOURCES

The databases Embase, Medline (OvidSP), Web-of-science, Scopus, Cinahl, Cochrane, PubMed publisher, and Google Scholar were used. Only articles written in English were included.

STUDY SELECTION

Articles published from 1985 until March 2015 describing the incidence, risk factors, or course of hearing loss after meningitis were used.

DATA EXTRACTION

The quality of the studies was assessed on three aspects: quality of audiometry, number of patients, and methodological quality.

DATA SYNTHESIS

For each publication, data were entered in spreadsheet software for analysis. The data were analyzed and interpreted using best evidence synthesis.

CONCLUSIONS

The overall quality of the included studies was poor. A major drawback was the quality of the (description of) audiometry, severity, and timing of hearing loss.A systematic review of the literature showed an incidence of hearing loss (>25 ± 5  dB) of 14% and an incidence of 5% for profound hearing loss (>90  dB). Patients with initial normal hearing after meningitis showed stable normal hearing over time. However, initial hearing loss related to meningitis can improve or deteriorate over time.We therefore recommend early audiological testing of all patients who suffered bacterial meningitis. However, long-term audiological follow-up is only needed for patients with early onset hearing loss and not for patients with normal hearing at the first hearing test.

Neurological sequelae of bacterial meningitis

OBJECTIVES

We reported on occurrence and impact of neurological sequelae after bacterial meningitis.

METHODS

We reviewed occurrence of neurological sequelae in children and adults after pneumococcal and meningococcal meningitis.

RESULTS

Most frequently reported sequelae are focal neurological deficits, hearing loss, cognitive impairment and epilepsy. Adults with pneumococcal meningitis have the highest risk of developing focal neurological deficits, which are most commonly caused by cerebral infarction, but can also be due to cerebritis, subdural empyema, cerebral abscess or intracerebral bleeding. Focal deficits may improve during clinical course and even after discharge, but a proportion of patients will have persisting focal neurological deficits that often interfere in patient's daily life. Hearing loss occurs in a high proportion of patients with pneumococcal meningitis and has been associated with co-existing otitis. Children and adults recovering from bacterial meningitis without apparent neurological deficits are at risk for long-term cognitive deficits. Early identification of neurological sequelae is important for children to prevent additional developmental delay, and for adults to achieve successful return in society after the disease.

CONCLUSIONS

Neurological sequelae occur in a substantial amount of patients following bacterial meningitis. Most frequently reported sequelae are focal neurological deficits, hearing loss, cognitive impairment and epilepsy.

The kynurenine pathway contributes to long-term neuropsychological changes in experimental pneumococcal meningitis

Pneumococcal meningitis is a lethal form of bacterial infection in the central nervous system that often causes lifelong neurological sequelae, despite therapeutic advances. The contemporary view is that the inflammatory response to infection contributes to the functional disabilities among survivors of this disease. We previously have established a mouse model of neurobehavioural deficits, using an automated IntelliCage™ system that revealed long-term behavioural and cognitive deficits in C57BL/6J female mice cured of meningitis by ceftriaxone treatment. We now have investigated the roles of two kynurenine pathway enzymes, indoleamine dioxygenase-1 (IDO1) and tryptophan dioxygenase-2 (TDO2), in the pathomechanisms of pneumococcal meningitis. Since tryptophan metabolism has long been implicated in behavioural and cognitive modulation through the production of neuroactive compounds, we hypothesised that preventing the actions of these enzymes through gene knockout would be beneficial in mice subjected to pneumococcal infection. We found no significant effect of IDO1 or TDO2 on mortality. Post-meningitic wild-type mice showed long-term diurnal hypoactivity and nocturnal hyperactivity when they were exposed to an Intellicage adaptation test throughout both the light and dark phases. These changes were not apparent in IDO1(-/-) survivors, but were present in the TDO2(-/-) survivors. Both IDO1(-/-) and TDO2(-/-) survivors were not protected against developing long-term cognitive deficits as measured in IntelliCage-based patrolling or reversal tasks. Collectively, these observations suggest (i) involvement of the kynurenine pathway in causing some behavioural sequelae of pneumococcal meningitis and (ii) that this pathway might operate synergistically with, or independently of, other pathways to cause other aspects of neurological sequelae.

Factors influencing neurological outcome of children with bacterial meningitis at the emergency department

We performed a cohort study of children who survived bacterial meningitis after the neonatal period at a single pediatric center in France over a 10-year period (1995-2004) to identify predictors of death and long-term neurological deficits in children with bacterial meningitis. We performed multivariate regression to determine independent predictors of death and neurologic deficits. We identified 101 children with bacterial meningitis of which 19 died during initial hospitalization. Need for mechanical ventilation [hazard ratio (HR) 11.5, 95 % confidence interval (CI) 2.4-55.5)] and thrombocytopenia defined as a platelet count <150 × 10(9) per liter (HR 0.6, 95 % CI 0.4-0.9) at presentation were associated with death during initial hospitalization. At final assessment, 42 of the 70 survivors had no neurologic deficits identified; 20 had a single deficit, and eight had multiple deficits. A delay in initiation of antibiotics (HR 1.3, 95 % CI 1.1-1.7) and hydrocephalus on computed tomographic scan (HR 2.6, 95 % CI 1.1-6.0) were associated with having one or more long-term neurologic deficits. Identification of children at risk of death or long-term neurologic sequelae may allow therapeutic interventions to be directed to children at the highest risk.

Long-term outcomes of pneumococcal meningitis in childhood and adolescence

A vaccine to prevent pneumococcal meningitis (PM) has recently been introduced. However, contemporary data to inform cost-effectiveness analysis and justify its routine use are sparse. We examined the cognitive, educational, psychological and social outcomes of PM in childhood. We completed a population-based case-control study in two regions of the UK. Children and young people currently between 3 and 20 years of age that had been diagnosed with PM ≤14 years of age were identified from active regional surveillance. Controls were siblings or neighbours of similar age. Standardised questionnaires and neuropsychological testing was administered to assess IQ, educational attainments, memory, psychological distress, quality of life and hearing impairment. Data were available on 97 patients and 93 controls. Eighty-four patients had a sibling/neighbour-matched control. Both matched and unmatched analyses were completed, and results of the 84 matched comparisons were highly similar to the unmatched. For the total sample, controls were similar in age, ethnicity and socioeconomic status. Median age at meningitis was 11 months. Median time between meningitis and assessment was 6.0 years. In the matched analysis, partial or profound hearing impairment was reported in 14% of patients and 1% of controls. Patients had significantly lower mean full-scale IQ (p = 0.05), verbal IQ (p = 0.0008), numeracy (p = 0.02), total quality of life (p = 0.04), school functioning (p = 0.005), psychosocial functioning (p = 0.001) and psychological difficulties (p = 0.01). Parents of patients reported greater functional disability (p = 0.008), impairment in all aspects of quality of life (p = 0.001) and psychological difficulties (p < 0.0006). Findings for IQ were not materially different when analyses were repeated only in those without hearing impairment. In multivariate regression analysis that included both case-control status and hearing status, both being a patient (p = 0.001) and having profound hearing impairment (p = 0.001) were independently associated with lower full-scale IQ. Conclusions Pneumococcal meningitis is associated with major sequelae. Our findings strongly support the introduction of pneumococcal conjugate vaccine as part of routine childhood vaccination programmes internationally.

Early prediction of postmeningitic hearing loss in children using magnetic resonance imaging

OBJECTIVE

To determine whether early gadolinium-enhanced magnetic resonance imaging (GdMRI) can reliably detect meningitic labyrinthitis and thereby predict which children are at high risk for hearing loss. Permanent sensorineural hearing loss (SNHL) remains a common sequela of bacterial meningitis, and early diagnosis of the associated suppurative labyrinthitis can be difficult, especially in critically ill, sedated patients and young children.

DESIGN

Retrospective cohort study.

SETTING

Tertiary pediatric hospital.

PARTICIPANTS

Twenty-three survivors of bacterial meningitis (median age, 15 months [range, 3 months-14 years]) who had undergone brain GdMRI during the acute disease and had subsequent ear-specific audiometric data.

MAIN OUTCOME MEASURE

Blinded to disease and outcome, a neuroradiologist rated the relative enhancement of each cochlea on T1-weighted images using a 4-point scale. Scores were then correlated with the degree of hearing loss on subsequent testing.

RESULTS

Sensorineural hearing loss occurred in 15 of 46 ears (8 of 23 patients). Enhancement on GdMRI was detected in 13 of the 15 ears that later developed SNHL but was absent in all 31 unaffected ears. Thus, GdMRI was 87% sensitive and 100% specific for predicting which ears would develop permanent SNHL. In the subgroup with pneumococcal meningitis (n = 15), GdMRI was 100% sensitive and 100% specific. Labyrinthine enhancement was detectable as early as 1 day after diagnosis.

CONCLUSION

Gadolinium-enhanced MRI detected meningitic labyrinthitis at early stages and accurately predicted which patients would later develop hearing loss.

Rebound in the incidence of pneumococcal meningitis in northern France: effect of serotype replacement

AIM

To determine whether the recent emergence of nonvaccine pneumococcal serotypes has affected the incidence of pneumococcal meningitis in children a few years after the introduction of the heptavalent pneumococcal conjugate vaccine.

METHODS

We conducted a multicentre retrospective cohort study from 2005 to 2008 in all hospitals with paediatric units in northern France. It included all patients < 18 years of age who were admitted for laboratory-confirmed pneumococcal meningitis during the study period. Data were collected from medical files and laboratory records at each hospital and compared with the regional hospital discharge codes. We assessed global and age-specific incidence rates of pneumococcal meningitis from 2005 through 2008, compared them with those from the prevaccine era (2000-2002) and evaluated pneumococcal serotypes.

RESULTS

In all, 41 cases were found during the study period. The incidence rate of pneumococcal meningitis varied from 0.8/100,000 children < 18 years in 2005 to 1.8/100,000 children in 2008 (2.2-fold increase, p = 0.06); and from 1.8 to 11.9/100,000 children < 2 years (6.5-fold increase, p = 0.004). This increase was caused by nonvaccine pneumococcal serotypes.

CONCLUSION

The incidence of pneumococcal meningitis in infants has rebounded in northern France during the pneumococcal conjugate vaccine programme, with the emergence of nonvaccine pneumococcal serotypes.

Epidemiology, diagnosis, and antimicrobial treatment of acute bacterial meningitis

The epidemiology of bacterial meningitis has changed as a result of the widespread use of conjugate vaccines and preventive antimicrobial treatment of pregnant women. Given the significant morbidity and mortality associated with bacterial meningitis, accurate information is necessary regarding the important etiological agents and populations at risk to ascertain public health measures and ensure appropriate management. In this review, we describe the changing epidemiology of bacterial meningitis in the United States and throughout the world by reviewing the global changes in etiological agents followed by specific microorganism data on the impact of the development and widespread use of conjugate vaccines. We provide recommendations for empirical antimicrobial and adjunctive treatments for clinical subgroups and review available laboratory methods in making the etiological diagnosis of bacterial meningitis. Finally, we summarize risk factors, clinical features, and microbiological diagnostics for the specific bacteria causing this disease.

Global and regional risk of disabling sequelae from bacterial meningitis: a systematic review and meta-analysis

Few data sources are available to assess the global and regional risk of sequelae from bacterial meningitis. We aimed to estimate the risks of major and minor sequelae caused by bacterial meningitis, estimate the distribution of the different types of sequelae, and compare risk by region and income. We systematically reviewed published papers from 1980 to 2008. Standard global burden of disease categories (cognitive deficit, bilateral hearing loss, motor deficit, seizures, visual impairment, hydrocephalus) were labelled as major sequelae. Less severe, minor sequelae (behavioural problems, learning difficulties, unilateral hearing loss, hypotonia, diplopia), and multiple impairments were also included. 132 papers were selected for inclusion. The median (IQR) risk of at least one major or minor sequela after hospital discharge was 19.9% (12.3-35.3%). The risk of at least one major sequela was 12.8% (7.2-21.1%) and of at least one minor sequela was 8.6% (4.4-15.3%). The median (IQR) risk of at least one major sequela was 24.7% (16.2-35.3%) in pneumococcal meningitis; 9.5% (7.1-15.3%) in Haemophilus influenzae type b (Hib), and 7.2% (4.3-11.2%) in meningococcal meningitis. The most common major sequela was hearing loss (33.9%), and 19.7% had multiple impairments. In the random-effects meta-analysis, all-cause risk of a major sequela was twice as high in the African (pooled risk estimate 25.1% [95% CI 18.9-32.0%]) and southeast Asian regions (21.6% [95% CI 13.1-31.5%]) as in the European region (9.4% [95% CI 7.0-12.3%]; overall I(2)=89.5%, p<0.0001). Risks of long-term disabling sequelae were highest in low-income countries, where the burden of bacterial meningitis is greatest. Most reported sequelae could have been averted by vaccination with Hib, pneumococcal, and meningococcal vaccines.

The risk of sequelae due to pneumococcal meningitis in high-income countries: a systematic review and meta-analysis

OBJECTIVES

To determine the risk of various kinds of sequelae in survivors of meningitis due to Streptococcus pneumoniae, as well as the influence of co-factors such as study design, study population and treatment on this risk.

METHODS

MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched from 1 September 1991 to 18 June 2009 for original articles on pneumococcal meningitis sequelae. Prevalence of sequelae was pooled using random effects meta-analysis. Studies were appraised for the influence of referral bias, external validity of study populations, testing procedure and publication bias.

RESULTS

Data were extracted from 63 studies involving 3408 pneumococcal meningitis survivors. The pooled prevalence of any reported sequelae from 48 studies was 31.7% (95% confidence interval 27.2-36.3%) using a random effects model (Cochran-Q = 277, p < 0.01). Differences in studies due to design, study population and treatment were not significant. The pooled prevalence of hearing loss, seizures, hydrocephalus, spasticity/paresis, cranial nerve palsies and visual impairment was 20.9% (17.1-24.7%), 6.5% (3.3-9.7%), 6.8% (3.3-10.2%), 8.7% (6.4-11.0%), 12.2% (5.3-19.1%) and 2.4% (0-5.7%) respectively.

CONCLUSIONS

The burden of sequelae due to pneumococcal meningitis remains high in the reviewed studies.

Strategies and new developments in the management of bacterial meningitis

The principles of antimicrobial therapy for acute bacterial meningitis include use of agents that penetrate well into cerebrospinal fluid and attain appropriate cerebrospinal fluid concentrations, are active in purulent cerebrospinal fluid, and are bactericidal against the infecting pathogen. Recommendations for treatment of bacterial meningitis have undergone significant evolution in recent years, given the emergence of pneumococcal strains that are resistant to penicillin. Clinical experience with use of newer agents is limited to case reports, but these agents may be necessary to consider in patients who are failing standard therapy.

[Corticosteroids in children with bacterial meningitis: indications and administration]

The use of dexamethasone (DXM) as adjunctive therapy for bacterial meningitis (BM) in infants and children has remained controversial for 20 years. In spite of solid pathophysiological arguments, the limited number of patients, methodological flaws in clinical studies taken individually and pooled into meta-analyses, and the emergence of pneumococcal cephalosporin-resistance did not allow to reach a consensus on the effectiveness of DXM in the prevention of neurological sequelae, in the course of non Haemophilus influenzae b (Hib) BM. A recent meta-analysis conducted with an adequate number of patients (2,750 patients including 2,074 infants and children below 15 years of age) demonstrated that DXM prevented mortality and sequelae in adults with pneumococcal meningitis and suggested that this efficacy could also apply to infants and children. Data from the active surveillance networks of pediatric BM and pneumococcal resistance in France suggested that DXM anti-inflammatory effect on antibiotic CSF penetration would not have a significant impact on the bactericidal efficacy if recommended dosages of cefotaxime (300 mg/kg per day) and vancomycin (60 mg/kg per day) were used. DXM could be considered in the early treatment of pneumococcal BM in infants and children in industrialized countries. But there is no proven efficacy of DXM in meningococcal meningitis in infants and children.

[Pneumococcal conjugate vaccine failure in HIV-infected child. Clinical case]

Despite appropriate antimicrobial therapy and vaccination, invasive pneumococcal infections remain associated with significant mortality, especially in selected high-risk groups (asplenic, humoral immunity deficient patients, etc.). We present a 13-year-old caucasian boy with HIV infection (vertical transmission). He received treatment with highly-active antiretroviral therapy (amprenavir, lamivudine and zidovudine) and vaccination with 23-valent vaccine (6 years old) and 7-valent pneumococcal conjugate vaccine (10 years old). His CD4 count and his viral load at these times were 2,063/microl and 13461 cop/ml, when he was 6 years old and 1,315/microl and 32400 cop/ml when he was 10 years old, respectively. The latest CD4 count (1,000/microl) and his viral load (3800 cop/ml) confirmed satisfactory control of the disease. He was referred to our emergency department presenting with fever, head and stomach-ache and vomiting. In the following hours his condition continued to deteriorate and depressed level of consciousness and meningismus were observed. Streptococcus pneumoniae, serotype 18 C, was detected in blood and cerebrospinal fluid cultures. Despite appropriate treatment with antibiotics (cefotaxime and vancomycin) and anti-oedema medications, brain-death was confirmed 24 hours after his admittance.

Mortalité par méningites à pneumocoque chez l'enfant. Facteurs pronostiques à propos d'une série de 73 observations

OBJECTIVE

Despite advances in antibiotic therapy strategies and pediatric intensive care, prognosis of Streptococcus pneumoniae meningitis remains very poor. To determine the factors associated with hospital mortality of children with pneumococcal meningitis.

METHODS

We conducted a retrospective study of 73 cases of childhood pneumococcal meningitis admitted in 4 teaching hospitals in the center of Tunisia during a 8-year period (1995-2002).

RESULTS

Hospital mortality was 13.7% (10 of 71 patients), and neurologic sequela were observed in 34.5% of survivors. Based on univariable analysis, five variables were associated with the outcome: Pediatric Risk of Mortality score (p < 0.001), coma (p=0.0009), use of mechanical ventilation (p=0.0001), convulsions (p = 0.0449), and shock (p=0.0085). In multivariable analysis, only 2 factors were independently associated with in-hospital mortality: Pediatric Risk of Mortality score and the use of mechanical ventilation. 11.8% of pneumococcal isolates were intermediate and resistant to penicillin. Non-susceptible pneumococcus strains to penicillin and the use of steroids were not associated significantly with the mortality rate.

CONCLUSIONS

Pneumococcal meningitis remains a devastating childhood disease. Two variables were independently associated with the in-hospital death in our series (high Pediatric Risk of Mortality score, and the use of mechanical ventilation). According to these data we may recommend the inclusion of vaccination against streptococcus pneumonia in the children's immunization program in Tunisia.

What is new in otitis media?

The "wait and see" approach in acute otitis media (AOM), consisting of postponing the antibiotic administration for a few days, has been advocated mainly to counteract the increased bacterial resistance in respiratory infections. This approach is not justified in children less than 2 years of age and this for several reasons. First, AOM is an acute inflammation of the middle ear caused in about 70% of cases by bacteria. Redness and bulging of the tympanic membrane are characteristic findings in bacterial AOM. Second, AOM is associated with long-term dysfunction of the inflamed eustachian tube (ET), particularly in children less than 2 years of age. In this age group, the small calibre of the ET together with its horizontal direction result in impaired clearance, ventilation and protection of the middle ear. Third, recent prospective studies have shown poor long-term prognosis of AOM in children below 2 years with at least 50% of recurrences and persisting otitis media with effusion (OME) in about 35% 6 months after AOM. Viruses elicit AOM in about 30% of children. A prolonged course of AOM has been observed when bacterial and viral infections are combined because viral infection is also associated with ET dysfunction in young children. Bacterial and viral testing of the nasopharyngeal aspirate is an excellent tool both for initial treatment and recurrence of AOM. Antibiotic treatment of AOM is mandatory in children less than 2 years of age to decrease inflammation in the middle ear but also of the ET particularly during the first episode. The best choice is amoxicillin because of its superior penetration in the middle ear. Streptococci pneumoniae with intermediary bacterial resistance to penicillin are particularly associated with recurrent AOM. Therefore the dosage of amoxicillin should be 90 mg/kg per day in three doses. In recurrent AOM with beta-lactamase-producing bacilli, amoxicillin should be associated with clavulanic acid at a dose of 6.4 mg/kg per day. The duration of the treatment is not established yet but 10 days is reasonable for a first episode of AOM. OME may be a precursor initiating AOM but also a complication thereof. OME needs a watchful waiting approach. When associated with deafness for 2-3 months in children over 2 years of age, an antibiotic should be given according to the results of the bacterial resistance in the nasopharyngeal aspirate. The high rate of complications of tympanostomy tube insertion outweighs the beneficial effect on hearing loss. The poor results of this procedure are due to the absence of effects on ET dysfunction. Pneumococcal vaccination has little beneficial effects on recurrent AOM and its use in infants needs further studies. Treatment with amoxicillin is indicated in all children younger than 2 years with a first episode of AOM presenting with redness and bulging of the tympanic membrane. Combined amoxicillin and clavulanic acid should be given in patients with beta-lactamase-producing bacteria. The duration of treatment is estimated to be at least 10 days depending on the findings by pneumo-otoscopy and tympanometry. Bacterial and viral testing of the nasopharyngeal aspirate is highly recommended particularly in children in day care centres as well as for regular follow-up. The high recurrence rate is due to the long-lasting dysfunction of the eustachian tube and the immune immaturity of children less than 2 years of age.