Validation of a Dutch risk score predicting poor outcome in adults with bacterial meningitis in Vietnam and Malawi

MeningiSSS: A New Predictive Score to Support Decision on Invasive Procedures to Monitor or Manage the Intracerebral Pressure in Children with Bacterial Meningitis

BACKGROUND

Knowing the individual child's risk is highly useful when deciding on treatment strategies, especially when deciding on invasive procedures. In this study, we aimed to develop a new predictive score for children with bacterial meningitis and compare this with existing predictive scores and individual risk factors.

METHODS

We developed the Meningitis Swedish Survival Score (MeningiSSS) based on a previous systematic review of risk factors. From this, we selected risk factors identified in moderate-to-high-quality studies that could be assessed at admission to the hospital. Using data acquired from medical records of 101 children with bacterial meningitis, we tested the overall capabilities of the MeningiSSS compared with four existing predictive scores using a receiver operating characteristic curve (ROC) analysis to assert the area under the curve (AUC). Finally, we tested all predictive scores at their cut-off levels using a Chi-square test. As outcome, we used a small number of predefined outcomes; in-hospital mortality, 30-day mortality, occurrence of neurological disabilities at discharge defined as Pediatric Cerebral Performance Category Scale category two to five, any type of complications occurring during the hospital stay, use of intensive care, and use of invasive procedures to monitor or manage the intracerebral pressure.

RESULTS

For identifying children later undergoing invasive procedures to monitor or manage the intracerebral pressure, the MeningiSSS excelled in the ROC-analysis (AUC = 0.90) and also was the only predictive score able to identify all cases at its cut-off level (25 vs 0%, p < 0.01). For intensive care, the MeningiSSS (AUC = 0.79) and the Simple Luanda Scale (AUC = 0.75) had the best results in the ROC-analysis, whereas others performed less well (AUC ≤ 0.65). Finally, while none of the scores' results were significantly associated with complications, an elevated score on the MeningiSSS (AUC = 0.70), Niklasson Scale (AUC = 0.72), and the Herson-Todd Scale (AUC = 0.79) was all associated with death.

CONCLUSIONS

The MeningiSSS outperformed existing predictive scores at identifying children later having to undergo invasive procedures to monitor or manage the intracerebral pressure in children with bacterial meningitis. Our results need further external validation before use in clinical practice. Thus, the MeningiSSS could potentially be helpful when making difficult decisions concerning intracerebral pressure management.

Prediction of Outcome From Adult Bacterial Meningitis in a High-HIV-Seroprevalence, Resource-Poor Setting Using the Malawi Adult Meningitis Score (MAMS)

Background.

Acute bacterial meningitis (ABM) in adults residing in resource-poor countries is associated with mortality rates >50%. To improve outcome, interventional trials and standardized clinical algorithms are urgently required. To optimize these processes, we developed and validated an outcome prediction tool to identify ABM patients at greatest risk of death.

Methods.

We derived a nomogram using mortality predictors derived from a logistic regression model of a discovery database of adult Malawian patients with ABM (n = 523 [65%] cerebrospinal fluid [CSF] culture positive). We validated the nomogram internally using a bootstrap procedure and subsequently used the nomogram scores to further interpret the effects of adjunctive dexamethasone and glycerol using clinical trial data from Malawi.

Results.

ABM mortality at 6-week follow-up was 54%. Five of 15 variables tested were strongly associated with poor outcome (CSF culture positivity, CSF white blood cell count, hemoglobin, Glasgow Coma Scale, and pulse rate), and were used in the derivation of the Malawi Adult Meningitis Score (MAMS) nomogram. The C-index (area under the curve) was 0.76 (95% confidence interval, .71–.80) and calibration was good (Hosmer-Lemeshow C-statistic = 5.48, df = 8, P = .705). Harmful effects of adjunctive glycerol were observed in groups with relatively low predicted risk of poor outcome (25%–50% risk): Case Fatality Rate of 21% in the placebo group and 52% in the glycerol group (P < .001). This effect was not seen with adjunctive dexamethasone.

Conclusions.

MAMS provides a novel tool for predicting prognosis and improving interpretation of ABM clinical trials by risk stratification in resource-poor settings. Whether MAMS can be applied to non-HIV-endemic countries requires further evaluation.

Risk scores for outcome in bacterial meningitis: Systematic review and external validation study

OBJECTIVES

To perform an external validation study of risk scores, identified through a systematic review, predicting outcome in community-acquired bacterial meningitis.

METHODS

MEDLINE and EMBASE were searched for articles published between January 1960 and August 2014. Performance was evaluated in 2108 episodes of adult community-acquired bacterial meningitis from two nationwide prospective cohort studies by the area under the receiver operating characteristic curve (AUC), the calibration curve, calibration slope or Hosmer-Lemeshow test, and the distribution of calculated risks.

FINDINGS

Nine risk scores were identified predicting death, neurological deficit or death, or unfavorable outcome at discharge in bacterial meningitis, pneumococcal meningitis and invasive meningococcal disease. Most studies had shortcomings in design, analyses, and reporting. Evaluation showed AUCs of 0.59 (0.57-0.61) and 0.74 (0.71-0.76) in bacterial meningitis, 0.67 (0.64-0.70) in pneumococcal meningitis, and 0.81 (0.73-0.90), 0.82 (0.74-0.91), 0.84 (0.75-0.93), 0.84 (0.76-0.93), 0.85 (0.75-0.95), and 0.90 (0.83-0.98) in meningococcal meningitis. Calibration curves showed adequate agreement between predicted and observed outcomes for four scores, but statistical tests indicated poor calibration of all risk scores.

INTERPRETATION

One score could be recommended for the interpretation and design of bacterial meningitis studies. None of the existing scores performed well enough to recommend routine use in individual patient management.

Dexamethasone treatment and prognostic factors in community-acquired bacterial meningitis: a Danish retrospective population-based cohort study

INTRODUCTION

The morbidity and mortality in community-acquired bacterial meningitis (CABM) remain substantial and treatment outcomes and predictors of a poor prognosis must be assessed regularly. We aimed to describe the outcome of patients with CABM treated with dexamethasone and to assess the performance of the Dutch Meningitis Risk Score (DMRS).

METHODS

We retrospectively evaluated all adults with CABM in North Denmark Region, 1998-2012. Outcomes included in-hospital mortality and Glasgow Outcome Scale (GOS) score. A GOS score of 5 was categorized as a favourable outcome and scores of 1-4 as unfavourable. We used logistic analysis to compute relative risks (RRs) with 95% confidence intervals (CIs) for an unfavourable outcome adjusted for age, sex, and comorbidity.

RESULTS

We identified a total of 172 cases of CABM. In-hospital mortality was unaffected by the implementation of dexamethasone in 2003 (19% vs 20%). Dexamethasone treatment was associated with a prompt diagnosis of meningitis and a statistically insignificant decrease in the risk of an unfavourable outcome (33% vs 53%; adjusted RR 0.64, 95% CI 0.41-1.01) and in-hospital mortality (15% vs 24%; adjusted RR 0.72, 95% CI 0.35-1.48). Of the risk factors included in the DMRS, we found age and tachycardia to be significantly associated with an unfavourable outcome in the multivariate analyses.

CONCLUSIONS

Patients treated with dexamethasone were more likely to have a favourable outcome, although statistical significance was not reached. Several parameters included in the Dutch risk score were also negative predictors in our cohort, although the entire risk score could not be validated due to a lack of data.

Unsuccessful validation of 2004 model for predicting academic or behavioural limitations after childhood bacterial meningitis

AIM

In 2004, a model identifying children at risk of academic or behavioural limitations after bacterial meningitis (BM) was presented. Risk factors were male gender, low birthweight, lower educational level of the father, Streptococcus pneumoniae, lower cerebrospinal fluid (CSF) leucocyte count, delay between admission and start of antibiotics, dexamethasone <2 days, seizures and prolonged fever. The aim of this study was to validate that prediction model in an independent cohort.

METHODS

Academic or behavioural limitations were determined in 93 Dutch school-age BM survivors. Risk factors for limitations were obtained from medical files. Validation was performed by applying the model in the cohort, then assessing discrimination and goodness of fit. Multiple imputation techniques were used to deal with missing values.

RESULTS

Although fit of the model appeared good when it came to similarity of expected and observed cases (p-value of the Hosmer-Lemeshow test 0.24-0.57), discrimination was poor. Area under the curve (AUC) of the receiver operated characteristics (ROC) curve of the model was 0.83 (95% CI: 0.77-0.89) in the development cohort and 0.53 (95% CI: 0.41-0.65) in the validation cohort.

CONCLUSION

External validation of the model was unsuccessful. It is not suitable for implementation in practice.