Podcast: Play in new window | Download

Subscribe: RSS

Date: February 12th, 2021

Guest Skeptic: Dr. Dennis Ren is a paediatric emergency medicine fellow at Children’s National Hospital in Washington, DC.

Reference: Mintegi S et al. Clinical Prediction Rule for Distinguishing Bacterial from Aseptic Meningitis. Pediatrics 2020

Case: A 4-year-old immunized girl presents to the emergency department (ED) with a fever and rhinorrhea for the past four days. Her parents report that she has been complaining of a headache and seems more tired and sleepy in the past day. On exam, she is febrile to 38.5 ºC, appears tired, with meningismus on examination but answers questions appropriately. She does not have any petechiae or purpura on skin exam. You explain that you must obtain some blood for laboratory work and perform a lumbar puncture (LP) because you are concerned that she has meningitis. Her nervous parents agree to the LP. Her cerebrospinal fluid appears clear and preliminary cerebrospinal fluid (CSF) results show a pleocytosis with 16 white blood cells per µL without any red blood cells. Her parents ask you whether or not she will have to stay in the hospital or receive antibiotics.

Background: Vaccines cause adults. Supporting this position is that since the introduction of conjugate vaccines the incidence of life-threatening bacterial meningitis has decreased. The first conjugate vaccine introduced was the haemophilus influenzae type b (Hib) vaccine. This vaccine has a reported efficacy of 98% (Makwana and Riordan 2007).

The success of conjugate vaccines is that most cases of pediatric meningitis are now aseptic (viral cause). It is important to distinguish between bacterial vs aseptic meningitis. This is because bacterial meningitis is associated with serious morbidity and mortality and requires prompt antibiotic treatment; aseptic meningitis is self-limited and requires only supportive care. Patients with suspected bacterial meningitis require hospital admission with empiric antibiotics pending culture results (Sáez-Llorens and McCracken 2003).

There is no single variable that can help discriminate between bacterial vs. aseptic meningitis.  Combinations of variables have been tried in the past as part of clinical scoring systems such as the Bacterial Meningitis Score (BMS) to identify children with CSF pleocytosis at low risk for bacterial meningitis (Nigrovic et al 2002).

However, BMS did not take into account C-reactive protein and procalcitonin levels that have shown promise in risk stratifying febrile children at risk for bacterial infection (Van den Bruel et al 2011).

Additionally, BMS has missed a few cases of bacterial meningitis. Specifically, 2 out of 1714 patients categorized as very low risk for bacterial meningitis had bacterial meningitis (sensitivity 98.3%, NPV 99.9%). Both patients missed were younger than 2 months old (Nigrovic et al 2007).

The study we are reviewing today aimed to develop and validate a more accurate scoring system called the Meningitis Score for Emergencies (MSE) to distinguish between bacterial vs. aseptic meningitis in children 29 days to 14 years old with CSF pleocytosis based on four objective lab criteria.

Clinical Question:  Can a clinical decision tool using laboratory data help distinguish between bacterial from aseptic meningitis in children 29 days to 14 years old with cerebrospinal fluid pleocytosis?

  • Pleocytosis- CSF WBC ≥10 cells per µL. Corrected for presence of CSF RBCS (1:500 leukocytes to erythrocytes in peripheral blood) and CSF protein (every 1000-cell increase on CSF RBCs per mm3, CSF protein increased by 1.1 mg/dL)
  • Bacterial meningitis defined as patient with either identification of bacterial pathogen in CSF culture and/or Neisseria meningitides or Streptococcus pneumoniae on polymerase chain reaction and either positive blood culture or blood PCR result for N meningitides or S pneumoniae
  • Aseptic meningitis defined as CSF pleocytosis and negative CSF and blood bacterial cultures and negative Neisseria meningitidesor Streptococcus pneumoniae on polymerase chain reaction

Reference: Mintegi S et al. Clinical Prediction Rule for Distinguishing Bacterial from Aseptic Meningitis. Pediatrics 2020

  • Population: Children between 29 days and 14 years old with a diagnosis of meningitis across 25 Spanish emergency departments.
    • Exclusion: Children <29 days old, critically ill, with purpura, not previously healthy or treated with antibiotics within 72 hours before lumbar puncture.
  • Intervention: Retrospective derivation and prospective validation of Meningitis Score for Emergencies (MSE) for distinguishing bacterial vs. aseptic meningitis using procalcitonin >1.2 ng/mL, CSF protein >80 mg/L, CSF absolute neutrophil count >1000 cells per mm3, and C-reactive protein >40 mg/L.

The four laboratory components were given different points if present and zero points if absent. So, if the procalcitonin was elevated you got 3 points, 1 point for elevated CRP, 1 point for elevated ANC and 2 points for elevated CSF protein (max score 7).

The BMS had 5 components: four are laboratory and one is clinical (seizure at or before presentation). Each of these components were also given a different number of points if present and zero points if absent. The BMS is available on MDCalc.

Outcome: Accuracy of clinical decision support tool in distinguishing bacterial vs aseptic meningitis in children with CSF pleocytosis. (sensitivity, specificity, negative predictive value, positive predictive value, and likelihood ratios).

Authors’ Conclusions: “The meningitis score for emergencies (MSE) accurately distinguishes bacterial from aseptic meningitis in children with CSF pleocytosis.”

Quality Checklist for Clinical Decision Tools:

  1. The study population included or focused on those in the ED. Yes
  2. The patients were representative of those with the problem. Unsure
  3. All important predictor variables and outcomes were explicitly specified. Yes
  4. This is a prospective, multicenter study including a broad spectrum of patients and clinicians (level II). No
  5. Clinicians interpret individual predictor variables and score the clinical decision rule reliably and accurately. Yes
  6. This is an impact analysis of a previously validated CDR (level I). No
  7. For Level I studies, impact on clinician behavior and patient-centric outcomes is reported.  N/A
  8. The follow-up was sufficiently long and complete. Unsure
  9. The effect was large enough and precise enough to be clinically significant. Unsure

Key Results: The final study included 1,009 patients between 29 days and 14 years of age. There were 819 patients assigned to the derivation group while 190 were assigned to the validation group. Slightly over 1/3 were female with mean age of 2 years. The vast majority (91%) had aseptic meningitis and only 9% had bacterial meningitis. Of those with meningitis, 80% had either N meningitides (41.3%) or S pneumonia (38.5%). Other organisms included: Group B Strep 5.5%, Strep pyogenes 4.3%, Enterococcus faecalis, H influenzae (2.2% each) and E. coli, Listeria monocytogenese, Salmonella typhimurium, Strep bovis, Kingella kingae, Fusobacterium necrophorum (1.1% each).

Validation Group MSE ≥1: Sensitivity 100%, specificity  77.4%, NPV 100%, PPV 46.3%, LR+ 4.4 and LR- 0

Derivation and Validation Group MSE ≥1: Sensitivity 100%, specificity 83.2%, NPV 100%, PPV 37.4%, LR+ 5.95 and LR- 0.

No patients with bacterial meningitis were missed with the MSE. Two patients with bacterial meningitis were missed with the BMS.

  • 1-month-old with Streptococcus agalactiae meningitis although BMS is used for patients >2 months
  • 3-year-old with meningococcal meningitis

1) Procalcitonin Testing: Procalcitonin is the cool new kid on the block when it comes to detecting bacterial infections. Procalcitonin testing may not be available at all hospitals limiting applicability of this new scoring tool.

2) Age: Similar to the Bacterial Meningitis Score (BMS) study, this study did not include any patients <29 days old. Unlike the BMS study that included patients up to age 19, this study does not include any patients over the age of 14. It is difficult to determine whether the incidence of aseptic vs bacterial meningitis in the age group >14 to 19 years would have affected the accuracy of the MSE.

3) Derivation vs Validation Sets: Out of a total of 1,509 eligible patients, 500 were excluded: 414 from derivation group, 86 from validation group. Proportionately, 50 patients (12%) with bacterial meningitis were excluded from the derivation group while 28 patients (32%) with bacterial meningitis were excluded from the validation set. This may suggest that the groups were uneven in their relative distribution of bacterial vs aseptic meningitis.

When comparing the characteristics between patients in the final derivation and validation groups, there was also a higher percentage of bacterial meningitis in the validation set compared to the derivation set (16.3% vs 7.4%).

4) CSF Gram-Stain: The authors made a decision to not include CSF Gram-stain in the score despite it being positive in 75% of cases of bacterial meningitis and positive CSF Gram-stain having specificity >97%. They state that this is due to limitations of availability of performing Gram-stain 24/7 in all EDs but recommend that a child with CSF pleocytosis and positive Gram stain should be put on antibiotics regardless of MSE score.

5) Geography: One of the limitations mentioned by the authors is that this study population is drawn purely from Spanish emergency departments. The prevalence of bacterial meningitis differs worldwide so this scoring tool would need external validation in different counties and in rural vs urban areas.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We generally agree with the authors’ findings. MSE ≥1 demonstrates high accuracy for distinguishing between bacterial vs. aseptic meningitis in children with CSF pleocytosis. It requires external validation prior to clinical application.

SGEM Bottom Line: MSE with addition of CRP and procalcitonin has improved accuracy compared to BMS in distinguishing bacterial vs aseptic meningitis and has potential to decrease admission and antibiotic administration in children with CSF pleocytosis but it needs to be externally validated.

Case Resolution: Additional labs for your patient return and show 5 red blood cells in the CSF sample, normal glucose, and protein of 50 mg/dL. Her initial Gram-stain does not show any organisms. CSF PCR testing is positive for enterovirus. The remainder of her bloodwork including CRP and procalcitonin are below the respective cut offs. Her headache and fever have improved after receiving acetaminophen. Her mental status remains stable off antibiotics. You call her primary care provider and arrange for the family to follow up in the office the next day. 

Dr. Dennis Ren

Clinical Application: The novel MSE scoring tool adds procalcitonin and CRP to distinguish bacterial vs aseptic meningitis in well-appearing children with CSF pleocytosis. It has better sensitivity and specificity compared to the Bacterial Meningitis Score. No patients with bacterial meningitis were missed.

This scoring tool needs external validation prior to clinical application, and it should only be applied in children >29 days up to and including 14 years of age.

We recommend waiting for results of CSF Gram-stain if available as a positive Gram-stain obviates the usefulness of the MSE. Patients with positive CSF Gram-stain should receive antibiotics and be admitted to the hospital.

Additionally, CSF PCR testing has high sensitivity and specificity and can test for a variety of viral, bacterial, and fungal pathogens and may be more accurate in comparison to meningitis scoring tools; results can be available in as quickly as 1-2 hours in some clinical practices.

It is important to be aware of the limitations of your institution’s meningitis/encephalitis PCR – while it may cover some of the most common bacterial and viral pathogens (E. coli K1, H. influenzae, Listeria monocytogenes, Neisseria meningitidis, Streptococcus agalactiae, Streptococcus pneumoniae, enterovirus, HSV-1, HSV-2), it is not all encompassing. For example, the PCR may not cover gut bacteria or may be limited to testing for one specific strain of E. coli. There is evidence to suggest that a positive enterovirus result on CSF PCR have very low risk of co-infection with bacterial meningitis.

What Do I Tell My Patient? Your daughter has meningitis which is caused by a virus called enterovirus. She clinically appears well. There is no antibiotic treatment for this type of meningitis, and she does not need to stay in the hospital at this time if she is at her baseline mental status and can keep herself hydrated. She will likely recover on her own. She can receive over the counter medication for fever or headache as needed. You will need to follow up closely with her primary care provider. If she is getting worse, can’t stay hydrated, or you are worried, please come back to the ED anytime.

Keener Kontest: Last weeks’ winner was…I owe another apology. For the second week in a row, we were not specific enough. The question was what was the first state to legalize cannabis. I was thinking recreational not medicinal. Cannabis was legalized in California for medical reasons first but it was legalized in Washington State first for recreational use.

Listen to the SGEM podcast to hear this weeks’ question. Send your answer to TheSGEM@gmail.com with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

Other FOAMed resources:

Contents is free, but there’s a small fee for the CME

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.

References:

  • Mintegi S, García S, Martín MJ, Durán I, Arana-Arri E, Fernandez CL, Benito J, Hernández-Bou S; Meningitis Group of the Spanish Society of Pediatric Emergencies. Clinical Prediction Rule for Distinguishing Bacterial From Aseptic Meningitis. Pediatrics. 2020 Sep;146(3):e20201126. doi: 10.1542/peds.2020-1126. Epub 2020 Aug 25. PMID: 32843440.
  • Nigrovic LE, Kuppermann N, Malley R. Development and validation of a multivariable predictive model to distinguish bacterial from aseptic meningitis in children in the post-Haemophilus influenzae era. Pediatrics. 2002;110(4):712-719.
  • Nigrovic LE, Kuppermann N, Macias CG, et al. Clinical prediction rule for identifying children with cerebrospinal fluid pleocytosis at very low risk of bacterial meningitis. JAMA. 2007;297(1):52-60.
  • Nigrovic LE, Malley R, Agrawal D, Kuppermann N, Pediatric Emergency Medicine Collaborative Research Committee of the American Academy of Pediatrics. Low risk of bacterial meningitis in children with a positive enteroviral polymerase chain reaction test result. Clin Infect Dis. 2010;51(10):1221-1222.
  • Sáez-Llorens X, McCracken GH Jr. Bacterial meningitis in children. Lancet. 2003 Jun 21;361(9375):2139-48. doi: 10.1016/S0140-6736(03)13693-8. PMID: 12826449.
  • Van den Bruel A, Thompson MJ, Haj-Hassan T, Stevens R, Moll H, Lakhanpaul M, Mant D. Diagnostic value of laboratory tests in identifying serious infections in febrile children: systematic review. BMJ. 2011 Jun 8;342:d3082. doi: 10.1136/bmj.d3082. PMID: 21653621.