By Philip R. Fischer, MD, DTM&H, and Sreelaxmi R. Ramesh
Dr. Fischer is Professor of Pediatrics, Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN; Department of Pediatrics, Sheikh Shakhbout Medical City, Abu Dhabi, United Arab Emirates.
Sreelaxmi Ramesh is a visiting medical student at Sheikh Shakhbout Medical City (in partnership with Mayo Clinic) in Abu Dhabi, United Arab Emirates.
SYNOPSIS: In children thought to have meningitis, the use of a molecular test pathogen panel reduced the time to optimal antimicrobial use and reduced the duration of intravenous antibiotic use. However, there was no difference in the time to effective antimicrobial administration or the length of hospital stay between those with treatable pathogens who were and were not evaluated using molecular testing.
SOURCE: Messacar K, Palmer C, Gergoire L, et al. Clinical and financial impact of a diagnostic stewardship program for children with suspected central nervous system infection. J Pediatr2022;244:161-168.
A presumptive clinical diagnosis of meningitis in children lacks specificity because of varying clinical features as well as the wide variety of pathogens that can cause meningitis and encephalitis in children. Often, no etiologic diagnosis is made. Thus, clinical care can involve broad antimicrobial treatments, sometimes with potential toxicity.
Molecular testing of cerebrospinal fluid for potential pathogens offers hope for improved etiologic testing and for more rapid use of appropriate treatment while avoiding excessive use of unnecessary treatments that can be costly and/or toxic. While molecular test panels now are used widely, both the clinical impact and the cost effectiveness of such testing in children has been incompletely characterized.
Thus, researchers in Colorado evaluated the usefulness of a molecular test panel for etiologic pathogens in children tested for meningitis who were considered to be at significant risk of serious disease. The test panel included 14 pathogens — likely bacteria causing meningitis in children (Escherichia coli, group B streptococcus, Listeria monocytogenes, pneumococcus, Haemophilus influenzae, and Neisseria meningitidis), common viral causes of meningitis in children (enterovirus, herpes 1 and 2), and less commonly seen causes of meningocephalitis in children (human parechovirus, varicella virus, human herpes virus 6, cytomegalovirus, and Cryptococcus neoformans). Children were included in the study if their cerebrospinal fluid samples were sent to the laboratory and if they were deemed to be at high risk of serious central nervous system infection (younger than 2 months of age, immunocompromised, evidence of encephalitis, and/or more than 5 white blood cells [WBC] per microliter of spinal fluid).
If children met the high-risk inclusion criteria, molecular testing was done. The results then were communicated to clinicians, along with antimicrobial recommendations. Clinical outcomes (time to appropriate treatment and discontinuation of unnecessary treatment, length of hospital stay) and cost effectiveness were evaluated by comparing children who received care prior to (1,124 children) and after (1,127 children) implementation of the molecular testing program. More single-pathogen polymerase chain reaction (PCR) tests were done in the pre-intervention period than after the use of the molecular test panel was instituted.
The average age of children in the study was 3.5 years; one-third of children had an underlying medical condition. Overall, half of children were younger than 2 months of age, 40% had more than 5 WBC/mL spinal fluid, and half had clinical findings (altered mental status, magnetic resonance imaging results, electroencephalogram results) suggestive of encephalitis; only 4% were immunocompromised.
Only 1% of spinal fluid samples had positive Gram stains, and only 2% yielded positive culture results. Only 2% had positive molecular test panel results for a bacterial pathogen. The most common viruses identified were enterovirus (8%), herpesvirus (4%), and human herpesvirus 6 (2%).
Clinically, 91% of children in the study were hospitalized, with a median length of stay of four days. Overall, 70% of children were started on antibiotic treatment, and the duration of antibiotic treatment was 36 hours prior to use of molecular panel tests and 24 hours after using molecular test panels. Twenty-six percent of children received intravenous acyclovir, and the duration of acyclovir therapy did not vary based on the use of the molecular test panel. For the 18% of children who had a treatable etiology for their illness identified, the use of the test panel did not alter the time to appropriate antimicrobial use. Overall, the time to optimal antimicrobial use (or non-use) was markedly reduced by use of the test panels (18 vs. 28 hours).
The costs of overall molecular testing doubled by adding the molecular test panel to ad-hoc use of individual molecular tests per clinician choice. However, the total hospital costs were not affected by use of the molecular test panels.
The authors wisely concluded their study noting that “newly available rapid molecular diagnostic technologies are revolutionizing our ability to rapidly and accurately detect infectious pathogens from biologic specimens. However, technology alone does not optimally improve care and may increase costs.” Thus, they advocated for “thoughtful, judicious” use of diagnostic tools in conjunction with principles of diagnostic stewardship.
COMMENTARY
During the past 11 years, several molecular test panels have become available, each targeting a different set of clinical bloodstream, respiratory, and gastrointestinal syndromes. The panel used in the Colorado study was the first to become commercially available in the United States that focused on central nervous system infection.1 These tests can be very useful, but there is a risk of overuse.1 Specifically, as seen in this study, negative tests for bacterial pathogens in children who did not receive antibiotics prior to cerebrospinal fluid testing were useful in shortening the duration of antibiotics that had been started presumptively. As also seen in this current study, though, the overall effects on children and on overall costs of care were minimal.
Of course, molecular testing for bacterial pathogens is very useful in children with meningitis who received antibiotic treatment prior to sampling of spinal fluid. This was relevant to the five children in the Colorado study who had positive molecular panel results despite negative culture results.
The molecular test panel for central nervous system infection, helpful as it may be, is associated with concerns for both the sensitivity and specificity of herpes simplex virus 1 detection.1There also is concern about the implications of positive human herpesvirus 6 infection and whether the positive result is due to central nervous system infection, self-limited primary infection, reactivation of a previous infection, or chromosomal integration of viral material.1
The vast majority of children in this study had neither a treatable pathogen nor a positive result that would lead to definitive cessation of antibacterial therapy. Even sensitive and specific tests do not always identify the actual cause of the clinical condition. Clinical interpretation of findings is necessary.
We must be aware of the problem of less-than-thoughtful use of tests. Messacar’s study included management recommendations from infectious disease specialists along with the communication of test results. Sometimes physicians less-thoughtfully interpret molecular panel results in ways that leave parents thinking their children have multiple concurrent serious infections (whether “both rhinovirus and enterovirus” from a single positive cross-reacting test result or “three pneumonias at the same time” in a child with a large panel result that identified multiple viruses of uncertain and unlikely clinical relevance in a single nasopharyngeal specimen). It is important to engage infectious disease specialists in diagnostic stewardship programs.
This Colorado study predetermined that only high-risk children would be included: 1) those younger than 2 months of age for whom we know that clinical presentations are far from definitive in ruling out serious bacterial infections, 2) immunocompromised children, although no cases of Cryptococcus were identified, 3) those with clinical or imaging or electroencephalogram findings suggestive of encephalitis who might be more likely to have viral infections that are different from the causes of meningitis, and 4) those with elevated numbers of white blood cells in spinal fluid samples. As the authors pointed out in their discussion and as mentioned by others,1 the clinical usefulness (and cost effectiveness) of molecular panel testing could be increased by being more selective in deciding for which children to do such testing. The majority of children in this study had normal spinal fluid white cell counts, so it could be argued that the meninges were not inflamed and that testing for causes of meningitis was unnecessary. Secondly, human herpesvirus 6 positivity is of uncertain meaning, and positive findings should not be interpreted to imply either causation or absence of other causes of the symptoms (such as in the three older immunocompetent children in the Colorado study who were incidentally positive for human herpesvirus 6 while also having serious bacterial infection). The claim that “it’s cheaper to do the whole panel” might be true in terms of test costs, but it does not account for the costs of physician and family confusion and of less-than-thoughtful interpretation of test results.
This Colorado study suggests that being more rigorous about patient selection for testing might reduce both costs and subsequent diagnostic confusion (perhaps prohibiting testing for meningitis in children without excessive spinal fluid white cells and/or prohibiting testing for human herpesvirus 6 in immunocompetent children, for example). Indeed, a European study randomized children to standard single-agent testing as per clinician preference vs. molecular panel testing but only included children with at least 15 WBC/mL of spinal fluid.2 In that setting with that more restrictive inclusion criterion, the use of the test was associated with reduced antibiotic use, reduced length of stay in the hospital, and reduced costs.2
REFERENCES
- Dien Bard J, McElvania E. Panels and syndromic testing in clinical microbiology. Clin Lab Med 2020;40:393-420.
- Posnakoglou L, Siahanidou T, Syriopoulou V, Michos A. Impact of cerebrospinal fluid syndromic testing in the management of children with suspected central nervous system infection. Eur J Clin Microbiol Infect Dis 2020;39:2379-2386.
In children thought to have meningitis, the use of a molecular test pathogen panel reduced the time to optimal antimicrobial use and reduced the duration of intravenous antibiotic use. However, there was no difference in the time to effective antimicrobial administration or the length of hospital stay between those with treatable pathogens who were and were not evaluated using molecular testing.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content
Already have an account? Log in