[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Views 216
Citations 0
Comment & Response
March 9, 2020

Neonatal Early-Onset Sepsis Calculator and Antibiotic Therapy

Author Affiliations
  • 1Department of Pediatrics, China-Japan Friendship Hospital, Beijing, China
  • 2Institute of Clinical Medical Sciences, China-Japan Friendship Hospital, Beijing, China
JAMA Pediatr. Published online March 9, 2020. doi:10.1001/jamapediatrics.2019.6260

To the Editor We read with great interest the systematic review and meta-analysis by Achten et al,1 who reported a significant association of using the neonatal early-onset sepsis (EOS) calculator with the substantially reduced use of empirical antibiotics for suspected EOS. The findings of this meta-analysis emphasize the importance of using the neonatal EOS calculator to enhance the accuracy of administrating empirical antibiotic to newborns with suspected EOS, thereby avoiding potential harms and adverse consequences owing to unnecessary empirical antibiotic therapy.2 However, we have 2 methodologic concerns for the meta-analysis by Achten et al,1 which may call to question the authors’ conclusion.

First, a major problem plaguing the interpretation and extrapolation of a meta-analysis relates to the proper handling of between-study heterogeneity.3 There is a clear recognition that as with all meta-analyses, the assumption of clinical and methodologic diversity across studies can often lead to statistical heterogeneity.4 In the meta-analysis by Achten et al,1 the presence of statistical heterogeneity in overall analyses was merely interrogated via grouping studies according to the exposure status to chorioamnionitis. Yet other possible sources of heterogeneity from clinical and methodologic aspects, such as quality of evidence and sample size, are left unexplored, which would be very interesting to the readers.

Second, similar to the problem of between-study heterogeneity, the effect of publication bias on the validity of meta-analytical findings has received considerable attention.5 However, Achten et al1 did not take the problem of publication bias into consideration when assessing the association between the neonatal EOS calculator and reduction in antibiotic therapy. In fact, publication bias is suspected in this meta-analysis1 because all 13 eligible studies were conducted exclusively in Western countries, and whether the conclusion can be generalized to non-Western countries, especially with different health care settings, remains an open question and calls for further research.

Back to top
Article Information

Corresponding Author: Wenquan Niu, PhD, Institute of Clinical Medical Sciences, China-Japan Friendship Hospital, No. 2 Yinghua E St, Chaoyang District, Beijing 100029, China (niuwenquan_shcn@163.com).

Published Online: March 9, 2020. doi:10.1001/jamapediatrics.2019.6260

Conflict of Interest Disclosures: None reported.

Achten  NB, Klingenberg  C, Benitz  WE,  et al.  Association of use of the neonatal early-onset sepsis calculator with reduction in antibiotic therapy and safety: a systematic review and meta-analysis.  JAMA Pediatr. 2019;173(11):1032-1040. doi:10.1001/jamapediatrics.2019.2825PubMedGoogle Scholar
Fjalstad  JW, Esaiassen  E, Juvet  LK, van den Anker  JN, Klingenberg  C.  Antibiotic therapy in neonates and impact on gut microbiota and antibiotic resistance development: a systematic review.  J Antimicrob Chemother. 2018;73(3):569-580. doi:10.1093/jac/dkx426PubMedGoogle ScholarCrossref
Yao  L, Niu  W.  Gastrointestinal cancer risk in cystic fibrosis: more exploration is needed.  Lancet Oncol. 2018;19(7):e332. doi:10.1016/S1470-2045(18)30384-XPubMedGoogle Scholar
Gagnier  JJ, Moher  D, Boon  H, Beyene  J, Bombardier  C.  Investigating clinical heterogeneity in systematic reviews: a methodologic review of guidance in the literature.  BMC Med Res Methodol. 2012;12:111. doi:10.1186/1471-2288-12-111PubMedGoogle ScholarCrossref
Sutton  AJ, Song  F, Gilbody  SM, Abrams  KR.  Modelling publication bias in meta-analysis: a review.  Stat Methods Med Res. 2000;9(5):421-445. doi:10.1177/096228020000900503PubMedGoogle ScholarCrossref