[Special comments--problems in systematic review for diagnostic accuracy of high-sensitivity C-reactive protein assay

Yuzuru Takemura1, Haku Ishida

  • 1Department of Laboratory Medicine, National Defense Medical College, Tokorozawa, 359-8513.

Rinsho Byori. the Japanese Journal of Clinical Pathology
|August 20, 2003
PubMed

Insights

High-sensitivity C-reactive protein assays show variable diagnostic accuracy for neonatal infection. The quality of primary studies included in meta-analyses was often low, potentially impacting reliability.

Area of Science:

  • Clinical Medicine
  • Biochemistry

Background:

  • Neonatal infections pose significant health risks.
  • Accurate diagnostic tools are crucial for timely treatment.
  • High-sensitivity C-reactive protein (hs-CRP) is a potential biomarker.

Purpose of the Study:

  • To systematically review the diagnostic accuracy of hs-CRP assays for neonatal infection.
  • To evaluate the quality of existing primary studies.

Main Methods:

  • Systematic review and meta-analysis of published reports.
  • Searched MEDLINE and Igaku-Chuo-Shi databases.
  • Included 30 primary studies after applying selection criteria.
  • Critically appraised study quality and extracted sensitivity/specificity data.

Main Results:

  • Only one of 30 studies met high-quality standards (Rank A).
  • Nineteen studies were classified as low-quality (Rank C).
  • Most studies lacked methodological rigor, particularly regarding blinding.
  • Poor study quality may have influenced meta-analysis results.

Conclusions:

  • The diagnostic accuracy of hs-CRP for neonatal infection is limited by the low quality of available primary studies.
  • Future research requires improved methodological standards.
  • Accessing raw data from studies could enhance meta-analysis reliability.