# A case of a misleading conclusion: a critical reassessment of the methodological and interpretive flaws of a recent meta-analysis of the vector role of cat fleas in feline hemotropic Mycoplasma species transmission

## TL;DR

This paper critiques a recent meta-analysis that questioned the role of cat fleas in spreading feline hemotropic Mycoplasma, pointing out flaws in methodology and interpretation.

## Contribution

The paper identifies critical flaws in a meta-analysis, including misreporting of data and flawed reanalysis design, which undermine the conclusion about flea vector roles.

## Key findings

- The meta-analysis misreported key data from source studies, including prevalence values and sample handling practices.
- The reanalysis experiment had design limitations like mismatched sampling units and unaddressed confounders.
- Alternative explanations for prevalence variation were ignored, and citations were selectively used.

## Abstract

A recent meta-analysis published in Parasites & Vectors [17(1):444, 2024] re-evaluated the prevalence of feline hemotropic Mycoplasma spp. in Ctenocephalides felis. The authors compared prevalence studies using different PCR primers: those employing Jensen or Manvell primers reported high prevalence (~ 33%), while others showed much lower rates (< 1%). To investigate potential Jensen/Manvell primer non-specificity, the authors used both primers and sequencing to reanalyze archived individual flea samples from pools previously reported as positive by Jensen primer-based PCR. Based on low prevalence and non-specific amplifications, they questioned primer specificity and concluded that prevalence was lower than previously reported, and that C. felis is less likely to be a vector of feline hemotropic Mycoplasma spp. This correspondence critically reassesses that conclusion by highlighting multiple methodological and interpretive flaws in the meta-analysis. Sequencing-confirmed Mycoplasma spp. detections from Jensen primer-based studies were disregarded and key data from source studies, including prevalence values and flea washing/pooling practices, were misreported in the meta-analysis. The reanalysis experiment suffered from design limitations, including mismatched sampling units (individual vs pools), unclear distribution of results across previously reported positive pools, and unaddressed confounders such as DNA degradation and contamination over ~ 12 years of storage. Grouping Jensen and Manvell primers, despite their distinct diagnostic behaviors, further undermines the analysis. Alternative explanations for prevalence variation, such as population heterogeneity, pooling flea samples, primer sensitivity, and strain variation, were not considered. The citations were used selectively to emphasize supporting studies, while contradictory evidence was omitted. Due to these limitations, the revised prevalence estimate is not supported, and the vectorial role of C. felis remains unresolved, necessitating further well-designed studies to establish its true epidemiological role.