FeLV triage model in hospitalized cats: associations and limitations
Veterinary sciences · 2025
In plain language
A study of hospitalized cats' medical records identified breed, other illnesses and a low hematocrit—a lower proportion of red blood cells in the blood—as the most important factors in a model predicting a positive FeLV result. It does not show that FeLV caused those other illnesses or measure how well the cats' immune defenses worked. Some FeLV infections may have been classified as negative. The model has not been validated outside the unit studied, and the source contains inconsistencies in its reported figures and procedures.
Scientific detail
This retrospective clinical study, published in Veterinary sciences in 2025 (doi: 10.3390/vetsci12090902), evaluated a statistical triage model in a veterinary isolation unit. It is a single-center proof-of-principle study without external validation; the claims provided are classified as evidence level C. The Methods describe 1211 hospitalized cats, 867 with retrovirus testing and 640 unique cats included in the analysis. Separately, the authors report 134 FIV-positive cats, 126 FeLV-positive cats and 504 retrovirus-negative controls after exclusions. The supplied text does not reconcile these final group counts with the reported unique-cat total.
In the final FeLV model, breed, concomitant disorders and low hematocrit were identified as the most important factors for predicting seropositivity. The model had an area under the curve (AUC) of 0.69, with a 95% confidence interval of 0.64–0.74. This estimate describes discrimination of recorded diagnostic status in the dataset, not the severity of immunosuppression, infection stage or readiness for clinical use. Only four FeLV-seropositive cats were purebred.
Leukocyte classification was not significantly associated with FeLV seropositivity. This result concerns a routine blood-count classification rather than direct assessment of immune function, and it does not establish preserved immune function in positive cats. Nor did the study establish that FeLV-induced immunosuppression caused concomitant disorders or that low hematocrit specifically measured immune dysfunction. The authors caution that, in some cats, concomitant disease rather than FeLV may have been the primary reason for hospitalization; these diseases could also influence hematological findings. Erythrocyte abnormalities could distort calculated hematocrit.
The authors acknowledge possible misclassification of regressive FeLV infections as negative controls because antigen-based assays do not detect antigen-negative, provirus-positive infections. The number of regressive infections potentially misclassified in this way was not measured, and test-negative controls cannot be interpreted as uninfected cats without this qualification. The Methods describe occasional PCR testing, whereas the cited Discussion passage describes a lack of confirmatory PCR.
Missing records restricted inclusion and the available predictors. Diagnostic testing varied over time and was influenced by clinical circumstances and owners' financial considerations. In addition to discrepancies in sample counts and confirmatory PCR use, the source contains inconsistencies between performance estimates in the abstract and the Results. These limitations constrain interpretation of the observational associations in this selected hospital population. The study did not characterize immune responses across FeLV infection stages.
What we know
In this selected hospital population, breed, concomitant disorders and low hematocrit were the most important factors in the final model of FeLV seropositivity. Its AUC was 0.69 (95% CI: 0.64–0.74), without external validation. Leukocyte classification showed no significant association with seropositivity, but it was not an immune-function test. These are preliminary observational findings classified as evidence level C, subject to limitations in selection, records and diagnostic classification, as well as inconsistencies in the source.
What we don't know
It has not been established whether the model retains its performance outside the unit studied, whether FeLV-induced immune dysfunction explains concomitant disorders or how immune responses vary across infection stages. The number of regressive infections that may have been classified as negative controls is unknown. The supplied text does not resolve discrepancies concerning sample totals, confirmatory PCR use or performance estimates.
Evidence
EvidenceLevel C — Preliminary evidenceConfidence 99/100
In the final FeLV triage model, breed, concomitant disorders and low hematocrit were identified as the most important factors for predicting seropositivity.
Excerpt verified in the source — verbatim, in the source's own language“The final logistic regression models suggested that the most important factors for triaging FeLV seropositivity were breed, concomitant disorders and a low hematocrit count.”
This is a preliminary observational association in a selected hospital population. It does not establish that FeLV-induced immunosuppression caused concomitant disorders or that low hematocrit specifically measures immune dysfunction.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
Leukocyte classification was not significantly associated with FeLV seropositivity in this dataset.
Excerpt verified in the source — verbatim, in the source's own language“It is important to note that while bone marrow suppression can affect other blood cell lines such as leukocytes, our study did not find a significant association between leukocyte classification and FeLV-seropositive cats.”
The null association concerns a routine blood-count classification, not direct immune-function testing. It does not establish preserved immune function in FeLV-positive cats.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
The authors caution that concomitant disease, rather than FeLV itself, may have been the primary reason for hospitalization in some cats.
Excerpt verified in the source — verbatim, in the source's own language“However, this association should be interpreted with caution, as in some cases the concomitant disease itself may have been the primary reason for hospitalization rather than FeLV infection per se.”
The association between FeLV positivity and concomitant disease cannot by itself establish a causal pathway through immune suppression. The authors explicitly recognize alternative explanations related to admission circumstances.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
The FeLV triage model had an AUC of 0.69, with a 95% confidence interval of 0.64–0.74.
Excerpt verified in the source — verbatim, in the source's own language“The final logistic regression model showed an AUC of 0.69 (95% CI = [0.64–0.74]), an estimate similar to the one obtained for predicting FIV infection status ( Figure 1 ).”
This measures discrimination of recorded diagnostic status in the study dataset, not severity of immunosuppression, infection stage or clinical readiness. The model was not externally validated.
EvidenceLevel C — Preliminary evidenceConfidence 100/100
The authors identify possible misclassification of regressive FeLV infections as negative controls because antigen-based assays do not detect antigen-negative, provirus-positive infections.
Excerpt verified in the source — verbatim, in the source's own language“A possible limitation is the misclassification of regressively FeLV-seropositive cats as FeLV-negative controls, due to the lack of confirmatory PCR testing. Regressive infections, which are antigen-negative but provirus-positive, cannot be detected by antigen-based assays such as ELISA or immunomigration tests [ 43 ].”
This is an acknowledged classification limitation, not a measured count of regressive infections. It limits interpretation of test-negative controls as uninfected cats. Methods report occasional PCR testing, whereas this Discussion passage describes a lack of confirmatory PCR; routine molecular confirmation was not established.
Provenance
Source: Statistical Triage Model for Feline Infectious Diseases in a Veterinary Isolation Unit: The Case of Feline Immunodeficiency and Leukemia Viruses.
AVA