Improving HLH Detection with Computable Phenotypes

Hemophagocytic lymphohistiocytosis (HLH) represents a life-threatening hyperinflammatory syndrome requiring immediate clinical intervention. However, identifying these patients within hospital systems remains notoriously difficult because diagnostic criteria are complex. A recent retrospective cohort study explored the utility of computable phenotypes for HLH to improve patient identification and outcome tracking within electronic health records (EHRs).

Evaluating Diagnostic Approaches in EHRs

Researchers compared three distinct methods to identify HLH patients using EHR data. These included searching for specific ICD-10 diagnosis codes, tracking HLH-specific treatment plans, and evaluating clinical criteria based on HLH-2004 guidelines. Consequently, the study highlighted significant discrepancies in how different data points capture the patient population. No single method successfully identified all potential cases.

Implementing Computable Phenotypes for HLH

The research team identified 388 patients across three different cohorts. Interestingly, using an HLH ICD-10 diagnosis (n=220) and meeting five or more clinical criteria (n=245) proved more common than identifying patients through a formal treatment plan (n=42). Moreover, among the 1325 patients presenting with fever and elevated ferritin, only 19% actually met five or more clinical criteria. This finding suggests that while ferritin is a sensitive marker, it lacks specificity for HLH in broader hospital populations.

Clinical Outcomes and Mortality Findings

The study also analyzed the impact of therapy on survival. Approximately 79% of patients who met the HLH-2004 criteria received directed therapy, such as dexamethasone or etoposide. Despite these interventions, in-hospital mortality remained alarmingly high. Specifically, mortality reached 15% in the treated group compared to 13.5% in those not receiving HLH-directed therapy. Therefore, the findings emphasize the urgent need for earlier identification strategies to improve survival.

Frequently Asked Questions

Why is ICD-10 coding for HLH often insufficient?

Diagnosis codes frequently fail to capture the full spectrum of HLH. Clinicians might not apply the specific ICD-10 code during the acute phase of illness. Furthermore, the overlapping symptoms with other inflammatory conditions can lead to coding variability across different departments.

How does the HLH-2004 criteria perform in EHR data?

While the HLH-2004 criteria remain the diagnostic gold standard, many parameters are not captured systematically. As a result, constructing accurate cohorts requires sophisticated computable phenotypes. These algorithms integrate disparate data points like lab values and clinical notes to provide a clearer diagnostic picture.

Disclaimer: This content is for informational and educational purposes only. It does not constitute medical advice, diagnosis, or treatment. Always seek the advice of a qualified healthcare provider regarding any medical condition. Refer to the latest local and national guidelines for clinical practice.

References

1. Ocak S et al. Identifying Hemophagocytic Lymphohistiocytosis and Describing Outcomes Using Computable Phenotypes: Retrospective Cohort Study. JMIR Cancer. 2026 Mar 26. doi: 10.2196/87347. PMID: 41886745.

2. Henter JI, et al. HLH-2004: Diagnostic and therapeutic guidelines for hemophagocytic lymphohistiocytosis. Pediatr Blood Cancer. 2007;48(2):124-131. doi: 10.1002/pbc.21039.

3. La Rosée P, et al. Recommendations for the management of hemophagocytic lymphohistiocytosis in adults. Blood. 2019;133(23):2465-2477. doi: 10.1182/blood.2018894618.