PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
May 9, 2026Journal of Clinical and Translational Science0 citationsOpen Access

27 Cachexia underdiagnosis and diagnostic delay in head and neck cancer: Insights from EHR weight data

RERonald EldridgeNSNabil F. SabaTZThomas R. Ziegler

Key Points

  • Assess the frequency and timing of weight loss in head and neck cancer patients using electronic health records to quantify cachexia underdiagnosis and diagnostic delay.
  • Identified patients using ICD-10 codes from EHR between 2016–2019.
  • Analyzed weight data for a year after the first qualifying code, focusing on those with ≥3 post-index weights.
  • Defined underdiagnosis by ≥5% weight loss without cachexia-related ICD coding present.
  • Among 561 patients, median weight measurements were 18, with a 7-day interval.
  • 11.2% median maximum weight loss was observed, occurring 5 months post-index date for 499 patients analyzed.
  • Only 23% of patients with ≥5% weight loss had ICD codes for cachexia, showing 55% were at risk for underdiagnosis.

Abstract

Objectives/Goals: Head and neck cancer (HNC) patients frequently experience cachexia (≥5% weight loss) that goes underrecognized in clinical practice. We assessed whether weights in electronic health records (EHR) were frequent and sustained enough to characterize weight-loss magnitude and timing, and to quantify cachexia underdiagnosis and diagnostic delay. Methods/Study Population: Patients were identified in the Emory Healthcare EHR (2016–2019) using ICD-10 codes for head and neck malignancies. Those with ≥2 codes within 90 days were eligible cases; the index date was the first code of the qualifying pair. Weights were extracted for all encounters six months before through one year after the index date; implausible values were removed. Feasibility analyses included all eligible patients while cachexia analyses were limited to those with ≥3 post-index weights. Baseline weight was the median pre-index value or the closest post-index value ≤14 days. Underdiagnosis was defined as ≥5% weight loss from baseline without a cachexia-related ICD code recorded 90 days prior to, or any time after, the index date: cachexia (R64), abnormal weight loss (R63.4), muscle wasting (M62.5x), or anorexia (R63.0). Results/Anticipated Results: EHR data demonstrated high feasibility for longitudinal analyses: among 561 HNC patients, the median number of weight measurements in the first year was 18 (IQR 8–27) with a median interval of 7 days (IQR 2–14). Among 499 patients analyzed for cachexia, the median max weight loss was 11.2% (IQR 4.9–18.2), occurring a median of 5 months after the index date (IQR 2.4–8.4). Overall, 373 patients (75%) lost ≥5% of baseline weight, yet cachexia-related ICD codes were present in only 113 (23%). Among uncoded patients, 274 met the ≥5% weight-based criterion, thus 55% of HNC patients were at risk for underdiagnosis. The median time to cachexia coding was 5.4 months after the index date (IQR 2.5–14.4). In 38% of HNC cachexia patients, coding occurred more than 6 months after ≥5% weight loss, indicating a meaningful diagnostic delay. Discussion/Significance of Impact: EHR data provide sufficient density and cadence to track cancer-related weight loss with precision and reveal that cachexia often goes unrecognized despite clear, measurable weight loss. These findings reveal a gap between data visibility and clinical action and support EHR-based monitoring to enable earlier recognition and timely care.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Eldridge et al. (2026) studied this question.

synapsesocial.com/papers/69fed10fb9154b0b828783fahttps://doi.org/10.1017/cts.2026.10260
Ask AI
Helpful
Bookmark
Share
View Full Paper