Nosocomial Infections in ICU · Journal article

Palliative care documentation in gastrointestinal bleeding hospitalizations: national trends and associated care patterns

Baylor University Medical Center Proceedings · August 10, 2026

Preliminary Why this matters to you Expert discussion

Early or partial results. Treat as a signal, not a conclusion.

AI Summary

This retrospective administrative study documents that palliative care coding occurred in 2.70% of gastrointestinal bleeding hospitalizations (2016–2022), rising from 2.13% to 3.43%, and was associated with markers of higher acuity and reduced procedural intensity. However, the study cannot establish whether documented palliative care caused these patterns or identify what clinical interventions were actually delivered, limiting interpretability for clinical decision-making.

What was studied

Retrospective serial cross-sectional study. Adult principal-diagnosis upper or lower gastrointestinal bleeding hospitalizations in the U.S. National Inpatient Sample, 2016–2022. Specific age range, exclusion criteria, or baseline comorbidity distribution not stated.. Intervention: Palliative care documentation (ICD-10-CM Z51.5 in any diagnosis field). Compared with: No palliative care documentation. n = 2,435,099. United States (National Inpatient Sample).

Key findings

Palliative care documentation present in 65,675 of 2,435,099 weighted hospitalizations (2.70%), increasing from 2.13% in 2016 to 3.43% in 2022 Metastatic cancer associated with highest adjusted odds ratio (aOR 8.276), followed by age ≥85 years (aOR 5.626), respiratory failure (aOR 3.159), and dementia (aOR 3.101) Black and Hispanic patients had lower adjusted odds of documentation relative to White patients (aOR 0.802 and 0.791 respectively)

Safety

Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.

Who this may help

Clinicians should recognize that palliative care coding remains uncommon in GIB hospitalizations despite rising trends, and that coded documentation correlates with higher-acuity presentations and reduced intervention intensity. However, because the code does not capture timing, specialty, or actual care delivered, these data cannot guide decisions about whether or when to offer palliative care consultation in individual GIB patients.

Plain-language summary

Retrospective observational study of administrative coding data that characterizes documentation prevalence and associations, but cannot establish causation, clinical impact, or validate whether documented palliative care drove observed care patterns.

Study design

As stated by the source record.

MethodsStudy population identified from the National Inpatient Sample (2016–2022) using ICD-10-CM code Z51.5 in any diagnosis field to define palliative care documentation. Survey-weighted logistic regression assessed factors associated with documentation and inpatient outcomes including endoscopy use, mechanical ventilation, renal replacement therapy, do-not-resuscitate coding, length of stay, and hospital charges. No explicit description of missing data handling or sensitivity analyses provided.

DesignRetrospective serial cross-sectional study

PopulationAdult principal-diagnosis upper or lower gastrointestinal bleeding hospitalizations in the U.S. National Inpatient Sample, 2016–2022. Specific age range, exclusion criteria, or baseline comorbidity distribution not stated.

Sample size2,435,099

InterventionPalliative care documentation (ICD-10-CM Z51.5 in any diagnosis field)

ComparatorNo palliative care documentation

SettingUnited States (National Inpatient Sample)

Reported figures

Quoted from the source exactly as published.

Sample size2435099

Palliative care documentation prevalence2.70%

Prevalence in 20162.13%

Prevalence in 20223.43%

aOR metastatic cancer8.276

aOR age ≥85 years5.626

aOR respiratory failure3.159

aOR dementia3.101

Clinical bottom line

Caveats

Administrative coding is a proxy for palliative care documentation and does not capture content, duration, provider specialty, or timing of consultation relative to outcomes.
Study design (cross-sectional) does not establish causation; associations may reflect confounding by disease severity, prognostication, or patient preference.
No confidence intervals, p-values, or statistical significance testing reported for outcome associations; adjusted vs. unadjusted comparisons not explicitly labelled.
Lack of validation data comparing ICD-10-CM Z51.5 coding to medical record review of actual palliative care services delivered; disparities in documentation by race/ethnicity may reflect coding variation rather than access differences.

Evidence assessment

Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.

Study design

moderate

Retrospective cross-sectional design using national administrative database; no randomization or control group.

Directness

limited

Uses billing codes (ICD-10-CM Z51.5) as proxy for palliative care; does not capture timing, provider specialty, or content.

Consistency

insufficient

No comparator studies cited; single-source analysis of U.S. national sample.

Risk of bias

limited

Administrative data subject to coding variation, missing data, and selection bias; authors acknowledge patterns are associations not effects.

Sample adequacy

strong

Large national sample of 2,435,099 weighted hospitalizations across 7 years.

Source abstract

BACKGROUND: Gastrointestinal bleeding (GIB) often affects older, medically complex patients, yet national patterns of inpatient palliative care documentation in GIB are unknown. We aimed to characterize trends and factors associated with palliative care documentation and examine associated inpatient care and outcome patterns. METHODS: We conducted a retrospective serial cross-sectional study of adult principal-diagnosis upper or lower GIB hospitalizations in the 2016 to 2022 National Inpatient Sample. Documentation was defined by ICD-10-CM Z51.5 in any diagnosis field. Survey-weighted regression assessed associated factors and inpatient outcomes. RESULTS: Among 2,435,099 weighted hospitalizations, palliative care documentation was present in 65,675 (2.70%), increasing from 2.13% in 2016 to 3.43% in 2022. Associated factors included metastatic cancer (adjusted odds ratio [aOR], 8.276), age ≥85 years (aOR, 5.626), respiratory failure (aOR, 3.159), and dementia (aOR, 3.101). Relative to the White category, Black (aOR, 0.802) and Hispanic (aOR, 0.791) categories had lower adjusted odds. Documentation was associated with lower odds of endoscopy but higher odds of mechanical ventilation and renal replacement therapy. Among in-hospital decedents, coded do-not-resuscitate status was more frequent (80.2% vs 40.5%), and several treatments were less frequent. Adjusted length of stay and hospital charges were 4.3% and 1.1% higher, respectively. CONCLUSIONS: Palliative care documentation was uncommon and concentrated in higher-acuity hospitalizations. Because the code does not identify provider specialty, timing, or content, patterns are associations rather than effects and require validation in clinically richer data.

Follow from this

Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.

Brian NohomovichKirsten HickokEric Martin SieloffAkram AlnounouFnu VikashNosocomial Infections in ICUMohammad Alabbas