Adentification of Heart Failure Patients Requiring Palliative Care Utilizing a High Risk Phenotype

Introduction

PC of End Stage HF (ESHF) patients uncommon due to lack of known high risk predictors of HF mortality.

Purpose

Examine referral frequency HF patients to PC. Identify high risk indicators hospitalized ESHF patients in tertiary and community setting. Identify specific high risk phenotype of ESHF patients requiring PC.

Subjects/Methods

Medical records of 1st 200 HF admissions in suburban tertiary center and 1st 100 HF admissions in community hospital with primary DCHF discharge diagnosis retrospectively reviewed in 2008 & 2010/11. ADHERE criteria (SBP &lt 115 mmHg, blood urea nitrogen &gt 43 mg/dl, creatinine &gt 2.75 mg/dl) used to predict in-hospital mortality. Framingham CHF criteria used confirm HF diagnosis. Seattle HF Model (SHFM) calculator used at discharge to predict life expectancy, 1 year mortality and survival using HF related lab values/clinical data. Cut Chart developed to define low/mod/high dose diuretics listed in SHFMC. Referrals to PC/Hospice, patient expirations noted.

Results

Purposive sample (n=300) Patients separated into 2 groups based on SHFM mortality calculated at DC (&lt 10% and &gt 10%), chi square analysis, independent t test performed. All three cohorts statistically analyzed. Cohort 1 comprised initial 200 patients from Tertiary Care Medical Center, Cohort 2 comprised 100 patients in Community Hospital, Cohort 3 combined Cohort 1 and 2 (n=300). Framingham criteria for diagnosis HF met in 252 patients. Framingham group (n=252): (n=85) had a &gt 10% mortality, and (n=167) had &lt 10 % mortality calculated by SHFM. Statistically significant findings in &gt 10% (n=85) group in all 3 cohorts: combined Cohort 3 (n=252) comprised of 62.4% male, 91.8% non-hispanic white, mean age 75 (SD=15). Identical statistically significant findings replicated in each cohort, and both cohorts combined. Combined cohort most functionally compromised (NYHA Class III 69.4%, IV 29.4%). 37.6% had HCPs. Death 1 year 9.4%, Life expectancy 5.6 years, PC referrals 4.7%. Phenotype of High Risk Indicators (PHRI) identified as DCHF patients at discharge with: Low discharge SBP/cholesterol/EF &lt 31%, BUN &gt 35 mg/dl, ischemic etiology, higher diuretic dose (defined in Diuretic Cut Chart), patients on no statin identified as being at higher risk of mortality

Conclusions

PHRI will provide objective method of determining high risk HF patients upon discharge, and increase referrals to PC.

Implications

Timely referral HF patients to PC not offered in most cases in both settings. Familiarizing health care providers with objective criteria for evaluation of the potential need for PC referrals is needed.