Is the palliative performance scale a useful predictor of mortality in a heterogeneous hospice population?

J Palliat Med. 2005 Jun;8(3):503-9. doi: 10.1089/jpm.2005.8.503.


Background: Hospices provide care to patients with a wide range of prognoses, and must develop care plans that anticipate each patient's likely illness trajectory. However, the tools available to guide prognostication and care planning in this population have limited data to support their use. For instance, one of the most widely-used prognostic tools, the Palliative Performance Scale (PPS), has been studied primarily in inpatient settings and in patients with cancer. Its prognostic value in a heterogeneous US hospice population is unknown.

Objective: The goal of this study was to evaluate the prognostic value of the PPS as a predictor of mortality in a heterogeneous hospice population, and to determine whether it performs equally well across diagnoses and sites of care.

Design: Prospective cohort study using existing medical records.

Setting/subjects: This study was conducted at a large community hospice program, and included all patients enrolled in hospice during the study period.

Measurements: Each patient's PPS score was recorded at the time of enrollment and patients were followed until death or discharge from hospice.

Results: A total of 466 patients enrolled in hospice during the study period. The PPS score was a strong independent predictor of mortality (log rank test of Kaplan Meier survival curves p < 0.001). Six-month mortality rates for 3 PPS categories were 96% (for PPS scores 10-20), 89% (for PPS scores 30-40), and 81% (for PPS scores > or =50). Evaluation of interaction terms in Cox proportional hazards models demonstrated a stronger association between PPS score and mortality among nursing home residents and patients with non-cancer diagnoses. Analysis of the area under receiver operating characteristic curves demonstrated strong predictive value overall, with somewhat greater accuracy for nursing home residents and patients with noncancer diagnoses.

Conclusion: The PPS performs well as a predictor of prognosis in a heterogeneous hospice population, and performs particularly well for nursing home residents and for patients with non-cancer diagnoses. The PPS should be useful in confirming hospice eligibility for reimbursement purposes and in guiding plans for hospice care.

MeSH terms

  • Aged
  • Cohort Studies
  • Disease / classification
  • Female
  • Hospice Care*
  • Humans
  • Karnofsky Performance Status*
  • Length of Stay
  • Male
  • Mortality*
  • Palliative Care*
  • Pennsylvania / epidemiology
  • Prognosis
  • Survival Analysis