Palliative Prognostic Index (PPI): score, thresholds, and limitations

What the PPI is

Morita and colleagues developed the Palliative Prognostic Index in terminally ill patients with cancer in a Japanese hospice inpatient unit. The score adds five current bedside findings and ranges from 0 to 15.¹ The original testing sample contained 95 patients and 233 assessments. This population and setting matter: PPI is a short-term prognostic aid for advanced or terminal cancer, not a general survival model for every cancer stage or setting.¹

A later prospective study assessed the score in hospital consultation, hospice home-care, and hospice inpatient settings, illustrating broader use while retaining the need for setting-aware interpretation.²

Exact component weights

| Component | Category | Points | | --- | --- | ---: | | PPS | >=60% | 0 | | PPS | 30-50% | 2.5 | | PPS | 10-20% | 4 | | Oral intake | Normal | 0 | | Oral intake | Reduced, but more than mouthfuls | 1 | | Oral intake | Mouthfuls or less | 2.5 | | Edema | Present | 1 | | Dyspnea at rest | Present | 3.5 | | Delirium | Present | 4 |

The PPI uses half-point as well as whole-point weights; it is inaccurate to call every weight an integer. An absent binary finding contributes 0 points.¹

Two overlapping thresholds, not three survival bands

The original publication reported two independent binary rules:¹

• PPI >4 predicts survival under 6 weeks: sensitivity 80%, specificity 77%.
• PPI >6 predicts survival under 3 weeks: sensitivity 80%, specificity 85%.

A score above 6 satisfies both rules. These are overlapping thresholds, not mutually exclusive categories. A score of 4 meets neither strict threshold; 4.5 and 6 meet only >4; 6.5 meets both. The calculator does not label <=4 as survival beyond 6 weeks or invent a deterministic 3-6 week interval.

Sensitivity and specificity describe cohort performance, not certainty for an individual. The ESMO guideline also describes PPI as a model for 3- and 6-week survival in advanced cancer and emphasizes understanding prognostic-tool limitations.³

PPSv2 operational caveat

The original PPI was published in 1999. Victoria Hospice identifies PPSv2 as a later revision introduced in 2001. Current PPSv2 use to choose the PPI band is therefore an operational adaptation, not evidence that the original PPI was derived with PPSv2.⁴

When PPSv2 is used, apply the official definitions, assess what the patient is capable of doing, read left to right, give leftward dimensions precedence, and select the best horizontal fit. Training matters because PPS can be easy to misapply.⁴

Using the calculator

1. Select one exact PPS band.
2. Select normal intake, reduced intake that remains more than mouthfuls, or mouthfuls or less.
3. Record edema, dyspnea at rest, and delirium as present or absent.
4. Review the 0-15 total, component trace, and each threshold independently.

The result supports, but does not replace, longitudinal assessment and uncertainty-aware communication. It does not independently determine treatment, hospice eligibility, transfer, or another care decision.