Patients’ Voices Predict Survival: The Missing Piece in Oncology

🧩 Can Patient-Reported Outcomes Predict Cancer Survival?

A cancer patient’s prognosis (disease course) is generally determined by stage, laboratory values and imaging . However, patient-reported outcomes (PRO – patient-reported outcomes), meaning the patient’s symptoms, quality of life and functioning , can also predict survival independently .

JAMA Oncologypublished a systematic review and meta-analysis on September 11, 2025 (source), which used data from more than 44 thousand patients and showed that PROs provide clinical outcomes with additional and distinct predictive power .

📋 What Are PROs and How Are They Used?

Patient-reported outcomes (PRO)are health data reported through patients’ own assessments. They include symptoms, quality of life, functional capacity and activities of daily living. Unlike physicians’ observations, PROs directly reflect the patient’s experience.

Today, PRO collection is mostly conducted electronically (ePRO) :

• Tablet- or kiosk-based systems: Questionnaires completed on arrival at the outpatient clinic
• Mobile apps: Daily/weekly symptom assessments
• SMS/email-based forms: Short questionnaires sent to patients with regular reminders
• Web portals: Quality of life and side effect reports submitted from home

PROs are completed by patients , but clinical researchers, physicians and healthcare managers actively use them. In clinical trials, they assess treatment effectiveness, while in routine care they serve as an early warning system .

Patients are generally asked questions at the start of treatment, at each follow-up visit and at specified intervals (e.g., every 1–4 weeks) . The most frequently assessed areas are pain, fatigue, nausea, sleep, appetite, daily functioning and perceived overall health.

This helps patients monitor themselves more effectively and gives the healthcare team an opportunity for early intervention when critical symptoms occur.

Image description: PRO (Patient Reported Outcomes) are collected through SMS, mobile apps or online forms electronically and transferred to a central PRO system. This flow provides early warnings and opportunities for personalized care for patients and the healthcare team.

• Clinical research: Provides researchers with reliable, patient-centered endpoint data.
• Personalized care: Physicians monitor symptom and functioning trends and intervene early.
• Self-management: Patients track their own symptoms, increasing awareness and engagement.
• Value-based assessment: Managers use PROs in reporting and performance measurement.

Note: Screening is generally performed before visits and in between, every 1–4 weeks ; the main areas are pain, fatigue, nausea, appetite, sleep, daily functioning and perceived overall health.

🚀 Quick Summary

• Study scope: 69 randomized controlled trials (RCTs), 44,030 patients; meta-analysis: 31 studies
• Strongest domains: Physical functioning (HR 0.94), role functioning (HR 0.96)
• Global health/Quality of life: Each +1 point → a 1% reduction in the risk of death (HR 0.99)
• Symptoms associated with poor prognosis: Nausea/vomiting (HR 1.12), pain (HR 1.07), fatigue (HR 1.05), appetite loss (HR 1.04), shortness of breath (HR 1.03)

Methods — What Is the Evidence Based On?

• Design: Systematic review of RCTs from 2000–2024; meta-analysis of 31 studies.
• Population: Adult cancer patients; main types: lung (20%), head and neck (12%), pancreatic (12%), colorectal (10%), prostate (10%).
• Measurements: Baseline PROs (including EORTC QLQ-C30 scales) and overall survival (OS).
• Analysis: Multivariable models to adjust for clinical/disease-related factors; random-effects meta-analysis, inverse-variance weighting; Egger’s test for publication bias (not significant).

📊 Main Findings — HR Values and Interpretation

Clinical Implications — PROs as the “Missing Variable”

Between two patients at the same stage, the one with higher baseline physical/role functioning may have better survival. This provides clinical justification for integrating PROs into risk stratification and decision support tools.

• Physician level: High symptom score → symptom management and supportive care: early intensification.
• Center level: Within E-nabız/EHR: a baseline PRO panel and automated alert thresholds.
• Study design: In clinical trials, PRO-based stratification and adaptive follow-up.

🛠️ 5 Practical Implementation Steps

1. Baseline screening: EORTC QLQ-C30 short form; complete it at the first outpatient visit.
2. Thresholds: Physical/role functioning <50 → nutrition, physiotherapy and the pain clinic: early involvement.
3. Digital monitoring: Monthly SMS/app questionnaires and an automated calling system for PRO monitoring at home.
4. Dashboards: On clinical dashboards: “PRO red flags” (pain ↑, nausea/vomiting ↑, fatigue ↑).
5. Reporting: PRO trend graphs in all visit summaries; always make a before/after comparison when treatment changes.

⚠️ Limitations

• Between-study heterogeneity (design, population, cancer types).
• RCT populations may be more selected than real-world patients.
• Among PRO domains, identifying the “strongest” marker by ranking was not possible (overlapping CIs).

🔭 Looking Ahead

In oncology, the combination of “biology + patient experience” could become the new standard. Integrating PROs into clinical decision support systems both sharpens survival predictionand strengthens patient-centered care.

Source: Huang RS et al. JAMA Oncology, published online: September 11, 2025. DOI: 10.1001/jamaoncol.2025.3153.