Streamlining patient data collection for lung cancer research
Lung cancer research needs more than scans and lab results. Symptoms, quality of life and lifestyle tell researchers how treatment affects real lives. Here is how to capture all of it without burdening patients or compromising data quality.
Lung cancer is the leading cause of cancer death in Hong Kong. Treatment has changed quickly, with targeted therapies and immunotherapy guided by genomic testing, and research has to keep pace. That means combining clinical data with what patients experience between visits.
Why lifestyle and patient-reported data matter
Clinical records show diagnoses, treatments and results. They rarely show fatigue, breathlessness, sleep, diet, smoking history or emotional wellbeing, all of which shape outcomes and treatment tolerance. Patient-reported outcome measures (PROMs) and lifestyle questionnaires fill that gap, helping researchers see correlations, risk factors and the real-world effect of treatment.
Where the data comes from
Patient questionnaires
Validated PROM instruments and lifestyle surveys, completed on a phone or tablet at home or in clinic.
Clinical records
Diagnosis, staging, treatment and laboratory results, ideally pulled from hospital systems rather than re-keyed.
Genomic reports
Biomarker and genomic profiling results that guide targeted therapy and define research cohorts.
Connected devices
Wearables that track activity and sleep, where the study design calls for continuous data.
Designing collection patients will actually complete
Patients with lung cancer are often tired, older and managing a demanding treatment schedule. Collection has to fit around them:
- Short, validated instruments, one question per screen, with save-and-resume.
- Bilingual Traditional Chinese and English, large text and plain language.
- Timed reminders that follow the treatment calendar, with a caregiver option where appropriate.
- Clinic-side capture for patients who prefer not to use their own phone.
Protecting data quality
Real-time capture reduces recall bias, but only if the data is clean. Build quality in from the start:
- Validation at entry: range checks, required fields and consistency rules.
- Standard formats and coding, so data from different sites and studies can be combined.
- Versioned instruments, so every answer can be traced to the exact questions the patient saw.
- Audit trails for every correction, recording who changed what, when and why.
Consent, privacy and ethics
Research data about cancer patients is among the most sensitive data there is. Consent should be layered and easy to understand, and stored as data the system checks before any record is used. Identifiable details should sit apart from research data, so analysts work only with pseudonymised records. In Hong Kong, collection and use must also follow the Personal Data (Privacy) Ordinance and the study’s ethics approval.
From data to insight
With clean, linked data, researchers can run longitudinal analyses of symptoms across treatment lines, compare quality of life between regimens, and study how lifestyle factors relate to outcomes. Larger, better-retained cohorts give studies more statistical power, and the same platform can feed findings back to clinicians caring for patients today.
For the full picture of how such a platform should be built, from experience design to architecture and validation, see Every data point is a patient: designing safe patient data systems.
Planning a patient-reported outcomes study?
Tell us about your cohort, instruments and timeline. We will show you how to collect better data with less burden on patients.
Get In Touch →