Table 1.
Summary of considerations across each reviewed pain methodology.
| Considerations | Methods and signals | |||||
|---|---|---|---|---|---|---|
| Patient-reported outcomes | Language proxies | Alternative visual reports | Physiological and environmental sensors | Neuroimaging | ||
| Dimensionality | Patient pain expression is more open or unlimited | |||||
| Captures short-term pain dynamics | ||||||
| Captures long-term pain dynamics | ||||||
| Captures multiple dimensions beyond intensity, quality, or location | ||||||
| Data has ecological validity (supports collection/analysis in multiple environments) | ||||||
| Data form is intended for pain-specific uses | ||||||
| More direct access to biological signals | ||||||
| Can more easily account for social influences | ||||||
| Data | Requires active participant engagement | |||||
| Collection | Supports passive data collection | |||||
| Internet/Connectivity/Smart Device-dependent | ||||||
| Data collection methods are scalable | ||||||
| Relatively easy to set-up data collection methods | ||||||
| Requires extensive researcher/clinician training | ||||||
| Prone to noise introduced from technical interfaces or environment | ||||||
| Prone to noise due to methodological and/or user errors | ||||||
| Data Analytics | Can be qualitatively analyzed | |||||
| Can be quantitatively analyzed | ||||||
| Data analysis is scalable | ||||||
| Requires advanced statistical analyses or complex processing | ||||||
| Existing analytical standards or benchmarks for reference | ||||||
| Accessibility | Flexible, participant-tailored collection methods possible | |||||
| May reduce patient burden (time, cost, or physical reqs) | ||||||
| May reduce researcher burden (time, costs, analysis reqs) | ||||||
| Utility | Method used in clinical settings or contexts to aid in therapeutic decisions | |||||
| Method itself can be therapeutic | ||||||
Color indicates extent to which each consideration applies. Orange—generally true of this methodology; blue—generally not true for this methodology; grey—varies often; and black—not applicable or unknown. Machine learning as a technique is not listed here as it requires data from the remaining methods and signals. Additionally, the problem of introducing human bias into data collection and analysis is also not listed here, as it's a consideration that applies to all methods and varies greatly.