The ePRO blueprint for pain trials: Optimizing data capture from protocol design to efficacy endpoints

The ePRO blueprint for pain trials: Optimizing data capture from protocol design to efficacy endpoints

A decision-making guide for Clinical Operations and Biostatistics teams selecting between eVAS and eNRS for decentralized and BYOD pain trials.

On this page

Executive summary

The migration of clinical pain assessments to decentralized, Bring Your Own Device (BYOD) electronic formats presents trial designers with a methodological choice that carries real statistical and regulatory consequences. For decades, the 100 mm Visual Analog Scale (VAS) served as the paper-era gold standard. The operational realities of modern chronic pain trials, including variable smartphone screen sizes, older patient demographics, and the demand for daily data capture, have challenged its digital role and driven the rapid adoption of the 11-point Numeric Rating Scale (NRS).

For chronic populations managing cognitive fatigue, diminished visual acuity, or compromised motor function, the discrete integers of the eNRS produce fewer missing data points and lower user burden than the continuous spatial task required by the eVAS. Statistically, in small Phase II cohorts (N<100), the eNRS protects statistical power by eliminating the mechanical measurement error introduced by touchscreen slider slip. In large cohorts, the Standard Error of the Mean diminishes sufficiently that residual slider noise becomes negligible, and either scale is statistically defensible. The eVAS remains the correct choice for large RWE registries, legacy-continuity Phase III extensions, and dose-finding studies where 100-point continuous sensitivity is the explicit design objective.

This blueprint gives sponsors the decision framework, clinical evidence, and statistical rationale to select, justify, and deploy the right pain scale for their next protocol.

eVAS vs. eNRS: at a glance

Feature Visual Analog Scale (eVAS) Numeric Rating Scale (eNRS)
Historical use Legacy gold standard for paper trials (physically measured with a 100mm ruler). Emerged as the preferred standard for digital, mobile, and verbal assessments.
Patient preference Lower. Often perceived as abstract; requires mental translation of physical sensation to spatial distance. Higher. Patients prefer the concrete, straightforward nature of discrete numbers. Over 60% preference in crossover trials.
Age and cognition Higher cognitive load. Can be confusing for elderly patients or those with cognitive decline, leading to higher error rates. Low cognitive load. Highly accessible across age ranges and educational backgrounds.
Motor function and vision High risk Continuous slider requires fine motor control. Difficult for patients with hand tremors (OA, RA) or poor visual acuity. Low risk Large, high-contrast discrete tap targets (buttons). Accommodates gross motor function.
Small cohorts (N<100) Higher variance: touchscreen slider slip introduces mechanical noise, inflating standard deviation and threatening statistical power. Lower variance: discrete choices eliminate mechanical noise, protecting standard deviation and statistical power.
Large cohorts (N≥1,000) Statistically valid: at large sample sizes, the Standard Error of the Mean decreases sufficiently that residual slider-slip noise becomes negligible. Appropriate for large RWE registries where NRS is not mandated by legacy protocol. Statistically valid: maintains all operational and usability advantages at scale.
BYOD application Requires proportional pixel-mapping software to overcome sub-100mm smartphone screens. Technically viable; adds engineering complexity. Inherently responsive. Immune to screen size distortions and aspect ratio changes. No pixel mapping required.
Clinical significance (MCID) Typically a ≥20mm absolute reduction (approximately 30% from baseline). Can create fractional decimal boundaries in responder analyses (e.g., 19.4mm vs. 20.0mm). Typically a ≥2.0 point absolute reduction (approximately 30% from baseline). Provides a clean integer boundary for responder analyses. Directly maps to FDA preferred responder thresholds.

Introduction and historical context

Pain is an inherently subjective human experience. Accurate quantification remains one of the most persistent challenges in clinical research. For a subjective symptom to serve as a primary or secondary efficacy endpoint in a clinical trial, it must be translated into reliable, standardized, and statistically analyzable data. The evolution of how the industry captures this data, from clinic-based paper assessments to decentralized, daily mobile inputs, determines which measurement scales are appropriate for modern study designs.

The paper era and the 100mm VAS legacy

For decades, the Visual Analog Scale (VAS) was the standard for pain measurement in clinical trials. First introduced in the early 20th century and widely adopted in clinical settings by the 1970s[8], the traditional VAS consists of a continuous horizontal line anchored at both ends by descriptors such as “no pain” and “worst pain imaginable.”

The defining characteristic of the legacy VAS was its strict physical dimension: it was universally printed as exactly 100 millimeters in length.

  • The manual ruler: This 100mm standard was not born from a biological requirement but a logistical one. In the paper era, clinical coordinators used physical millimeter rulers to measure the distance from the left anchor to the patient’s pen mark, calculating a score from 0 to 100.
  • Continuous data appeal: Statisticians historically favored the VAS because it produced continuous, parametric data. It allowed for granular measurement of treatment effects, capturing micro-variations in pain intensity that discrete categories could not. This granularity remains a genuine advantage in specific study designs today.

The digital paradigm shift: ePRO and BYOD

As the industry moved into the 21st century, the operational realities of clinical trials shifted substantially. The demand for higher data integrity, the need to eliminate retrospective recall bias[15], and pressure to reduce patient burden accelerated the adoption of electronic Patient-Reported Outcomes (ePRO) and decentralized clinical trials (DCTs).

The widespread implementation of the Bring Your Own Device (BYOD) model, where patients use their personal smartphones to log daily pain diaries, created an immediate operational problem for the legacy VAS[13].

  • The screen size bottleneck: The physical 100mm requirement became logistically unenforceable across thousands of varying smartphone and tablet screen sizes.
  • The rise of the NRS: This bottleneck prompted the rapid adoption of the 11-point Numeric Rating Scale (NRS). Because the NRS relies on discrete conceptual numbers rather than a continuous physical distance, it is immune to screen size distortions and works on any device with a simple button tap.

As a result, while the eVAS can still be deployed via proportional pixel mapping, the eNRS has become the preferred digital standard for modern, mobile-first pain trials[2]. The IMMPACT consensus group formally recommended the 11-point NRS as the preferred self-report pain intensity measure for clinical trials, citing its superior reliability, ease of administration, and psychometric equivalence to longer rating scales[12].

The head-to-head comparison: clinical and demographic factors

When designing a protocol for chronic pain populations, such as those with osteoarthritis (OA), rheumatoid arthritis (RA), or chronic lower back pain, the demographic profile of the patient is as important as the statistical plan. Chronic pain populations skew older, often present with comorbidities, and may experience cognitive or motor fatigue. Evaluating the eVAS against the eNRS through a clinical lens reveals meaningful differences in usability, accessibility, and data completeness[6].

1. Usability and preference across age ranges

Age is one of the strongest predictors of scale preference and accuracy[7].

  • The elderly population: Studies consistently show that older adults find the VAS confusing and difficult to complete accurately. The abstract nature of the scale leads to higher frustration and error rates. The eNRS is highly preferred by geriatric populations due to its concrete, straightforward structure.
  • Pediatric and adolescent populations: While FACES scales are standard for young children, adolescents transitioning to adult scales grasp the 0-10 concept of the eNRS faster than the spatial mapping required by the eVAS.
  • Patient preference: In crossover trials where patients use both scales, preference favors the eNRS (often above 60%)[7], primarily because it is faster to complete and requires less mental translation on a mobile device.

2. Cognitive load and educational background

Pain scales demand varying levels of cognitive processing.

  • The cognitive burden of eVAS: The VAS requires a two-step cognitive task. First, the patient must quantify their internal pain experience. Then, they must map that abstract sensation onto a proportional physical distance. Research shows that patients with lower educational attainment or cognitive decline struggle with this spatial mapping, producing clustered or erratic data[3].
  • The simplicity of eNRS: The eNRS reduces this to a single, widely understood concept: selecting a number from 0 to 10. This lowers the cognitive load substantially, making the eNRS a more equitable tool across diverse socioeconomic and educational backgrounds.

3. Motor function and visual acuity challenges

In chronic musculoskeletal trials, the physical interface of the BYOD app can become a barrier to data entry.

  • Fine motor control and tremors: Patients with severe OA, RA, or neurological conditions often experience joint stiffness, hand tremors, or diminished fine motor control. Operating an eVAS requires placing a finger on a digital slider and dragging it to a precise pixel location. A tremor can turn an intended score of 60 into a 75, introducing mechanical noise into the dataset.
  • Visual acuity: Chronic pain patients, particularly the elderly, frequently have diminished visual acuity. A continuous line on a small smartphone screen can be difficult to read precisely. The eNRS resolves both issues with large, high-contrast, discrete tap targets meeting the WCAG 2.1 minimum 44×44px requirement for accessible touch interfaces[16]. Tapping a large “7” requires gross motor function, not fine motor precision.

4. Historical error and missing data rates

The cumulative effect of cognitive, motor, and visual barriers directly impacts clinical database integrity.

  • Unscorable data: The VAS historically yields higher rates of missing or unscorable entries compared to the NRS. Patients either abandon the slider out of frustration, fail to register the tap correctly, or leave the assessment blank because the spatial mapping is unclear[7].
  • Compliance impact: Literature indicates that the VAS failure rate can be 5% to 10% higher than the NRS in older chronic pain cohorts[7]. In a tightly powered small cohort, losing 10% of daily diary entries to user error can fatally compromise the trial’s statistical power.

The head-to-head comparison: statistical and BYOD considerations

While clinical demographics dictate how a patient interacts with a scale, the statistical design of the trial determines how that interaction translates into a p-value. When deploying a BYOD strategy, trial designers must account for the mechanical realities of smartphone screens and the mathematical vulnerabilities of their specific sample size.

1. BYOD application and screen real estate

The defining feature of a BYOD trial is hardware variance. An ePRO platform must deliver a standardized psychometric instrument across hundreds of different screen dimensions.

  • eVAS and pixel mapping: To deploy an eVAS on a screen smaller than 100mm, the software uses proportional pixel mapping. The tap position is translated to a floating-point value in the 0-100 range, producing continuous decimal outputs (such as 19.4mm) rather than integer values. Validation frameworks for BYOD screen scaling confirm that this approach can produce results psychometrically comparable to paper administration, provided the screen rendering is accurately proportional[1][14], but it still requires the patient to interact with a continuous spatial mechanism on a restricted footprint.
  • eNRS and discrete integers: The eNRS bypasses spatial limitations entirely. Because it relies on conceptual integers (0-10) presented as distinct tap targets (buttons), it does not require pixel mapping. The eNRS is inherently responsive and immune to physical screen size distortions, making it the lower-risk option for BYOD deployment from both an engineering and data-integrity perspective[2].

2. Statistical power: small vs. large cohorts

The choice between an eVAS and eNRS directly affects the standard deviation of the collected data, which in turn determines statistical power.

  • Small cohorts (N<100): In tightly powered Phase II studies (e.g., N=70), reducing data variance is paramount. The continuous slider of an eVAS introduces mechanical noise. A patient intending to log a 60 might record a 64 due to touchscreen sensitivity. The eNRS eliminates this mechanical measurement error by forcing a discrete integer choice, protecting statistical power in small samples[4]. The underlying SD of the pain outcome is not inherently tighter with NRS, but removing slider-slip error reduces the contamination of that SD with mechanical noise.
  • Large cohorts (N≥1,000): In Phase III or Real-World Evidence mega-trials, mechanical noise becomes statistically negligible. At large sample sizes, the Standard Error of the Mean decreases sufficiently that residual slider-slip variance contributes only trivially to the group-level estimate. In large cohorts, the eVAS and eNRS are virtually indistinguishable in their ability to detect treatment effects, allowing the sponsor to choose based on historical preference or legacy protocol requirements.

When eVAS continuous granularity is a genuine statistical asset

The eNRS’s advantage is most pronounced in small chronic pain cohorts. There are specific study types where the eVAS’s 100-point continuous resolution is a deliberate design choice rather than a liability:

  • Analgesic dose-finding studies (Phase II dose-response): When a study is designed to detect incremental differences between adjacent dose levels of an analgesic, the inter-dose effect size may be 5-8mm on the VAS scale, which translates to only 0.5-0.8 points on the 11-point NRS and falls below NRS resolution. In these designs, the 100-point continuous VAS scale captures between-dose sensitivity that the discrete NRS cannot. This is one of the clearest clinical rationales for selecting eVAS in a small cohort, despite the mechanical noise concern.[17][18]
  • Studies following published VAS precedent: Some established pain research domains have a published body of VAS data that informs power calculations. If the sponsor’s power assumptions are derived from VAS effect size estimates in that literature, switching to NRS requires re-deriving those assumptions. Where this constraint applies, eVAS is the more defensible choice.
  • Post-surgical acute pain in controlled settings: Unlike chronic pain populations, post-surgical patients typically have intact motor function and are often assessed in a clinic or provisioned-device context rather than BYOD. The usability barriers that make eVAS problematic in chronic, elderly, decentralized populations are substantially reduced when patients are younger, have normal hand function, and are using a standardized provisioned device.

3. The Minimum Clinically Important Difference (MCID)

Regulatory bodies evaluate efficacy not just on statistical significance but on clinical relevance. The protocol must define the Minimum Clinically Important Difference (MCID), the smallest change in a treatment outcome that a patient would identify as meaningful.

  • Proportional equivalency: The MCID for pain reduction is consistent across both scales due to proportional equivalency. In foundational analyses of chronic pain trials, a clinically meaningful reduction corresponds to roughly a 30% drop from baseline[4][9], or approximately 2.0 points on the 11-point NRS. The IMMPACT consensus group specifically identified the 11-point NRS as the preferred self-report pain intensity measure for clinical trial primary endpoints, citing its psychometric properties and ease of use[12].
  • Translating to the VAS: This 2.0-point eNRS reduction translates proportionally to a 20mm reduction on the 100mm eVAS. Both represent approximately a 30% reduction from a mid-range baseline score.
  • The responder analysis advantage: While the MCID is proportionally equal, the eNRS makes the FDA-preferred responder analysis mathematically cleaner at the point of capture. It is simpler to program a database lock to flag a discrete integer drop of 2 or more than to manage fractional decimal outputs (e.g., 18.4mm vs. 21.1mm) generated by a continuous eVAS pixel map. A patient recording 19.4mm technically misses a 20mm responder threshold despite clinically equivalent relief. One important nuance: in chronic pain trials using a 7-day rolling average as the primary endpoint, the average of multiple integer NRS scores will itself produce fractional values regardless of scale. The practical advantage of the eNRS is at the point-of-capture (patient reliability and discrete input) rather than at the final endpoint computation level, where rolling averages will generate decimals either way.

The decision framework: when to use which scale

Both the eVAS and eNRS are scientifically validated and accepted by global regulatory bodies. They are not universally interchangeable in practice. Trial designers must weigh the specific parameters of their protocol, from patient demographics to statistical constraints and regulatory strategy, to make the right choice.

Recommended for most modern trials

When to select the eNRS

  • The sample size is small (N<100): Discrete integers eliminate mechanical slider variance, protecting statistical power in tightly powered Phase II studies.
  • The cohort is elderly or cognitively burdened: As seen in osteoarthritis, RA, or Alzheimer’s-related pain studies, concrete integer selection reduces missing data and user error significantly.
  • The trial relies on patient BYOD smartphones: Large, discrete tap targets are immune to screen size distortions and accommodate compromised motor function.
  • FDA responder analysis is the primary efficacy endpoint: Clean integer boundaries prevent fractional ambiguity at the responder threshold.
  • The trial spans multiple countries and languages: While both scales require anchor localization, the eNRS’s verbal administration option provides additional operational flexibility in multilingual settings.

Specific scenarios where eVAS is justified

When to select the eVAS

  • Legacy continuity is strictly mandated: If a Phase III extension study must perfectly mirror a legacy Phase II paper VAS endpoint without data transformation.
  • The sample size is massive (N>1000): In large RWE registries or mega-trials, mechanical slider variance is washed out by data volume, making the continuous scale statistically safe.
  • Dose-response sensitivity is the design objective: When a study is designed to detect fine inter-dose differences (e.g., 5-8mm between adjacent dose arms), the 100-point continuous resolution of the eVAS captures sensitivity below NRS resolution. This is most relevant in analgesic dose-finding Phase II studies.[17][18]
  • Post-surgical acute pain with controlled device conditions: Non-chronic populations with normal motor function and provisioned (non-BYOD) devices reduce the primary barriers that make eVAS problematic in decentralized chronic pain settings.

Strategic consideration 1: the FDA responder analysis

When submitting a medical product for labeling claims, the FDA increasingly prefers a responder analysis over a simple evaluation of mean differences. A responder analysis evaluates the proportion of patients who achieve a predefined, clinically meaningful reduction in pain score, for example a 30% or 2-point drop from baseline.

  • The eNRS advantage: The eNRS makes defining and calculating the responder threshold mathematically clean. A patient either drops from an 8 to a 6 (a definitive responder) or they do not.
  • The eVAS challenge: The continuous nature of the eVAS can create ambiguous boundaries. If the responder threshold is a 20mm reduction, a pixel-mapped slider might record a drop of 19.4mm. From a clinical perspective, the patient experienced meaningful relief. From a strict database programming perspective, they fail the responder criteria. The eNRS eliminates this fractional ambiguity.

Strategic consideration 2: phase migration strategy

A common hesitation among sponsors is the fear that switching from VAS in Phase II to NRS in Phase III will invalidate comparative data.

Psychometric literature addresses this directly. Research consistently demonstrates a high correlation (often r > 0.85) between the VAS and NRS[7], and validated transformation approaches exist (such as dividing a 0-100mm VAS score by 10 as an approximation). This is not, however, a regulatory rubber stamp. FDA and EMA may require formal equivalence justification or a bridging study before accepting a scale switch within a development program, particularly when the Phase II endpoint forms the basis of Phase III power calculations. Sponsors should engage with regulators early and document the scientific rationale thoroughly before executing the transition. When those conditions are met, sponsors are not permanently locked into a legacy instrument if the operational realities of the next trial phase justify the change.

Operationalizing the decision framework: the Castor platform

Translating a rigorous psychometric framework into a functional clinical trial requires a technology partner capable of balancing scientific precision with a consistent patient experience. Whether a protocol demands the legacy continuity of a continuous eVAS or the statistical protection of a discrete eNRS, the ePRO platform must execute that choice reliably across a decentralized patient population.

Castor has built specialized infrastructure for musculoskeletal (MSK), spine, and pain research, supporting complex, longitudinal ePRO protocols on consumer hardware.

Proven scale in pain and MSK research

The advantages of BYOD data capture only materialize when deployed on a platform with production experience in the therapeutic area. Castor actively supports 9 commercial MSK and pain studies (out of 457 total in the therapeutic area), spanning both tightly controlled Phase II cohorts and large, decentralized registries.

  • Orthofix Rotator Cuff Stimulation Study: Castor powers this 1,021-participant trial, which relies on longitudinal, ePRO-driven daily pain diary designs required to meet modern regulatory standards at scale.
  • Artialis Biomarker Programs: Partnering with Artialis, a commercial MSK biomarker company, Castor supports controlled programs that depend on high-fidelity, ePRO-based patient-reported endpoints to correlate with precise biological changes.

Cross-device rendering

As established by ISPOR guidelines, the validity of a BYOD deployment depends on presenting a consistent instrument across diverse consumer devices. Castor’s platform supports both scale options without operational compromise.

  • eNRS deployment: For protocols using the eNRS, Castor delivers responsive, high-contrast discrete buttons. This ensures accessibility for elderly patients or those with diminished motor function, and eliminates screen-size distortion entirely.
  • eVAS pixel mapping: If a protocol requires an eVAS for legacy continuity or specific study design reasons, Castor’s rendering engine uses validated proportional pixel mapping. Whether the patient is operating a 4-inch smartphone or a 12-inch tablet, the continuous line maintains its structural integrity and proportional scoring, ensuring conceptual equivalence across all hardware[1].

Protecting statistical power through patient-centric design

In a daily chronic pain diary protocol, survey fatigue is the primary threat to study completion. Castor uses patient-centric features to maximize compliance and protect the trial’s standard deviation.

  • Conditional skip logic: To reduce daily cognitive burden, Castor uses conditional branching. Patients are only presented with relevant follow-up questions based on their initial pain scores. For example, a prompt regarding breakthrough rescue medication will only appear if the patient logs an eNRS score above a defined threshold. This streamlines the daily diary to seconds, reducing participant fatigue and dropout.
  • Proactive compliance monitoring via real-time dashboards: In a 70-patient Phase II trial, every data point matters. Castor equips clinical operations teams with real-time compliance dashboards. Study coordinators are flagged immediately when a patient becomes disengaged or misses a daily entry, enabling intervention before data is permanently lost.

By combining validated cross-device rendering with proactive compliance monitoring, Castor provides the operational infrastructure needed to capture regulatory-grade pain data in modern decentralized clinical trials.

Common questions on pain scale selection

The eVAS requires patients to place a finger on a continuous digital slider and drag it to a precise pixel location, which introduces mechanical noise from touchscreen sensitivity. The eNRS presents discrete tap targets (buttons labeled 0-10), eliminating slider slip entirely. In BYOD settings, the eNRS is also immune to screen size variation, while the eVAS requires proportional pixel mapping to render correctly across devices.

The eVAS is most appropriate when: (1) a Phase III extension study must maintain continuity with a legacy Phase II paper VAS endpoint; (2) sample sizes are large (N>1000) where mechanical slider variance becomes statistically negligible; (3) a dose-finding study requires continuous 100-point sensitivity to detect fine inter-dose differences below NRS resolution; or (4) the study involves post-surgical acute pain in a controlled device setting where chronic pain usability barriers do not apply.

For NRS, the Minimum Clinically Important Difference (MCID) is approximately 2.0 points, which corresponds to roughly a 30% reduction from baseline. For VAS, the equivalent MCID is approximately 20mm on the 100mm scale, also representing approximately 30% from a mid-range baseline. The NRS MCID provides a cleaner integer boundary for FDA-preferred responder analyses, while the VAS MCID can produce fractional decimal values that complicate database programming.

Psychometric literature consistently demonstrates a high correlation (r > 0.85) between VAS and NRS scores, and validated transformation approaches exist, such as dividing the 0-100mm VAS score by 10 as an approximation. However, this is not a regulatory rubber stamp. FDA and EMA may require formal equivalence justification or a bridging study before accepting a scale switch within a development program, particularly when the Phase II endpoint forms the basis of Phase III power calculations. Sponsors should engage with regulators early and document the scientific rationale thoroughly before executing the transition.

References

  1. Bakhshi, S., et al. (2022). Measurement Comparability of Electronic and Paper Administration of Visual Analogue Scales: A Review of Published Studies. Therapeutic Innovation & Regulatory Science, 56(4), 684-695.
  2. Byrom, B., et al. (2018). Measurement Equivalence of Patient-Reported Outcome Measures Migrated to Electronic Formats: A Review of Evidence and Recommendations for Clinical Trials and Bring Your Own Device. Therapeutic Innovation & Regulatory Science, 52(4), 425-431.
  3. Chiarotto, A., et al. (2019). Measurement properties of visual analogue scale, numeric rating scale, and pain severity subscale of the brief pain inventory in patients with low back pain: a systematic review. The Journal of Pain, 20(3), 245-263.
  4. Farrar, J. T., Young, J. P., LaMoreaux, L., Werth, J. L., & Poole, R. M. (2001). Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain, 94(2), 149-158.
  5. Gwaltney, C. J., et al. (2008). Equivalence of electronic and paper-based patient-reported outcome measures: a meta-analysis. Value in Health, 11(2), 322-333.
  6. Hawker, G. A., Mian, S., Kendzerska, T., & French, M. (2011). Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire, and others. Arthritis Care & Research, 63(S11), S240-S252.
  7. Hjermstad, M. J., et al. (2011). Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. Journal of Pain and Symptom Management, 41(6), 1073-1093.
  8. Huskisson, E. C. (1974). Measurement of pain. The Lancet, 304(7889), 1127-1131.
  9. Salaffi, F., Stancati, A., Silvestri, C. A., Ciapetti, A., & Grassi, W. (2004). Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale. European Journal of Pain, 8(4), 283-291.
  10. Suso-Ribera, C., et al. (2020). Validation of a Smartphone-Based eVAS for Acute Pain Assessment. Journal of Medical Internet Research, 22(2), e13468.
  11. U.S. Food and Drug Administration (FDA). (2009). Guidance for Industry: Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. U.S. Department of Health and Human Services.
  12. Dworkin, R. H., et al. (2005). Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain, 113(1-2), 9-19. [IMMPACT consensus recommendations identifying the 11-point NRS as the preferred self-report pain intensity measure for clinical trials.]
  13. U.S. Food and Drug Administration (FDA). (2023). Digital Health Technologies for Remote Data Acquisition in Clinical Investigations: Guidance for Industry, Investigators, and Other Stakeholders. U.S. Department of Health and Human Services. [Current FDA guidance on BYOD, mobile apps, and continuous data capture in clinical trials.]
  14. Zbrozek, A., et al. (2013). Patients’ acceptance of tablet computers for the collection of patient-reported outcome data: a review of published literature. Value in Health, 16(3), 551-558. [ISPOR/ePRO Consortium validation frameworks for BYOD screen scaling and electronic administration of patient-reported outcomes.]
  15. Stone, A. A., Shiffman, S., Schwartz, J. E., Broderick, J. E., & Hufford, M. R. (2002). Patient non-compliance with paper diaries. BMJ, 324(7347), 1193-1194. [The foundational “parking lot study” demonstrating fabrication of paper diary entries and establishing the need for electronic, time-stamped ePRO collection to eliminate retrospective recall bias.]
  16. W3C Web Accessibility Initiative (WAI). (2018). Web Content Accessibility Guidelines (WCAG) 2.1. World Wide Web Consortium. Success Criterion 2.5.5: Target Size — minimum touch target size of 44×44 CSS pixels for interactive elements. Available at: https://www.w3.org/TR/WCAG21/
  17. Jensen, M. P. (2003). The validity and reliability of pain measures in adults with cancer. The Journal of Pain, 4(1), 2-21.
  18. Breivik, H., Borchgrevink, P. C., Allen, S. M., et al. (2008). Assessment of pain. British Journal of Anaesthesia, 101(1), 17-24.

Related Posts

To read the rest of this content, please provide a little info about yourself

EDC For Researchers, Designed By Researchers

Discover all the features offered by Castor EDC

Discover Now
Back to top