Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 34 of the 34 references it cites were found in a registry or an index.

Outcome measures based on digital health technology sensor data: data- and patient-centric approaches

Open access

Abstract

Digital health technology tools (DHTTs) are digital tools that remotely and frequently acquire health and disease-related data from individuals (see Table 1 for a list of terms and definitions used here). DHTTs are becoming a standard part of clinical research studies in which patients collect two basic modalities of sensor data 1 , 2 , 3 , 4 , 5 , 6 , 7 : (1) sensor data can be captured while patients perform specific tasks designed to measure key symptoms (“active tests”) and (2) sensors may monitor human behavior passively (“passive monitoring”). Active tests can be administered, e.g., using tablets, smartphones, and to a limited extent also smartwatches. Passive data can be collected using a wide range of devices such as smartphones, smartwatches, wearables, beacons, nearables, and smart patches. Digital sensors can sample behavior at a frequency many times higher than is possible during traditional in-clinic testing. Thus, there are many advantages to the use of this technology in clinical research. Firstly, information is collected from patients much more frequently than would be possible at infrequent clinic visits, with the potential to increase the signal to noise and increasing reliability and validity 8 and the measures’ ability to detect change over time 9 . Secondly, the acquired data are objective: while trained healthcare professionals have a wealth of knowledge with which to characterize and treat patients’ symptoms, the reliance on such ratings necessarily adds a subjective component to assessments including possible increased interrater variability and decreased reliability (e.g., see ref. 10 ). Thirdly, data are collected in the environment in which individuals actually live and work, generating ecologically valid measurements, i.e., ones which approximate the real world. A fourth key advantage is the amount of sensor data generated (e.g., via continuous 50 Hz sampling), promising the phenotypic characterization of patients with an unprecedented breadth and precision. These advantages have led to DHTTs being used with increasing frequency in clinical trials and observational studies to complement traditional clinical outcome assessments (COA) and biomarkers. However, one of the chief open questions now facing the field is what to do with this vast amount of sensor data, i.e., how can patients’ sensor data be integrated into one or several outcome measures which meaningfully recapitulate the clinical status of the patient 5 , 11 ?

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41746-020-0305-8

Cites

34 of 34 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Procedure to develop digital outcome measures. Signal processing generates sensor features which may be selected in a data-driven or patient-centered procedure to generate maximally sensitive or patient-relevant digital health technology outcome measures, respectively.
Fig. 1 Procedure to develop digital outcome measures. Signal processing generates sensor features which may be selected in a data-driven or patient-centered procedure to generate maximally sensitive or patient-relevant digital health technology outcome measures, respectively.CC BY · PubMed Central
Fig. 2 A roadmap to build and validate a data-centric digital outcome measure to monitor disease progression. A build dataset is used to select robust features for model building. The model building procedure is iterative, beginning with the generation of preliminary models tailored to the concept of interest, followed by tests for statistical robustness and preliminary validation. An independent validate dataset is used to confirm the statistical robustness of the model, validate that the model performs as expected, and support preliminary interpretation of the score magnitudes via comparisons with reference clinical assessments. Leftward arrows denote iterative processes.
Fig. 2 A roadmap to build and validate a data-centric digital outcome measure to monitor disease progression. A build dataset is used to select robust features for model building. The model building procedure is iterative, beginning with the generation of preliminary models tailored to the concept of interest, followed by tests for statistical robustness and preliminary validation. An independent validate dataset is used to confirm the statistical robustness of the model, validate that the model performs as expected, and support preliminary interpretation of the score magnitudes via comparisons with reference clinical assessments. Leftward arrows denote iterative processes.CC BY · PubMed Central
Fig. 3 A roadmap to develop a patient-centric sensor-based outcome measure for an existing DHTT. Patient input provides the foundation for the development of this outcome measure. A conceptual model describes symptoms and impacts that are important to patients. A conceptual framework depicts the relationships between the concepts measured via the DHTT and associated items and domains. Additional qualitative research seeks to establish the impacts relevant to patients and their relationships with the DHTT features. Clinical expert opinions are required to ensure that the features represented in the final conceptual framework are neurobiologically relevant. The ensuing stages of outcome development require two independent datasets: the build dataset is used to select statistically robust and appropriate features, build a scoring algorithm, confirm psychometric properties of the scoring algorithm, and provide a preliminary validation; the independent validate dataset confirms the robustness of the psychometric properties, provides further validation data related to the DHTT outcome measure, and generates minimally clinically important difference (MCID) estimates and responder definitions.
Fig. 3 A roadmap to develop a patient-centric sensor-based outcome measure for an existing DHTT. Patient input provides the foundation for the development of this outcome measure. A conceptual model describes symptoms and impacts that are important to patients. A conceptual framework depicts the relationships between the concepts measured via the DHTT and associated items and domains. Additional qualitative research seeks to establish the impacts relevant to patients and their relationships with the DHTT features. Clinical expert opinions are required to ensure that the features represented in the final conceptual framework are neurobiologically relevant. The ensuing stages of outcome development require two independent datasets: the build dataset is used to select statistically robust and appropriate features, build a scoring algorithm, confirm psychometric properties of the scoring algorithm, and provide a preliminary validation; the independent validate dataset confirms the robustness of the psychometric properties, provides further validation data related to the DHTT outcome measure, and generates minimally clinically important difference (MCID) estimates and responder definitions.CC BY · PubMed Central
Fig. 4 Path towards identifying the sensor features of an existing DHTT which correspond best to patient relevant symptoms and impacts. Traditional semi-structured interviews (i.e., concept elicitation (CE) involving the identification of activities of daily living (ADLs) that patients feel are impacted by their disease), followed by the identification of the disease symptoms related to each impaired ADL. In a third step, patients interact with the DHTT and complete a given task (e.g., draw a shape) and are asked to identify which aspects of the DHTT task present challenges, and finally, the relationships between the symptoms and challenges presented by the DHTT task are identified. The last two steps may include a cognitive debriefing (CD) interview to determine the patients’ understanding of the task instructions and further establish the relevance of the DHTT assessment. All other potential ADLs related to patients’ functions and DHTT sensor features should be discussed, represented by a return to the beginning of the interview procedure. The output of these interviews may lead to a refinement of the conceptual framework and associated DHTT if necessary, which should then be reviewed by experts (denoted by asterisk) to ensure a robust neurobiological mapping between symptoms and impacts and associated sensor features.
Fig. 4 Path towards identifying the sensor features of an existing DHTT which correspond best to patient relevant symptoms and impacts. Traditional semi-structured interviews (i.e., concept elicitation (CE) involving the identification of activities of daily living (ADLs) that patients feel are impacted by their disease), followed by the identification of the disease symptoms related to each impaired ADL. In a third step, patients interact with the DHTT and complete a given task (e.g., draw a shape) and are asked to identify which aspects of the DHTT task present challenges, and finally, the relationships between the symptoms and challenges presented by the DHTT task are identified. The last two steps may include a cognitive debriefing (CD) interview to determine the patients’ understanding of the task instructions and further establish the relevance of the DHTT assessment. All other potential ADLs related to patients’ functions and DHTT sensor features should be discussed, represented by a return to the beginning of the interview procedure. The output of these interviews may lead to a refinement of the conceptual framework and associated DHTT if necessary, which should then be reviewed by experts (denoted by asterisk) to ensure a robust neurobiological mapping between symptoms and impacts and associated sensor features.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 34 references

Research works printed but not linked: 2

  1. Yang, J. & Honavar, V. In: Liu, H., Motoda, H. (eds) Feature Extraction, Construction and Selection 453 (Springer, 1998).
  2. Food and Drug Administration. Patient-Focused Drug Development Guidance Public Workshop: Incorporating Clinical Outcome Assessments into Endpoints for Regulatory Decision-Making, Silver Spring, MD, 6 December (FDA, 2019). https://globalforum.diaglobal.org/issue/june-2019/regulatory-perspective-digital-health-technology-tools/.

Cited by

See all 85

Metadata

Authors
Taylor, K. I., Staunton, H., Lipsmeier, F., Nobbs, D., & Lindemann, M.
Affiliations
  • F. Hoffmann-La Roche AG (CH)
  • Roche (United Kingdom) (GB)
Corresponding author
Taylor, Kirsten I.
Venue
npj Digital Medicine
Published
2020-07-23
Volume
3
Issue
1
Pages
97-97
ISSN
2398-6352 (ISSN-L)
OpenAlex
W3044978152
Cited by
85
References
34
Citations by year
2026: 9 · 2025: 16 · 2024: 16 · 2023: 12 · 2022: 20 · 2021: 10 · 2020: 1
FWCI
11.21
Type
journal article
Language
en
Open access
Licence
Locations
  • npj Digital Medicine (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • F. Hoffmann-La Roche
  • Prothena
Topics
  • Health Sciences › Health Professions › General Health Professions › Mobile Health and mHealth Applications
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Mental Health Research Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
Keywords
digital health technologies, sensor data, outcome measures, passive monitoring, wearable devices, smartphone sensors, clinical outcome assessment, clinical trials, phenotypic characterization

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Outcome measures based on digital health technology sensor data: data- and patient-centric approaches. https://app.cochat.ai/papers/10.1038/s41746-020-0305-8
The paper
Taylor, K. I., Staunton, H., Lipsmeier, F., Nobbs, D., & Lindemann, M. (2020). Outcome measures based on digital health technology sensor data: data- and patient-centric approaches. npj Digital Medicine. https://doi.org/10.1038/s41746-020-0305-8

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.