Collecting Real-World Data via an In-Home Smart Medication Dispenser: Longitudinal Observational Study of Survey Panel Persistency, Response Rates, and Psychometric Properties
Bibliographic record
Abstract
Background: A smart medication dispenser called "spencer" is a novel generator of longitudinal survey data. The patients dispensing medication act as a survey panel and respond to questions about quality of life and patient-reported outcomes. objectives: Our goal was to evaluate panel persistency, survey response rates, reliability, and validity of surveys administered via spencer to 4138 polychronic patients residing in the United States and Canada. Methods: Patients in a Canadian health care provider's program were included if they were dispensing via spencer in the June 2021 to February 2024 time frame and consented to have their data used for research. Panel persistency was estimated via discrete survival methods for 2 years and survey response rates were computed for 1 year. Patients were grouped by mean response rates in the 12th month (<90% vs ≥90%) to observe differential response rate trends. For reliability and validity, we used a spencer question about recent falls with ternary responses value-coded -1, 0, and 1. For reliability, we computed Pearson correlation between mean scores over 2 years of survey responses, and transitions between mean score intervals of [0, 0.5), [-0.5, 0.5), and [0.5, 1]. For validity, we measured the association between the falls question and known factors influencing fall risk: age, biological sex, quality of life, physical and emotional health, and use of selective serotonin reuptake inhibitors or serotonin-norepinephrine reuptake inhibitors, using repeated-measures regression for covariates and Kendall τ for concomitant spencer questions. Unlabelled: From 4138 patients, dispenser persistency was 68.3% (95% CI 66.8%-69.8%) at 1 year and 51% (95% CI 49%-53%) at 2 years. Within the cohort observed beyond 1 year, 82.3% (1508/1832) kept surveys enabled through the 12th month with a mean response rate of 84.1% (SD 26.4%). The large SD was apparent in the subgroup analysis, where a responder versus nonresponder dichotomy was observed. For 234 patients with ≥5 fall risk responses in each of the first 2 years, the Pearson correlation estimate between yearly mean scores was 0.723 (95% CI 0.630-0.798). For mean score intervals [0, 0.5), [-0.5, 0.5), and [0.5, 1], self-transitions were the most common, with 59.8% (140/234) of patients starting and staying in [0.5, 1]. Fall risk responses were not significantly associated with sex (P=.66) or age (P=.76) but significantly related to selective serotonin reuptake inhibitor or serotonin-norepinephrine reuptake inhibitor usage, quality of life, depressive symptoms, physical health, disability, and trips to the emergency room (P<.001). Conclusions: A smart medication dispenser, spencer, generated years of longitudinal survey data from patients in their homes. Panel attrition was low, and patients continued to respond at high rates. A fall risk measure derived from the survey data showed evidence of reliability and validity. An alternative to web-based panels, spencer is a promising tool for generating patient real-world data.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.041 | 0.007 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".