MultiCenter Interrupted Time Series Analysis: Incorporating Within and Between-Center Heterogeneity
Bibliographic record
Abstract
Joycelyne E Ewusie,1,2 Lehana Thabane,1 Joseph Beyene,1 Sharon E Straus,3 Jemila S Hamid1,2,4 1Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, Ontario, Canada; 2School of Epidemiology and Public Health, University of Ottawa, Ottawa, Canada; 3Li Ka Shing Knowledge Institute of St Michael’s Hospital, Toronto, Ontario, Canada; 4Children’s Hospital of Eastern Ontario, Ottawa, Ontario, CanadaCorrespondence: Jemila S HamidChildren’s Hospital of Eastern Ontario, Ottawa, Ontario, CanadaTel +1 613 737-7600 x 4194Email jhamid@uottawa.caBackground: Segmented regression (SR) is the most common statistical method used in the analysis of interrupted time series (ITS) data. However, this modeling strategy is indicated to produce spurious results when applied to aggregated data. For multicenter ITS studies, data at a given time point are often aggregated across different participants and settings; thus, conventional segmented regression analysis may not be an optimal approach. Our objective is to provide a robust method for analysis of ITS data, while accounting for two sources of heterogeneity, between participants and across sites.Methods: We present a methodological framework within the segmented regression modeling strategy, where we introduced weights to account for between-participant variation and the differences across multiple sites. We empirically compared the proposed weighted segmented regression (wSR) with the conventional SR as well as with a previously published pooled analysis method using data from the Mobility of Vulnerable Elders in Ontario (MOVE-ON) project, a multisite ITS study.Results: Overall, the wSR produced the most precise estimates, where they had the narrowest 95% CI, while the conventional SR method resulted in the least precise estimates. Our method also resulted in increased power. The pooled analysis method and the wSR had comparable results when there were ≤ 4 sites included in the overall analysis and when there was moderate to high between-site heterogeneity as measured by the I2 statistic.Conclusion: Incorporating participant-level and site-level variability led to estimates that were more precise and accurate in determining the magnitude of the effect of an intervention and led to increased statistical power. This underscores the importance of accounting for the inherent variability in aggregated data. Extensive simulations are required to further assess the methods in a wide range of scenarios and outcome types.Keywords: aggregated data, weighted segmented regression, pooled analysis, interrupted time series, multisite studies
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.147 | 0.298 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.005 |
| Bibliometrics | 0.003 | 0.005 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.005 | 0.003 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.004 | 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 source (direct Gemma or distilled Codex), 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".