Making health technology assessment more dynamic – Temporal trend analysis to capture performance trajectories
Health technology assessment (HTA) has an increasing role in evaluating not only drugs, but also medical devices. Assessing medical devices is more challenging as outcomes tend to improve substantially over time. This paper analyzes clinical outcomes over time of insulin pump therapy in adult type 1...
Gespeichert in:
Veröffentlicht in: | Health policy and technology 2017-09, Vol.6 (3), p.328-338 |
---|---|
Hauptverfasser: | , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Health technology assessment (HTA) has an increasing role in evaluating not only drugs, but also medical devices. Assessing medical devices is more challenging as outcomes tend to improve substantially over time. This paper analyzes clinical outcomes over time of insulin pump therapy in adult type 1 diabetes. Meta-regression was used to capture outcome trends while considering innovation theory.
A systematic review of 4.297 studies was conducted covering a 35 year time period. To meet the need for a more dynamic evaluation method, temporal trend analysis was applied based on meta-regression of three extracted outcome parameters: HbA1c, incidence of severe hypoglycemia and ketoacidosis.
The systematic review included 150 studies. Performance improvement in HbA1c and ketoacidosis was rapid during the 1980 and 90s. Thereafter the improvement trajectory of insulin pumps measured by HbA1c, ketoacidosis or severe hypoglycemia was essentially flat. Meta-regression of several covariates was performed showing publication year to be statistically significant. Retrospective recruitment and the percentage of female patients were also statistically significant. A technology cycle model analysis revealed convergence to a dominant design by the end 1990s, followed by slower progress in outcomes.
Insulin pump technology currently does not offer an improving performance trajectory according to key indicators HbA1c, incidence of severe hypoglycemia and ketoacidosis, but compares well to manual insulin injections in terms of quality of life. Applying temporal trend analysis is especially valuable in an early technology cycle stage when uncertainty is high, and when predicted improvements in future performance can influence the choice of technology.
•Capturing the moving target of medical devices outcomes by trend analysis.•Combined meta-regression and technology cycle analysis allow trend prediction.•Insulin pump therapy outcomes progress fast before dominant design then slowly. |
---|---|
ISSN: | 2211-8837 2211-8845 |
DOI: | 10.1016/j.hlpt.2017.04.005 |