Open AccessDecision thresholds and minimal important difference estimates for evidence-based practice and policy (Part 2)
Author(s) -
Beth McDougall,
Mike Reid,
Souvik Mitra,
Bree Johnston
Publication year - 2021
Publication title -
dalhousie medical journal
Language(s) - English
Resource type - Journals
eISSN - 1492-000X
pISSN - 1488-9994
DOI - 10.15273/dmj.vol47no1.10719
Subject(s) - biostatistics , core (optical fiber) , clinical practice , outcome (game theory) , clinical decision support system , evidence based medicine , psychology , actuarial science , computer science , management science , medline , medicine , epidemiology , decision support system , family medicine , data mining , political science , economics , telecommunications , mathematical economics , law
Understanding core concepts in epidemiology and biostatistics is crucial for evidence-based clinical practice and policy. In this second installment of our two-part series on threshold concepts, we transition from understanding the ubiquitous p-value to tools and measures for decision making among clinicians-in-training, highlighting the growing importance of utilizing explicit and evidence-based approaches to make appropriate and efficient decisions. We review two related decision-making concepts: (1) Minimal Important Difference (MID) estimates and (2) Decision Thresholds, focusing specifically on patient-reported outcome measures (PROMs). These terms and many other related expressions are used regularly, and often interchangeably, but what are they? Why are they valuable? And how can they be used to support evidence-based decision-making in clinical contexts and develop strong clinical practice guidelines? We conclude our brief review on the utility of these measures with a spotlight on a local example of how the theory underlying MID estimates and decision thresholds is currently being embedded in electronic platforms in primary care contexts targeting depression in Nova Scotia.