When it comes to predicting outcomes, one of the most important questions you can ask is, "how much variance is explained?" The savvy person focuses on this question because the accuracy of the prediction depends on the response to this question. If a prediction is used for future decisions, like quality payment programs, then you want the most accurate prediction possible.
To have 100% variance explained is quite improbable. Teams creating prediction models are focused on as high amount of variance explained as possible.
How much information should patients be providing to help explain variance?
I recently read a study focused on factors that affect functional ability. I was impressed that the investigators provided a high explanation of variance in the abstract. At the same time, I believe this is one of those studies that requires full text access to really take a look at the methods and tables to critically think.
One of the questionnaires the investigators included was the revised illness perception questionnaire. This questionnaire was originally created in 1996. It was revised in 2001. This questionnaire is really, really long. The topics covered with about about 6 items required to respond include: timeline acute/chronic, timeline cyclical, consequences, personal control, treatment control, illness coherence, emotional representations, psychological attributions, risk factors, immunity, and accident or chance.
In order to be clinically helpful, we need to learn which areas had the largest impact. There has to be balance between the ability to explain variance, time burden along with how the the information gained will be used to impact and improve care.
Below you will find a quick view of the abstract.
Illness perceptions explain the variance in functional disability, but not habitual physical activity, in patients with chronic low back pain: A cross-sectional study.
Although the importance of psychosocial factors has been highlighted in many studies in patients with chronic low back pain (CLBP), there is a lack of research examining the role of illness perceptions in explaining functional disability and physical activity in patients with CLBP.
The aim of the study was to explore the value of illness perceptions in explaining functional disability and physical activity in patients with CLBP.
Eighty-four participants with CLBP (> 3 months) completed a battery of questionnaires investigating psychosocial factors (pain catastrophizing scale (PCS), illness perceptions questionnaire revised (IPQ-R) and SF-36 mental health scale (SF-36_MH)) as well as the perceived pain intensity (VAS), Oswestry Disability Index (ODI) and Baecke questionnaire. The latter two were entered separately as dependent variables in a regression analysis.
The combined variables (VAS, PCS, SF-36_MH, IPQ-R) accounted for 62% of the variance in functional disability (ODI). The IPQ-R significantly increased the explained variance of ODI scores in CLBP patients, on top of the other 3 variables (VAS, PCS, SF-36_MH) with 18% (p<.01). Only 5% of the variance in the Baecke questionnaire was explained by the 4 combined variables. None of the single variables alone made a significant contribution to R².
Illness perceptions are an important factor for explaining functional disability, but not explaining habitual physical activity in CLBP patients.
Pain Pract.2017 Sep 15. doi: 10.1111/papr.12642. [Epub ahead of print]