TITLE:
Investigating the minimally important difference of the Diabetes Health Profile (DHP-18) and the EQ-5D and SF-6D in a UK diabetes mellitus population
AUTHORS:
Brendan Mulhern, Keith Meadows
KEYWORDS:
Minimally Important Difference; DHP-18; EQ-5D; SF-6D; Diabetes Mellitus; PROMS; Psychometrics
JOURNAL NAME:
Health,
Vol.5 No.6,
June
18,
2013
ABSTRACT:
Objectives: It is important to know what
patient reported outcome measure (PROM) scores relate to a meaningful change
in health status across time. The aim of this study was to investigate the
minimally important difference (MID) of the Diabetes Health Profile (DHP-18),
EQ-5D and SF-6D in a Type 1 and Type 2 diabetes patient sample. Methods: A longitudinal dataset
including a UK community sample of people with Type 1 and Type 2 diabetes was
used for the analysis. A combination of anchor and distribution methods was
used to investigate the MID. For the anchor based method, a global health
change indicator was used if it correlated with the PROM scores at baseline and
follow up. To calculate the anchor based MID, the change in PROM score for
those reporting no change on the anchor was subtracted from those reporting
small change. For the distribution based estimation, the 1 Standard Error of
Measurement, 0.5 and 0.33 standard deviation methods were used. Results: The anchor was not correlated
with the DHP-18 dimensions so was only used to estimate MID values for the
EQ-5D and SF-6D. For the DHP-18, MID estimates for the Psychological Distress
domain range from 6.99 to 10.59, the Barriers to Activity domain range from
6.48 to 9.89, and the Disinhibited Eating domain range from 7.52 to 11.39. The
EQ-5D estimations range from 0.058 to 0.158, and the SF-6D estimations range
from 0.038 to 0.081. The 0.5 SD and 1SEM estimations are of a similar magnitude
across the three measures. Conclusions:This study
has derived a range of values for each measure that may correspond to an important change in health status. The MID
values may guide researchers who are using the measures as part of their
assessment of both Type 1 and Type 2 patients with diabetes mellitus.