International Classification of Functioning, Disability and Health (ICF)

Validation of the ICF Comprehensive Set for Patients with Low Back
Pain: Background: The International Classification of Functioning,
Disability and Health (ICF) is a multipurpose classification to
describe functional states associated with health conditions. To
ensure practicability the ICF Checklist was developed, a short form
of the ICF which only contains the most important categories
irrespective of the present diagnoses. Furthermore ICF
Comprehensive Sets were developed which contain the most important
categories concerning a specific disease. Objectives: The general
objective is to examine the explanatory power of the ICF Checklist
in order to explain the PHI-score and the MHI-score of the SF-36.
The specific aims are 1) to explore the percentage of variance of
the SF-36 parameters accounted for by the ICF categories, 2) to
identify the ICF categories which explain most of the variance of
the two SF-36 parameters, 3) to assess the importance of the four
components of the ICF Checklist for the SF-36 parameters. Methods:
Cross sectional analysis of n=200 inpatients of rehabilitation
centres suffering from low back pain. The International
Classification of Functioning, Disability and Health (ICF) belongs
to the WHO family of international classifications. At present in
the ICF the following components are included: 1) Body Functions 2)
Body Structures 3) Activities and Participations 4) Environmental
Factors. Patients’ health status was assessed by the SF-36 Health
Survey, a generic instrument to measure health status. Analyses
were focused on the two summary measures Physical Health Index
Score (PHI-score) and Mental Health Index Score (MHI-score).
Statistical Analysis was conducted in four steps: In step 1 a first
selection of potential predictor variables of health status was
performed by the use of descriptive statistics. Analysis of
regression in step 2 was conducted for each component of the ICF.
In step 3 the variables selected in the four analyses of regression
in step 2 were integrated into one multiple linear regression
model. In the fourth step the model constructed in step 3 was
verified and optimized. Finally three control variables were
included into the model (gender, age and number of concomitant
diseases). Results: The first model accounts for 44.6% of the
variance of the PHI-score with F= 16.36 (p