Skip to main content

Table 6 An in-depth analysis of the covariates' (likelihood) ratio profiles (Figure 1) suggests a way to redefine the ADLCAP.

From: Using machine learning algorithms to guide rehabilitation planning for home care clients

initialize score = 0; threshold = 15.8

if (c3 = 0)

then score = score + 1.5;

if (h2a = 0)

then score = score + 1.1;

if (h2b = 0)

then score = score + 1.1;

if (h2c = 0)

then score = score + 1.1;

if (h2d = 0)

then score = score + 1.2;

if (h2e = 0)

then score = score + 1.3;

if (h2f = 0)

then score = score + 1.3;

if (h2g = 0)

then score = score + 1.3;

if (h2h = 0)

then score = score + 1.3;

if (h2i = 0)

then score = score + 1.4;

if (h2j = 0)

then score = score + 3.1;

if (h3 = 1)

then score = score + 1.7;

if (h7a = 1)

then score = score + 3.0;

if (h7b = 1)

then score = score + 2.9;

if (h7c = 1)

then score = score + 4.1;

if (k8b = 0)

then score = score + 1.3;

if (k8c = 1)

then score = score + 1.1;

if (k8d = 1)

then score = score + 2.0;

if (p6 = 1 or 2)

then score = score + 1.9;

  1. if score > threshold;
  2. then (alternative ADLCAP) = 1;
  3. else (alternative ADLCAP) = 0.