Table 3. Comparison of the mean scores used in the mean scores used in the Logistic Regression and Decision Tree Models between the validation and development cohorts.
| PTLD symptoms | Recovered | |||||
|---|---|---|---|---|---|---|
| Development | Validation | p-value | Development | Validation | p-value | |
| SF-36 MHC score (SD) | 47.71 (-10.11) |
42.19 (-13.24) |
0.28 | 58.84 (-2.16) |
53.08 (-6.42) |
0.008 |
| SF-36 PF subscale score (SD) | 47.59 (-8) |
45.87 (-11.32) |
0.68 | 56.31 (-2.45) |
56.07 (-1.94) |
0.78 |
| Neuro-QoL Fatigue t-score (SD) | 51.1 (-6.53) |
51.18 (-8.48) |
0.98 | 37.19 (-5.72) |
39.09 (-7.28) |
0.46 |
A higher score in the Neuro-QoL Fatigue represents more symptoms and a worse score of the concept being measured (e.g., more fatigue). For both the Physical Functioning (PF) subscale and Mental Health Component (MHC) scores, a lower score represents worse symptoms. Fisher’s exact test was used to compare proportions between groups. PTLDs: post-treatment Lyme disease. SF-36: Short Form-36 version 2.