Robertson et al developed a score for identifying an HIV patient who is likely to experience the immune reconstitution syndrome (IRS) after receiving highly-active antiretroviral therapy (HAART). The authors are from the University of Cincinnati.
Analysis: classification and regression tree (CART)
Parameters:
(1) number of opportunistic infections
(2) CD8 positive T-cell count
(3) hemoglobin in g/dL
|
Number of Opportunistic Infections |
CD8 positive T cell count |
Hemoglobin |
Score |
|
<= 1 |
NA |
NA |
0 |
|
> 1 |
> 468.5 per µL |
> 10.2 g/dL |
1 |
|
> 1 |
> 468.5 per µL |
<= 10.2 g/dL |
2 |
|
> 1 |
<= 468.5 per µL |
> 13.6 g/dL |
3 |
|
> 1 |
<= 468.5 per µL |
<= 13.6 g/dL |
4 |
|
Score |
Percent with IRS |
|
0 |
4% |
|
1 |
0% |
|
2 |
66% |
|
3 |
0% (0 of 2) |
|
4 |
69% |
Interpretation:
• minimum score: 0
• maximum score: 4
• The higher the score the more likely the patient was to experience IRS.
• The Youden index indicates using a cutoff >= 2 to identify a person at risk for the IRS.
Performance:
• A cutoff of = 4 had a sensitivity of 79% and specificity 88%.
• A cutoff of >=2 had a sensitivity of 93% and specificity of 80%.
Purpose: To identify a patient at risk for the Immune Reconstitution Syndrome (IRS) based on the score of Robertson et al.
Specialty: Infectious Diseases
Objective: risk factors, selection
ICD-10: B24,
