 |
|
|
| |
Less than 95% adherence to ritonavir-boosted protease inhibitors leads to viral suppression: an update to address currently prescribed antiretroviral regimens
|
| |
| |
DR Bangsberg, Judy Hahn, Sheri Weiser, K Ragland, R Clark and E Riley
Epidemiology and Prevention Interventions Center, Division of Infectious Diseases, and Positive Health Program
at San Francisco General Hospital, AIDS Research Institute at University of California, San Francisco
Study Objective
Compare viral suppression within adherence strata
--Unboosted PI
--NNRTI
--Ritonavir-boosted PI
Compare adherence level
--Unboosted PI
--NNRTI
--Boosted PI
Lopinavir-R co-formulation
Atazanavir/Ritonavir
Limitations
- Regimens not randomly assigned
- Uncontrolled confounders
- Limited power to detect differences
AUTHOR CONCLUSIONS
While adherence remains a strong predictor of viral suppression, most individuals with moderate levels of adherence have viral suppression with currently used regimens
Levels of adherence are comparable for all three regimens, despite differences in pill burden, dosing frequency, and side effect profile (see graphs below)
Background
- Unboosted PIs in treatment experienced patients require high levels of adherence (95%) (Patterson Ann Int Med 2001)
- NNRTIs viral suppression at moderate (70-90%) adherence (Bangsberg CID 2006)
- Better suppression by adherence strata with boosted PIs than unboosted PI (King JID 2005, Gross IAS 2006)
- Pill burden and side effects have been associated with adherence (Carrieri JAIDS 2006, Stone JAIDS 2001, Ammasarri JAIDS 2001, Portsmouth HIV Med 2005)
- No cross comparison of viral suppression by adherence for NNRTI, boosted PI and PI using detailed objective adherence measures
Methods
- Participants: REACH Cohort
- Eligibility: HAART >3months
- Adherence: unannounced pill counts at participants usual place of residence
- Outcomes
--Viral suppression <50 copies
--Adherence
- Analyses
--Viral suppression by regimen and adherence
- Chi square
- Logistic regression
--Adherence
- Savage score (non-parametric comparison continuous variable across multiple categories)
|
| |
|
|
|
|
|
