Background: Long-term retention is a crucial component of HIV care because treatment
success can only be measured among retained patients. Understanding determinants of retention will
inform retention strategies. We evaluated the correlates of retention in a large HIV program in Nigeria.
Methods: We reviewed quality of care data for 5320 randomly selected HIV-positive adults aged ≥15
years enrolled in 37 treatment facilities in Nigeria between 2005 and 2009. Retention was described as
having one or more clinic visits in the one year (2010) review period. Patient-related correlates of
retention were determined using logistic regression.
Results: 144 patients exited the program through deaths or transferrals. Of the 5176 with no documented exits, 3231
(62.4%) were retained (65.6% female; median age: 35.6 years). 2938 (75.8%) patients on ART, and 286 (23.4%) pre-ART
patients were retained. Being on ART (OR=10.3, p<0.001), Age 30-60years (30 – 45 years: OR=1.36, p<0.001 and >45 –
60 years: OR=1.47, p<0.001) compared to patients <30 years; Female gender (OR=1.18, p=0.006), baseline CD4 cell
count (100-350 cells/mm3: OR=1.24, p=0.006) vs <100cells/mm3 and lower WHO stage at baseline (WHO Stage IV, III,
II: OR=0.50,0.51,0.77 respectively) vs Stage I were associated with retention. Among patients on ART, recent ART
initiation 2008-09 (OR=1.73, p<0.001) vs 2005-07, being on ART for >6months (p<0.001) vs <6month and initiating
ART on non-Stavudine based regimen (p<0.001) were also associated with retention.
Conclusion: 3 out of 4 pre-ART patients and 1-in-4 ART patients were not retained in 37HIV treatment facilities in
Nigeria. These findings provide insight that enables HIV programs integrate retention strategies at all stages of the HIV