Methods: Clinic and provincial health records of adults who received HIV care through a sole regional provider in southern Alberta, Canada, from August 2009 to 2014 were retrospectively reviewed. Statistical comparison of demographic and HIV specific variables was done between those LTFU and engaged in care. Lost to follow-up was defined as more than a year since last assessment, having had two previous visits and care related blood work within the preceding year and without documentation of transferring care or death.
Results: 1928 individuals received care during the study period of which 176 (9%) were LTFU, 15 (9%) of whom subsequently returned to care. Age, ethnicity, and city of residence were significantly different between groups. Those LTFU were younger (43.9 +/- 10.1 yr vs. 46.5 +/- 11.2 yr, P=0.003), more likely to be Native Canadian (21 of 176 (12%) vs. 144 of 1735 (8%), P=0.002) and more likely to reside outside of the city where the centralized HIV clinic was located (88 of 154 (57%) vs. 422 of 1746 (24%), P=0.009). Comparing HIV specific variables, those LTFU had lower absolute CD4 count (448.4 +/- 249.4 cells x 106/l vs. 527 +/- 285.8 cells x 106/l, P=0.005) and were less likely to have had HIV viral suppression (148 (84%) vs. 1583 (90%), P<0.001) on last available testing. Of those LTFU who returned to care, previous duration of engagement was longer for those returning to care than those that did not return (10.6 +/- 6.9 yr vs. 6.7 +/- 5.9 yr, P=0.006). Route of infection, relationship status, employment, time since diagnosis and location of diagnosis were not significantly different between those LTFU and retained in care (P>0.05).
Conclusion: Novel targetted interventions are needed to optimize retention in HIV care. Specific age, ethnic and geographic groups appear to be at greater risk of disengagement. Amongst our cohort, greater previous engagement in care predicted return to care reinforcing the importance of retention interventions.
M. J. Gill, None