Pin142 Impact of Adherence on Viral Suppression and Determinants of Adherence in Patients With Hiv

Abstract

A systematic search was conducted using PubMed, Embase, and Cochrane for real-world studies (January 2014 to February 2019) following PRISMA guidelines. Cross-references were checked to identify studies post 2009, with ongoing status. Results: Of 2,838 screened articles, 42 met the inclusion criteria. Observed adherence ranged between 69%-98% in the US, 57%-96% in European, 62%-93% in African, 79%-80% in Latin and 66%-98% in Asian population. Most of the studies indicated a positive correlation between adherence and VS, consistent across all geographies, irrespective of adherence-thresholds (90-95%) and treatment class. Across all ARTs, the odds of detectable viral load (.400 copies/ml) for different adherence levels range as follows: 95-99%: 1.5 (1.2-1.9); 90-94%: 2.0 (1.5-2.8); ,90%: 4.1 (3.1-5.4). The impact of adherence also varies as per the treatment class. For unboosted-PIs, patients with adherence $95% were 1.6 times and 7.7 times more likely to achieve viral-suppression than those with adherence between 75%-95% and ,75% respectively. For INSTIs, patients with adherence $95% were 1.1 times and 4 times more likely to achieve viral-suppression than those with adherence between 75%-95% and ,75% respectively. Main determinants of non-adherence were alcohol/ substance use (p,0.005) apart from gender, age, comorbidities and polypharmacy. Naïve patients seemed to be more adherent than the experienced patients. Conclusions: Adherence levels of 90-95% seem to maximise the possibility of achieving sustained viral suppression with clinically significant improvements. Interventions to improve adherence in HIV patients with alcohol/substance use are deemed desirable.