While reported adherence rates vary dramatically from study to study, there is a consistent pattern of non-adherence that warrants further exploration and development of more effective interventions. Given the need for nearly perfect levels of adherence for successful suppression of viral load, research has discovered alarming rates of non-adherence. In a study conducted by Gordillo et al (1999), overall rates of adherence among 366 people was 57.6% when adherence was defined as taking 90% or more of medications. Bakken et al (2000) found that only 33% of 39 people were adherent over a week when adherence was defined as taking 95% or more of medications as prescribed. Among 445 respondents in a study conducted by Spire et al (2002), 73.3% considered themselves "totally" adherent. Of 39 participants in a study conducted by Murphy et al (2000), only 33% reported being adherent over a week. In addition, Williams et al (2006) found that out of 2088 children and adolescents with HIV, 84% reported complete adherence to medication using a validated three-day recall tool. Although evidence from research recognizes a range of reported adherence rates, it is evident that adherence rates have much room for improvement.