icon-folder.gif   Conference Reports for NATAP  
 
  ID Week
Oct 8-12 2014
Philadelphia
Back grey_arrow_rt.gif
 
 
 
Risk Score and Race Predict Missed HIV Appointments in US Group
 
 
  IDWeek 2014, October 8-12, 2014, Philadelphia
 
Mark Mascolini
 
Black race and appointment-keeping risk determined by a published virologic failure risk tool were linked to higher odds of missing the next HIV appointment in a 510-person observational study at the Vanderbilt Comprehensive Care Clinic in Nashville, Tennessee [1].
 
Current US antiretroviral guidelines encourage HIV clinicians to assess retention in care and to take steps to increase retention. But assessing and improving retention can fall to the bottom of the priority list in a busy HIV practice. Vanderbilt researchers conducted this study to see if a tool developed to predict virologic failure [2] would also predict poor appointment attendance in their HIV clinic.
 
The study involved all adults scheduled for a routine HIV appointment from August 2013 through March 2014. All had a viral load above 200 copies, regardless of whether they were taking antiretroviral therapy. The Vanderbilt team adapted the virologic failure tool to include seven predictors: suboptimal adherence, CD4 count below 100, substance abuse, high antiretroviral experience, missing two or more appointments in past 12 months, prior virologic failure, and most recent viral load above 200 copies [2].
 
For this study, the investigators determined each person's RPT score before their HIV appointment and classified risk into three groups: low risk (RPT 0 to 1), medium risk (RPT 2 to 3), high risk (RPT 4 or higher). The primary outcome was completing the next HIV appointment. The researchers used logistic regression adjusted for age, race, gender, HIV risk factor, year of entry to care, baseline CD4 count, and baseline viral load to determine adjusted odds ratios for missing the next appointment.
 
The study involved 510 people, 74% of them men and 55% non-Hispanic black. There were 193 people who did not complete their next appointment. Completers and noncompleters did not differ significantly in median age (38 and 40), CD4 count (355 and 291, P = 0.12), or viral load (21,192 and 25,426 copies). But noncompleters included a higher proportion of non-Hispanic blacks (69% versus 47%, P < 0.001), a lower proportion of men (68% versus 77%, P = 0.03), a lower proportion of heterosexuals (46% versus 57%, P = 0.01), a higher proportion of drug injectors (10% versus 5%, P = 0.03), and an earlier year of entering care (2009 versus 2012, P < 0.001).
 
The risk tool classified 126 people (25%) as low risk, 244 (48%) as medium risk, and 140 (27%) as high risk. Proportions of blacks rose from the low- to medium- to high-risk groups (46%, 58%, 59%), while proportions of men fell (83%, 74%, 66%). Median age rose across the three risk groups (32, 38, 44).
 
Logistic regression analysis identified three independent predictors of missing the next HIV visit at the following adjusted odds ratios (aOR) (and 95% confidence intervals):
 
-- Black race: aOR 2.45, 1.62 to 3.72, P < 0.001
-- Medium vs low risk score: aOR 3.94, 2.14 to 7.25, P < 0.001
-- High vs low risk score: aOR 9.51, 4.56 to 19.83, P < 0.001
 
The Vanderbilt team cautioned that results of their study are specific to this patient population, though demographics of this group are similar to those in many US clinics. They noted that the link between RPT score and appointment keeping must be validated before the tool can be used in other settings. And they suggested that clinics without electronic medical records may not be able to use this tool.
 
With these limitations in mind, the researchers proposed that RPT scoring may allow HIV clinics to stratify large numbers of patients for appointment-keeping risk. Such stratification could help HIV practices "target resources to improve appointment attendance to those groups of patients most at risk."
 
References
 
1. Woodward B, Person A, Kheshti A, Raffanti S, Pettit A. Prediction tool for missed healthcare provider visits among HIV-infected persons. IDWeek 2014. October 8-12, 2014, Philadelphia. Abstract 86.
 
2. Robbins GK, Johnson KL, Chang Y, et al. Predicting virologic failure in an HIV clinic. Clin Infect Dis. 2010;50:779-786. http://cid.oxfordjournals.org/content/50/5/779.long