Deborah Horowitz

Columbia University, New York, New York, United States

Are you Deborah Horowitz?

Claim your profile

Publications (6)19.55 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Despite the extraordinary scale up of HIV prevention, care and treatment services in sub-Saharan Africa (SSA) over the past decade, the overall effectiveness of HIV programs has been significantly hindered by high levels of attrition across the HIV care continuum. Data from ¿real-life¿ settings are needed on the effectiveness of an easy to deliver package of services that can improve overall performance of the HIV care continuum.Methods/DesignWe are conducting an implementation science study using a two-arm cluster site-randomized design to determine the effectiveness of a combination intervention strategy (CIS) using feasible, evidence-based, and practical interventions¿including (1) point-of-care (POC) CD4 count testing, (2) accelerated antiretroviral therapy initiation for eligible individuals, and (3) SMS reminders for linkage to and retention in care¿as compared to the standard of care (SOC) in Mozambique in improving linkage and retention among adults following HIV diagnosis. A pre-post intervention two-sample design is nested within the CIS arm to assess the incremental effectiveness of the CIS plus financial incentives (CIS¿+¿FI) compared to the CIS without FI on study outcomes. Randomization is done at the level of the study site, defined as a primary health facility. Five sites are included from the City of Maputo and five from Inhambane Province. Target enrollment is a total of 2,250 adults: 750 in the SOC arm, 750 in the CIS cohort of the intervention arm and 750 in the CIS¿+¿FI cohort of the intervention arm (average of 150 participants per site). Participants are followed for 12 months from time of HIV testing to ascertain a combined endpoint of linkage to care within 1 month after testing and retention in care 12 months from HIV test. Cost effectiveness analyses of CIS compared to SOC and CIS¿+¿FI compared to CIS will also be conducted.DiscussionStudy findings will provide evidence on the effectiveness of a CIS and the incremental effectiveness of a CIS¿+¿FI in a ¿real-life¿ service delivery system in a SSA country severely impacted by HIV.Trial, NCT01930084.
    Full-text · Article · Oct 2014 · BMC Infectious Diseases
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Generalizable data are needed on the magnitude and determinants of adherence and virological suppression among patients on antiretroviral therapy (ART) in Africa. We conducted a cross-sectional survey with chart abstraction, patient interviews and site assessments in a nationally representative sample of adults on ART for 6, 12 and 18 months at 20 sites in Rwanda. Adherence was assessed using 3- and 30-day patient recall. A systematically selected sub-sample had viral load (VL) measurements. Multivariable logistic regression examined predictors of non-perfect (<100%) 30-day adherence and detectable VL (>40 copies/ml). Overall, 1,417 adults were interviewed and 837 had VL measures. Ninety-four percent and 78% reported perfect adherence for the last 3 and 30 days, respectively. Eighty-three percent had undetectable VL. In adjusted models, characteristics independently associated with higher odds of non-perfect 30-day adherence were: being on ART for 18 months (vs. 6 months); younger age; reporting severe (vs. no or few) side effects in the prior 30 days; having no documentation of CD4 cell count at ART initiation (vs. having a CD4 cell count of <200 cells/µL); alcohol use; and attending sites which initiated ART services in 2003-2004 and 2005 (vs. 2006-2007); sites with ≥600 (vs. <600 patients) on ART; or sites with peer educators. Participation in an association for people living with HIV/AIDS; and receiving care at sites which regularly conduct home-visits were independently associated with lower odds of non-adherence. Higher odds of having a detectable VL were observed among patients at sites with peer educators. Being female; participating in an association for PLWHA; and using a reminder tool were independently associated with lower odds of having detectable VL. High levels of adherence and viral suppression were observed in the Rwandan national ART program, and associated with potentially modifiable factors.
    Full-text · Article · Jan 2013 · PLoS ONE
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: To utilize routinely collected service delivery data from HIV care and treatment clinics in Mozambique to describe the patient population and programmatic outcomes from 2003 to 2009. Data from patient charts were entered into an electronic database at 28 clinics in 5 Mozambican provinces. Patients' characteristics at enrollment in HIV care and at antiretroviral therapy (ART) initiation were examined. We calculated a corrected 12-month mortality estimate using a recently developed nomogram for sub-Saharan African ART patients. A total of 154,188 HIV-infected individuals (10,164 children <15 years old) were enrolled in HIV care services between 2003 and 2009. Of the 51,269 (36%) adults who started ART, 35% initiated ART with CD4 count <100 cells per microliter and 14.4% with World Health Organization stage IV. Just more than 10% (10.5%) of women were documented to be pregnant at enrollment. One-third of the 3,745 (37%) children who initiated ART were <2 years old, and 53% of those <5 years initiated ART severely immunosuppressed (CD4% <15%). Thirty-five percent of all children and 30% of those initiating ART met the definition of severe malnourishment (weight-for-age Z score <-3). Among those who initiated ART, the median estimated 12-month mortality rate across sites was 13.1% (interquartile range: 11.5%-16.0%) and 13.5% (interquartile range: 11.4%-17.4%) for adults and children, respectively. A substantial number of HIV-infected patients have been enrolled in HIV care and initiated on ART, with many patients having advanced HIV disease. With the release of new guidelines for ART use for adults, pregnant women, and children, extensive efforts are needed to ensure more timely initiation of ART.
    Full-text · Article · Jun 2011 · JAIDS Journal of Acquired Immune Deficiency Syndromes
  • [Show abstract] [Hide abstract]
    ABSTRACT: To understand pregnancy intentions and contraception knowledge and use among HIV-positive and negative women in the national prevention of mother-to-child transmission (PMTCT) program in Rwanda. A cross-sectional survey of 236 HIV-positive and 162 HIV-negative postpartum women interviewed within 12 months of their expected delivery date in 12 randomly selected public-sector health facilities providing PMTCT services. : Bivariate analyses explored fertility intentions, and family planning knowledge and use by HIV status. Multivariate analysis identified socio-demographic and service delivery-related predictors of reporting a desire for additional children and modern family planning use. HIV-positive women were less likely to report wanting additional children than HIV-negative women (8 vs. 49%, P < 0.001), and although a majority of women reported discussing family planning with a health worker during their last pregnancy (HIV-positive 79% vs. HIV-negative 69%, P = 0.057), modern family planning use remained low in both groups (HIV-positive 43% vs. HIV-negative 12%, P < 0.001). Condoms were the most commonly used method among HIV-positive women (31%), whereas withdrawal was most frequently reported among HIV-negative women (19%). In multivariate analysis, HIV-negative women were 16 times more likely to report wanting additional children and nearly 85% less likely to use modern family planning. Women who reported making two or less antenatal care visits were 77% less likely to use modern family planning. Our results highlight success in provision of family planning counseling in PMTCT services in Rwanda. As family planning use was low among HIV-positive and negative women, further efforts are needed to improve uptake of modern methods, including dual protection, in Rwandan PMTCT settings.
    No preview · Article · Nov 2009 · AIDS (London, England)
  • Source
    Denis Nash · Evie Andreopoulos · Deborah Horowitz · Nancy Sohler · David Vlahov
    [Show abstract] [Hide abstract]
    ABSTRACT: We assessed the impact of differing laboratory reporting scenarios on the completeness of estimates of people living with human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) (PLWHA) in the U.S., which are used to guide allocation of federal Ryan White funds. We conducted a four-year simulation study using clinical and laboratory data on 1,337 HIV-positive women, including 477 (36%) who did not have AIDS at baseline. We estimated the completeness of HIV (non-AIDS) case ascertainment for three laboratory reporting scenarios: CD4 < 200 cells/microL and detectable viral load (Scenario A); CD4 < 500 cells/microL and no viral load reporting (Scenario B); and CD4 < 500 cells/microL and detectable viral load (Scenario C). Each scenario resulted in an increasing proportion of HIV (non-AIDS) cases being ascertained over time, with Scenario C yielding the highest by Year 4 (Year 1: 69.0%, Year 4: 88.1%), followed by Scenario A (Year 1: 63.3%, Year 4: 84.5%), and Scenario B (Year 1: 43.0%, Year 4: 67.7%). Overall completeness of PLWHA ascertainment after four years was highest for Scenario C (95.8%), followed by Scenario A (94.5%), and Scenario B (88.5%). Differences in laboratory reporting regulations lead to substantial variations in the completeness of PLWHA estimates, and may penalize jurisdictions that are most successful at treating HIV/AIDS patients or those with weak or incomplete HIV/AIDS surveillance systems.
    Full-text · Article · Sep 2007 · Public Health Reports
  • Source
    Denis Nash · Christina Chan · Deborah Horowitz · David Vlahov
    [Show abstract] [Hide abstract]
    ABSTRACT: Breast and cervical cancer screening both are routinely recommended for women. However, data are sparse on factors associated with joint screening behaviors. Our objective to describe the factors associated with receiving both, one, or neither screening test among women aged > or = 50. Using data from the New York City Community Health Survey (NYC CHS), we compared the characteristics of women > age 50 (n = 2059) who missed (1) a Pap smear only, (2) mammography only, or (3) both screening procedures with the characteristics of women who received both tests. Analyses were performed using multiple logistic regression. Seventy-three percent of women had both screening tests, 6.7% needed a Pap smear only, 10% missed mammography only, and 10% missed both tests. After multiple logistic regression, missing a Pap smear only was more likely among women > 70 years compared with younger women and among women from Queens than from Manhattan. Missing mammography only was more common among women not reporting a personal doctor than among those with a doctor and among uninsured women relative to the privately insured. Missing both tests was more common among women > 74 years, current smokers compared with never smokers, women without a personal doctor, and the uninsured. This was less common among women from the Bronx than women from Manhattan and among racial/ethnic minorities compared with non-Hispanic white women. The predictors of each screening outcome appear to be qualitatively different. Changes in provider practices and targeted education may improve Pap smear screening rates, whereas policy initiatives and increased access for the uninsured may raise mammography rates. To achieve optimal preventive care, coscreening should be considered.
    Full-text · Article · Jan 2007 · Journal of Women's Health

Publication Stats

101 Citations
19.55 Total Impact Points

Top co-authors View all


  • 2009-2013
    • Columbia University
      • • Department of Epidemiology
      • • International Center for AIDS Care and Treatment Programs
      New York, New York, United States