• Loading...

    DSHS HIV/STD Program

    Post Office Box 149347, MC 1873
    Austin, Texas 78714

    Phone: (512) 533-3000

    E-mail the HIV/STD Program

    E-mail data requests to HIV/STD Program - This email can be used to request data and statistics on HIV, TB, and STDs in Texas. It cannot be used to get treatment or infection history for individuals, or to request information on programs and services. Please do not include any personal, identifying health information in your email such as HIV status, Date of Birth, Social Security Number, etc.

    For treatment/testing history, please contact your local Health Department.

    For information on HIV testing and services available to Persons Living with HIV and AIDS, please contact your local HIV services organization.

Linkage to Care Estimates for Newly Diagnosed Individuals in Texas

Loading...

The National HIV/AIDS Strategy has set the goal of linking 85% of newly diagnosed HIV patients to clinical care within 3 months of diagnosis. After individuals are diagnosed with HIV, positive health outcomes are achievable when individuals are both immediately linked into medical care and remain adherent to medical treatment.

This measure is developed using the data collected by DSHS and used for estimating unmet need in 2010. For this analysis, however, the focus is shifted towards individuals diagnosed with HIV for the first time in 2010 and measuring the time span between the first diagnosis date and their first service date in Texas involving a CD4 count, a viral load test, antiretroviral therapy, or an outpatient/ambulatory medical care visit. A newly diagnosed individual in 2010 is said to be linked into care within three months if they have a CD4 count, a viral load test, antiretroviral therapy, or an outpatient/ambulatory medical care visit 90 days or less starting from their first date of diagnosis (excludes out of state cases).

Statewide Trends for 2010 Linkage to Care

In 2010, over two-thirds (69%) of newly diagnosed people in Texas (n=2,180) were linked into care within three months of their diagnosis. Currently, Texas has to make a 16% gain in order to meet the National HIV Strategy goal (85% linkage to care rate). Increasing the linkage rate at both the state and local level is achievable if targeted interventions are developed for groups with a linkage rate below the average and/or groups representing the largest proportions of new diagnoses in Texas. Both strategies would guarantee increased, but equitable, gains in linkage to care.

Linkage to care estimates vary by region (for this analysis, we focus on the five Ryan White Program Part A eligible metropolitan area/transitional grant areas (abbreviated EMA and TGA),), race/ethnicity, sex , age), and major exposure category, late HIV diagnosis (whether one progressed to an AIDS diagnosis within 12 months of their initial HIV diagnosis) and a 2010 HIV/STD co-morbidity (whether the individual had a reported sexually transmitted disease (STD) within 2010).

Demographic Group Estimates for Linkage to Care (Table 1-4)

Populations with larger-than-average proportions of people out of care are important to identify because they may uncover systematic difficulties or obstacles to care. Groups with linkage to care rates below the state average level (69%) include (looking across Table 1-4):

  • Black men in almost all exposure categories (53-62% with timely linkage)
  • IDU (60%)
  • MSM/IDU (64%)
  • Hispanic men IDU and heterosexuals (51-64%)
  • Youth (13-24 years old) (64%)
  • Individuals diagnosed with HIV in a timely manner (not an AIDS case or having progressed from HIV to AIDS in more than a year as of June 30, 2011) (62%)
  • Newly diagnosed individuals from the Houston EMA region (65%) and areas of Texas outside of metropolitan epicenters (66%)

In addition to focusing on the abovementioned groups, it is important to acknowledge that targeting MSM, especially Black and Latino MSM, those in the 13-34 age groups, and those diagnosed in the Houston/Dallas EMA could also bring significant gains to Texas’ timely linkage to care rate because these group characteristics are observed in more than half of all newly diagnosed people in 2010. Among the individuals included in the linkage to care analyses:

  • Black and Latino males account for 55% of newly diagnosed individuals
  • Black and Latino MSM account for 42% of newly diagnosed individuals
  • Adolescents (and young adults) account for over half of newly diagnosed indviduals in 2010
  • Individuals diagnosed in Houston (or Dallas EMA) account for 57% of all newly diagnosed individuals in 2010

Regional Estimates of Linkage to Care by Demographic Group (Table 1-4)

When interpreting the linkage to care rates within each EMA/TGA region, consumers of these estimates are encouraged to develop an understanding of the aggregate trends (Table 1 – 3) before delving into the cross tabulations shown in Table 4 especially across the EMA/TGAs. Because cell sizes become even smaller in Table 4 and Table 3 when compared to Table 1 and 2, caution is warranted when interpreting the proportions based on a very small cell size.

  • The Austin TGA exhibited the highest levels of linkage (80%, 138 new cases in 2010). Almost all the demographic subgroups in the Austin EMA area met the state average (69%) or regional average (80%) with the exception of a couple of very small subgroups in Table 4. The following subgroups in the Dallas EMA exhibited timely linkage to care below the regional average: Blacks (73%), Hispanics (74%), IDU (69%), individuals diagnosed with HIV in a timely manner (70%), and those with HIV and a STD co-infection in 2010 (70%). Black and Hispanic males and to a lesser degree Black and Hispanic females lagged behind their White male and female counterparts (Table 2) in getting linked into care in a timely manner.
  • The following subgroups in the Forth Worth TGA exhibited timely linkage to care below the regional average (72%): Blacks (64%), Black men (64%), Black MSM (63%), those in the 35-44 age group (65%), those identified as MSM/IDU (61%), and those with HIV and a STD co-infection in 2010 (55%). Black males and to a lesser degree Black females lagged behind their White male and female counterparts in getting linked into care in a timely manner (Table 2).
  • The following subgroups in the San Antonio TGA exhibited timely linkage to care below the regional average (69%): males (67%) which is mostly attributed to Whites males (60%) and Blacks males (57%), those in the 25-34 age group (64%), and individuals diagnosed with HIV in a timely manner (progressed from HIV to AIDS in more than a year) (62%).
  • Timely linkage to care rates in the Houston EMA area were below the regional average (65%) for: Blacks (61%), Blacks males (59%), Black MSM (57%), White heterosexuals (59%), those in the 13-24 age group (57%), IDU (51%) and MSM/IDU (58%) and individuals diagnosed with HIV in a timely manner (not an AIDS case or progressed from HIV to AIDS in more than a year) (57%).

 

Table 1

Percentage of Newly Diagnosed Individuals Linked into Care within Three Months of Diagnosis
Table 1 (PDF : 12 kb)

Table 2

Percentage of 2010 Newly Diagnosed Individuals Linked into Care within Three Months of Diagnosis by Race/Ethnicity and Sex
Table 2 (PDF : 11 kb)

Table 3

Percentage of 2010 Newly Diagnosed Individuals Linked into Care within Three Months of Diagnosis by Race/Ethnicity and Mode of Transmission
Table 3 (PDF : 13 kb)

Table 4

Percentage of 2010 Newly Diagnosed Individuals Linked into Care within Three Months of Diagnosis by Mode of Transmission, Race/Ethnicity and Sex
Table 4 (PDF : 13 kb)

 


Data Sources and Matching

Matching methods varied depending on the type of information that was available for matching.
For data sets where names and other personal identifiers (e.g. date of birth) were available, Link King or other linking algorithms (ELR data only) were used for matching. When only unique record numbers or limited data elements were available (e.g. first and third initial of first and last name combined with date of birth) were available, exact matching using SAS 9.2 was used.

The following data sets were matched against HIV/AIDS cases in eHARS to determine if a client had a met medical need:

  • Electronic HIV/AIDS Reporting System (eHARS) -This is the data source that is used as the universe of HIV/AIDS cases for estimating unmet need, retention to care for PLWH, linkage to care for newly diagnosed individuals and continuity of care for outpatient/ambulatory medical care visits, CD4 labs, and viral load labs. 
  • Texas AIDS Drug Assistance Program (ADAP) or State Pharmacy Assistance Program (SPAP) – If ADAP/SPAP provided antiretroviral (ARV) medications for a client, then that person was considered to have met medical need for the year in which the medication was provided. Name based matching was performed to determine persons with a met medical need during 2010.
  • Electronic Lab Reporting (ELR) – The largest providers of laboratory services for the state report CD4 and viral load labs to DSHS. Name based matching of these reports was used to determine if individuals received a CD4 count or viral load test during 2010. Please note that most paper-based labs and labs reported directly to the Houston health jurisdiction were not available at the time these measures were developed and are not reflected in the estimates.
  • AIDS Regional Information and Evaluation System (ARIES) – Services provided to Ryan White eligible clients by funded service providers are reported in ARIES. If a client received a viral load, CD4 count, laboratory service, ARV medication, or an outpatient/ambulatory visit medical care during 2010, the client was reported as having a met medical need during that year. When available, name based matching was used to determine persons with a met medical need during 2010. When client names were not available, matching was based on a unique record number generated in ARIES and eHARS.
  • Medicaid/Children's Health Insurance Program (CHIP) – If a client received a viral load, CD4 count, laboratory service, ARV medication, or an outpatient/ambulatory medical visit through Medicaid/CHIP during 2010, the client was reported as having a met medical need during that year. Name based matching was performed to determine persons with a met medical need during 2010. Please note that at the time of the project, the fourth quarter of the 2010 Medicaid/CHIP data was not available for release at the time these measures were developed and are not reflected in the estimates.
  • Private Insurers – For this analysis, a few of the largest private providers in Texas extracted relevant procedures (CD4 counts, viral load measurements, ARV, or an outpatient/ambulatory medical visit) from their claims systems. Matching was based on available data elements such as first and third initial of first and last name, date of birth and sex.

Methods

The midyear 2010 eHARS dataset (6/30/2011) was used for the 2010 unmet need estimates, 2007-2009 unmet need updated estimates, retention to care estimates, linkage to care estimates, and the continuity of care measures (OAMC visit, CD4 labs, and viral load labs). Diagnosed HIV/AIDS cases that had been entered and were living on 12/31/2010 were included for the total population for unmet need in 2010. Using the datasets and matching methods described above, a newly diagnosed individual in 2010 is said to be linked into care within three months if they have a CD4 count, a viral load test, antiretroviral therapy, or an outpatient/ambulatory medical care visit 90 days or less starting from the first date of diagnosis in Texas as identified in eHARS (excludes out of state cases).

For cases diagnosed in 2010, the linkage to care estimate were calculated by dividing the number of new cases in 2010 with evidence of HIV-related medical care within three months of their first HIV date of diagnosis by the total number of new cases in 2010 (multiplied by 100). Individuals diagnosed during the last quarter of the year were excluded from the linkage to care measure because data for the required follow-up period (January 2011-March 2011) were not available during the time that this measure was calculated. A categorical variable was coded as 0 for new cases linked into care within 3 months, 1 for new cases linked into care in four or more months and 3 for new cases with no evidence of medical services in 2010. Results examined here focus on timely linkage (with 3 months of diagnosis), but results for the proportion of newly diagnosed individuals linked in four or more months (7%) and the proportion of newly diagnosed individuals with no evidence of being linked into care in 2010 (24%) are available by contacting the .

Diagram 1 shows the linkage to care measure before and after all the data exclusions were applied. Please note that the method selected is reflected in outcome D.

Diagram 1

Diagram 1
Diagram 1

Limitations

The estimates of unmet need and all the measures computed using the 2010 data call for unmet need should be considered liberal estimates for a number of reasons: 

  1. They do not include all the HIV-related care provided by the VA, Medicare, and all private providers in the state. 
  2. Matches conducted between eHARS and some of the cases in ARIES and between eHARS and private payer data were based on a unique identifier or limited data elements rather than client name; this may underestimate the true number of clients with met need from these data sources. 
  3. There are persons reported in eHARS who have since moved out of the state (out-migrated cases) and since we do not have a systematic way of identifying and removing these out-migrated cases they remain in the denominator and inflate our unmet need estimate.

DSHS did not exclude TDCJ cases when estimating unmet need for 2010 and did not exclude TDCJ cases when recalculating unmet need estimates for previous years. They were also not excluded in the linkage to care analysis. When TDCJ cases are excluded, the statewide unmet need estimates increased by approximately a one or two percentage points. While including TDCJ cases in this year’s unmet need analysis did not impact overall estimates, it is evident that DSHS is not receiving all medical service data for cases diagnosed in TDCJ. In addition, University of Texas Medical Branch paper labs received by DSHS were not available at the time these measures were developed and are not reflected in the estimates.


Back to the dashboard


Last updated December 01, 2011