Measures and Methodology

Measures and Methodology2021-10-05T09:57:11-04:00

Data Source

The IeDEA Treat All Dashboard aggregates data from HIV care and treatment clinics in sub-Saharan Africa that participate in the IeDEA research consortium. Aggregate data are generated from patient-level databases, using the IeDEA Data Exchange Standard (DES). All ART-naïve patients enrolling in HIV care at sites participating in IeDEA from 2010 to present are included in aggregate measures, with the exception of: patients with incomplete data related to age, sex, and site of enrollment; patients with evidence of ART initiation more than 30 days prior to the date of enrollment in HIV care; and patients with insufficient potential follow-up time for the ascertainment of key IeDEA Treat All Dashboard metrics of interest (see Definitions).

Definitions

HIV Treatment Outcomes among Patients Enrolling in Care at Sites Participating in IeDEA

Data and metrics reflect the HIV treatment initiation and outcomes among ART-naïve patients enrolling in HIV care at sites participating in IeDEA who have at least 15 months of potential follow-up time between care entry and the earliest of transfer or database close. The following measures are also presented as trends over time on the Trends in HIV Care at sites participating in IeDEA page.

Definitions

Care enrollment

Year of enrollment in HIV care at a site participating in the IeDEA consortium, based on the recorded date of enrollment in the participating site’s database.

Timely ART initiation

Initiation of a combined regimen of antiretroviral therapy (ART) medications within 7 days, 14 days, 30 days, and 12 months after enrollment, or more than 12 months after enrollment.

Viral load test recorded

Any record of a viral load test at 6 months (+/- 3 months) and 12 months (+/- 3 months) after enrollment.

Viral suppression

Viral load <1,000 copies/mL among patients with a recorded viral load test at 6 months (+/- 3 months) and 12 months (+/- 3 months) after the date of enrollment in care at a site participating in IeDEA. Patients with no viral load test result within the ascertainment windows are excluded from viral suppression measures. For patients with >1 viral load test result within the ascertainment window, the lowest recorded viral load measure is used.

Treatment Outcomes among Patients Initiating Antiretroviral Therapy (ART) at Sites Participating in IeDEA

Data and metrics reflect the HIV treatment outcomes among ART-naïve patients who initiated antiretroviral therapy at a site participating in IeDEA and who have at least 15 months of potential follow-up time between treatment initiation and the earliest of documented transfer out or database close. The following measures are also presented as trends over time on the Trends in HIV Care at sites participating in IeDEA page.

Definitions

Treatment initiation

Initiation of a combined regimen of antiretroviral therapy (ART) at a site participating in the IeDEA consortium.

Viral load test recorded

Any record of a viral load test at 6 months (+/- 3 months) and 12 months (+/- 3 months) after treatment initiation.

Viral suppression

Viral load <1,000 copies/mL among patients with a recorded viral load test at 6 months (+/- 3 months) and 12 months (+/- 3 months) after the date of treatment initiation at a site participating in IeDEA. Patients with no viral load test result within the ascertainment window are excluded. For patients with >1 viral load test result within the ascertainment window, the lowest recorded viral load measure is used.

CD4 Count Trends: Among Patients Enrolling in Care at Sites Participating in IeDEA

Data and metrics reflect trends in CD4 counts at the time of enrollment in HIV care at clinics participating in IeDEA among patients aged 5 years and older with at least 3 months of potential follow-up time between care entry and the earliest of documented transfer out or database close before ART initiation, or at least 30 days potential follow-up time after ART initiation among those initiating treatment within 3 months of care entry.

Definitions

Enrollment in HIV care

Year of enrollment in HIV care at a site participating in the IeDEA consortium, based on the recorded date of enrollment in the participating site’s database.

Enrollment CD4 count

CD4 count measure recorded within 3 months (+/-) of the date of enrollment, or no more than 30 days after ART initiation among those initiating treatment within 3 months of care entry. For patients with >1 CD4 count within the ascertainment window, the CD4 measure closest to enrollment date is used.

Median CD4 count at enrollment

Median CD4 count among patients with a recorded CD4 count measure within 3 months of enrollment and no more than 30 days after ART initiation. For patients with >1 CD4 count within the ascertainment window, the CD4 measure closest to enrollment date is used.

Data Stratification

The IeDEA Treat All Dashboard allows viewers to explore metrics related to HIV treatment at sites participating in IeDEA across the four Africa regions of IeDEA, as well as by region and by country. Subject to data suppression rules outlined below (see Data Suppression), these metrics can also be viewed by year of enrollment in HIV care or ART initiation, as well as by sex and age group within each unit of geography.

Definitions

IeDEA region

IeDEA regions within sub-Saharan Africa include:

  • East Africa: Kenya, Tanzania, and Uganda.
  • Central Africa: Burundi, Cameroon, Democratic Republic of Congo, Republic of Congo, and Rwanda.
  • Southern Africa: Lesotho, Malawi, Mozambique, South Africa, Zambia, Zimbabwe.
  • West Africa: Benin, Burkina Faso, Cote d’Ivoire, Ghana, Mali, Senegal, and Togo.

Year

Year of enrollment in HIV care or ART initiation at a site participating in the IeDEA consortium.

Sex

Sex as documented in the database of each site participating in the IeDEA consortium.

Age group

Age groups are defined as under 10 years, 10-19 years, 20-29 years, 30-39 years, 40-49 years, and 50+ years at the time of enrollment in care or ART initiation at a site participating in IeDEA.

Data Suppression and Stability of Estimates

To protect the privacy of persons ever enrolled in HIV care at sites participating in IeDEA, the IeDEA Treat All Dashboard does not display metrics based on cell size counts less than 20. Metrics based on cell size counts less than 20 for any stratification variable (e.g. year, country, age-group or sex) are suppressed. While metrics based on cell size counts over 20 are displayed, users should interpret estimates based on small cell size counts with caution.

The stability of all estimates, particularly those based on small cell size counts, is reflected by 95% confidence intervals and interquartile ranges. 95% confidence intervals were calculated using the normal approximation to the binomial.

IeDEA DATA LIMITATIONS

IeDEA data do not necessarily represent the wider HIV care experience of all people living with HIV in any given country or region. In some countries, all data displayed are from a single programmatic cohort, and sites participating in IeDEA are not representative of all sites providing HIV care and treatment in a country. Metrics presented on the IeDEA Treat All Dashboard should not be interpreted as nationally representative, and they only represent the data reported by sites participating in IeDEA in a given country.

IeDEA data are abstracted from records of routine HIV care at participating sites. As with all routinely collected real-world service delivery data, IeDEA data may be incomplete for several reasons, including missing or incomplete reporting of clinical and/or laboratory data, as well as incomplete ascertainment of transfers of patients to a different treatment site, death, and patient loss to care. Metrics presented on the IeDEA Treat All Dashboard do not account for incomplete/missing data, or for undocumented transfers, deaths, or other types of ‘loss to follow-up’ among those ever enrolling in HIV care at participating sites.

In addition, at some sites participating in IeDEA, viral load testing data may be incomplete or missing in the database, despite tests having been conducted. For viral load metrics, viral load tests performed outside the +/- 3 month ascertainment windows are not included in the outcome measures. Therefore viral load testing metrics may not accurately reflect viral load monitoring practices among persons enrolled in HIV care at sites participating in IeDEA.

Finally, lags in data abstraction and processing differ across sites and programs that participate in IeDEA. As some metrics shown on the Dashboard require 15 months of potential follow-up, there can be a minimum data lag of 15 months for these metrics, with longer lags for certain countries and programs.

Limited Comparability of data Across Sites and Countries

Sites participating in the global IeDEA research consortium comprise a diverse group of real-world HIV care and treatment clinics and programs whose routine service delivery data may not be directly comparable due to differences in data collection practices. For example, in many settings, patients’ date of enrollment in HIV care reflects the timing of their engagement in care at the site participating in IeDEA. However, some clinics and programs participating in IeDEA are limited to ART patients—i.e., only enroll patients in HIV care at the time they initiate antiretroviral treatment, making it appear that 100% of patients initiate ART on the day of enrollment. The diversity of the programs and sites that contribute data to IeDEA may limit the comparability of data across countries.

Public Use and Suggested Citation

The information presented on this website, www.iedeadashboard.org, has been provided for public use. You may access the information freely and may use, copy, or distribute this information without explicit permission. If referencing any specific data, please cite as you would any other document, resource, or data on the internet:

  • Recommended citation of the website: IeDEA Treat All Dashboard. September 2021. Retrieved from www.IeDEAdashboard.org.