Antiretroviral therapy coverage among people living with HIV
- Rows
- 3,647
- Columns
- 4
- Data range
- 2000-01-01 to 2024-01-01
- Auto-refresh
- Re-checks its source every 30 days
Time-series on ART coverage rates by country, from UNAIDS data. Covers treatment progress over years.
AI analysis
Dataset Analysis: Antiretroviral Therapy (ART) Coverage Among People Living with HIV
1. Dataset Overview & Value
This dataset, sourced from UNAIDS via Our World in Data (https://ourworldindata.org/grapher/antiretroviral-therapy-coverage-among-people-living-with-hiv.csv), captures a critical global health metric — the percentage of HIV-positive individuals receiving antiretroviral therapy across 152 countries over 25 years. Our World in Data is a well-regarded, continuously updated publisher, and the versioned URL (v=1) suggests a stable snapshot, though the 25-year time span implies coverage through at least the early 2020s. This dataset is exceptionally valuable for correlation studies because ART coverage is both a cause and effect variable: it reflects healthcare system capacity, funding flows (e.g., PEPFAR, Global Fund), and policy decisions, while simultaneously driving outcomes like AIDS-related mortality and HIV transmission rates. Researchers can use it to explore whether treatment access has kept pace with prevalence, and to identify which country clusters have made the most or least progress.
---
2. Data Quality Observations
Overall data quality is good but not pristine. Across 4 columns and 3,647 rows, there are 114 null cells — all concentrated in the Antiretroviral therapy coverage column (3.1% of rows), which is the sole quantitative measure. This is an acceptable missing-data rate for a global health time-series, likely reflecting countries with no reporting infrastructure or suppressed data in early years. The Year, Entity, and Code columns report zero nulls, indicating a structurally clean backbone. Duplicate row counts are pending final recomputation (Phase D), but the clean join key columns suggest deduplication is unlikely to be a major issue. No type mismatches are flagged, though the ART coverage column is typed as Integer — a minor concern, since percentage values are often reported to one decimal place and rounding may introduce small precision loss at the margins.
---
3. Key Column Distributions
The Antiretroviral therapy coverage column is the analytical core and reveals a richly heterogeneous global story. The mean of 35.3% sits above the median of 33%, indicating a slight rightward skew (skew = 0.343) — most countries cluster in the lower-to-mid range, with a tail of high-performing nations pulling the average up. The interquartile range (Q1=9, Q3=57) is strikingly wide, spanning nearly 48 percentage points, which signals enormous global disparity in treatment access. The minimum of 0% and maximum of 98% represent the full spectrum from crisis-level gaps to near-universal coverage. The standard deviation of 27.7 further confirms high variance. With 99 distinct values out of a 0–98 range and 114 nulls, the distribution is well-spread rather than clumped, making it suitable for regression and segmentation analysis. The Year column covers exactly 25 distinct time points, suggesting annual data — ideal for trend analysis — while the 152 distinct entities in Entity and Code provide solid global representation.
---
4. Recommended Join Key Columns
For cross-dataset correlation, the optimal join strategy uses a compound key of Code + Year: - Code (ISO 3-letter country code, 152 distinct, zero nulls) is the preferred geographic join key — standardized, machine-readable, and widely used across global health, economic, and demographic datasets. - Year (25 distinct values, zero nulls) enables temporal alignment with other annual time-series. - Entity (country name) is useful as a human-readable fallback but should be treated as secondary to Code due to naming inconsistencies across datasets (e.g., "Congo" vs. "Democratic Republic of Congo").
---
5. Suggested Pairing Datasets for Correlation Discovery
| Dataset / Topic | Rationale | |---|---| | HIV prevalence by country (UNAIDS) | Directly complements coverage — correlate treatment access with prevalence trajectories | | AIDS-related mortality rates | Tests the causal hypothesis: does higher ART coverage reduce deaths? | | Health expenditure per capita (WHO/World Bank) | Examines whether funding drives coverage improvements | | PEPFAR & Global Fund disbursements | Links donor aid flows to country-level ART scale-up | | GDP per capita (World Bank) | Tests wealth as a predictor of treatment access | | Healthcare workforce density (physicians/nurses per 1,000) | Explores supply-side constraints on treatment delivery | | New HIV infections (incidence) | Evaluates whether treatment-as-prevention reduces transmission |
A particularly high-value analysis would be a panel regression joining this dataset with AIDS mortality and health expenditure data on Code + Year, to isolate the independent effect of ART coverage on survival outcomes while controlling for economic development.
Columns
- Year (date)
- Entity (string)
- Code (string)
- Antiretroviral therapy coverage (% of people living with HIV) (integer)