5,676 is the cumulative number of diagnoses recorded by Fiji’s Ministry of Health from 1989 through 2025. UNAIDS estimated 9,100 people living with HIV in 2025. The latter is a modelled estimate that includes people who have not been diagnosed. Cumulative diagnoses and current estimated prevalence have different definitions. Neither figure should be substituted for the other.
In 2021, Fiji recorded 151 new HIV diagnoses. Four years later, that figure had reached 2,016. The scale of this increase gives context to the national crisis declaration in September 2026. Yet the figures also raise an important question: does the increase reflect more infections, more diagnoses through testing, or a combination of both?
02 / THE TURNING POINT
Looking more closely at the period after 2021, the acceleration becomes particularly apparent. However, a diagnosis does not tell us when someone acquired HIV. Testing also matters. In 2025, the Ministry recorded 92,377 screening tests across laboratory and point-of-care services. Consequently, the trend captures the growing number of infections identified by surveillance, rather than providing a direct measure of new infections. Source: Ministry of Health and Medical Services (2026b), Fiji HIV Surveillance Report 2025, Table 13 and discussion.
03 / THE POLICY RESPONSE
The urgency of the response becomes clearer when considering estimates beyond the Ministry’s diagnosis records. On 16 September 2026, UNAIDS welcomed Fiji’s declaration of HIV as a national crisis, reporting an estimated twelvefold rise in new infections between 2010 and 2025. Unlike the chart, this estimate concerns infections rather than recorded diagnoses. Read the UNAIDS announcement.
UNAIDS estimates that annual new HIV infections in Fiji rose about twelvefold between 2010 and 2025. Across the Pacific, estimated new infections rose by 65% over the same period. Papua New Guinea has the region’s largest HIV epidemic, although Fiji’s rapid increase is a different warning sign.
These are modelled estimates of new infections, not the recorded diagnoses shown in Graph 1. The comparison describes change over time, not infection rates per person or the number of people living with HIV.
Source: Joint United Nations Programme on HIV/AIDS, The Pacific and HIV: A chance to turn the tide (23 October 2026). UNAIDS statement (16 September 2026).
The national trend establishes the scale of the increase. It does not, however, tell us who is represented in those figures. Looking at age groups provides another perspective on the epidemic, particularly whether diagnoses are concentrated among younger or older people. The heatmap compares recorded diagnoses from 2022 to 2025. Lighter blue represents lower counts and darker blue higher counts, with exact figures available by hovering over each cell.
Young adults account for a substantial share of the recorded
diagnoses. To understand this pattern, it is important to distinguish
the number of diagnoses from the likelihood of infection. The chart uses
the Ministry workbook’s Age worksheet, which breaks annual
diagnoses into age groups. This differs from the Cases
worksheet used for the national trend. In 2023, the age-group counts
total 419, while the national series reports 415. The original values
have been retained rather than adjusted, and the discrepancy is
discussed further in the methods.
02 / WHY AGE MATTERS
Ages 20–24 accounted for 17.6% of age-recorded diagnoses in 2022 and 32.0% in 2025. These are shares of diagnoses, not population-adjusted infection rates.
The change extends beyond the total number of diagnoses. Between 2022 and 2025, people aged 20–24 accounted for a growing share of the age-recorded cases. This suggests a changing profile among those diagnosed, although it does not establish that young adults face a higher infection risk. That comparison would require population figures and information about testing across age groups.
03 / CHILDREN IN THE CRISIS
Beyond the national surveillance figures, Movono and Evans (25 September 2026) report on the challenges surrounding mother-to-child HIV transmission in Fiji. Their reporting draws attention to stigma and barriers to treatment, highlighting how the consequences of the epidemic extend beyond the number of recorded diagnoses.
Source: Ministry of Health and Medical Services (2026a), Fiji HIV Surveillance Data 2025 Public, Age worksheet (2022–2025). Counts are not age-standardised rates.
Age is only one part of the picture. The location of recorded diagnoses also matters for understanding where testing and treatment services may be needed. Fiji is an island nation in the South Pacific, east of Vanuatu and north of New Zealand. To examine the geographic pattern, the map groups provincial boundaries into four reporting divisions. The colours distinguish divisions, rather than representing differences in infection risk.
of Fiji’s recorded HIV diagnoses were in Central Division.
The Ministry reported 1,359 diagnoses in Central, compared with 556 in Western, 84 in Northern and 17 in Eastern. Central and Western together accounted for 95% of reported diagnoses.
The geographic concentration identifies where the largest recorded caseload occurred and where demand for testing, follow-up and treatment services may be substantial. It does not establish that Central Division has the highest per-person infection risk. The permanent key lists each division’s count using the source table’s division names. Select a division to read its share of national diagnoses.
02 / WHAT THE MAP CANNOT ESTABLISH
At first glance, the concentration in Central Division may suggest that the epidemic is more severe there. However, the number of recorded diagnoses also depends on how many people live in each division and how easily they can access testing. Without population-adjusted rates, the map identifies the distribution of reported cases rather than the risk faced by an individual.
Sources: Ministry of Health and Medical Services (2026b), Fiji HIV Surveillance Report 2025, Table 1 (four-division counts); United Nations SALB, Fiji administrative boundaries (geographic shapes).
Interpretation: Counts may reflect population size, testing coverage and access to services. The map compares recorded caseloads, not population-adjusted diagnosis rates. The SALB shapefile lists Rotuma as Northern in its division-name field, while the division-code field identifies it as Eastern. The map follows the division-name field. This discrepancy limits the boundary interpretation. The map uses local geographic boundaries without external map tiles.
While the map identifies where diagnoses were recorded, it leaves another important question unanswered: what happens once someone receives a diagnosis? The Ministry’s surveillance report provides a closer look by recording both the reported mode of transmission and whether a person was engaged in care. ART (antiretroviral therapy) is medication used to treat HIV. This chart compares the share receiving ART with the share not enrolled in care within each transmission classification. People in care without ART and recorded deaths are excluded from these paired bars and appear in Graph 5, therefore the two percentages do not add to 100%.
One category stands out. People whose mode of transmission was recorded as unknown also accounted for a large proportion of those not enrolled in care. This connects two challenges for the health response: understanding how cases are documented and ensuring people can access treatment. However, an unknown classification is not itself a route of transmission, and the records do not explain why care enrolment was missing.
These are the transmission categories recorded by the Ministry, not treatment groups. The bars display the share of people in each category receiving ART or not enrolled in care.
ART is defined above the chart. The paired percentages do not total 100%, since the categories In Care Not on ART and Deceased appear in Graph 5. Hover over each bar for the original Ministry classification, care-status wording and its count.
Of the 2,016 new diagnoses, 1,167 (57.9%) had no recorded transmission classification. Within that group, 815 people (69.8%) were not enrolled in care, compared with 249 (21.3%) receiving ART. The association between missing transmission information and non-enrolment warrants attention to both documentation and follow-up. These records cannot establish a causal relationship or explain why care was not recorded.
Among those classified as people who inject drugs, 316 of 362 (87.3%) were on ART. Among men who have sex with men, 127 of 144 (88.2%) were on ART. These percentages describe recorded care status within each classification, not population-wide treatment coverage.
02 / A RESPONSE BEYOND THE CHART
The UNAIDS announcement discusses expanded prevention and testing, sterile needle-and-syringe access and stigma reduction. In reporting about Fiji, bluetoothing refers to the practice of sharing blood after drug injection. It is a serious potential exposure route, not a confirmed explanation for cases with an unknown transmission classification. The Guardian’s reporting describes barriers faced by people who inject drugs. Neither source establishes the route of transmission for the unknown cases in this dataset.
Source: Ministry of Health and Medical Services (2026b), Fiji HIV Surveillance Report 2025, Table 6. The chart retains the Table 6 labels in the imported data and hover details, while shortening some axis labels for readability. The source groups transmission classifications with care status. Categories may not represent independently verified infection routes.
The differences between transmission classifications raise a broader question: how many people diagnosed in 2025 were receiving treatment? Examining all four recorded care statuses places the earlier comparison in context and reveals the gap between a recorded diagnosis and enrolment in care.
Across every transmission classification, 845 people newly diagnosed in 2025 were recorded as not enrolled in care. A further 128 (6.3%) were in care, although not yet on ART. Of the 2,016 diagnoses, 976 (48.4%) were on ART and 67 (3.3%) were recorded as deceased. These four statuses sum to the 2025 newly diagnosed cohort. They are not sequential stages that every individual passes through.
Beyond this newly diagnosed cohort, the wider treatment picture uses a different denominator. The Ministry separately reported 3,535 people living with HIV in its records at the end of 2025, of whom 1,979 were receiving ART and 288 had documented viral suppression. These figures describe all registered people, not just those newly diagnosed in 2025. Documented suppression is also affected by whether viral-load tests were completed and recorded. It should not be treated as the true biological suppression rate among everyone living with HIV.
02 / THE BIGGER DENOMINATOR
UNAIDS estimated 9,100 people living with HIV in Fiji in 2025, of whom 39% knew their status and 22% accessed ART. These are national estimates, whereas the chart displays the 2,016 people newly diagnosed in 2025. The percentages cannot be directly compared.
03 / WHAT TREATMENT CAN PREVENT
The ABC’s reporting describes how treatment can prevent mother-to-child transmission, while stigma, travel costs and limited privacy can make treatment harder to access. This illustrates why connecting people to care matters beyond recording a diagnosis.
Source: Ministry of Health and Medical Services (2026b), Fiji HIV Surveillance Report 2025, Tables 5–6 and Section 6 (cascade of care). Chart denominator: 2,016 newly diagnosed people in 2025.
Taken together, the five visualisations trace a crisis that extends beyond the rise in diagnoses. Fiji recorded 2,016 new HIV diagnoses in 2025, with the largest age-group count among people aged 20–24 and roughly two-thirds of cases recorded in Central Division. These findings provide insight into who and where the surveillance system is identifying. However, they do not establish which populations face the greatest individual risk, nor can the records separate changes in transmission from changes in testing and reporting.
The consequences of the epidemic are also evident beyond the diagnosis figures. In a separate measure, the Ministry recorded 117 HIV-related deaths in 2025, including 58 among people aged 25–39 and 17 among children younger than 15. The annual total was 126 in 2024. The report cautions that reporting delays and incomplete capture limit interpretation of the apparent decline. These mortality figures cover deaths recorded in 2025, not exclusively people newly diagnosed that year. Source: Ministry of Health and Medical Services (2026b), Fiji HIV Surveillance Report 2025, Section 8, Figure 24 and Table 8.
Ultimately, recording a diagnosis is only one part of responding to the crisis. Of the 845 newly diagnosed people not enrolled in care, 815 were also recorded with an unknown transmission classification. This overlap brings the story back to the importance of both accurate surveillance and access to treatment. It identifies an area requiring closer investigation, rather than providing an explanation for any individual’s circumstances.
UNAIDS reported a 12-fold rise in estimated new infections from 2010 to 2025, and outlined Fiji’s plans to expand prevention, treatment and harm reduction.
Official announcement ↗Reporting examines mother-to-child transmission and the role of stigma, privacy and access to treatment. The article provides context rather than data for the visualisations.
Read reporting ↗Reporting examines unsafe injecting, drug use and the practical challenges of providing sterile equipment and accessible HIV services.
Read reporting ↗Fiji’s HIV crisis cannot be understood through the national diagnosis count alone. Examining age, geography, transmission classification and care status reveals where the surveillance records provide useful insight and where important questions remain. The findings point to the need for continued testing, reliable reporting and timely access to treatment, while recognising that recorded cases and estimated infections measure different aspects of the epidemic.
Scope of sources. The visualisations draw on three independent source organisations: the Fiji Ministry of Health and Medical Services (surveillance report and workbook), UNAIDS (regional and national estimates used for context) and the United Nations SALB programme (map boundaries). News reports are cited for journalistic context and are not used to construct the charts. The news articles provide qualitative context; no additional hospital statistics are incorporated into the analysis.
Primary data sources. The Ministry of Health and Medical Services (2026b) Fiji HIV Surveillance Report 2025: National epidemiological analysis of HIV testing, diagnosis and care cascade provides the source tables, care outcomes and surveillance context. The Ministry’s (2026a) Fiji HIV Surveillance Data 2025 Public workbook supplies the annual and age-series observations.
Data and calculations. Annual diagnoses (1989–2025)
and age-group counts (2022–2025) were read from the public workbook’s
Cases and Age worksheets. The geographic
counts (Central 1,359, Western 556, Northern 84 and Eastern 17) came
from the Ministry’s four-division surveillance report. The
transmission-by-care matrix came from Table 6 of the Ministry’s Fiji
HIV Surveillance Report 2025. The 2025 care-status distribution was
calculated by summing Table 6 columns and is consistent with the
published totals. Percentages use the stated chart-specific
denominators. Counts are newly recorded diagnoses, not
estimates of infection incidence.
Data preparation. Annual diagnoses and age counts
are imported from the Ministry surveillance workbook. Division counts
and the transmission-by-care table are transcribed from the Ministry
reports into geography_2025.csv and
transmission_care_2025.csv, then imported into R.
Care-status totals are derived by summing the latter table rather than
entering them separately. Source-table values and full category labels
are retained in the imported data and hover text; shorter labels are
used on axes where needed, and totals are checked against the published
2,016 diagnoses. The Rmd automatically checks that division and
care-status totals equal the reported 2,016 diagnoses. The four
geographic values were compared individually against Table 1 of the
original surveillance report, and all 20 transmission-by-care cells were
compared against its Table 6. These source-table checks supplement the
automated total checks in the R code.
Regional context. The UNAIDS comparison uses reported modelled estimates of new HIV infections between 2010 and 2025. It is contextual reporting, not a sixth chart or a dataset merged with Ministry diagnosis records.
Software and reproducibility. Data processing and
visualisations were completed in R using readxl,
dplyr, tidyr, ggplot2,
plotly, scales, htmltools,
knitr, sf and leaflet. The
relevant package publications and documentation are listed in the
references. Package-version references describe the cited software, not
necessarily the exact versions installed on every computer. The design
uses sequential colour for quantitative heatmap counts, distinct colours
for geographic categories, and interactive hover labels to provide
precise values. The age heatmap and transmission comparison both
represent multiple variables.
Visual methods. The five interactive visualisations
are (1) annual time-series line chart, (2) age-by-year heatmap, (3)
interactive division map, (4) clustered column comparison of care
statuses by transmission classification and (5) stacked distribution of
2025 care status. Graphics are generated in R using
ggplot2, plotly, leaflet,
sf, dplyr, tidyr,
readxl, scales and htmltools. The
heatmap relates age group, year and diagnosis count. The grouped columns
compare care status within transmission classifications using
category-specific denominators. The time series and map establish the
historical and geographic context. The final stacked bar summarises the
2025 care-status distribution. Hover details provide exact counts while
keeping the graphics readable.
Data quality and interpretation. The public workbook
contains separate Cases (annual national totals) and
Age (age-group breakdowns) worksheets. For 2023, the
Age values sum to 419 diagnoses, whereas Cases
reports 415, a difference of four. Each graph retains its respective
worksheet values, and neither total has been silently adjusted. The
source material does not establish the reason for the discrepancy. The
four division totals represent counts, not population-standardised
risks. The United Nations SALB shapefile labels Rotuma as Northern by
division name, while its code corresponds to Eastern. The map follows
the division-name field. The 2025 care-status chart refers to newly
diagnosed cases (n = 2,016), whereas the broader treatment cascade
refers to registered people living with HIV (n = 3,535). Unknown
transmission classification represents missing surveillance information,
not a transmission mechanism. Results should not be used to infer
individual causation or comparative risk.
Contemporary reporting (contextual links). UNAIDS (16 September 2026), UNAIDS to support Fiji as it declares HIV a national crisis and puts in new measures to strengthen the response. ABC News (25 September 2026), reporting on mother-to-child transmission and treatment barriers. Duff (29 September 2026), ‘Children are dying’: Fiji hopes declaring an HIV emergency will help it tackle crisis. Linked at their relevant passages and above. These contextual articles are not used as replacements for the Ministry’s chart data.
Publication note. The 2023 age-sheet discrepancy is disclosed rather than silently corrected. The two transcribed CSVs were checked cell by cell against Tables 1 and 6 of the supplied surveillance report. The workbook and two CSVs are imported by the visualisation code.
Cheng, J., Schloerke, B., Karambelkar, B., Xie, Y., & Aden-Buie, G. (2025). leaflet: Create interactive web maps with the JavaScript ‘Leaflet’ library (Version 2.2.3) [R package]. https://doi.org/10.32614/CRAN.package.leaflet
Cheng, J., Sievert, C., Schloerke, B., Chang, W., Xie, Y., & Allen, J. (2025). htmltools: Tools for HTML (Version 0.5.9) [R package]. https://doi.org/10.32614/CRAN.package.htmltools
Duff, M. (2026, September 29). ‘Children are dying’: Fiji hopes declaring an HIV emergency will help it tackle crisis. The Guardian. https://www.theguardian.com/world/2026/sep/29/children-fiji-hiv-emergency-testing-drug-raid
Joint United Nations Programme on HIV/AIDS. (2026a, September 16). UNAIDS to support Fiji as it declares HIV a national crisis and puts in new measures to strengthen the response. https://www.unaids.org/en/resources/presscentre/pressreleaseandstatementarchive/2026/september/20260916_Fiji_national_crisis
Joint United Nations Programme on HIV/AIDS. (2026b, October 23). The Pacific and HIV: A chance to turn the tide. https://unaids-ap.org/2026/10/23/the-pacific-and-hiv-a-chance-to-turn-the-tide/
Ministry of Health and Medical Services. (2026a). Fiji HIV surveillance data 2025 public [Data set]. https://www.health.gov.fj/fiji-hiv-surveillance-report-2025/
Ministry of Health and Medical Services. (2026b). Fiji HIV surveillance report 2025: National epidemiological analysis of HIV testing, diagnosis and care cascade [Surveillance report]. https://www.health.gov.fj/fiji-hiv-surveillance-report-2025/
Movono, L., & Evans, K. (2026, September 25). Babies pay the price as HIV epidemic surges in Fiji. ABC News. https://www.abc.net.au/news/2026-09-25/babies-pay-price-as-hiv-epidemic-surges-in-fiji/107168510
OpenAI. (2026). ChatGPT [Large language model]. https://chatgpt.com/
Pebesma, E. (2018). Simple features for R: Standardized support for spatial vector data. The R Journal, 10(1), 439–446. https://doi.org/10.32614/RJ-2018-009
Sievert, C. (2020). Interactive web-based data visualization with R, plotly, and shiny. Chapman and Hall/CRC. https://plotly-r.com
United Nations. (2023). Fiji, SALB, United Nations [Geographic data set]. Second Administrative Level Boundaries (SALB) Programme. https://salb.un.org/en/data/fji
Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer. https://ggplot2.tidyverse.org
Wickham, H., & Bryan, J. (2025). readxl: Read Excel files (Version 1.4.5) [R package]. https://doi.org/10.32614/CRAN.package.readxl
Wickham, H., François, R., Henry, L., Müller, K., & Vaughan, D. (2026). dplyr: A grammar of data manipulation (Version 1.2.0) [R package]. https://doi.org/10.32614/CRAN.package.dplyr
Wickham, H., Pedersen, T. L., & Seidel, D. (2025). scales: Scale functions for visualization (Version 1.4.0) [R package]. https://doi.org/10.32614/CRAN.package.scales
Wickham, H., Vaughan, D., & Girlich, M. (2025). tidyr: Tidy messy data (Version 1.3.2) [R package]. https://doi.org/10.32614/CRAN.package.tidyr
Xie, Y. (2025). knitr: A general-purpose package for dynamic report generation in R (Version 1.51) [R package]. https://yihui.org/knitr/
This article was developed for RMIT University’s Data Visualisation and Communication assignment. Data processing and all five interactive visualisations were completed in R. The project folder contains the R Markdown source code, styling, the original surveillance report and workbook, two CSV files transcribed from the Ministry’s report, and the map boundary files. The published HTML is self-contained and does not require these files to display the visualisations.
Generative AI acknowledgement: OpenAI ChatGPT (OpenAI, 2026) was used to assist with suggestions for refining written content and providing technical guidance on R, R Markdown, CSS, JavaScript, Plotly and Leaflet.