Thinking · Evidence
Dengue in India 2026: the death count measures the counting before it measures access.
TL;DR
- India recorded 23,952 dengue cases and 31 deaths between January and June 2026. Five southern states carried 63.2 per cent of the cases.
- Tamil Nadu had the most cases and three deaths. Kerala had fewer cases and 19 deaths. The obvious reading is that Kerala's care chain is weaker. The evidence says otherwise.
- India keeps two dengue death registers, and they disagree by a factor of nearly seven nationally. Kerala is one of the few states where surveillance counts more deaths than death certificates do.
- For a CSR or duty-of-care programme, the lesson is practical: never choose districts off surveillance deaths, and measure the chain you control instead.
The dengue India 2026 climate story arrived in July as a table in a parliamentary answer. Between January and June, India recorded 23,952 dengue cases and 31 deaths, according to Health Ministry figures tabled in the Rajya Sabha on 28 July and reported by The South First. Tamil Nadu led the country with 7,511 cases and three deaths. Kerala followed with 4,458 cases and 19 deaths, more than six times Tamil Nadu's toll from a little over half its caseload.
I nearly wrote the piece that table invites. Kerala's patients reach care too late, the fatality gap is an access gap, here is the completion spine in one dataset. It is a clean story. It is also the wrong one, and the reason it is wrong matters more for anyone deciding where health money goes than the story would have.
Southern states, January to June 2026
| State | Cases | Deaths | Deaths per 1,000 cases (derived) |
|---|---|---|---|
| Tamil Nadu | 7,511 | 3 | 0.40 |
| Kerala | 4,458 | 19 | 4.26 |
| Karnataka | 1,601 | 0 | 0 |
| Andhra Pradesh | 968 | 0 | 0 |
| Telangana | 599 | 0 | 0 |
Chart reads from the table above. Three states reported no deaths at all from over 3,100 cases between them.
Look at the bottom three rows before the top two. Karnataka, Andhra Pradesh and Telangana reported 3,168 cases between them and not one death. That is possible. It is not the pattern a reader should accept without asking how deaths get counted there.
India keeps two dengue death registers
The numbers in the parliamentary answer come from surveillance: states report cases and deaths to the National Centre for Vector Borne Diseases Control. A second register exists. Every death certified by a doctor is recorded, with its cause, in the Registrar General of India's annual report on medical certification of cause of death.
In December 2025, The South First set the two side by side for 2023. Surveillance recorded 485 dengue deaths from 2,89,235 cases that year. Medical certification recorded 3,254. On certified deaths, the national fatality rate rises from 0.17 per cent to about 1.13 per cent.
The same year, two registers
| State | Surveillance deaths | Certified deaths |
|---|---|---|
| Delhi | 19 | 538 |
| Maharashtra | 55 | 487 |
| Karnataka | 11 | 243 |
| Tamil Nadu | 12 | 94 |
| Andhra Pradesh | 0 | 90 |
| Kerala | 153 | 40 |
Chart reads from the table above. Kerala is the only state shown where surveillance counts more deaths than death certificates.
Now reread the 2026 table. Karnataka's zero sits beside 243 certified deaths three years earlier. Tamil Nadu's surveillance counted one death for every eight on certificates. Andhra Pradesh counted none against 90. Kerala went the other way, with its surveillance recording nearly four times as many deaths as its certificates did, which The South First reads as an active surveillance system catching deaths the formal certification process misses.
The pattern is not new. In December 2025 the Health Ministry told the Rajya Sabha that India recorded 1,13,440 cases and 94 deaths up to November 2025, with Tamil Nadu at 20,866 cases and 12 deaths, and Kerala at 10,239 cases and 49 deaths. Same shape, different year. A stable ratio across years is what you would expect from a stable difference in counting.
A state that finds its deaths looks deadlier than a state that loses them. Rank districts on surveillance deaths and you fund the ones that stopped counting.
The climate part is real, and it sits underneath
None of this makes dengue a smaller problem. The Lancet Countdown's 2025 India data sheet reports that the average basic reproduction number for dengue transmission by Aedes albopictus rose from 0.86 in 1951-1960 to 1.60 in 2015-2024. Below one, an outbreak burns out. Above one, each infection seeds more than one more. India crossed that line on transmission suitability over the same decades its heat and rainfall patterns shifted.
So transmission conditions have worsened, the southern states carry most of the reported load, and the count of who dies of it is unreliable in a direction that differs by state. That combination is exactly where a programme can do harm with good intentions: aim at the loudest number, and the loudest number is partly an artefact.
Where the chain breaks, and which chain
Figure 1 · Two chains, one patient
A published fatality rate is the product of both chains. Only the first is about access.
Dengue kills through a specific window. A patient who is tested early and whose warning signs are watched through the critical days can be admitted before they deteriorate. The access failures that matter sit in the middle of that chain: the days between fever and first test, and whether anyone is watching during the days when it turns. Rain that cuts a road, heat that keeps a daily-wage worker on site through a fever, a clinic that closes its outreach in the season it is needed most. These are climate-access failures in the exact sense of the definition: reachability and continuity breaking while availability looks fine on paper.
The count chain is different. A death reaches a certificate or it does not. A certified death reaches surveillance or it does not. Neither step changes what happened to the patient. Both change the table a CSR committee reads.
What this means if you fund or employ in the south
Choosing and measuring, a checklist
- SelectionRead reported cases, not surveillance deaths, as the burden signal, and set certified deaths beside them where the district data exists. A zero in the surveillance column is a question, not a finding.
- BaselineBefore the season, record where people in the target population go first when they have a fever, and how far that is in the monsoon.
- MeasureDays from fever onset to first test, for every case the programme touches. It is the one number that moves when access improves, and the programme can collect it itself.
- MonitorWhether warning signs were checked during the critical days, and by whom. A test with no follow-up is enrolment, not care.
- ReportOutcomes against your own chain, not against the state's fatality rate. A programme cannot claim credit for, or be blamed for, a number set by the count chain.
The same logic applies to an employer. A plant or field workforce in Tamil Nadu, Karnataka or Kerala meets dengue season as an absence spike that nobody links to a cause. A fever protocol with a same-day test route and a named person watching the days after is a small, specific duty-of-care measure, and it produces the days-to-test figure without a survey. The legal side of workforce health duties under the new labour code is set out in the workplace heat duty of care, and the rules for booking a vector-borne disease programme as CSR in Section 135 and climate-health.
One limit. The 2026 half-year figures come from reporting of the parliamentary answer, and the two-register comparison from a newspaper's compilation of two official datasets. Each is specific, dated and consistent with the Ministry's earlier answer. A programme using them for allocation should pull the Registrar General's report and the state's own surveillance tables directly, district by district, before committing money.
Common questions
Why is Kerala's dengue death rate higher than Tamil Nadu's?
Kerala records more dengue deaths per reported case than any other state, but the difference is mostly in counting, not in care. In 2023, Kerala's surveillance system logged 153 dengue deaths against 40 in the Registrar General's medical certification data, while Tamil Nadu logged 12 in surveillance against 94 certified. Kerala's surveillance appears to capture deaths other states miss. A higher recorded fatality rate there is evidence of a stronger counting system before it is evidence of weaker access.
How many dengue cases and deaths has India recorded in 2026?
Between January and June 2026, India recorded 23,952 dengue cases and 31 deaths, according to Health Ministry figures tabled in the Rajya Sabha on 28 July 2026. Five southern states, Tamil Nadu, Kerala, Karnataka, Andhra Pradesh and Telangana, accounted for 15,137 cases, or 63.2 per cent of the national total.
Is climate change making dengue worse in India?
The transmission conditions are measurably worse. The Lancet Countdown's 2025 India data sheet reports that the average basic reproduction number for dengue transmission by Aedes albopictus mosquitoes rose from 0.86 in 1951-1960 to 1.60 in 2015-2024, crossing from below one to above it. Above one, each infection can be expected to cause more than one further infection.
How should a CSR programme choose where to work on dengue?
Not from surveillance death counts alone, because they rank states partly by how well they count. Read reported cases alongside medically certified deaths, and measure the programme's own chain: days from fever to first test, whether warning signs are monitored, and whether admission happens before a patient deteriorates. Those are numbers a programme controls and can report.
Where to start
The free climate-access exposure assessment gives a directional read on where climate is already reaching your access, workforce and supply chain, benchmarked against FY2024-25 BRSR disclosures from 59 listed Indian companies. Under three minutes, no sign-up.
Test your exposure →The position, for citation
- India recorded 23,952 dengue cases and 31 deaths between January and June 2026, with five southern states accounting for 63.2 per cent of cases (Health Ministry, Rajya Sabha, 28 July 2026).
- In 2023, India's surveillance system recorded 485 dengue deaths while medical certification recorded 3,254. State-level differences between the two registers run in both directions.
- Kerala's higher recorded dengue fatality rate is better read as a sign of more complete death capture in surveillance than as evidence of weaker access to care.
- Syntropy Earth's position: surveillance death counts should not be used on their own to select districts for CSR or duty-of-care health programmes. Programmes should measure days from fever to first test instead.
Sources
- The South First, "South India dominates India's dengue burden; Tamil Nadu reports most cases, Kerala most deaths", 31 July 2026, reporting Health Ministry figures tabled in the Rajya Sabha on 28 July 2026. thesouthfirst.com
- Sumit Jha, "Government records show 571% more dengue deaths than Health Ministry data", The South First, 27 December 2025, comparing NCVBDC surveillance with the Registrar General's Report on Medical Certification of Cause of Death 2023. thesouthfirst.com
- Office of the Registrar General of India, Report on Medical Certification of Cause of Death 2023. censusindia.gov.in
- Health Ministry written reply in the Rajya Sabha, 16 December 2025, on dengue cases and deaths in 2025, as reported by IANS. ianslive.in
- The Lancet Countdown on Health and Climate Change, 2025 Report: India Data Sheet, Indicator 1.3.1. lancetcountdown.org
- National Center for Vector Borne Diseases Control, dengue situation in India. ncvbdc.mohfw.gov.in
This page reads published data. It is not clinical guidance; dengue case management follows national treatment guidelines and the advice of a qualified doctor.
Last updated: 21 September 2026