“The best time to plant a tree was twenty years ago. The second best time is now.”— Commonly given as a Chinese proverb. I can find no primary source for that attribution, so take it as an adage of unknown origin, which does not make it less true.

Earlier this month I wrote about the haze — HAZE-OC II, published on 26 August, the sequel to a column I ran twenty years earlier, in which I had observed that nothing much had changed since 1997. Nothing much had changed since 2006 either. This piece is the other half of that argument, and it is the half I find harder to write, because the haze at least arrives on a schedule.

Climate change does not. It arrives as a hotter Tuesday, a dengue season that starts early, a flood that costs a percentage point of GDP. Individually each is deniable. Together they change what walks through the door, and they will keep changing it for as long as any of us are still working.

I want to set out three things: what the evidence now says, what our government has and has not done about it, and what — specifically — the profession could do that it is not doing.

The numbers have caught up with the rhetoric

The 2025 report of the Lancet Countdown on health and climate change — the ninth, drawing on 128 experts across 71 institutions and 57 indicators — records that 2024 was the first year in which mean annual temperatures exceeded 1.5°C above pre-industrial levels. Of the 20 indicators tracking health risks, 12 reached unprecedented levels. Heat-related deaths are up roughly 63% since the 1990s, averaging some 546,000 a year over 2012–2021.

For the region, a 2025 meta-analysis of Asia-Pacific cohort studies gives the figure worth carrying into a consultation: each 10 µg/m³ increment in long-term PM2.5 exposure is associated with a relative risk for all-cause mortality of 1.11, for cardiovascular disease 1.13, for ischaemic heart disease 1.13 and for stroke 1.12. That is not a projection. It is a dose-response relationship, measured in populations like ours.

Speaking in October 2025, the acting NRES minister set out the domestic case: the December 2021 floods displaced thousands and cost RM6.1 billion, equivalent to 0.4% of GDP; Malaysia recorded 45 heatwave warnings in 2024, up from just two in 2023; and on current trends the country could face up to 200 heatwave days a year by 2050. He also gave the figure that makes the argument uncomfortable — Malaysia contributes 0.69% of global carbon emissions and is disproportionately exposed to the consequences.

Heat, and the wrong signal

Those are the headline numbers. The clinically interesting one is smaller, and comes from Selangor.

Yong and colleagues, The Scorching Truth (Int J Environ Res Public Health 2023;20(10):5910), examined elderly hospital admissions in Selangor from 2010 to 2020. Declared heatwaves produced no significant increase in admissions among those aged 60 and over. But a 1°C rise in mean apparent temperature raised admission risk by 12.9%, with the effect delayed by nought to three days, and women were relatively more vulnerable than men.

Read that twice, because it cuts against the intuition. The categorical alert did not predict who came through the door. The continuous temperature did, several days later. A health response built on threshold warnings alone is watching the wrong variable — and watching it too late.

Dengue, and the danger of overclaiming

As at 22 August, the end of the thirty-third epidemiological week, Malaysia had recorded more than 61,000 dengue cases and 58 deaths — a 60.8% rise on the same period in 2025. The trajectory is the part worth seeing: 33,367 cases by mid-June, 42,848 by 10 July, 61,000 by 22 August. Malaysia’s entire 2025 total was around fifty thousand. We passed a full year’s burden in July, with four months still to run. The early clustering was in the Klang Valley, with Kuala Lumpur and Putrajaya up 106% and Selangor 61%; by mid-August Perlis was up 207% and Johor 109%.

Here I want to be careful, because this is exactly where a column like this overreaches — and because the Ministry has already made the point better than I would. Its own explanation is that the rise is consistent with dengue’s cyclical epidemiological pattern, under which endemic countries typically see major outbreaks every three to five years, driven by changes in population immunity, shifts in the dominant serotype, and weather conditions such as rainfall, temperature and humidity that favour Aedes breeding — alongside urbanisation, population movement and the effectiveness of vector control. Climate is one factor among several, and the Ministry says so itself.

Our epidemics have long run in those four-to-five-year cycles, and the dominant serotype has moved to DENV-3. Warmer nights shorten the extrinsic incubation period; unpredictable rainfall creates breeding sites; poorly maintained water containers do the rest.

That nuance matters. If we oversell climate as the cause of every dengue case, we will be rightly disbelieved on the things where the evidence is strong — and on heat and fine particulate, the evidence is now very strong indeed.

Meanwhile, in Parliament

Here the déjà vu becomes acute.

The National Climate Change Bill — RUUPIN — was to be tabled in September 2025. Then after Budget 2026. In January 2026 the deputy minister told the Dewan Rakyat it was expected to reach at least a first reading in that sitting. By May the expectation had moved to July, alongside a companion National Climate Change Authority Bill. Former NRES minister Nik Nazmi Nik Ahmad, who says the legislation was effectively ready before he left office in May 2025 after extensive consultation with state governments and industry, has called the continued delay a missed opportunity — noting that what businesses fear more than carbon taxes or restrictive legislation is uncertainty.

Some of that gap has a concrete institutional explanation, at least on paper. NRES’s own consultation paper on RUUPIN sets out what happens after feedback closes: a Regulatory Impact Assessment conducted under Malaysia Productivity Corporation procedures, followed by formal review at the Attorney General’s Chambers, before the Bill is ever tabled. That is a real pipeline, not an excuse invented after the fact, and it takes time by design. It does not explain a year and a half of missed dates on its own, but it is worth knowing the delay is not entirely political inertia -- part of it is a process with defined, sequential steps that nobody has published a timeline for completing.

The carbon tax has its own history of slippage. It was announced in Budget 2025, tabled in October 2024, to apply first to the iron, steel and energy sectors, and its introduction was meant to align with the European Union’s Carbon Border Adjustment Mechanism. A year later Budget 2026 confirmed it as slated for the following year. In Nik Nazmi’s summary of the pattern: we keep on delaying.

It has since delayed again, this time apparently by design rather than accident. Cabinet approved a National Carbon Market Policy, launched in April 2026, on the reasoning that credits must exist before the penalty — the market framework first, the tax afterwards, and RUUPIN after that to make the whole structure enforceable, since RUUPIN’s own consultation paper states its carbon market provisions are to follow the National Carbon Market Policy’s lead. The stated logic for the sequencing is that pricing carbon before the market infrastructure to trade it exists would land as pure cost on industry rather than incentive. Reporting at the time described the carbon tax review as being put on hold as a direct consequence of the NCMP launch.

That does not sit easily against what came two months later. By June the Deputy Finance Minister told the Dewan Rakyat the carbon tax was in its final phase of formulation, with rate, threshold and scope all still to be announced, and phased implementation so as not to burden industry amid the conflict in West Asia. A policy reported as paused in April is a hard fit with a policy described as nearly finished in June, and I have not found anything from NRES or the Ministry of Finance reconciling the two accounts. Perhaps the pause was narrower than reported, or resumed quietly, or the minister’s language did not mean what it appeared to. Either way, the sequencing may be defensible. The elapsed time, and the unexplained gap between two official accounts of it, is not.

The National Adaptation Plan, MyNAP — funded by a Green Climate Fund grant, delivered with GIZ, led by NRES with NAHRIM and MGTC — is still in development and anticipated to take effect by the end of 2026. Its five priority sectors are water and coastal resources, agriculture and food security, infrastructure and cities, forestry and biodiversity — and public health, which in the coordinator’s own account comes last of the five.

It is also worth noting what was promised. In October 2025 the acting minister told a Securities Commission conference that MyNAP would be launched in January 2026, developed with over a hundred stakeholders across government, industry and civil society — an announcement that remains on the National Disaster Management Agency’s own portal, unedited. January came and went. So did the first half of the year. Nobody appears to have said so out loud: no correction, no revised date announced as such. The language simply changed underneath the promise — by late February, Malaysia’s own Senior Climate Change Adaptation Coordinator, writing for the NAP Global Network, was describing MyNAP as anticipated to take effect by the end of 2026 instead.

Malaysia would be roughly the sixtieth country to enact a specific climate law. Sixtieth.

To be fair, I would rather the Bill be tabled well than tabled fast, and the addition of an adaptation chapter after public feedback was a real improvement. But a Bill deferred through five announced dates is not being refined. It is being avoided.

The sharpest line on all this is not mine. It is the acting minister’s own, from that same October keynote: we cannot simply pledge figures without ensuring implementation. Quite so.

A floor that is not a floor

There is a further point, which I made in HAZE-OC II and which belongs here too because it is not really about haze at all. The New Malaysia Ambient Air Quality Standard sets PM2.5 at 35 µg/m³ over 24 hours and 15 µg/m³ annually, against the WHO 2021 guideline of 15 and 5. Our annual figure is three times the WHO’s.

But as SUHAKAM records, that standard is policy guidance with no explicit relation to a legal basis. Malaysian law regulates industrial point sources; it does not regulate ambient air. So the comparison understates the problem. It is not that our legal limit is lenient. It is that there is no legal limit — only a number the Department publishes and nobody is obliged to meet. Whatever the Climate Change Bill eventually contains, a country without a binding air quality standard is measuring its climate-health performance against nothing.

What it is costed at

Malaysia’s Fourth National Communication to the UNFCCC sets out what we say we need. For improving the greenhouse gas inventory, USD 2.25 million. For the mitigation actions identified, major financial support to the iron and steel industry — USD 47 million. And then, in a single sentence: for adaptation, a preliminary estimate of USD 63.6 million is required for various initiatives to enhance resilience measures throughout the country. One line, for the whole of adaptation, in a 311-page document.

The NAP Global Network, which has assessed the plans other countries have submitted, puts the median cost of implementing a national adaptation plan at USD 2 billion. Malaysia’s estimate is a thirtieth of that. The coordinator of our own NAP process calls it fairly conservative, which is generous.

Set it against what the climate already costs us. The World Bank puts Malaysia’s annual economic losses from floods and storm surges at over USD 1.3 billion, affecting around a tenth of the land area. So the sum we have costed for adapting to climate change is under five per cent of what a single year of flooding takes out of the economy — and flooding is only one of the exposures.

Numbers of that shape usually mean the exercise was scoping rather than costing. That may well be what it was. But it is the figure we have submitted to the UNFCCC, and until it is replaced by a serious one, nobody can say what adaptation in Malaysia is supposed to cost. That is why a doctor should care what happens to a Bill: without the framework there is no plan, and without a plan, no budget worth the name.

So what has this to do with doctors?

Rather more than we have been willing to own. Five things, in ascending order of difficulty.

1. Count what actually kills people. MOH does track heat-related illness and reports the figures — 90 cases as at 23 August, of which 71 heat exhaustion, 14 heatstroke and five heat cramps, with eight of the heatstroke cases arising at the Malaysia Games in Selangor. That is genuine surveillance and credit where it is due. The last confirmed death count I can find is six weeks older than that case figure: as of 7 July, Dr Hazlina Yahaya, a Public Health Medicine Specialist, put the cumulative toll at four — two men, 42 and 22, in Penang and Kedah, and two children, aged two and four, in Negeri Sembilan and Perak. Nothing published since has updated that number against the higher August case count, so four is a floor, not a current total. Those are the directly attributed cases: the collapsed athlete, the child left in a car. What is missing is the excess — the myocardial infarctions, the renal failures, the decompensated COPD that cluster in a hot week and get coded as something else.

It is not that this cannot be done here. It has been. And the same pattern of rising counts and a handful of deaths, mostly children left in vehicles, recurs every year I checked back to 2023 — which makes this less an anomaly to report once than a predictable seasonal harm nobody treats as one.

A Klang Valley study covering 2006 to 2015 found that days at the 99th temperature percentile — 30.2°C — carried a 9% higher risk of natural mortality over the following three days, across nearly seventy thousand deaths, adjusted for particulate matter and ozone. That is the same nought-to-three-day tail Yong found in Selangor, reached by a different method. A national study of under-five deaths found no such effect in that age group. And during the pandemic, Universiti Malaya and the Ministry of Health together produced national excess all-cause mortality estimates from Department of Statistics data.

So we have the registration data, the method and the people who know how to use it. What we do not have is a current national estimate of how many Malaysians heat kills, or any attempt to produce one since the pandemic ended. The analysis was mobilised for one emergency. It has not been mobilised for this one.

1. Submit formally on RUUPIN and MyNAP. The consultation paper was issued on 4 October 2024 with a one-month window. In that time law firms, business networks and think tanks wrote in; Climate Governance Malaysia argued that adaptation should be central to the Bill rather than an afterthought, and the minister later acknowledged the extensive, even passionate discussion the draft had generated. NRES has not published a list of respondents, so I cannot say who wrote in and who did not — only that the submissions made public came from lawyers, business networks and think tanks. Did any come from a medical body? I could not find one. A wider search turns up the same names every time: the Malaysian Bar arguing for a rights-based approach, Climate Governance Malaysia, Klima Action Malaysia with three separate submissions, corporate law firms treating it as an ESG compliance matter. No medical body appears among them, in any source I can find. It is not for want of thinking: Malaysian doctors have been arguing in print since at least 2022 that the profession should take a leadership role on climate and health, embed planetary health in training, and advocate for more ambitious national policy. And the MMA’s own objects already commit it to the view that environmental degradation threatens human health and survival. The argument is made. If it was submitted, I would like to know. If it was not, MyNAP is still open.
2. Watch the right variable. Malaysian researchers have built and validated weather-based dengue forecasting models for Kelantan, Selangor and Sabah. Their own conclusion is the uncomfortable part: weather variables are yet to be fully utilised for prevention and control activities. The models exist; the operational system does not. The same is true of heat. The Selangor finding says the categorical alert is a blunt instrument and the admissions arrive on a lag. A useful heat-health alert would track apparent temperature continuously and warn clinics of a rise before the elderly patient on diuretics presents three days later. The dengue hotspot map on iDengue shows what an operational tool looks like; there is no equivalent for heat.
3. Our own footprint, and what to make of MyReHF. Hospitals are energy- and waste-intensive institutions and the health sector is a measurable emitter in its own right — anaesthetic gases and single-use plastics being the obvious starting points. Something is now under way, and it turns out to be further along than a single pilot. A four-day workshop from 23 to 26 June 2025, led by the Ministry of Health with WHO international best practices as guidance, field-tested a consolidated draft toolkit — Malaysia Resilient Healthcare Facilities, MyReHF — at four sites: Hospital Serdang, Hospital Putrajaya, Klinik Kesihatan Putrajaya and Klinik Kesihatan Nilai. A further round in April 2026, this time described by WHO as its own pilot of a toolkit it calls CLEAR+ (Climate and Emergency Adapted for Readiness), tested the Malaysian adaptation at two more sites, Hospital Banting in Selangor and Klinik Kesihatan Seremban 2 in Negeri Sembilan. Between the two accounts, six facilities across four states are confirmed assessed. I cannot tell from either source whether the two rounds tested the same instrument or a revised one, or who is actually driving this — MOH describes itself as leading with WHO guidance; WHO describes Malaysia as piloting its toolkit. Both may be true. Neither source gives a roll-out timeline, and I would ask my own hospital or clinic whether it has heard of MyReHF before assuming the six sites are the only ones. Two of CLEAR+’s three named component tools are themselves properly documented, if you want to see what “assessing readiness” actually involves. WHO’s guidance for climate-resilient and environmentally sustainable health care facilities (2020, 92 pages) sets out 24 intervention tables across four areas — health workforce; water, sanitation and waste; energy; infrastructure and technology — each downloadable individually. Its companion checklists to assess vulnerabilities in health care facilities (2021) does the harder work: seven hazard-specific checklists — floods, storms, sea-level rise, drought, extreme heat, extreme cold, wildfires — each cross-tabulated against the same four areas, asking concrete questions such as whether a facility has emergency backup generation able to cover all critical service areas during a cold-wave event. That is the actual instrument. What I cannot find published anywhere is the merged CLEAR+ tool itself, or a scoring method that would let two facilities, or two countries, compare a result.
4. Put it in the curriculum. Climate and health belongs in training, not in the special-interest folder. My generation trained without it. The house officers on the wards today will practise their entire careers inside it — and I would be interested to know how many have had any formal teaching on the subject. I suspect the answer is few, but I do not know it, and somebody should.

In short

The evidence is no longer the weak link. We have dose-response figures for particulate mortality in populations like ours, two Malaysian studies agreeing on which temperature signal to watch and how long it takes to arrive, and a national adaptation plan that lists public health last among its five priorities.

What is missing is not information. It is follow-through, at every level this piece has looked at. A Bill through five tabling dates, now waiting on a Regulatory Impact Assessment and Attorney-General’s review nobody has put a date on. A carbon market policy launched in April and a carbon tax described as nearly finished in June, with no account of how both can be true. A climate-resilience toolkit tested twice, at six facilities, with no public word on what happens next.

None of this moves by itself. It moves when someone writes the submission, asks the RIA’s own timeline out loud, or simply notices that six months of monsoon-season deaths keep arriving in the same shape every year and stops treating each one as news. That is not a large ask. It has just not yet been made.

The tree, as the adage has it, should have gone in twenty years ago. The second best time is the sitting after next.

On the haze specifically, see HAZE-OC II; the standing resource page, with monitoring links and the peer-reviewed evidence, is at Vads Corner: Haze.