We Don’t Know Yet

Cyberdoc — writing on medicine since 1995

SARAWAK, 1997

In the mid-1990s, the Internet was still a novelty to many doctors. I was writing about MEDLINE, email, online journals and medical websites, convinced that this new technology would transform access to medical information.

Then came Sarawak.

On 6 June 1997, the World Health Organization reported an outbreak of “as yet, undetermined etiology.” Seventeen children, aged five months to four years, had died since 14 April. They had developed fever, sometimes neurological manifestations and poor circulation, followed by cardiogenic shock. Myocarditis was suspected. Coxsackie virus was being investigated.

There was another clue. Some children had oral lesions and rashes involving their hands and feet. Something resembling hand, foot and mouth disease — HFMD — appeared to be part of the picture.

The numbers continued to rise.

By 19 June, 26 children had died. Enteroviruses had been detected in some specimens, but further identification was awaited. By 26 June there were 29 deaths. On 3 July, WHO reported that two isolates had been identified as enterovirus 71 — EV71.

We can now compress those weeks into a paragraph. At the time, there was no such luxury.

THE QUESTIONS BEFORE THE ANSWERS

I started an online Outbreak page to collect whatever credible information I could find.

Those old pages remain interesting because they preserve the questions before hindsight supplied the answers.

Were all these deaths part of the same disease? Was this really viral myocarditis? Was Coxsackie responsible? Another enterovirus? A toxin? Could some deaths have been incorrectly classified?

In my contemporary article Information Online, I wrote:

“Information to the public should be rapid to allay fears and prevent misconception.”

I was equally concerned about doctors. Parents would naturally turn to them, and doctors needed enough information to answer their questions correctly.

But there was a difficulty that remains familiar today.

What should a doctor say when medicine itself does not yet know the answer?

“STILL GROPING IN THE DARK”

My colleague and Malaysian Medical Journal (MMJ) editor, Dr David Quek, was asking similar questions.

Writing during the outbreak, he described the public fear, conflicting reports and sometimes sensational coverage surrounding the deaths.

One newspaper characterised the authorities as “still groping in the dark.”

But was uncertainty itself evidence of failure?

An unfamiliar outbreak does not arrive with a diagnosis attached. Clinicians recognise patterns. Epidemiologists compare cases. Pathologists examine tissues. Virologists identify organisms. Hypotheses are proposed and tested.

Some survive. Others do not.

observation → hypothesis → testing → contradiction → revision → better explanation

What the public sees is the changing explanation. What it may not see is the evidence changing underneath it.

GOOD SCIENCE AND HARD FACTS

A year later I interviewed Professor Lam Sai Kit, then Head of Medical Microbiology and Senior Consultant Virologist at University Hospital, University of Malaya.

By then Taiwan was facing an outbreak disturbingly similar to Malaysia’s.

I asked him about the continuing debate over EV71, adenovirus and other possible pathogens.

“I think good science and hard facts must apply in such a situation.”

His approach was straightforward: define the clinical syndrome, examine the pathology and determine whether the laboratory findings could explain what happened.

In four fatal cases investigated at University Hospital, post-mortem examination demonstrated brainstem encephalomyelitis, with EV71 infection demonstrated. The findings were subsequently published in The Journal of Pediatrics.

This changed the interpretation of the dramatic cardiopulmonary collapse. Perhaps this was not primarily myocarditis. Perhaps severe neurological injury was producing pulmonary oedema and cardiovascular failure.

Later investigation of the Sarawak deaths supported that interpretation: cardiac tissue examined showed normal myocardium, while central nervous system tissue showed inflammatory changes, including severe brainstem pathology in the specimens available.

The explanation had evolved from apparent primary cardiac disease towards severe neurological disease with neurogenic cardiopulmonary dysfunction.

The early doctors had not been foolish to consider myocarditis. They were interpreting what they could see with the evidence then available. Later evidence allowed a better explanation.

SCIENCE IS SELDOM THAT TIDY

Hindsight can make the eventual answer appear cleaner than it really was.

Adenovirus was also isolated during the Sarawak outbreak. A subsequent Lancet report raised the possibility that EV71 was not the only agent involved. Later investigators regarded EV71 as central while acknowledging possible contributions from adenovirus, other enteroviruses or unidentified cofactors.

Prof. Lam showed similar caution when I asked whether Malaysia had discovered a new, unusually virulent EV71 strain.

“We did not claim that the EV71 isolates are new strains.”

Sequence differences had been found. Whether they were significant or explained increased virulence, he said, “remain to be seen.”

A laboratory observation had been made. Its meaning had not yet been established.

There is a large distance between those two statements — and that distance is where misinformation can easily arise.

The uncertainty was reflected in contemporary medical discussion too. In EV71 or Adenovirus? Hypothesis Arising from Two Epidemics Spanning 1997/98, T. G. Yap explored competing explanations involving EV71, adenovirus and possible viral interaction. He explicitly presented these as hypotheses rather than conclusions. Looking back, the article is valuable not because every hypothesis proved correct, but because it captures medicine trying to make sense of incomplete and sometimes conflicting evidence.

TAIWAN, 1998 — WE HAD SEEN SOMETHING LIKE THIS BEFORE

Beginning in the spring of 1998, Taiwan experienced a major epidemic of HFMD and herpangina.

Again, young children developed neurological disease, pulmonary oedema and rapidly progressive cardiopulmonary failure.

But this time Malaysia’s experience provided an important reference point.

The completed Taiwan epidemic analysis subsequently recorded 129,106 reported cases of HFMD or herpangina, 405 severe cases and 78 deaths. EV71 was strongly associated with severe and fatal disease.

The Malaysian and Taiwanese experiences, together with subsequent pathological and epidemiological work, established severe EV71 neurological disease — particularly brainstem involvement — as an important explanation for the catastrophic cardiopulmonary deterioration seen in young children.

What had been a frightening Malaysian clinical puzzle was becoming a recognisable regional disease pattern.

COULD WE HAVE DONE BETTER?

On 18 July 1998, while Taiwan was still dealing with its epidemic, David Quek wrote Outbreaks and Poor Communication: Could We Have Done Better?

He questioned whether Malaysia had adequately shared what it had learned and asked:

“Could more children have been saved?”

Nearly thirty years later, I would be cautious about answering that question. We cannot know the counterfactual, and the evidence does not allow us to conclude that better Malaysian communication would necessarily have prevented deaths in Taiwan.

But his question exposes an important distinction.

There is scientific uncertainty because the necessary evidence does not yet exist. And there is information uncertainty because useful evidence exists but has not reached the people who need it.

The first is an unavoidable part of medicine. The second, wherever possible, should not be.

AN OUTBREAK IS NOT FINISHED WHEN THE CASES STOP

I was still frustrated by this two years later.

In a June 1999 Cybermed article, Enterovirus, Haze, Nipah and Dioxin. What’s Next?, I noted that Taiwan had already published several reports from its epidemic while I was still waiting for a comprehensive account of Malaysia’s experience.

I asked:

“How can we learn if this information is not available?”

The historical record did not end there. Malaysian investigators subsequently published important clinical, pathological, virological and molecular studies, including a major 2000 analysis of the fatal Sarawak cases.

But the question I asked in 1999 still matters.

An outbreak is not finished when the cases stop.

Observations made during a crisis have to become usable knowledge — analysed, challenged, published and shared. Otherwise the next clinician, health authority or country may have to rediscover what has already been learned.

That is the bridge between outbreak response and institutional memory.

WHAT UNCERTAINTY LOOKED LIKE

The old Outbreak Guestbook provides another perspective.

Parents were asking immediate questions. Should I send my child to kindergarten? Should schools close? Is travel safe? What symptoms should I look for? How can I protect my family?

Some parents were already keeping their children indoors.

On 20 June 1997, one reader made an observation that captured the changing information landscape:

“The newspapers seem to run about three days behind the updates on the Net.”

When reliable information is incomplete, the vacuum is rapidly filled — by personal experience, rumours, newspapers, friends and, increasingly then, the Internet.

The wider story of those early online conversations is covered separately in Thirty Years of Someone Writing Back.

“ONE UP FOR THE INTERNET!”

When I asked Prof. Lam about sharing Malaysia’s outbreak experience, he replied:

“Open communication is so vital and we should learn to share our information and experiences quickly. One up for the Internet!”

For 1998, that was an important insight.

ProMED connected clinicians and scientists across continents. WHO outbreak reports could be read directly. Researchers could exchange observations without waiting months for conventional publication.

Prof. Lam recalled being contacted online by an Australian paediatrician about children who had left Sarawak with HFMD and positive EV71 isolation. During the Taiwan epidemic, correspondence reaching my website showed people thousands of kilometres away trying to understand what travel to and from Taiwan might mean for their families.

The virus crossed borders. So did the uncertainty.

FROM INFORMATION SCARCITY TO INFORMATION EXCESS

Today the information environment is almost unrecognisable.

Genomic sequences can be shared internationally within hours. Preprints circulate before peer review. Clinicians retrieve papers at the bedside.

But an unverified observation can also become a social-media post, headline, video, forwarded message or AI-generated explanation almost immediately.

The World Health Organization uses the term “infodemic” for an overabundance of information — including false or misleading information — accompanying an outbreak.

A modern outbreak therefore has two parallel trajectories.

THE DANGER IS NOT UNCERTAINTY — IT IS FALSE CERTAINTY

Return to Prof. Lam’s EV71 sequence observation.

The scientist says: We have found sequence differences. We do not yet know whether they are biologically important.

Remove the qualification and it becomes: A dangerous new mutant virus has been discovered.

The evidence has barely changed. The certainty has.

A provisional diagnosis is not misinformation. A hypothesis is not misinformation. Changing an interpretation when new evidence emerges is not misinformation.

What misleads is presenting provisional evidence as established fact — or concealing the uncertainty attached to it.

The Internet democratised access to information. It did not democratise expertise.

Artificial intelligence adds another complication: an answer may sound coherent and authoritative even when the underlying evidence is incomplete.

HOW SHOULD WE SAY “WE DON’T KNOW”?

Simply saying “We don’t know” may be scientifically honest, but it is inadequate public-health communication. People still have decisions to make.

Don’t wait for certainty to communicate. Communicate the uncertainty.

TRUST DOES NOT REQUIRE CERTAINTY

Trust should not depend upon pretending to know more than we do. It depends upon being clear about the boundary between evidence and inference.

Health authorities should communicate established information promptly and acknowledge important gaps. Scientists should resist turning preliminary findings into conclusions. Journalists should distinguish hypothesis from fact. Doctors should be prepared to tell patients when the evidence does not justify a confident answer.

And all of us should remember that a scientific explanation changing in response to better evidence does not necessarily mean science has failed. Sometimes it means science is working exactly as it should.

WHAT CHANGED?

Nearly thirty years later, EV-A71 is no longer the mysterious virus it appeared to be during those early outbreaks. Its association with severe neurological disease and cardiopulmonary complications is much better understood, while surveillance, molecular virology and vaccine development have advanced enormously.

The progress in vaccines is striking. In 2012, researchers reviewing the field were still comparing competing vaccine platforms and described formalin-inactivated EV71 vaccines as the most promising candidates then being evaluated in human clinical trials. Production of EV71 vaccine candidates documents that stage of development. By 2026, a large randomized phase III trial involving 4,011 children aged 2–71 months in Taiwan and Vietnam reported 99.2% efficacy for the inactivated EnVAX-A71 vaccine against laboratory-confirmed EV-A71-associated HFMD or herpangina. There were no EV-A71-associated hospitalizations in vaccine recipients, compared with 19 in the placebo group. Hwang et al., npj Vaccines, 2026.

There is also a Malaysian connection to the continuing research. Dr Chua Kaw Bing, who was closely involved in identifying EV71 during the 1997 Malaysian outbreak, later joined Temasek Life Sciences Laboratory (TLL) in Singapore. His laboratory has pursued the prevention, control and treatment of HFMD caused by EV-A71 and coxsackievirus A16, including experimental live-attenuated vaccine approaches. A cold-adapted, temperature-sensitive EV-A71 candidate was evaluated preclinically in monkeys, and subsequent work included attenuated and chimeric EV-A71/CVA16 candidates. These experimental approaches should be distinguished from licensed or clinically evaluated inactivated EV-A71 vaccines.

Treatment has advanced less dramatically than prevention: supportive care remains central, and there is still no licensed direct antiviral agent for severe EV-A71 infection.

The questions being asked in 1997 eventually generated better pathology, better molecular epidemiology and vaccines — but not all the unanswered questions disappeared. Science moved forward not because the first explanation was necessarily right, but because evidence continued to accumulate.

FROM CYBERMED TO KERBSIDE CONSULT

Looking back at the old pages now is revealing.

The outbreak reports record what was happening. The scientific papers tell us what was eventually discovered. But the old Cybermed pages preserve something different: the questions we were asking before we knew the answers.

In 1997, I wanted medical information to reach doctors more quickly. Parents wanted practical guidance. David Quek was worrying about communication and trust. Prof. Lam was insisting on good science, hard facts and open exchange.

The technology surrounding those conversations has changed beyond recognition. The fundamental problem has not.

THE PROBLEM WE SOLVED — AND THE ONE WE DIDN’T

In 1997, I worried that doctors could not get information quickly enough.

Nearly thirty years later, that particular problem has largely disappeared.

A genome can cross continents in seconds. A scientific paper can reach millions before it has been peer reviewed. A patient’s video can circle the world before the local health department has completed its investigation. Artificial intelligence can produce a confident explanation almost instantly.

We solved the problem of getting information quickly.

We did not solve the harder problem of knowing how much confidence that information deserves.

Perhaps that is the enduring lesson of Sarawak and Taiwan.

And good doctors must sometimes be prepared to look at an anxious patient — or an anxious public — and say:

That is not an admission of failure.

It is an acknowledgement that the evidence is incomplete — and a promise to keep looking for the answer.

SELECTED SOURCES & FURTHER READING

CONTEMPORARY OUTBREAK RECORD — 1997–1999

• Information Online — Vads Corner
• Outbreak (Enterovirus) Revisited… Taiwan — Vads Corner
• Cyber-Interview — Prof. Lam Sai Kit
• Outbreaks and Poor Communication: Could We Have Done Better? — Dr David Quek
• Enterovirus, Haze, Nipah and Dioxin. What’s Next? — June 1999
• Outbreak Guestbook — Vads Corner
• Yap TG. EV71 or Adenovirus? Hypothesis Arising from Two Epidemics Spanning 1997/98 — Vads Corner
• WHO Disease Outbreak News — Malaysia, 6 June 1997
• Thirty Years of Someone Writing Back — Kerbside Consult

SCIENTIFIC EVIDENCE AND SUBSEQUENT UNDERSTANDING

• Lum et al. Fatal enterovirus 71 encephalomyelitis. J Pediatr. 1998
• Ho et al. EV71 epidemic in Taiwan. N Engl J Med. 1999
• Cardosa et al. Adenovirus during EV71-associated HFMD outbreak. Lancet. 1999
• Chan et al. Fatal Sarawak HFMD cases. Clin Infect Dis. 2000
• WHO — Infodemic
• WHO — Communicating uncertainty in health emergencies, 2025

WHAT CHANGED — VACCINES, TREATMENT AND CURRENT UNDERSTANDING

• WHO — Enterovirus 71 vaccines: licensure, efficacy and vaccine standardization
• Chong et al. Production of EV71 vaccine candidates. Hum Vaccin Immunother. 2012
• Temasek Life Sciences Laboratory — Dr Chua Kaw Bing
• Chua et al. Safety and Immunogenicity of a Cold-Adapted EV-A71 Candidate in Monkeys. Viruses. 2021
• Chua et al. Chimeric EV-A71/CVA16 vaccine-candidate development. Virologica Sinica. 2022
• Non-Polio Enterovirus Inhibitors: Scaffolds, Targets, and Potency — antiviral development review
• Hwang et al. A randomised trial of a bioreactor-produced EV-A71 vaccine for endemic control in children. npj Vaccines. 2026