Medical certificate confirming death · SIIEASC · The RO-STEMI registry, 1997–2009
The Hour the State Loses
A doctor writes the hour of death by hand, on the certificate, every time. It never makes it into the record the family receives, let alone into a statistic.
In Romania, a medical document requires the doctor confirming a death to write, alongside the year, month and day, the exact hour it happened. That document has existed for decades, and the state has just finished digitising a full century of deaths, back to 1921. No one has ever published what that hour says, at the scale of the whole country. This page shows exactly which step of the bureaucratic chain the figure vanishes on, and what is known instead about the clock people die by.
Verified directly against official forms and reports on 2 August 2026.
8–12 · RO-STEMI, 1997–2009
The doctor writes the hour, by hand, every time
The medical certificate confirming death, CMCD for short, is the first official document in the chain that follows a person’s death. It has a field for the hour.
A begins with an external examination of the body and the confirmation of signs of actual death. The doctor then fills in the form: name, national identification number, sex, date of birth, address, then the date of death, year, month, day, hour. The hour sits there, on paper, next to the day, every single time someone dies in Romania.
It does not matter where it happened. If the person dies in hospital, the certificate falls to the pathologist, after the autopsy, or to the attending physician, if the family obtains a waiver. If the person dies at home, it is filled in by the family doctor or, if none is available, by the doctor on the ambulance crew. The route changes case by case. The field stays the same.
year ...... month .......... day ...... hour ..........
Reproduced from the form’s standard structure, with the blanks left blank.
Source: Certificat medical constatator al decesului (model de lucru)
Three steps later, the hour no longer exists
From the doctor’s hand to the record the family holds, a few days usually pass. The hour does not survive even that.
The death certificate, the civil-status record issued through , the system that replaced town halls’ paper registers, has a “Date of death” section with three boxes: year, month, day. That is all. The official template, published by the Ministry of the Interior, is printed trilingually, Romanian, French, English, so it is recognised abroad too. In none of the three languages is there a field for the hour.
The civil-status officer copies over from the CMCD what is needed to fill in the record, and the hour stays on the doctor’s paper, unused. From there also starts the , the document the county public health department codes medically and sends on to the National Institute of Statistics. I could not see that bulletin’s form directly. It was not necessary: if the hour no longer exists in the official record everything downstream depends on, it has nowhere to come back from three steps later.
The path of an hour, from certificate to statistic
- CMCD The doctor confirming the death Writes the hour by hand in the form's field, right after year, month and day. The hour exists
- The death certificate The civil-status officer, through SIIEASC The record the family receives. The official template has only year, month, day for date of death. The hour disappears here
- BSD The civil-status officer and the county public health department The bulletin that travels on to INS. It starts from a record that no longer has the hour. Form not directly verified; there is nowhere left for the hour to come from
- INS The National Institute of Statistics Aggregates by month, county, cause, age, sex. No publication contains an hour
- Tempo-Online The public What finally reaches anyone looking for the numbers. The field has never existed here
What the National Institute of Statistics publishes, through Tempo-Online and through the annual “Demographic Events” reports, confirms this from the other direction. The finest granularity is the month: how many deaths, in which county, from which coded cause, at what age, of which sex. No INS publication, checked directly, contains an hour.
Source: Ministerul Afacerilor Interne · Institutul Național de Statistică
The state digitised a century of deaths, missing hour included
SIIEASC is not a small project.
The Integrated Information System for Issuing Civil-Status Records digitised the birth, marriage, divorce and death records issued in Romania between 1 January 1921 and 31 December 2020. Nearly a full century, in a single database, operational nationwide. It is the kind of infrastructure that could, in principle, answer almost any question about Romanian mortality patterns, if that question had been asked at the moment the system was built.
It was not asked. The system digitises exactly what the records it replaces contain, and those records, as the official template shows, never had an hour field to begin with. SIIEASC lost nothing along the way. It faithfully digitised an absence that already existed on paper, decades before anyone thought of an information system.
Source: Ministerul Afacerilor Interne · Direcția Generală pentru Evidența Persoanelor
RO-STEMI, the only Romanian clock that exists
Not the hour of death. The hour the heart attack began. But it is something.
Between 1997 and 2009, the Romanian Society of Cardiology collected, from hospitals across the country, data on 19,510 patients admitted with acute ST-elevation myocardial infarction, for short, the most severe form. For 11,743 of them, doctors managed to record the exact hour the chest pain began, the onset of the heart attack. It is the only Romanian national registry in which the hour of a major cardiac event has genuinely been counted.
The distribution is not even. Nearly a quarter of onsets, 23.73%, fall in the four hours between 8 and 12 in the morning, significantly more than in any other four-hour window of the day. The least likely moment is late night and early morning, between 20:00 and 04:00, at just over 15% of cases in each of those two windows.
The hour heart attacks began, in 11,743 RO-STEMI patients
The hour of hospital admission shows something different, because it depends on city traffic and emergency department schedules, not just biology: the peak shifts to the afternoon, between 12 and 16. And what the tables do not show is the link between onset hour and who survived. The registry’s overall in-hospital mortality was 11.79%, falling from 13.21% in 2004 to 8.39% in 2009, but the report does not cross that with the hour, so it cannot be said whether a heart attack at four in the morning was, in Romania, more often fatal than one at noon.
It is the hour of a symptom, in patients who made it to hospital, between 1997 and 2009. It is not the hour of death, and it is not the whole population. It is, still, the only Romanian hour anyone bothered to count.
Source: Societatea Română de Cardiologie
Where Romania is silent, other countries have spoken
The circadian rhythm of sudden death is not a Romanian hypothesis. It is an established fact in other health systems, with other data.
Cardiologists have known for decades, since the American Framingham study, a clear pattern: sudden cardiac death peaks in the morning, in the first hours after waking. The explanation is physiological, not mysterious. On waking, blood pressure spikes, the adrenal glands release cortisol, and the sympathetic nervous system switches on all at once, after the night’s rest. For an already fragile heart, that transition is a test.
The most recent and largest study of the time of day cardiac arrest strikes comes from the United States, from the CARES registry, which included 874,415 adults with out-of-hospital cardiac arrest between 2013 and 2024, in an article published in JAMA Network Open. Arrests occurring between 23:00 and 06:59 had 16% lower adjusted odds of a neurologically favorable outcome than daytime ones, 14% lower odds of sustained return of circulation, and, among those with a pulse restored, 7% lower odds of survival. The median ambulance response interval was longer at night, 7 minutes versus 6.2 during the day, but it explained only 12.6% of the difference. The nighttime disadvantage remained constant across all twelve years studied.
United States, the CARES registry, 2013–2024: the odds, day versus night
Return of pulse
Survival without severe neurological damage
Survival to discharge
None of this is a Romanian measurement. It is the context in which the gap described in the sections above should be read: not because the world has no way of knowing what hour people die, but because other systems chose to measure it, and ours has not gotten there yet.
Source: JAMA Network Open
What it would take to know
- Why did the hour never make it onto the civil-status record?Because the standardised form, inherited from before digitisation, never included it. SIIEASC digitised the old structure, it did not redesign it.
- Could the National Institute of Statistics publish an hourly breakdown of deaths?Not from civil-status records as they are built today. It would take either a new field on the death certificate, or a direct, system-level link to the hour already written on the CMCD, which exists, with every doctor, every single time.
- What would the figure show, if it existed?Probably something close to what RO-STEMI and the international literature show: a morning peak, a nighttime trough. But “probably” is all that can be said without the real figure, and this page exists to show exactly where it went missing.
Until then, the only certain clock is the one on the certificate the doctor writes, in the room where someone just died, and that no one, ever, reads again after that.