During Greece’s second wave of the pandemic, seven out of ten COVID-19 related deaths occurred outside the ICU according to data first published by the iMEdD Lab, while an analysis of ELSTAT data shows a 7% increase in mortality for 2020 compared to the statistically expected mortality for the same year.
An estimated 8300 additional deaths than statistically expected were recorded in 2020, with the second wave of the pandemic seeing the largest increase in statistically unexpected mortality from any cause. In fact, the majority of COVID-19 patients who died in the hospital at the peak of the wave were outside the ICU and died faster compared to patients who succumbed to the disease during the following weeks.
According to an analysis of data on weekly deaths from any cause for the year 2020 published by the Hellenic Statistical Authority (ELSTAT), the mortality directly or indirectly associated with COVID-19 is higher than believed.
Between 23 November and 6 December 2020, a 41% increase in excess deaths is estimated to have occurred than the statistically expected number of deaths.
Excess mortality is the most reliable and impartial indicator for monitoring COVID-19-related deaths, given how it helps us address an important problem: deaths indirectly attributable to the coronavirus go unrecorded. International scientific studies suggest the pandemic is associated with an increase in mortality from causes other than those directly linked to the virus. For example, patients suffering from various diseases may neglect their treatment during the lockdown, heart attack patients may avoid emergency care for fear of contracting the coronavirus, while overloaded health systems cancel surgeries or prioritize COVID-19 cases, resulting in non-COVID-19 patients not receiving timely treatment in many instances.
In 2020, Greece recorded a rate of 7% (1st – 52nd week) in excess mortality. Of the approximately 8300 statistically unexpected deaths, it is estimated that 61% are confirmed COVID-19 deaths. The remaining ones are presumed to be indirectly related to the pandemic.
Seven out of ten deaths during the second wave of the pandemic were recorded outside the ICU.
To estimate the statistically expected deaths for 2020, we calculated the average weekly deaths that occurred between 2015 and 2019. We then compared the recorded deaths for 2020 with the statistically expected ones. Of the 52 weeks (the 53rd week is not included in the analysis, as there were 52 weeks in the previous four years), 15 saw an excess mortality of more than 10%. At the peak of the second wave, from 16 November to 6 December, the highest rates were recorded, reaching 35% during the 47th week (16/11/2020-22/11/2020) and 41% between 23 November and 6 December. An excess mortality of 41% means that – assuming the statistically expected deaths amounted to 100 – 141 people ended up dying. As far as the excess deaths between 23 November and 6 December are concerned, it is estimated that about seven out of ten are confirmed COVID-19 deaths, with the remaining three being indirectly related to the pandemic.
The regions with the highest excess mortality rates are Central Macedonia, Eastern Macedonia and Thrace, Western Macedonia, and Thessaly. In Central Macedonia, 7 out of 10 deaths are directly related to COVID-19, while in Attica the rate is 45.3%. In contrast, Epirus and the Ionian Islands, which are less affected by the pandemic compared to other regions, had a negative excess mortality rate – i.e. fewer deaths than statistically expected.
| Region | Avg 2015-2019 | 2020 | Excess mortality (%) | Estimated deaths COVID-19 |
|---|---|---|---|---|
| Attica | 40402 | 42483 | 5.2 | 943 |
| Eastern Macedonia and Thrace | 7497 | 8638 | 15.2 | 557 |
| Central Macedonia | 21240 | 24757 | 16.6 | 2276 |
| Western Greece | 3337 | 3781 | 13.3 | 216 |
| Thessaly | 7750 | 9683 | 24.9 | 470 |
| Epirus | 5002 | 4205 | -15.9 | 64 |
| Ionian Islands | 2647 | 2637 | -0.4 | 20 |
| Western Macedonia | 7765 | 8019 | 3.3 | 76 |
| Central Greece | 6674 | 6793 | 1.8 | 66 |
| Peloponnese | 7420 | 7530 | 1.5 | 38 |
| North Aegean | 2419 | 2425 | 0.2 | 44 |
| South Aegean | 2831 | 2953 | 4.3 | 13 |
| Crete | 6184 | 6328 | 2.3 | 31 |
Deaths outside ICUs and increased clusters of cases in workplaces
Focusing on the second wave of the pandemic, the data first published by the iMEdD Lab (concerning the whole country, for the period between 12 October 2020 and 3 January 2021), paint a more complete picture of the situation in the course of those dramatic months, during which the coronavirus claimed more than 4 200 lives.
Elderly patients were dying fast, before ICU admission
According to data, we obtained from EODY’s weekly reports, which are sent to the Ministry of Health for the briefing of the Committee of Health Experts, almost seven out of ten patients (2954) who died of COVID-19 were being treated outside the ICU, while 1322 deaths of intubated patients were recorded inside the ICU, i.e. 30.9% out of a total of 4276 deaths during that period.
The median age of patients who died outside the ICU is 82.8 years (average of the median ages for weeks 43-53, i.e. from 19 October 2020 to 3 January 2021), while the median age of patients who died inside the ICU is 70.7 years. Regarding the deaths of ICU-admitted patients, we observe that, while the median age was 79 years during week 42 (12-18 October), when Greece was nearing the peak of the second wave (week 49-50, i.e. 30 November-13 December), the median age decreased to 67 years. During the same period, however, the median age of those who died outside the ICU ranged from 85 to 82 years.
Majority of new cases detected in workplaces
Looking at the time from hospital admission to death, we see that, during the first few weeks, the majority of patients would die much faster than patients who died throughout the last few weeks. In particular, during the 46th week (9-15 November), when 316 new deaths were recorded, the median number of days from admission to death was 4 days, while the mean was 10. This difference suggests that there were many patients who died very quickly, while fewer were hospitalized for a longer period of time. “Most of the deaths in Northern Greece, at 75%, occurred outside the ICU (…) patients were going to the hospital very late, they were in a very serious condition and did not manage to make it to the unit”, Anastasia Kotanidou, President of the Hellenic Society of Intensive Care, had stated in an iMEdD Lab article on Greece’s National Health System’s preparedness/response to the pandemic and nursing shortages. According to the data available to the iMEdD Lab, during the 47th week, when 536 new deaths were recorded, the average duration from admission to death was reduced to 8 days, while the median number of days remained the same. This potentially means that there was an increase in the number of patients who died quickly. From the following week until the end of the year, the time from admission to death gradually increased – at week 53 it amounted to 13 days (median).
Nikos Kapravelos, Department Manager of the ICU and Coordinating Director of the 2nd ICU at Papanikolaou Hospital in Thessaloniki, noted in the same iMEdD Lab article that “these people who came in, were very old and this is the variant between the second and the third wave: the patients were very old, which means that you couldn’t afford a single hour to support the patient, to find an empty intensive care bed –old age, rapid death. The deterioration was very rapid. Obviously, there was a lack in ICUs because the needs were far greater than the capabilities we had. This is well-known. Therefore, many were not given the opportunity to go through an ICU.”
The health system’s pressure is also reflected in the data available to the iMEdD Lab: in the 46th week, there were 2.45 intubations for every extubation. In the following week, the number of extubations almost doubled (90% increase), in order for the system to respond to the imperative need for new intubations. “Also, many left the ICU and died in the department: the cases were so serious and such was the pressure put on us for someone who was a little better to leave (editor’s note: the ICU)… However, the conditions were not such (editor’s note: to move to an acute care bed). They had to go to an IMCU, which does not exist in Papanikolaou. It helped us a lot that some of these cases were diverted to a private center that, after a deal with the state, was taking such cases for intermediate care. And many people were saved that way,” Kapravelos noted. By the end of the year, the intubations-extubations ratio had gradually been reversed.
| Week | Date | Intubation | Extubation | Avg days | Median days | New deaths | Avg days | Median days | Deaths inside ICUs | Deaths outside ICUs | Age inside ICUs | Age outside ICUs |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ISO week | Start of the week | New intubations | New extubations | From intubation to extubation | From intubation to extubation | From hospitalization to death | From hospitalization to death | Median age of deaths | Median age of deaths | |||
| 42 | 12/10/20 | 61 | 28 | 33 | 79 | 85 | ||||||
| 43 | 19/10/20 | 63 | 31 | 32 | 77 | 84 | ||||||
| 44 | 26/10/20 | 57 | 25 | 32 | 69 | 82.5 | ||||||
| 45 | 02/11/20 | 161 | 36 | 125 | 72 | 82 | ||||||
| 46 | 09/11/20 | 277 | 113 | 10 | 8 | 316 | 10 | 4 | 69 | 247 | 70 | 82 |
| 47 | 16/11/20 | 351 | 215 | 11 | 9 | 536 | 8 | 4 | 110 | 426 | 69 | 82 |
| 48 | 23/11/20 | 308 | 272 | 10 | 652 | 8 | 6 | 163 | 489 | 67 | 82 | |
| 49 | 30/11/20 | 282 | 295 | 10 | 664 | 10 | 8 | 183 | 481 | 67 | 82 | |
| 50 | 07/12/20 | 238 | 295 | 14 | 12 | 576 | 13 | 9 | 202 | 374 | 67 | 83 |
| 51 | 14/12/20 | 196 | 254 | 14 | 12 | 500 | 13 | 10 | 175 | 325 | 69 | 82 |
| 52 | 21/12/20 | 173 | 224 | 14 | 12 | 385 | 15 | 14 | 164 | 221 | 70 | 84 |
| 53 | 28/12/20 | 124 | 173 | 17 | 16 | 305 | 16 | 13 | 136 | 169 | 72 | 83 |
Majority of new COVID-19 clusters linked to workplaces
In terms of outbreaks by setting type, it appears that new cases detected in the workplace far exceed those associated with other types of settings, such as closed facilities, educational institutions, or health care facilities. In particular, from 2 November to the end of the year, 758 new cases were detected (this figure does not reflect the total number of cases in the above-mentioned environments), 434 of which were recorded in work settings, such as industries and private entities. When it comes to closed settings (nursing homes, mental hospitals, camps, rehabilitation centers, etc.), 166 new cases were detected.
Regarding the correlation between new cases and the total number of cases, it appears that there may have been a greater spread of the virus in closed facilities, resulting in 2476 cases out of a total of 4769 reported cases. Of the 2476 cases in closed facilities, 1155 (46%) were detected in care homes, nursing homes, and rehabilitation centers. Although countries such as the UK, record, and report on mortality in closed facilities, Greece keeps no registry of people in closed facilities who eventually succumb to the disease.
According to Eurostat, between March and December 2020, an estimated 580 000 additional people than statistically expected died in Europe, with April and November recording the highest excess mortality rates (25.1% and 40.7% respectively).
Compared to other European countries, Greece has had a relatively lower (but not negligible) excess mortality rate, with about four out of ten statistically unexpected deaths being indirectly related to COVID-19.
However, the third wave of the pandemic has taken a toll on Greece, with public hospitals being under extreme pressure and experiencing staff shortages. The European platform EUROMOMO, which monitors mortality from any cause in 27 European Union (EU) member states, has recorded a slight increase in excess mortality in Greece in recent weeks, even though the rate does not exceed the normal range. Various reports in the press suggest dozens of patients remain outside the ICU, waiting for a bed to be emptied, while healthcare staff is fighting a daily battle, under adverse conditions, to keep these patients alive. Unfortunately, fourteen months into the pandemic, EODY has yet to disclose any official detailed data, preventing us from getting a clear picture of the situation during the third wave of the pandemic.