Abstract
People living in prisons have higher mortality rates compared to the general population. We undertook a retrospective analysis of deaths recorded between 2010 and 2018 at the sole prison hospital in Greece (Korydallos Prison Special Health Centre for men) to assess the causes of death overall and by type of offence (drug-related or other), sociodemographic characteristics by cause of death, and mortality trends over time. Permission to access forensic reports and criminal files was obtained from the relevant authorities. Deaths were categorized as either non-natural (drug overdose, suicide, and homicide) or natural (cardiovascular disease, cancer, and others). Between 2010 and 2018, 236 deaths were reported; 80.9% were natural deaths, and 19.1% were non-natural deaths. The primary causes of death were circulatory disease (34.7%), cancer (17.8%), suicide (10.2%), respiratory disease (8.9%), and overdose (6.4%). Suicide and overdose accounted for 53.3% and 33.3% of non-natural deaths, respectively. The mean (SD) age at death was 52.4 (16.2) years, with individuals experiencing non-natural deaths being significantly younger than those experiencing natural deaths [39.1 (10.5) vs. 55.5 (15.7), p < 0.001]. Among individuals incarcerated for drug-related offences, 23.8% died from non-natural causes, with drug overdose accounting for 60% of non-natural deaths. A significant peak in mortality was observed in 2013. This study emphasizes the need to closely monitor mortality rates, including drug-related fatalities, to implement suicide prevention training as well as measures to prevent deaths by overdose, including comprehensive harm reduction strategies, overdose education, and naloxone training.
Introduction
The global prison population in 2023 was estimated to be 11.5 million, with the United States, China, Brazil, India, and Russia having the highest rates of incarceration. 1 Compared with the wider community, people living in prisons (PLIP) experience a higher burden of infectious and chronic diseases, mental health problems, substance use issues, and mortality.2–5 Historically, suicide has been the leading cause of death in prisons 6 and remained the most common cause reported in 35 countries within the WHO European Region in 2020. 7 People who use drugs are disproportionately represented in prison populations, 8 and drug overdose is another frequent cause of death, ranking third among causes in 33 European countries reporting to the WHO in 2020. 7
Among 25 EU countries with available data in 2022, Greece ranked 14th in terms of the prisoner rate, with 97.3 prisoners per 100,000 inhabitants. 9 The share of PLIP with foreign citizenship varies considerably across EU member states, with Greece holding the second-highest percentage (57.2%) after Luxembourg. 10 In Greece, there are approximately 10,000 to 12,000 people detained across 35 facilities (11,364 in 2010 and 10,371 in 2024), with occupancy exceeding 100% in several of them. 11 There is only one prison hospital for male detainees, the Korydallos Prison Special Health Centre, located in Athens next to the largest Greek prison (Korydallos prison).
The proportion of individuals incarcerated in Greece for drug offences has significantly declined, dropping from 38.4% in January 2010 to 18.3% in January 2024. 11 This decrease is attributed to a law enacted in 2013 that distinguishes between drug possession for personal use and commercial use, establishing more lenient sanctions. According to data from a large intervention implemented among community-recruited people who inject drugs in Athens, Greece (2012–2013), 50.5% of men and 39.2% of women reported a history of imprisonment. 12 Notably, a large HIV outbreak occurred among people who inject drugs in Athens in 2011, with the prevalence of HIV increasing to 16.5% by 2013. 13
The mortality rate in Greek prisons ranged from 307.3 to 606.4 deaths per 100,000 PLIP between 2013 and 2020.7,14 In 2020, 66.7% of recorded deaths were attributed to natural causes, while suicides accounted for 13.0%. 7 However, there are no detailed published data regarding mortality rates in Greek prisons and the characteristics of deceased individuals. This study aims to utilize data from the only prison hospital in the country over the period from 2010 to 2018 to assess: (1) Specific causes of death; (2) sociodemographic characteristics of deceased individuals by cause of death; (3) causes of death according to the type of offence (drug-related or other); and (4) trends in mortality over time.
Methods
Study design and data collection
We retrospectively collected data using forensic reports for all deaths recorded at the Korydallos Special Health Centre between 2010 and 2018. This facility is the sole prison hospital in Greece. According to Greek legislation, only male PLIP who have acute or chronic illnesses are admitted to this facility. These individuals may come from prisons located in Athens or from other correctional institutions throughout Greece. The institutional protocol mandates that autopsies be performed for all prisoners who die in the prison hospital.
Information on the cause of death, age at death, country of origin, and type of offence (drug-related or other) was collected from forensic reports, medical records, and criminal files.
Causes of death
We analysed the causes of all deaths recorded at the Korydallos Prison Special Health Centre during the study period. The primary cause of death was determined on the basis of autopsy findings, review of terminal medical events, and toxicology results when applicable. We defined comorbidities as chronic conditions documented in medical records that required ongoing treatment and were distinct from the terminal condition. These included cardiovascular diseases, diabetes, chronic viral infections (HIV, viral hepatitis), respiratory conditions, and mental health disorders. The presence of comorbidities was verified through medical documentation of the diagnosis and specialist consultations.
We categorized all causes of death as either natural or non-natural, as described in Table S1. Natural deaths result from medical conditions or diseases and were defined as ICD-10 chapters A through Q (including cardiovascular disease, infections, cancers, etc.). Non-natural deaths are caused by external factors such as violence, accidents, or substance use and were defined as ICD-10 chapters V through Y (including homicide, suicide, accidental causes, and drug-related deaths). 15 This classification follows coding conventions for mortality statistics, which distinguishes between natural causes (disease processes) and non-natural mechanisms of death (external causes).
Calculation of mortality rates
Obtaining mortality rates in this setting is not straightforward. Τhe high degree of churn of inmates and the lack of data on the precise entry and exit dates (or average length of stay) do not allow estimation of the population at risk. Thus, we calculated annual mortality rates using two approaches to determine the denominator. In the first approach, we divided the number of deaths in a year by the total number of inmates entering the facility that year (ranging from 957 to 1624 receptions per year; data obtained from the Ministry of Justice). Since this method tends to overestimate person-time at risk, we also calculated annual mortality rates using the facility's census population on January 1 of the respective year as the denominator.
Statistical analysis
We described the number and cause of death by year, as well as the sociodemographic characteristics of the deceased individuals, using counts and proportions or means and standard deviations as appropriate. We assessed differences in the characteristics of the deceased persons by cause of death (natural vs. non-natural) and type of offence using chi-squared test or t-test, as appropriate.
Ethics
The study protocol was approved by the Institutional Review Board of the Medical School of the National and Kapodistrian University of Athens (Protocol number/Date: 573/19 November 2021). All procedures were carried out in accordance with ethical guidelines and regulations to safeguard the privacy and dignity of individuals. Permission to access these records was granted by both the Director of the facility and the Ministry of Justice. The first author visited the facility to collect and record the data on-site. The information gathered was then recorded anonymously in an Excel file using a unique participant code.
Results
Number/type of deaths and characteristics of the deceased persons
Between 2010 and 2018, 236 deaths occurred at the Korydallos Special Health Centre for male prisoners. Among those deaths, 191 (80.9%) were classified as natural deaths, whereas 45 (19.1%) were classified as non-natural deaths. The annual number of deaths is depicted in Figure 1.

Number of deaths recorded in the prison hospital for male detainees in Greece (Korydallos prison special health centre) from 2010 to 2018, by year and type of death (natural or non-natural).
The mean (SD) age at death was 52.4 (16.2) years, with individuals experiencing non-natural deaths being significantly younger than those experiencing natural death [39.1 (10.5) vs. 55.5 (15.7), p < 0.001] (Table 1). Most deaths occurred between 2013 and 2015 (40.3%), followed by the periods from 2010 to 2012 (30.1%) and from 2016 to 2018 (29.7%), with similar proportions of natural and non-natural deaths (p = 0.247). The ethnic composition of the deceased was predominantly Greek (73.3%), and there was no statistically significant difference in the type of death by ethnicity. Comorbidities were present in 55.9% of the total deceased individuals, with a significantly higher prevalence in natural deaths compared to non-natural deaths (67.5% vs. 6.7%, respectively, p < 0.001) (Table 1).
Characteristics of deceased individuals overall and by type of death (Korydallos prison special health centre, Greece, 2010–2018).
Student's t-test.
Pearson's chi-square test.
Causes of death
Table 2 presents the causes and mechanisms of death according to type (natural vs. non-natural). With respect to natural deaths, circulatory system-related deaths were the most prevalent, accounting for 82 cases (42.9% of natural deaths and 34.7% of all deaths). Within this category, acute myocardial infarction was the leading cause, with 60 cases (73.2% within the circulatory category and 25.4% of all deaths).
Causes and mechanisms of deaths recorded in Korydallos prison special health centre, Greece (2010–2018).
Cancer-related deaths constituted 42 cases (22.0% of natural deaths and 17.8% of all deaths). Lung cancer was the most common type, with 17 cases (40.5% within the cancer category and 7.2% of total deaths), followed by generalized cancer with 5 cases (11.9% within the cancer category and 2.1% of total deaths). Other cancers included brain and bladder cancer (4 cases each), liver cancer (3 cases), and non-Hodgkin lymphoma (2 cases), as well as leukaemia, prostate cancer, myelodysplastic syndrome, nasopharyngeal cancer, oesophageal cancer, penile cancer, and stomach cancer (1 case each).
Respiratory-related deaths were reported in 21 cases (11.0% of natural deaths and 8.9% of all deaths). Respiratory tract infection was the most frequent cause within this category, with 11 cases (52.4% within the respiratory category and 4.7% of all deaths), followed by pneumonitis due to solids and liquids with 6 cases (28.6% within the respiratory category and 2.5% of all deaths).
Digestive system-related deaths were recorded in 11 cases (5.8% of natural deaths and 4.7% of all deaths). Gastrointestinal haemorrhage was the leading cause in this category, with three cases (27.3% within the digestive category and 1.3% of all deaths): one due to bleeding oesophageal varices in the context of liver cirrhosis, one from a gastric ulcer haemorrhage in a patient with hepatitis B and C with associated thrombocytopenia and coagulation disorders, and one from haemorrhagic gastritis attributed to non-steroidal anti-inflammatory drugs abuse. Hepatic encephalopathy, acute liver failure, cirrhosis of the liver, and peritonitis contributed to two cases each.
With respect to non-natural deaths, suicide was the predominant mechanism, accounting for 24 cases (53.3% of non-natural deaths and 10.2% of all deaths). All suicides were by hanging. Drug overdose resulted in 15 deaths (33.3% of non-natural deaths and 6.4% of all deaths), whereas homicide accounted for six cases (13.3% of non-natural deaths and 2.5% of all deaths).
Characteristics of the deceased individuals by type of offence (drug-related or other)
Approximately one fourth (63/236, 26.7%) of the deceased individuals were incarcerated for drug-related offences. The mean age at death for those incarcerated for drug-related offences was significantly lower than that of individuals imprisoned for other types of offences (mean (SD): 46.8 (13.0) years vs. 54.5 (16.8) years, p = 0.001) (Table 3). While non-natural deaths were more common in this subgroup than in those imprisoned for other reasons, this difference was not statistically significant (23.8% vs. 17.3%, respectively, p = 0.263) (Table 3).
Characteristics of deceased individuals by offence type (drug-related versus other) in Korydallos prison special health centre, Greece (2010–2018).
Student's t-test.
Pearson's chi-square test.
Fisher's exact test.
The main mechanism of non-natural deaths among individuals incarcerated for drug-related offences was drug overdose (9/15, 60.0%), followed by suicide (3/15, 20.0%) and homicide (3/15, 20.0%). The primary cause of natural death in this group was circulatory disease (18/48, 37.5%), followed by cancer (10/48, 20.8%), infections (7/48, 14.6% with six being AIDS-related deaths) and respiratory disease (5/48, 10.4%).
Mortality trends over time
Mortality rates differ substantially depending on the approach used to estimate the denominator (Figure 2). Nevertheless, the time trends were similar and the statistically significant increase in mortality observed in 2013 was evident under both approaches.

Trends in annual mortality rates at the prison hospital for male detainees in Greece (Korydallos Prison Special Health Centre) from 2010 to 2018. Mortality rates were calculated using as the denominator the total number of unique people entering the facility in the respective year (solid line) or the census population of the facility on January 1st of the respective year (dashed line). PLIP: people living in prisons; CI: confidence interval.
Discussion
Statement of principal findings
In this study, we analysed deaths that occurred at the sole prison hospital in Greece between 2010 and 2018 and identified distinct patterns in both natural and non-natural deaths, with important implications for prison healthcare policy. One-fifth of these deaths were attributed to non-natural causes, with suicide accounting for half and drug overdose for one-third of these cases. Among natural deaths, the majority were due to circulatory issues and cancer. Individuals who died from non-natural causes were significantly younger than those who died from natural causes. Non-natural deaths were more prevalent among inmates who were incarcerated for drug-related offences compared to those who were imprisoned for other reasons, although this difference was not statistically significant. Overdose was the primary mechanism of non-natural death within this subpopulation. We identified a statistically significant increase in all-cause mortality in 2013.
Strengths and weaknesses of the study
This study provides valuable data regarding mortality and causes of death over a nine-year period from a facility where the majority of deaths within the Greek prison system are recorded. This period includes significant events, such as austerity measures implemented following an economic crisis, a large HIV outbreak among people who inject drugs beginning in 2011, and reforms related to drug legislation. Our data allowed for the first time to compare the characteristics of deceased individuals by cause of death and type of offence (drug-related or not), assess age at death, and observe an increase in mortality occurring with a one-year lag compared with trends seen within the general population in Greece.
Our study has several limitations. First, as we assessed deaths within a prison hospital setting, individuals may have had more comorbidities than those within the overall prisoner population in Greece; thus, the proportion of natural deaths may be overestimated. Similarly, discrepancies observed regarding age at death compared with that of males in the general population may also be overestimated. However, it should be noted that our data include a substantial proportion of deaths reported within the Greek prison system (e.g., approximately two-thirds of all recorded deaths during 2013–2018).13,16 Second, cause-of-death data were collected retrospectively from forensic reports maintained at the prison hospital; more detailed information from histopathological and/or toxicological examinations (if performed) at other institutions (e.g., forensic services) were not available. We also lacked comprehensive information regarding the circumstances surrounding homicide cases, and for suicide cases, information regarding psychiatric status and the use of illicit psychotropic substances and/or alcohol was limited. Similarly, for deaths categorized as drug overdose, we were unable to determine with certainty whether those were accidental or intentional. Third, it was not possible to estimate mortality rates using appropriate denominators for the population at risk—a common issue within prison settings characterized by high turnover rates among prisoners. 17 However, while our annual mortality estimates per se may not necessarily reflect the mortality within this population accurately, they do adequately capture trends over time.
Strengths and weaknesses in relation to other studies
The proportion of natural deaths (80.9%) was relatively high compared with other studies reporting lower numbers, including studies conducted in prisons in Texas, the United States (1985–2010) 18 ; New South Wales, Australia (1988–2002) 19 ; England and Wales, the UK (1978–1997) 20 ; Ontario, Canada (1990–1999) 21 ; Milan, Italy (2003–2017) 22 ; and Lagos, Nigeria (2008–2019). 23 Notably, our study assessed deaths among individuals imprisoned in a prison hospital due to acute or chronic illnesses, which could have contributed to the higher proportion of natural deaths. Nevertheless, data from all prison establishments in Greece in 2020 indicated relatively similar proportions (natural deaths, including COVID-19: 72.5%, suicide: 8.7%, homicide: 1.4%), 7 suggesting that these trends represent broader patterns in the Greek correctional system. Cardiovascular diseases and cancer were the most prevalent causes of natural deaths, which is consistent with findings from other prison studies,15,18–20,24 including those conducted in low- and middle-income countries. 25
The mean age at natural death (52.4 years) was substantially lower than the life expectancy of men in Greece (exceeding 76 years during the period from 2000 to 2021). 26 A similar finding was reported by a review of healthcare in UK prisons, where prisoners who died from natural causes did so at a median age of 67.5 years compared with 86.7 years in the general population. 27
Suicide accounted for one out of ten deaths and was the predominant mechanism of non-natural death, which is consistent with the literature. 7
Meaning of the study: Possible mechanisms and implications for clinicians or policymakers
Our study highlights the predominant role of cardiovascular diseases and cancer in natural deaths; these are also the leading causes of death among men in the general population in Greece. 28 However, we identified a gap of at least 20 years between the mean age at natural death in prison and the life expectancy of men in the general population in Greece. Despite the fact that the prisoner population in this study was incarcerated in a prison hospital—thus potentially having increased comorbidities—the significant gap in age at death suggests that the healthcare needs of prisoners are not adequately met, highlighting the urgent need for improvements.
Suicide was the main cause of non-natural deaths, a finding supported by a previous study of mortality in the Greek prison system (1977–1996, 20% of all deaths). 29 Suicide and drug overdose accounted for 27% of the deaths recorded during 1999–2010 in major correctional facilities across the country. 30
Our analysis also provides data on the causes of death according to the type of offence (drug-related or not) and underscores the risk of overdose death among individuals incarcerated for drug-related offences. People who use drugs are overrepresented in prisons, and deaths resulting from drug overdose are frequently reported in this setting.7,31 Drug use may also contribute to suicide; for example, a study conducted in Irish prisons reported that half of self-inflicted deaths tested positive for illicit drugs. 32 Therefore, access to opioid agonist/drug treatment programmes, overdose education, and naloxone delivery are crucial for reducing mortality rates in prisons. In Greece, the proportion of prisoners with drug-related offence ranged between 34% and 38% during 2010–2013 to 19% and 27% during 2014–2018, following the implementation of a national drug law in 2013 that established more lenient sanctions. 11 However, naloxone is not available, and drug treatment and opioid agonist treatment programmes are offered only in some prison establishments.
The substantial increase in the mortality rate observed in 2013 is particularly noteworthy. Greece has been severely impacted by the global economic crisis and entered into an austerity programme in 2010 that significantly affected health expenditures. Studies on mortality rates in Greece have identified a direct effect of health spending on mortality following the economic crisis. 33 A spike in the age-standardized mortality rate was observed in the general population of Greece a year earlier, in 2012. 34 Additionally, it should be noted that during 2011–2013, a large HIV outbreak was recorded among people who inject drugs in Athens.12,13 This outbreak likely impacted mortality rates within the prison hospital; we recorded three AIDS-related deaths (out of 6 recorded in the total period 2010–2018) and four tuberculosis deaths (including three drug-resistant cases) in 2013, highlighting the effects of the HIV outbreak during this period. That year also saw an increased number of deaths from cardiovascular causes and suicides, possibly reflecting broader impacts of the economic crisis on healthcare access and mental health. Overall, the mortality peak observed in 2013 appears to reflect a convergence of factors: the HIV epidemic among people who inject drugs, austerity measures affecting healthcare delivery, and broader social stressors impacting the prison population.
Unanswered questions and future research
Our findings underscore the need to collect accurate data on prison deaths and to closely monitor mortality rates, including drug-related fatalities, in order to implement targeted policies to address the most common causes of deaths and extend the life expectancy of PLIPs. To prevent drug-related deaths, it is important to provide access to drug treatment and comprehensive harm reduction measures, introduce naloxone within the prison setting, and educate both staff and inmates about overdose and naloxone administration. Suicide prevention training, through systematic staff training and improved mental health services, is essential to address the most prevalent cause of non-natural death in this setting.
Supplemental Material
sj-docx-1-msl-10.1177_00258024251363671 - Supplemental material for Causes of death among people living in prison in Greece: A 9-year retrospective study (2010–2018) at the national prison hospital
Supplemental material, sj-docx-1-msl-10.1177_00258024251363671 for Causes of death among people living in prison in Greece: A 9-year retrospective study (2010–2018) at the national prison hospital by Sofia Boundi, Sotirios Roussos and Vana Sypsa in Medicine, Science and the Law
Footnotes
Acknowledgments
The authors would like to thank the Ministry of Justice and the Director of the Korydallos Prison Special Health Centre for providing permission to access the records as well as Dr Stergios Georgoulas, physician of the Korydallos Prison Special Health, for his support.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Contributorship statement
SB conducted the data collection. SB and SR analysed the data and wrote the first draft of the manuscript. VS conceived and supervized the study, and critically revized the manuscript for important intellectual content. All authors approved the final version of the manuscript.
Guarantor for the work
SB accepts full responsibility for the work and the conduct of the study, had access to the data, and controlled the decision to publish.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
