Article Text

Original research
Retrospective cross-sectional analysis of concurrent VTE diagnosis in hospitalised socially excluded individuals in Ireland
  1. Chloe Carpenter1,2,
  2. Anne O’ Farrell3,
  3. Fionnuala Ní Áinle1,2,
  4. Clíona Ní Cheallaigh4,5,
  5. Barry Kevane1,2
  1. 1Irish Network for VTE Research, University College Dublin, Dublin, Ireland
  2. 2School of Medicine, University College Dublin, Dublin, Ireland
  3. 3Department of Statistics and Epidemiology, Health Intelligence Unit, HSE, Dublin, Ireland
  4. 4School of Medicine, Trinity College, Dublin, Ireland
  5. 5Inclusion Health Service, St James's Hospital, Dublin, Ireland
  1. Correspondence to Dr Clíona Ní Cheallaigh; nicheacm{at}


Objective Social exclusion (such as that experienced by people who are homeless, incarcerated or use drugs) increases morbidity across a range of diseases but is poorly captured in routine data sets. The aim of this study was to use a novel composite variable in a national-level hospital usage dataset to identify social exclusion and to determine whether social exclusion is associated with concurrent venous thromboembolism (VTE) in hospitalised patients in Ireland. Identifying and characterising this association in people who are socially excluded will inform VTE prevention and treatment strategies.

Design Retrospective cross-sectional study.

Setting Irish Hospital Inpatient Enquiry (HIPE) system, which collects diagnostic information by International Classification of Diseases Tenth Revision code on all hospital admission episodes in the Ireland.

Participants All hospital admission episodes involving a VTE diagnosis (in a primary ‘Dx 1’ or secondary ‘Dx 2–30’ coding position) during a 12-month period in the Ireland were identified from consolidated, national-level datasets derived from the Irish HIPE system. Social exclusion was defined as the presence of one or more indicators of homelessness, drug use, incarceration, health hazards due to socioeconomic status or episodes of healthcare terminated prematurely.

Results Of 5701 admission episodes involving a VTE diagnosis (in a primary or secondary position) during the study period, 271 (4.8%) related to an individual affected by social exclusion. Among hospitalised individuals identified as being socially excluded based on the novel composite variable, the likelihood of having a concurrent VTE diagnosis was over twofold greater than that observed in the general population (OR 2.14, 95% CI 1.79 to 2.26; p<0.001).

Conclusion These data suggest that VTE (primary and secondary) is over-represented in hospitalised socially excluded persons in Ireland and that the development of strategies to address this potentially life-threatening accompanying condition in this vulnerable patient group must be prioritised.

  • haematology
  • health equity
  • thromboembolism

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • In this study we have determined, for the first time in a large national dataset, that concurrent venous thromboembolism (VTE) during hospitalisation is over-represented in people who are socially excluded.

  • The datasets used for this study were derived from data describing all admissions to acute public hospitals in the Ireland during a 12-month period and so represent a complete account of inpatient admission episodes at a national level.

  • As these data were gathered in a single country, the association observed should be tested in other settings (particularly in those where groin injecting is not a frequent practice among people who are socially excluded).

  • Due to the nature of Hospital Inpatient Enquiry datasets, it is not possible to exclude potential duplication in some cases or to capture cases of VTE where patients were diagnosed with VTE but not admitted to hospital.


Illness does not affect all equally. Lower socioeconomic status is associated with worse health outcomes across all age groups.1 Social exclusion can be conceptualised as an extreme form of poverty or deprivation and is associated with extreme health inequality.2 Several definitions for social exclusion exist, however it is generally agreed that it represents a complex and multi-dimensional state characterised by the inability of the individual to participate in the normal economic, social, cultural and political activities of society.1 3 Social exclusion is experienced by people with substance-use disorders, homeless people, people with severe and enduring mental illness, prisoners and certain minoritised ethnic groups (eg, Irish Traveller community and the Roma community). Social exclusion is frequently characterised by family breakdown, exclusion from education, unemployment, low income, poor housing, high-crime environments.4 Although financial poverty is a major determinant of health among people who are socially excluded, there are numerous other additional biological, psychological and social factors which contribute to the severe adverse health outcomes which are observed in this specific group. Healthcare services are generally not designed to meet the complex needs of socially excluded people, and they frequently have difficulty in accessing healthcare resources and services.5 Standardised mortality ratios may be up to 10-fold increased among socially excluded persons when compared with the general population.1 Challenges in the identification of socially excluded people from routinely collected national or international level health data compounds inequality. This negatively impacts on the ability of healthcare policy makers to develop and prioritise interventions to improve health equity.

Venous thromboembolism (VTE) comprises deep vein thrombosis (DVT) and pulmonary embolism (PE) and is a major contributor to global disease burden, affecting millions of individuals worldwide every year.6 It is recognised as being the third-leading cause of cardiovascular mortality globally, with an estimated 500 000 VTE-related deaths occurring in Europe alone annually.7 High rates of chronic morbidity have also been reported among survivors of VTE, including debilitating post-thrombotic syndrome, which can dramatically affect long-term quality of life and the ability to work.8–12 VTE risk is known to be increased in the setting of lower socioeconomic status but has not been studied in the context of social exclusion.13–15 In Ireland, social exclusion is strongly associated with injecting drug use, with the femoral veins in the groin representing a frequent target for vascular access. In Dublin, up to 70% of homeless individuals report having used illegal drugs with over half reporting injecting drug use.16 Groin injecting with opiates, such as heroin, and other drugs is recognised as a strong risk factor for lower limb DVT.13

Although highly effective strategies directed at the prevention and treatment of VTE in the general population have been developed in recent decades, limited data exist to inform the approach to VTE risk assessment and prevention in socially excluded persons.17 Although an increased prevalence of VTE in lower socioeconomic groups has been described,14 the impact of social exclusion (encompassing experiences of homelessness, substance use disorders, sex work and incarceration) has never been characterised in detail, despite its large impact on the health and lives of these people. Addressing this knowledge gap has been identified as a priority by the Acute Hospitals Division of the Irish Health Services Executive (HSE), the agency responsible for planning and commissioning all acute public hospital care in the Ireland18 and by the National VTE Clinical Programme.

We developed a novel composite variable utilising routinely collected clinical and demographic data to identify social exclusion in order to test our hypothesis that VTE may be over-represented in socially excluded persons admitted to hospital in Ireland.


Data sources

All inpatient discharges with a VTE diagnosis during 2017 coded in positions 1–30 (where position 1 (‘Dx 1’ generally indicates the primary reason for admission and positions 2–30 (‘Dx 2–30’) generally indicate that the coded condition was a secondary diagnosis) were extracted from the Hospital Inpatient Enquiry (HIPE) database. HIPE is a national administrative database containing patient-level records of all admissions to acute public hospitals in Ireland. All acute public hospitals participate in HIPE.19 In 2017, HIPE reported on over 1.7 million hospital episodes, with a coverage rate of over 98%. The HIPE database contains information on the patient’s age, sex, area of residence, date of admission and discharge, together with their principal diagnosis and up to 20 other diagnoses, coded by trained clinical coders using the International Classification of Diseases Tenth Revision (ICD-10-AM). HIPE data are entered and validated by trained HIPE coders at hospital level at the time of hospital discharge, extracted from standardised discharge forms completed by the responsible discharging clinician. After entry of data by HIPE coders at hospital level, it is further validated by the centralised HSE HIPE Healthcare Pricing Office when uploaded from each hospital. For the purposes of this study, datasets generated through the consolidation of HIPE data were analysed.

VTE case identification

We identified VTE cases (in a primary or secondary position) from HIPE datasets on the basis of ICD-10 discharge diagnostic codes for VTE (online supplemental table 1).

Composite variable for identification of socially excluded persons

Previous investigators have demonstrated that experiences such as homelessness, sex work, incarceration and substance use disorders can be conceptualised as social exclusion.1 For the purposes of this study, a composite variable, based on data which could be extracted from HIPE records, was created which comprised variables which were known to be over-represented in socially excluded persons in Ireland.1 5 Any individual with one of more of these features, listed in table 1, were categorised as socially excluded for the purposes of this study.5 The composite variable consisted of factors such as a drug misuse diagnosis (excluding alcohol); a temporary place of residence or no fixed abode as source of discharge; a hepatitis C diagnosis; those in prison and those who absconded or self-discharged from hospital against medical advice. The rationale for including hepatitis C in this composite variable was that drug use is a risk factor in at least 80% of cases of hepatitis C virus infection in Ireland.20 In addition, ICD-10 codes which reflect potential health hazards associated with the spectrum of psychological and social factors which can contribute to disease, and which are hallmarks of social exclusion (such as problems relating to education/literacy, problems related to social environment and upbringing), were also included. We elected to exclude alcohol due to the overwhelming over-representation of this ICD-10 code in patients NOT in a socially excluded group.

Table 1

Derived composite variable for social exclusion

With respect to the definition of homelessness, for the purpose of this study we included individuals who would be considered to be homeless based on the European Typology of Homelessness and Housing Exclusion criteria which defines a person as roofless or homeless if they have an identification deficit in at least two of the following: no dwelling, no legal title to a place for exclusive possession, and no private and safe space for social relations. The definition of homeless also includes those who are sleeping rough (ie, those sleeping in the open air); those living in emergency accommodation such as a hostel, night shelter or B&B accommodation; those living with family and friends, or in a squat.4 Homelessness may be chronic (lasting >1 year), intermittent or short-term/crisis-related.

Statistical analysis

Pearson χ2 tests and Wilcoxon rank-sum tests were carried out to assess the significance of frequency and continuous variable differences between those with VTE diagnosis and those without a VTE diagnosis. Logistic and multiple regression modelling were carried out using discretionary backward elimination to identify factors independently and significantly associated with having a VTE diagnosis including the ‘socially excluded’ variable. The Bonferroni correction was used to reduce the likelihood of type I errors from multiple hypothesis testing. All analysis were carried out in JMP statistical package.

Patient and public involvement



There were 420 466 emergency inpatient admissions identified in HIPE/Health Atlas Ireland in patients aged over 16 years during this 12-month period, of which 5701 (1.4%) had a VTE diagnosis. Fifty-five per cent of admission episodes with a concurrent VTE diagnosis involved DVT of the lower limb veins (either isolated or with concurrent PE; PE was reported in 49% of all VTE admission episodes). Thrombosis involving unusual sites (such as portal vein, renal vein, vena cava) was rare, reported in less than 5% of inpatient admission episodes (across a range of possible ICD-10 codes) (table 2). In total, 11 648 patient admissions occurred where individuals met criteria for social exclusion based on the composite variable. A history of drug use was reported in 4003 of these patient admission episodes (34.4%) (table 3). Using the proposed composite variable, 271 of the 5701 hospitalisations during 2017 in which VTE was either a primary or secondary diagnosis were in patients who met criteria as socially excluded. The contribution of each component of this variable to the overall composite variable is outlined in table 4. Strong collinearity was observed among individual components of the composite variable, suggesting that each component is likely to represent a marker of the phenomenon of social exclusion.

Table 2

Characteristics of hospitalised patients with a concurrent venous thromboembolism (VTE) diagnosis

Table 3

Patterns of drug-misuse among socially excluded patients

Table 4

Components of socially excluded composite variable and risk of venous thromboembolism (VTE)

Social exclusion and VTE

On multivariate analysis, several variables were found to be independently associated with a hospitalisation in which VTE was either a primary or secondary diagnosis (table 5). These included well-established risk factors for VTE including cancer, cardiovascular disease and a high Charlson Comorbidity Index Score. The Charlson Comorbidity Index is a methodology used to categorise comorbidities of patients based on ICD-10 code; each comorbidity category has an associated weight, based on the adjusted risk of mortality or resource use. The higher the score, the more likely the predicted outcome will result in mortality or higher resource usage. For this study, a Charlson score of >10 is indicative of high morbidity.12 After controlling for these and other factors, those patients captured using the social exclusion composite variable were significantly more likely to have a VTE diagnosis compared with those without a social exclusion diagnosis (OR 2.14, 95% CI 1.79 to 2.26, p<0.001).

Table 5

Factors associated with concurrent venous thromboembolism during hospitalisation*


The absence of a widely used metric for identifying social exclusion in large datasets makes studying the effect of social exclusion on health more challenging. However, given the profound health inequality experienced by people who are socially excluded, it is of critical importance that methods are developed to delineate their disease profiles. In this study, we created a composite variable using variables present in a national dataset of routinely collected demographic and clinical data to identify socially excluded persons based on the presence of specific characteristics known to be elements of, or associated with experience of social exclusion including homelessness, incarceration and drug use. Of note, we used local disease patterns (eg, current or previous drug use in more than 80% of people with hepatitis C in Ireland) and local behavioural patterns (rates of more than 15% of episodes of care being terminated prior to treatment completion in people living in social exclusion) and the presence of health hazards due to socioeconomic status to build our model.5 20 21 The novel composite variable incorporating these indicators of social exclusion was found to be associated with an increased likelihood of an individual being affected with VTE. During a 12-month period, a total of 5701 patients were admitted to hospital with VTE. Two hundred and seventy-one of these patients were socially excluded. Thus, our findings suggest that almost 1 in 20 hospital admissions associated with a VTE diagnosis in this country during a 12-month period were for an individual affected by social exclusion. In the patient admission episodes where any component of the socially excluded composite variable was present, the likelihood of that individual having a concurrent VTE diagnosis was greater than twofold higher than that observed in patient admission episodes where none of the criteria for social exclusion were met.

People affected with social exclusion have poorer health outcomes than unaffected individuals1 yet health services are rarely designed to meet their needs. We have previously demonstrated high rates of emergency department (ED) attendance but a greater than 50% rate of leaving the ED before treatment completion and a greater than 15% rate of premature termination of inpatient care.5 These data clearly demonstrate the challenges encountered by socially excluded persons in accessing care and the challenges faced by healthcare providers in adequately addressing their needs.

VTE is recognised as a leading cause of cardiovascular morbidity and mortality globally.6 7 22 Numerous risk factors have been identified in the general population which are associated with an increased risk of VTE.9 Strategies for the prevention of VTE have been developed in recent years and are widely implemented among certain at-risk groups, such as pregnant people, hospitalised patients and patients with cancer.23 24 Similarly, studies which aim to investigate the optimal treatment of VTE in these high-risk populations, continue to be prioritised by clinicians involved in the care of these patients.25 26 We have demonstrated that VTE appears to represent a major source of morbidity among people who are socially excluded, including homeless people and people who inject drugs. In contrast to other at-risk groups, socially excluded persons are poorly represented in clinical trials and observational studies and consequently evidenced-based recommendations for the prevention and treatment of VTE in this population are difficult to formulate.3

The precise mechanisms underlying the overrepresentation of a concurrent VTE diagnosis observed among socially excluded persons hospitalised in Ireland remain to be fully determined. Vascular injury following injecting drug use is thought to represent a major risk for DVT among socially excluded persons who engage in drug use.27 Factors related to some of the broader determinants of health (such as nutrition, smoking status) which are often adversely affected in the setting of social exclusion, may also contribute to VTE in this population, given the association between these factors and vascular complications in general.28 29 Emerging evidence suggests that the interplay between various other factors associated with social exclusion (such as hepatitis C infection, liver cirrhosis, premature ageing) may elicit a pro-inflammatory state which might also be implicated in adverse vascular outcomes in this patient group.30 31 Additional studies investigating thrombo-inflammatory and fibrinolytic pathway activity in socially excluded populations could provide new insights into the role of coagulation and vascular dysfunction in this vulnerable group.

The interpretation of our findings is limited by several factors. This study consisted of the evaluation of comprehensive, national level hospital admission data during a complete 12-month period; however, a multi-national study—particularly in settings in which social exclusion is less strongly associated with injecting drug use—would have permitted the generation of more generalisable findings which could perhaps be more easily extrapolated to other, more heterogenous populations outside of the Ireland. Additionally, due to the nature of HIPE data (which describe patient admission episodes and not individual, patient level data with unique identifiers), it is not possible to eliminate the risk of duplication in some of these reported cases. Due to the nature of HIPE, it is also not possible to always definitively differentiate between hospital-acquired and non-hospital acquired VTE events in this dataset. Patients with VTE who were not admitted to hospital would also not have been captured in this dataset. Finally, data used in the study are reliant on the accuracy of capture of diagnoses and variables by coders using HIPE. However, the accuracy of HIPE data is quality checked, via a built-in software system, which queries coding decisions and runs quality regular checks. Data entry including input of diagnoses must conform to expected values.

In summary, we have demonstrated the use of developing a composite variable, based on local disease and behaviour patterns, to identify social exclusion and have demonstrated over-representation of a concurrent VTE diagnosis (primary or secondary) in people experiencing social exclusion who are hospitalised. With respect to future implications of this work, increased awareness of VTE risk and burden in the socially excluded population is clearly warranted.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.


The authors would like to thank the Health Intelligence Unit of the Health Service Executive for permitting to conduct this study.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • CC and AO’F are joint first authors.

  • Twitter @clionani

  • CC and AO’F contributed equally.

  • CNC and BK contributed equally.

  • Contributors CC, AOF, FNÁ, CNC and BK conceptualised and designed the study. AOF prepared and analysed the study data. All authors contributed to the interpretation of study findings. CC, AOF, FNÁ, BK and CNC wrote the original draft. All authors contributed to reviewing and editing the manuscript. All authors approved the final manuscript. CNC and BK are joint final authors. CNC is the guarantor for the study.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests FNÁ reports: IIS awarded to institution (current): Bayer, Daiichi-Sankyo. Consultancy (paid to institution): Boston Scientific.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.