BackgroundA wide-ranging debate has taken place in recent years on mediation analysis and causal modelling, raising profound theoretical, philosophical and methodological questions. The authors build on the results of these discussions to work towards an integrated approach to the analysis of research questions that situate survival outcomes in relation to complex causal pathways with multiple mediators. The background to this contribution is the increasingly urgent need for policy-relevant research on the nature of inequalities in health and healthcare.MethodsThe authors begin by summarising debates on causal inference, mediated effects and statistical models, showing that these three strands of research have powerful synergies. They review a range of approaches which seek to extend existing survival models to obtain valid estimates of mediation effects. They then argue for an alternative strategy, which involves integrating survival outcomes within Structural Equation Models via the discrete-time survival model. This approach can provide an integrated framework for studying mediation effects in relation to survival outcomes, an issue of great relevance in applied health research. The authors provide an example of how these techniques can be used to explore whether the social class position of patients has a significant indirect effect on the hazard of death from colon cancer.ResultsThe results suggest that the indirect effects of social class on survival are substantial and negative (-0.23 overall). In addition to the substantial direct effect of this variable (-0.60), its indirect effects account for more than one quarter of the total effect. The two main pathways for this indirect effect, via emergency admission (-0.12), on the one hand, and hospital caseload, on the other, (-0.10) are of similar size.ConclusionsThe discrete-time survival model provides an attractive way of integrating time-to-event data within the field of Structural Equation Modelling. The authors demonstrate the efficacy of this approach in identifying complex causal pathways that mediate the effects of a socio-economic baseline covariate on the hazard of death from colon cancer. The results show that this approach has the potential to shed light on a class of research questions which is of particular relevance in health research.

BackgroundMore people die in hospital than in any other setting which is why it is important to study the outcomes of hospital care at end of life. This study analyses what influenced outcomes in a sample of patients who died in hospital in Ireland in 2008/9. The study was undertaken as part of the Irish Hospice Foundation’s Hospice Friendly Hospitals Programme (2007–2012).MethodsOutcomes of care were assessed by nurses, doctors and relatives who cared for the patient during the last week of life. Multi-level modelling was used to analyse how care outcomes were influenced by care inputs.ResultsThe sample of 999 patients represents 10% of acute hospital deaths and 29% of community hospital deaths in Ireland in 2008/9. Five care outcomes were assessed for each patient: symptom experience, symptom management, patient care, acceptability of the way patient died, family support. Care outcomes during the last week of life tended to be better when: the patient had cancer; admission to hospital was planned rather than emergency; death occurred in a single room or where privacy, dignity and environment of the ward was better; team meetings were held; there was good communication with patients and relatives; relatives were facilitated to stay overnight and were present at the time of death; nursing staff were experienced and had training in end-of-life care; the hospital had specific objectives for developing end-of-life care in its service plan.ConclusionsThe study shows significant differences in how care outcomes, including pain, were assessed by nurses, doctors and relatives. Care inputs operate in a mutually reinforcing manner to generate care outcomes which implies that improvements in one area are likely to have spill-over effects in others. Building on these findings, the Irish Hospice Foundation has developed an audit and review system to support quality improvement in all care settings where people die.Electronic supplementary materialThe online version of this article (doi:10.1186/s12904-015-0014-2) contains supplementary material, which is available to authorized users.

It is frequently argued that changes in the occupational structure and labour markets of European cities have the potential to undermine social cohesion. The term ‘social polarisation’ has been widely employed to characterise this effect, either in a broadly descriptive manner or in line with specific hypotheses. In the first part of this article, alternative definitions are reviewed and the results of empirical research on social polarisation are summarised. Some of its limitations are discussed and its theoretical origins explored. In the second part, attention is turned to the ‘mechanisms’ driving change in workplaces and urban labour markets in Europe. It is argued that an accurate account of changing occupational structures and labour markets in European cities—and a balanced assessment of their consequences for social cohesion—can only be obtained by building up a complex and carefully contextualised analysis of the ways in which these ‘mechanisms’ interact in different cities

Emergency presentation of rectal cancer carries a relatively poor prognosis, but the roles and interactions of causative factors remain unclear. We describe an innovative statistical approach which distinguishes between direct and indirect effects of a number of contextual, patient and tumour factors on emergency presentation and outcome of rectal cancer. All patients diagnosed with rectal cancer in Ireland 2004-2008 were included. Registry information, linked to hospital discharge data, provided data on patient demographics, comorbidity and health insurance; population density and deprivation of area of residence; tumour type, site, grade and stage; treatment type and optimality; and emergency presentation and hospital caseload. Data were modelled using a structural equation model with a discrete-time survival outcome, allowing us to estimate direct and mediated effects of the above factors on hazard, and their inter-relationships. Two thousand seven hundred and fifty patients were included in the analysis. Around 12% had emergency presentations, which increased hazard by 80%. Affluence, private patient status and being married reduced hazard indirectly by reducing emergency presentation. Older patients had more emergency presentations, while married patients, private patients or those living in less deprived areas had fewer than expected. Patients presenting as an emergency were less likely to receive optimal treatment or to have this in a high caseload hospital. Apart from stage, emergency admission was the strongest determinant of poor survival. The factors contributing to emergency admission in this study are similar to those associated with diagnostic delay. The socio-economic gradient found suggests that patient education and earlier access to endoscopic investigation for public patients could reduce emergency presentation.

The acute phase proteins (APP) form part of a non-specific host response to inflammation. They may be induced by a range of different causes, including infection, inflammation, cancer and trauma. As they form part of the earliest response to such insults, they have potential for early identification of disease. In people, APP levels have been shown to correlate both with the extent of disease and also the prognosis in several forms of neoplasia, including prostate, oesophageal and colorectal cancer. As such, they can be used as prognostic and monitoring tools. To date, similar studies in veterinary patients have been limited, largely retrospective in nature and many are non-specific for tumour type. The purpose of this study was to evaluate a panel of four APPs in dogs with naturally occurring mast cell tumours (MCTs) and sarcomas to identify in the first instance whether increased levels of individual APPs, or identifiable combinations of APPs, was linked with the presence of disease. In the patients with MCTs, C-reactive protein (CRP) and α-1 acid glycoprotein levels increased, with a concurrent drop in serum amyloid A levels. In the sarcoma patients, CRP, α-1 acid glycoprotein and haptoglobin were increased. These findings suggest that specific solid tumour types in dogs may be associated with specific changes in APP profiles