Authors:
Ivana Durcinoska | The University of Sydney | Australia
Kon Shing Kenneth Chung | The University of Sydney
Jane M Young | The University of Sydney
Michael Solomon | Sydney Local Health District
Aim: Providing coordinated care is a key priority for health service improvement. Given the interpersonal nature of health care provision, social networks have emerged as an innovative approach to improving health care. However, little is known about how social network ties and structure affect patient care coordination and navigation through the health system. We sought to i) describe the personal networks of patients receiving treatment for colorectal cancer and ii) explore the role of personal network attributes in the patient experience of cancer care coordination.
Method: The study utilised a mixed-method egocentric network analysis approach. Eligible patients with colorectal cancer were identified from a representative state-wide study the NSW Bowel Cancer Care Survey. Participants initially completed a self-report questionnaire at baseline assessing care coordination experiences 6-8 months following diagnosis. Network data was subsequently collected in semi-structured telephone-based interviews. Four name generator questions were utilised to identify key alters involved in their decision making, information provision, emotional and practical support, and negative relationships. Participants were also asked to report on the frequency of contact, closeness of relationships, time known, and demographic characteristics of all alters named. We used descriptive social network measures (density, degree centrality, tie strength, efficiency, constraint, functional diversity) to characterise the networks in NETDRAW, and multivariate regression models to examine the association between network properties and experience of care coordination.
Results: A total of 126 patients participated, in which 875 alters were identified. Respondents had mean age of 67 years, 57% male, 24% living alone, and care coordination scores (mean 75.9, 10.11 SD, range 20-100) were normally distributed. Mean participant network size was 6.9 (SD 2.9) ranging from 2 – 18 alters. Participant network composition was approximately evenly distributed among family, friends and healthcare professionals. However, family members were more likely than other individuals to be identified as a close network and the vast majority of the negative ties identified (n=46) were health professionals. Multivariate analyses showed that higher care coordination scores were associated with higher density (β 5.5, p=0.001) and degree centrality (β 1.9, p<0.001).
Conclusion: This study offers a novel approach to exploring patient care experiences and navigation through the health system. More specifically it provides insight into care coordination processes and how relational networks influence this dimension of care. These results can inform evidence-based interventions aimed at improving coordination of patient care.