In part 2 on Social Network Analysis (SNA), I highlighted how data became more accessible with the launch of the Internet and widespread adoption of home computers. Unfortunately, the healthcare industry has lagged. Until 2009, with the launch of the Health Information Technology for Economic and Clinical Health Act (HITECH), only 10% of hospitals had adopted electronic health records. Many still had only paper-based records. Even today a growing collection of applications collecting patient information have little or no connection to the existing EHR ecosystem. This affects the creation of a true longitudinal patient record. Yet there is a growing use of Data Aggregators, Enterprise Data Warehouses and Patient Registries to consolidate different patient data sources and data types. This trend may allow wider use of Social Network Analysis to examine the value of coordinated care AND coordinated communications when treating chronically ill patient populations.
Why? Because, at its core, SNA analyzes how efficiently information flows through a network of people or groups. A critical element in treating chronically ill patients, as they typically deal with a wider variety of clinical personnel. They also generate the greatest amount of healthcare cost. SNA represents a way to document the effect communication has on a process. And the current process for treating chronically ill patients certainly represents an opportunity for improvement. Even small improvements could generate billions of dollars in healthcare savings while improving the quality of life for millions of chronically ill patients.
As I mentioned in my initial SNA post, there is a fair amount of research using SNA to analyze the healthcare industry. Unfortunately, most of it appears to represent a snapshot of a current state. There is very little comparative research to quantify savings or outcomes.
I spent some time digging around the Internet to see what I could find. I certainly understand this topic is specific – and the following studies offer a narrow topic of interest. My goal is to highlight some of the previous work to show the limited scope of SNA research in healthcare and highlight how this might be changing.
Clinical is the focus
I found a variety of research and have provided links so you may review in more detail. Each link provides the ability to download a PDF of the actual study.
To start, I’ll share a 2012 U.K. based paper focused on finding as many studies as possible that used SNA in healthcare settings. This paper documented 52 different studies using SNA to assess clinical personnel. Even with this number of studies, their conclusion was: most used SNA methods to describe social interaction at one point in time in a variety of healthcare settings! In other words, they didn’t document any outcomes. They documented the current state.
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0041911
Conclusions: We found very little evidence for the potential of SNA being realised in healthcare settings. However, it seems unlikely that networks are less important in healthcare than other settings. Future research should seek to go beyond the merely descriptive to implement and evaluate SNA-based interventions.
Published in 2017, the second study again focused on a large-scale review of existing studies. This study was positioned as a follow-up to the 2012 study listed above. They reviewed 5,970 articles looking for studies that analyzed healthcare worker professional communication, network methods used, and patient outcomes. They ended up finding a total of six, published between 2011 and 2016, that met their criteria for inclusion! None of the studies documented any savings.
https://link.springer.com/article/10.1186/s13643-017-0597-1
Conclusion: Network methods are underutilized for the purposes of understanding professional communication and performance among healthcare providers. The paucity of articles meeting our search criteria, lack of studies in middle- and low-income contexts, limited number in non-tertiary settings, and few longitudinal, experimental designs, or network interventions present clear research gaps.
Yet some hope!
Here are two studies that attempted to document savings. The first analyzed chronically ill patient sharing within practices to determine how this affected healthcare cost.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3579968/
Conclusion: Patients treated by sets of physicians who share high numbers of patients tend to have lower costs. Future work is necessary to validate care density as a tool to evaluate care coordination and track the performance of health care systems.
The second study analyzed the level of care density (a proxy measure for how frequently a patient’s doctors collaborate) to see how it affected adverse events and readmissions.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4384684/
Conclusion: In some settings, patients whose doctors share more patients had lower odds of adverse events and 30-day readmissions.
Both studies focused on the communication and information sharing between clinical personnel who treat chronically ill patient populations. In other words, coordinated care. It’s important to note that both documented a variety of limitations, including the inability to determine the structural relationships between providers. This is an important point. The research couldn’t document whether the providers who shared patients more efficiently were part of the same practice/health system – or from independent providers. This clearly represents an opportunity for future study.
More to follow?
This research will likely expand for the following reasons:
- Cost Savings: Any work documenting key elements in communication (and leading to lower costs and better outcomes when treating chronically ill populations) is a high priority, given the size of the current spend and government focus on reducing costs.
- Expansion of Population Health Data: The growth of healthcare data is astounding and the ongoing consolidation of chronically ill patient data through Aggregators, Enterprise Data Warehouses and Patient Registries will present more opportunities to examine data flow in healthcare organizations.
The goal is to efficiently move relevant patient information into the hands of providers who treat chronically ill populations. I’ll also add that we need to figure out how to efficiently move information to patients in ways to accommodate their preferences in language, education and culture. Any use of SNA that helps us achieve success in both areas will help. My next post on Social Network Analysis will examine areas where we might find more success.
No Comments