The Camden Coalition of Healthcare Providers (CCHP) works to improve the health of Camden, New Jersey residents through innovative approaches to increase quality, capacity, and accessibility within the city’s health care system. At the core of CCHP’s work is the concept of “hotspotting” — a data-driven process for timely identification of extreme patterns of health care use. Hotspotting is paired with targeted interventions and follow-up to better address patient needs, reshape ineffective utilization patterns, and reduce costs.

CCHP’s hotspotting method has evolved over the organization’s 13-year history. In its earliest forms, CCHP collected and merged claims data from across the city’s hospitals to better understand how violent crime was impacting the health of Camden’s residents and the related financial tolls to the city. While examining this newly created citywide dataset, CCHP’s analysis revealed that a small subset of patient population was responsible for an overwhelming majority of health care spending in the city. This striking pattern — where 10 percent of hospital patients account for nearly 75 percent of total hospital spending — is mirrored in communities across the nation.

Using data to identify and target interventions in the most efficient manner has become a mantra of CCHP. It has spurred the creation of a diverse coalition of stakeholders aligned around the common goal of using data to implement community-based solutions to improve quality and reduce the unsustainable cost of health care delivery in Camden.

Hotspotting Core Philosophies

The concept of hotspotting builds on several core philosophies that have shaped how CCHP approaches and works with data.

Information must flow between delivery systems in real-time.

CCHP operates its own health information exchange (HIE) that pulls real-time data from the city’s three health systems. The exchange serves as the engine for hotspotting and allows the organization to conduct real-time population health surveillance. CCHP’s community-based care teams receive daily inpatient admissions and emergency department visit reports. This information allows the care teams to triage and identify appropriate candidates for intervention, which involves visiting them at bedside and working with them in the community upon discharge. The organization has used the HIE to build reports enabling primary care practices across Camden to understand the hospital utilization patterns of their patient panels and receive real-time alerts when their patients visit one of the city’s hospitals.

Successful interventions need to take the community into consideration.

Local perspective and nuanced understanding of a community are critical to building a successful community-based intervention. Unfortunately, there is typically a paucity of high-quality, granular-level health and social data publicly available to help communities understand their unique landscapes. Data sets often encompass too large of a geographic area or are simply not updated frequently enough to be relevant.

Non-traditional data sources, such as administrative data, can be mined to provide insight at both the community and individual-level. For example, CCHP is currently building an integrated data system that will link hospital, housing, and criminal justice administrative data. This will expand the organization’s data lens and bring together a broad set of stakeholders to advocate for important policy initiatives, like housing first and other multi-sector collaborations. The more that communities can partner with disparate entities to encourage collaboration and data sharing, the better equipped the social service community  will be to generate well-tailored solutions that address the full spectrum of physical, behavioral, and social health and well-being.

Strong reciprocal relationships with entities that generate data are critical.

Building relationships with data generators is a long-term process that requires trust and a history of protecting data. Patient-level confidentiality is a critical concern, as is ensuring that data-sharing is a collaborative process focusing on productive use of data. CCHP is an independent non-profit with membership comprised of many health care organizations working in Camden (hospitals, primary care offices, social service agencies, etc.). The organization has worked hard to position itself as a neutral, honest broker—in effect, a hub with links to other sectors—engendering trust across the community. CCHP has sought to demonstrate that it offers a value-add to data partners. Many data providers (hospitals in particular) receive countless requests to share data externally. It is important to establish a process for soliciting input from data contributors that in turn provides them with reciprocal benefits.

Ongoing and fearless evaluation is vital to long-term success and sustainability.

While several studies have shown that better post-discharge care can improve the quality of care, lower mortality, and improve costs, there remains no solid understanding of what works and does not work. To address this issue, CCHP is currently administering a randomized controlled trial (RCT) with the Massachusetts Institute of Technology’s Abdul Latif Jameel Poverty Action Lab (J-PAL). The RCT will shed light on whether, and how well, CCHP’s flagship intervention works.

Additionally, CCHP is building the Camden Health Explorer, an interactive website that will de-identify hospital data and allow the public to see real-time Camden health trends on issues like readmission rates and the superutilizer population. The tool will serve as a population health “weather map” for the city, increasing the visibility of CCHP’s work in Camden and setting a strong precedent for data stewardship.

For more information, please explore the Hotspotting Toolkit, which was developed by CCHP through support from The Commonwealth Fund, to share practical how-to lessons for hotspotting with a rich community of stakeholders across the country.

Notify me about

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Inline Feedbacks
View all comments