INTRODUCTION

The quest to reduce costs and readmissions has given rise to system- and practice-level approaches to identifying and managing the care of high-cost, high-risk populations.1 Care management programs (CMPs) have been developed in US health systems to connect nurses, social workers, and/or other staff with patients to coordinate care, reduce utilization, and lower costs, but evidence of success in cost reduction is variable.2, 3 Early models featured registered nurses as care managers, although more recent reports show models led by social workers.4 We aimed to develop approaches to classifying models of care management in Medicare ACOs that included staffing, CMP process, and outcomes.

METHODS

The data reported here come from a survey completed in a secure web-based survey tool by medical directors or clinical leaders of 15 care management programs in 2017. The sites were members in the Great Plains Collaborative (GPC) and the Scalable Collaborative Infrastructure for Learning Healthcare System (SCILHS), in the Patient-Centered Outcomes Research Network (PCORnet). CMP leaders were recruited by email through referral by site principal investigators to identify a “care or case management program for high cost or high utilizing Medicare beneficiaries or other populations.” Our questionnaire was developed through interviews with clinicians, patients, and administrators in CMPs; key domains include program structure, staffing, services, and outcomes. Question and response wording are shown in the tables.

RESULTS

Of 20 sites contacted, 18 responded; 3 sites did not have CMPs; at 2 non-respondent sites, a website search did not identify any existing care management programs. Table 1 shows program characteristics. Most programs provide a range of services; all have an RN or APRN on staff; most also have a social worker. Care managers in 12 programs interact directly with patients in hospitals or ambulatory offices, and 8 make home visits. High variability is seen in the annual and current number of cases served by the program, as well as in the typical caseloads per full-time-equivalent care manager.

Table 1 Program Characteristics (n = 15 Programs)
Full size table

Table 2 shows respondent reports of program outcomes as well as responses to attitudinal measures about program communication. Most programs use cost and utilization data to identify patients and track outcomes. Only one respondent “strongly” agreed their system had “finely tuned” their case-selection methodology. The most common program metrics include 30-day readmission rates and emergency utilization. While most agreed on the importance of “direct interaction of care managers with primary care providers,” only 2 rated care managers’ interactions with PCPs as “excellent.” Other ratings of “excellent” were rare for communications and work relationships.

Table 2 Data for Identifying Patients and Measuring Quality (n = 15 Programs)
Full size table

DISCUSSION

We surveyed medical directors of care management programs for high cost, high utilizing patients in 20 ACOs in two large networks that collectively cover approximately 25 million patients in 13 states. We cannot be certain about the generalizability of these findings outside of these two extensive networks. We surveyed medical directors of programs; we acknowledge that care managers, nurse, or social work leaders might report differently on attitudinal measures. These data show widely variable CMP process and structures that made classification challenging. Care management programs are a rapidly expanding as part of population health management strategy; few have realized cost savings.2 More research to understand the role of staffing, duration of service, and caseloads might inform future analyses of program effectiveness5 and return on investment. The relatively low level of use of patient reported measures and lack of uniformity in other outcome measures across programs was discouraging. Improving the range of outcomes beyond utilization to include patient and stakeholder experience is essential to understanding the best processes to achieve those outcomes and the value of these programs.