The GHFP is a hospital-funded outpatient program for heart failure patients following a hospitalization at Grady Memorial Hospital, a public, safety-net hospital in Atlanta, GA. Most GHFP patients are African American (92%) with an average age of about 60 years. Most patients are covered by Medicaid or Medicare (77%) and some patients do not have health insurance (17%). A third of the patients have no high school diploma or equivalent.¹⁴

The GHFP is an advanced practice provider-led heart failure clinic.¹⁵ Nurse practitioners work with patients during the inpatient visit, during the transition to outpatient, and in outpatient settings following hospital discharge. During the initial patient intake session, the GHFP provides patient education, assesses clinical treatment barriers, provides recommendations and referrals, and plans the transition to outpatient care.^16,17

The GHFP includes activities to address socioeconomic barriers to treatment plan adherence, and these activities have expanded over time (Figure 1). Since 2011, the GHFP provided financial counseling and 30-day medication supply at no cost to the patient.^18,19,20 In 2016, the GHFP received grant funding to provide transportation. Grady Heart Failure Program worked with CommonCourtesy (CommonCourtesy, Inc., Atlanta, GA) to provide Uber (Uber Technologies, Inc., San Francisco, CA) or Lyft (Lyft, Inc., San Francisco, CA) rides to outpatient appointments; this activity was added for patients with transportation barriers.^21,22 In 2018, the GHFP hired a full-time CHW as a permanent member of the clinical team and also added a patient liaison. In addition, the GHFP worked with Grady Mobile Integrated Health to provide home health visits for patients who were most in need of these services.²³ In mid-2018, the GHFP also began tracking social needs through Epic’s Healthy Planet (Epic, Verona, WI), an electronic health record system module used for accountable care and population health management.

The GHFP CHW conducts needs assessments and provides counseling and health education. In general, a GHFP CHW time allocation for addressing patients’ social needs is: 5% towards financial counseling, 15% towards transportation needs (e.g., assistance in arranging rides, navigating public transportation), 10% towards mobile home visits, and 40% towards linkages to other services.The patient liaison assists with follow-up calls, scheduling appointments, and addressing other socioeconomic barriers. For example, the patient liaison spends approximately 10% of their time arranging transportation for health care appointments. It is important to note that other key clinical staff also address patients’ social needs.¹⁴

This evaluation estimated both program costs and program effectiveness to calculate the short-term program cost-effectiveness of the expanded 2018 GHFP, compared with the previous year without the expanded activities, from the perspective of the hospital-provider’s budget.

Program costs were modeled and estimated using the best available information from the GHFP, health system, and hospital records using an activity-based micro-costing approach described in detail in a cost study published in 2021.¹⁴ The cost inputs, values, and data sources are presented in Table 1. The cost analysis is presented in Table 2. Cost included both labor and non-labor components. A majority of the GHFP total cost is for labor.¹⁴ Labor included personnel time of senior managers to operate the program (e.g., medical director, population health director) as well as other important program staff (clinical staff, CHW, patient liaison) and additional part-time staff (e.g., phlebotomy, front desk support). Non-labor costs included facilities, equipment, supplies, program-related materials, transportation ride service, and mobile integrated health home visits.¹⁴

Unit costs for the labor and non-labor components are from the original cost study that used cost data collected during 2018 and 2019.¹⁴ The unit cost for labor is the sum of salary, benefits (if applicable), and cost of training for full-time staff. Key staff were reported for each year by the GHFP program manager in 2020. In both 2017 and 2018, key staff included 4.5 full time equivalent (FTE) nurse practitioners, one FTE registered nurse, and one FTE medical assistant. In 2018 a CHW and patient liaison were added and unit costs for each staff were inputs. Labor costs were estimated using the number or proportion of full-time staff multiplied by the unit costs. Non-labor costs included facilities and 144 mobile integrated health home visits based on hospital records, available from the previous cost analysis.¹⁴ Other non-labor costs included equipment based on an inventory, supplies based on invoices, and program-specific costs reported by GHFP staff.¹⁴ Total cost for the transportation ride service for each year is based on invoices. We summed all labor and non-labor cost components by calendar year to derive the total program cost for 2017 and 2018 assuming the number of GHFP patients in 2018 for both years (Table 2).

A primary outcome for the GHFP is heart failure-related hospital readmission. To lower the risk of readmission, the GWTG^® (Get With the Guidelines) program for heart failure recommends that patients have a follow-up visit within seven days of hospital discharge.²⁴ Therefore, a program output is a GHFP follow-up visit within seven days of heart failure hospital discharge and a short-term program outcome is related hospital readmission within 30 days. The percentage of follow-up visits conducted within seven days and 30-day hospital readmissions were obtained from Grady Memorial Hospital from their electronic medical record data dashboard in 2020 (Table 1).

To estimate the expected number of follow-up visits within seven days of a hospital discharge and 30-day heart failure hospital readmissions, we multiplied the rates each year (2017, 2018) by the number of unique GHFP patients in 2018 based on program intake conducted during the inpatient encounter. We estimated what would have occurred among 2018 GHFP patients in the expanded (program) year (2018) compared with the baseline year (2017). The expected number of visits and readmissions in the program/baseline year estimates the number of visits and readmissions that would have occurred in the program/baseline year with the same number of unique patients across both years (Tables 1 and 3).

We compared the expanded 2018 GHFP with the previous year without a CHW, patient liaison, and mobile integrated health home visits to evaluate how program costs, follow-up, and related hospital readmission changed. For program output and outcome evaluation, we calculated the difference in expected number of follow-up visits within seven days of hospital discharge,and in heart failure-related hospital readmissions within 30 days of hospital discharge, in the expanded program year (2018) compared with the baseline year (2017). This calculation was based on 2018 GHFP patients to assess pre-post program effectiveness.²⁴ The difference in total program costs between the two years is estimated as the incremental program costs in 2018 compared with 2017 if the number of unique heart failure patients in 2017 were the same as in 2018.

The program cost-effectiveness metric is the incremental program cost per 30-day hospital readmission averted. The program cost-effectiveness metric was calculated by dividing incremental program cost by program effectiveness, which was reported as incremental program cost per heart failure-related hospital readmission in 2018 compared with 2017 (Table 3).

All program costs are reported in 2019 U.S. dollars as part of the original data collection. For a discussion of the costs, all non-labor costs, including heart failure inpatient costs from the literature, were standardized to 2024 U.S. dollars using the personal consumer expenditures health care price indices.^25,26 Program labor costs were standardized to 2024 U.S. dollars by using the Employment Cost Index for wage and salaries for private industry hospital workers.²⁷ The CDC, Morehouse School of Medicine Institutional Review Board, and Grady Research Oversight Committee reviewed and approved this activity as program evaluation. This activity was reviewed by CDC and was conducted consistently with applicable federal law and CDC policy.