Susan Hazelett, BSN, RN, MS, has been the manager of the Seniors Institute research department at Summa Health System since 1999, collaborating in the design and conduct of federally funded chronic illness management trials. She is adjunct faculty at Northeastern Ohio Medical University (NEOMED), where she collaborates with other faculty and assists students conducting research projects.
ElderBranch interviewed Ms. Hazelett about her paper, “Results of the Promoting Effective Advance Care Planning for Elders (PEACE) Randomized Pilot Study,” which she wrote with Dr. Steven “Skip” Radwany, medical director of hospice and palliative care services at Summa Health System.
What led you to conduct the PEACE pilot study? Why were you interested in assessing the effectiveness of in-home interdisciplinary chronic illness care management intervention for enrollees in Ohio’s PASSPORT program?
Our research group began studying how to improve care of the frail elderly beginning in the 1990s with a randomized trial of our Acute Care for Elders (ACE) Unit. That study led us to understand that it is not enough to improve inpatient care for frail elders – you must also provide interventions in the home after discharge to ensure that the plan of care is implemented.
As a result we conducted an NINDS-funded post-stroke care management study that involved sending a care manager into the home to perform an in-home comprehensive geriatric assessment and report back to an interdisciplinary team. The team would then make personal care plans to be implemented by the care manager.
This study led us to understand that this chronic illness management approach could be expanded to multiple chronic illnesses, so we conducted an ARHQ-sponsored randomized trial where we enrolled frail chronically ill patients during an acute hospitalization and provided post-discharge care management. We connected to the community long-term care providers – the Area Agency on Aging (AAA) – by including a representative from the AAA on the interdisciplinary team.
This study showed us that we needed to further integrate our medical services with the community-based social services to optimize efficiencies. It also showed us that we needed to take the chronic illness care/palliative care model further upstream in the chronic illness trajectory in order to have a greater impact on consumer outcomes.
Thus, in the PEACE trial we enrolled consumers at the time of enrollment into the Ohio’s Medicaid waiver program, PASSPORT, since all PASSPORT enrollees have functional impairment and virtually all have at least one chronic illness. We trained AAA care managers to perform the assessments and bring them back to the hospital-based interdisciplinary team so that individualized care plans could be generated.
Please describe your study. It was not powered to test hypotheses, but rather to generate hypotheses, yes? Please explain.
PEACE was a randomized pilot study that included 80 new enrollees into Ohio’s PASSPORT program. PEACE utilized AAA care managers, who normally concentrate on consumers’ psychosocial needs, to perform more in-depth in-home assessments of consumers’ medical, symptom management, and advance-care planning needs.
These assessment findings were reported back to a hospital-based interdisciplinary team consisting of a geriatrician, palliative care physician (PCP), pharmacist, social worker, and spiritual advisor. This team generated an individualized care plan based on explicitly stated patient goals, which was communicated to the PCP during a consumer visit that the care manager attended.
The care plan was then implemented by the care manager in collaboration with the PCP, with special emphasis on advance care planning. Consumers were followed by the team for 12 months.
The sample size of 80 is too small to have enough power to detect potentially important differences between groups. As a result, we looked instead at differences in mean values for a variety of variables, along with the associated 95 percent confidence intervals, comparing the intervention and control groups.
By looking at effect sizes rather than performing a hypothesis test, we can examine the effect of the intervention across a number of variables without risk of committing a Type 1 error. We can look at trends in the data that suggest an effect of the intervention on specific variables and use the size of the observed effect to calculate a sample size that will be sufficient to determine whether this effect is statistically significant.
What were the key findings from your research?
Our findings are currently in press with the journal Population Health Management so I am unable to provide specific details until that is published. As a general statement, we saw less healthcare utilization by the intervention group.
Would you make any specific recommendations to PASSPORT care managers as a result of your findings?
Our care managers were hesitant, naturally, at the start of the study. Indeed, it added to their work load. However, they did feel that the intervention gave them a much deeper knowledge of their consumers and felt that the program was worth the extra effort.
Are there implications of your findings that could extend beyond Ohio’s PASSPORT program?
Yes. An effective way to improve outcomes for frail elders is by targeting consumers upstream in the chronic illness process and by using a combined geriatrics/palliative care approach to address issues that cross the spectrum from geriatric syndromes to disease management to symptom management to advance-care planning.
This intervention requires only collaboration between community-based, long-term care providers and minimal time commitment from hospital-based medical personnel, both of whom share the common goal of keeping the consumer living comfortably and safely in their own home for as long as possible, according to consumer goals. This study provides a model for Accountable Care Organizations that must ensure that care is coordinated across settings.
What are the next steps to further your work in this area?
We are planning a fully powered trial that will provide the sample size necessary to definitively answer whether the intervention has a statistically significant effect on outcomes related to improved consumer outcomes and lower costs.