Heart failure disease management in skilled nursing facilities

Abstract

Heart failure disease management (HF-DM) in skilled nursing facilities (SNF) is challenging because of the medical complexity of the patients and quality problems in SNFs, including high staff turnover rates, frequent transitions of care, heterogenous patient populations, and a lack of protocols to help staff members identify and monitor patients with HF. Recommendations developed specifically for HF-DM in SNFs, detailed in a 2015 scientific statement written by the American Heart Association and the Heart Failure Society of America, recommend two key non-pharmacological HF-DM practices that alert clinicians to worsening HF: weight monitoring to detect fluid retention and symptom monitoring. Signs and symptoms helpful in detecting worsening HF are shortness breath, chest pain, lower extremity edema, paroxysmal nocturnal dyspnea (PND), orthopnea (shortness of breath while lying flat), and jugular vein distention (JVD). Guidelines also recommend sodium restriction to avoid decompensated HF, characterized by fluid overload. The 2015 scientific statement called for further research to improve our understanding of SNF HF-DM and ultimately improve care. The aims of this dissertation are to 1) Measure the frequency of non-pharmacological HF-DM practices and estimate the association of patient- and facility-level factors with the frequency of HF-DM practices; 2) Identify barriers to successful HF-DM in SNFs and develop a theory that explains variation in care; and 3) Identify the information needs of SNF staff that are not met in SNF information systems and propose design recommendations to address these needs. To achieve the first aim, we used quantitative data from the SNF Connect Trial, a randomized cluster trial of a HF-DM intervention in Colorado SNFs. To accomplish the second aim, we conducted semi-structured interviews with SNF staff, patients, and caregivers to identify patient and facility causes for care inconsistency. Staff interview data were then used to identify HF-DM tasks, information needs, and design recommendations to improve care. The SNF Connect sample (n=320) was majority female (66%), white (92%), with mean age 80 ± 10 years. Seventy percent had HF with preserved ejection fraction (EF) with a mean EF of 50 ± 16% and a New York Heart Association (NYHA) class III-IV (40%). On average, patients were weighed 32% of their days in the SNF and had documentation of at least one HF-related symptom assessment 55% of their days in the SNF. In linear regression analysis, patient-level factors (demographics, baseline weight, comorbidity, type of HF, NYHA class) were not associated with frequency of HF-DM practices. Health Inspection Star Rating, a measure of overall SNF quality, was positively associated with frequency of weight monitoring (p <0.05) but not associated with frequency of symptom assessment. Semi-structured in-depth interviews with 5 patient-caregiver dyads (n=10 individuals) and 15 staff members were conducted at 9 facilities from September 2017 to October 2018. HF-DM is often inconsistent because HF is not prioritized over other competing health conditions, HF-DM practices interfere with patient preferences, and unfamiliarity with symptomatology of HF interferes with patients’ ability to report symptoms and staff’s ability to assess and monitor them. Lastly, staff operate with misconceptions and gaps in HF knowledge, and do not utilize protocols or other tools to prompt HF care. Director of Nursing oversight of HF-DM practices by floor nurses and certified nursing assistants seems to improve HF management. Based on interviews with staff, we identified four design recommendations to improve HF-DM in SNFs: 1) nurse education and training that address information needs of staff, 2) printable materials to improve patient and caregiver participation, 3) a charting feature to monitor longitudinal trends, and 4) personal notetaking space for staff. Unsupported HF-DM information needs could be met through a user-centered, HF-specific module embedded within SNF electronic health records to overcome barriers and improve HF outcomes. Results from this study shows that SNF HF-DM practices have high variability and differ from recommended clinical guidelines. Variation in care is due in part to barriers faced by SNF staff, including information needs not yet supported by SNF information systems. Quantitative findings of low and variable HF-DM practice motivates opportunities to support HF-DM in keeping with the qualitative findings that describe challenges and information needs as described by SNF staff.