Implementation of Early Follow-up Care After Heart Failure Hospitalization

ABSTRACT

Objectives: The evidence supporting early postdischarge hospital follow-up is limited. We implemented a new, multidisciplinary, multistrategy heart failure (HF) team approach that included new clinic slots, predischarge nurse visit, providing a blood pressure cuff and scale, and cardiologist supervision.

Study Design: Pre- vs postintervention evaluation of outcomes in patients hospitalized with HF between September 1, 2010, and May 30, 2013. We utilized the RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework to evaluate the intervention.

Methods: For the quantitative evaluation, we compared the proportion of patients in both groups who were scheduled for and completed a cardiology appointment within 7 days after hospitalization (“reach”). We created a Cox model to evaluate the “effectiveness” of the intervention period on a 30-day composite outcome (all-cause emergency department [ED] visit, all-cause hospitalization, or death). In qualitative evaluation, we describe the adoption, implementation, and maintenance of the intervention.

Results: Data for 261 patients were analyzed (preintervention, n = 142; post intervention, n = 119). The postintervention period was associated with a higher proportion of patients who were referred to (40% vs 12%; P < .001) and completed (24% vs 10%; P = .003) cardiology follow-up within 7 days of hospital discharge (reach) compared with the preintervention period. After adjustment, the postintervention period was associated with a reduced hazard of the 30-day composite end point (HR, 0.59; 95% CI, 0.37-0.96; P = .04) (effectiveness).

Conclusions: The intervention succeeded in increasing referral to and completion of cardiology appointments within 7 days of discharge. In adjusted analysis, the intervention was associated with lower risk of 30-day all-cause ED visits, all-cause hospitalizations, or death.

Am J Manag Care. 2021;27(2):e42-e47. https://doi.org/10.37765/ajmc.2021.88588

Takeaway Points

Heart failure (HF) rehospitalizations occur in 1 in 4 patients within 30 days of discharge and remain a significant burden on patients and hospitals. Outpatient follow-up within 7 days of discharge is recommended, but real-world implementation experience and outcomes are rarely reported.

• We created an early hospital follow-up intervention that was associated with 3.3 times higher rate of referral for 7-day HF follow-up.
• The intervention period was associated with a 41% lower risk of emergency visits, rehospitalizations, or death.
• An early follow-up intervention can be successfully implemented, but there needs to be an emphasis on long-term maintenance.

Inpatients hospitalized for heart failure (HF), all-cause rehospitalizations occur in 1 in 4 patients within 30 days of discharge and, after 1 year, 1 in 3 are dead.¹ In response to these poor outcomes, the American College of Cardiology and the Veterans Health Administration launched the Hospital to Home Initiative nationally with a goal of reducing cardiovascular-related readmissions and improving care transitions.² The premise of this initiative was that results of observational studies have associated early hospital follow-up with lower 30-day readmission rates. However, there have been few real-world descriptions in the literature regarding outcomes and implementation experience of early hospital follow-up interventions. Phoenix Veterans Affairs (VA) Medical Center created an early postdischarge outpatient intervention for veterans with HF. In this retrospective single-center study, we conducted a pre- and postintervention analysis, as informed by the RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework,³ to determine if the transitional intervention (1) improved the timeliness of cardiology care (“reach”) and (2) was associated with an improved 30-day HF composite outcome, defined as 30-day all-cause emergency department (ED) visit, all-cause rehospitalization, or all-cause death (“effectiveness”). We also conducted a qualitative evaluation (adoption, implementation, maintenance) of this intervention.

METHODS

Reporting Framework

We utilized the RE-AIM Implementation Model^3,4 to describe the impact (reach, effectiveness, adoption, implementation, and maintenance) of the intervention. Reach was defined as the proportion of patients with primary HF discharge diagnosis who are referred for outpatient cardiology follow-up within 7 and 14 days after hospital discharge. It is important to recognize that reach is unlikely to ever reach 100%, as the patients hospitalized with HF (the denominator) were identified from administrative records, which are dependent on hospital coders for assigning discharge diagnoses; in contrast, the intervention is requested by clinicians who are referring patients based on their assessment of individual patient needs (the numerator). We also report the proportion of patients who completed an appointment in the same respective intervals; this outcome incorporates all possible reasons why patients did not complete appointments, ranging from rescheduling to nonadherence. Effectiveness was defined as the hazard of the 30-day composite outcome in the postintervention period compared with the preintervention period, adjusted for a priori clinical predictors associated with the outcome. We hypothesized that the intervention would be associated with a decreased hazard of the composite outcome. Adoption was defined as a description of the setting and staff who participated in intervention and reasons for adoption or nonadoption. Implementation was defined as the consistency, time involved, costs, adaptions made, and implementation strategies utilized. Maintenance was defined as the extent to which the intervention was sustained as part of routine organizational practice.

Data Collection

This study was approved by the Phoenix VA Medical Center Institutional Review Board with a waiver of informed consent. The study sample consisted of patients with an International Classification of Diseases, Ninth Revision primary discharge diagnosis of HF who were hospitalized between September 1, 2010, and March 1, 2011, inclusive (“preintervention”), or between November 30, 2012, and May 30, 2013, inclusive, after the intervention began (“post intervention”). Discharge diagnoses were determined by hospital coding staff. For all patients, clinical variables including demographics, vital signs, hemoglobin, blood urea nitrogen (BUN), creatinine, estimated glomerular filtration rate (eGFR), gender, ethnicity, length of hospital stay, brain natriuretic peptide (BNP), and left ventricular ejection fraction (LVEF) were collected by chart review (S.A.). We excluded patients who were discharged to hospice, enrolled in hospice during 30-day follow-up, left against medical advice, or were discharged to a skilled nursing facility (SNF) or acute care.

Description of the Intervention

Preintervention/control period. Patients hospitalized with HF were treated by the inpatient hospitalist team. There was no specific hospital policy recommending cardiology follow-up. If the team requested outpatient cardiology clinic follow-up, they submitted an electronic request, and patients were then contacted by telephone and/or letter after hospital discharge. The cardiology department consisted of nurse practitioners, cardiologists, and cardiology fellows; however, there was no cardiologist who specialized in the care of patients with HF. Providers had access to all Veterans Health Administration (VHA) services, including home nursing services and a home telehealthmonitoring program that included remote nursing support.

Intervention period. In December 2011, we created a transitional care pathway in which the hospitalist team was asked to refer all hospitalized patients with a primary diagnosis of HF for follow-up within 7 to 14 days. The triage procedure was as follows: Once a HF clinic consult was received for these hospitalized patients, the patient was met by the cardiology nurse prior to hospital discharge. The nurse provided the patient with a blood pressure (BP) cuff and scale. She also provided HF clinic contact information and limited instructions regarding self-care, scheduled the patient for a HF clinic appointment, and ordered laboratory tests. If the patient had already been discharged from the hospital (eg, weekend or after-hours discharge), the nurse completed clinic scheduling by phone. The nurse also was available during daytime hours after the patient’s discharge to answer questions, triage concerns, and receive medication refill requests. When necessary, she facilitated clinic scheduling or instructions to patients residing in SNFs or being followed by a community home care service. However, the nurse did not do any prospective case management or active patient surveillance.

Patients were seen by a newly hired HF cardiologist or 1 of the 3 existing nurse practitioners. Similar to the preintervention period, providers had access to the VHA home nursing services as well as home telehealth monitoring. Providers were encouraged to use all available resources including telehealth monitoring, although this resource was not typically available until at least 3 to 4 weeks after hospital discharge. In this new clinic, 12 clinic slots per week were available.

Statistical Analysis

Descriptive statistics and cumulative incidence (within 30 days of discharge) of individual and composite outcomes were calculated. The composite outcome was defined as occurrence of all-cause death, all-cause hospitalization, or all-cause ED visit within 30 days of index hospitalization discharge. Hospital utilization at our VA hospital was captured as well as ascertainment of death, regardless of location; however, hospital utilization at outside hospitals was not available in the VHA electronic record.

We created a Cox proportional hazards model to evaluate the effect of the postintervention period compared with the preintervention period on 30-day composite outcome, adjusted for a priori clinical predictors associated with the outcome (P < .1, without adjusting for multiple comparisons). We evaluated all collected clinical variables as potentially associated with the outcome. These variables were age, gender, a composite of race and ethnicity, systolic BP, pulse rate, hemoglobin, BUN, serum creatinine, eGFR, BNP, length of index hospital stay, and LVEF. We chose to evaluate the intervention period as the predictor variable because this follows an intent-to-treat principle and is less biased than a comparison of only individual patients who were referred for cardiology follow-up vs those who were not referred.

RESULTS

Study Cohort

Of 282 patient records, 21 were excluded due to nonqualifying index hospitalization, admission from or discharge to hospice, discharge to SNF, or refusal to follow up; thus, data for 261 patients were analyzed (preintervention, n = 142; post intervention, n = 119). Patients were predominantly male (97%) and white (71%) (Table 1). Mean (SD) age was 70 (11) years. Sixty-five percent of patients had HF with systolic dysfunction (LVEF ≤ 40%). Of baseline variables, only BNP differed between the pre- and postintervention periods (median [interquartile range], 1212 [613-2752] vs 990 [426-1726], respectively; P = .04). There was no difference in eGFR, BUN, hemoglobin, LVEF, systolic BP, or pulse rate (eAppendix Table [available at ajmc.com]).

Reach: Timeliness of Posthospital Care

Forty percent of the postintervention group had a scheduled cardiology appointment within 7 days of hospital discharge vs 12% of the preintervention group (P < .001) (Table 2 and Figure 1). At 14 days post discharge, 55% of the postintervention group had a scheduled appointment vs 31% in the preintervention group (P < .001). Similarly, at 7 and 14 days post discharge, 24% and 41% of the postintervention group completed a cardiology appointment vs 10% and 23% in the preintervention group, respectively (P = .003 and P = .002).

Effectiveness

Unadjusted cumulative incidence of clinical outcomes (all-cause ED visit, rehospitalization, or death). Compared with the preintervention group, the postintervention group did not have a significantly lower cumulative incidence of the composite outcome at 30 days (35% vs 27%; P = .149) (Table 3 and Figure 2).

Adjusted 30-day clinical outcomes. In the RE-AIM framework, we defined effectiveness as the hazard of the 30-day composite outcome in the postintervention period compared with the preintervention period, adjusted for a priori clinical predictors associated with the outcome. Before adjustment, there was a suggestive but not statistically significant postintervention effect (HR, 0.71; 95% CI, 0.45-1.10; P = .12). However, after adjustment for significant clinical predictors (non-White ethnicity, LVEF ≤ 40%, hemoglobin, systolic BP, BNP [log]; P < .1 with outcome), the postintervention period was associated with a reduced hazard of the 30-day composite end point (HR, 0.59; 95% CI, 0.37-0.96; P = .04). Non-White race was also associated with a reduced hazard of the composite outcome (HR, 0.50; 95% CI, 0.27-0.89; P = .02) in this model.

Adoption

Adoption was defined as a description of the setting and staff who participated in intervention and reasons for adoption or nonadoption. This intervention was tailored to facility hospitalist physicians and resident physicians who care for hospitalized inpatients with HF. The inpatient providers did not voice any concerns about referring patients to the clinic because it ensured adequate postdischarge care for their patients. Some patients with HF were likely not referred due to physicians’ unawareness of the clinic’s availability or lack of perceived need for cardiology specialty care. This quality initiative was facilitated by a representative from the hospital’s quality management department who helped to conduct monthly meetings. The group’s progress was communicated by the HF cardiologist/physician champion via approximately quarterly 20-minute huddles with hospital leadership in an informal setting. The cardiology nurse practitioners received supervisory assistance and consultation from the HF cardiologist for patients with complicated HF. This supervisory assistance may have facilitated their engagement and self-efficacy to manage complex patients with HF. The cardiology nurse proved an important link not only for patients but also to support the cardiology providers during clinic (eg, obtaining emergency oxygen supply, transporting unstable patients to ED).

Implementation

Implementation was defined as the consistency, time involved, costs, adaptions made, and implementation strategies utilized. The physician champion lobbied for creation of new clinic capacity by a “carve-out” in which appointment slots were designated for posthospital follow-up patients with HF, and the cardiology service chief was supportive of this effort. At that time, the intervention was launched in the context of national efforts to reduce HF readmissions. An additional enabler was the presence of a HF physician champion from within cardiology rather than the quality management department. The nursing service was supportive of the new role of the cardiology nurse as a transition navigator, but this support required lobbying for these additional nursing duties. In addition, our team presented at internal VA conferences, which helped us to gain organizational visibility and reward the team’s efforts. After the intervention was implemented, we added the in-person nurse visit to the patient’s bedside to create a more personalized introduction to cardiology clinic, instruct on self-care measures, and dispense BP cuffs and scales. We met some resistance to dispensing medical equipment at the bedside, and we resolved this by meeting with the prosthetics department. Another challenge was that the volume of referrals dropped at times. This was resolved after the HF physician met with hospitalist physicians to remind them about the program, and the cardiology nurse began monthly announcements with rotating medical residents.

Maintenance

Maintenance was defined as the extent to which the intervention was sustained as part of routine organizational practice. The intervention’s costs consisted primarily of HF provider clinic time (12 slots per week), nurse support (approximately 0.25 full-time equivalents), and patient self-monitoring equipment (BP cuff, scale). Our original nurse transferred to another hospital department and the second nurse transferred to another department after completing an advanced degree. The program did effectively meet the capacity demands in that available clinic slots were adequate for the number of referrals and HF discharges. Unfortunately, after 5 years of the program’s operation, the HF cardiologist and 3 nurse practitioners either moved to other institutions or retired. The new cardiology leadership closed the clinic and reassigned the cardiology nurse to other roles.

DISCUSSION

The creation of a transitional HF care intervention was successful in improving the proportion of patients who received and completed HF cardiology appointments within 7 and 14 days of discharge. At 7 days post discharge, for example, twice as many patients in the intervention group had completed a cardiology appointment. In adjusted analysis, the intervention appeared effective in reducing the risk of significant clinical outcomes (ie, all-cause ED visit, rehospitalization, death). The intervention was readily accepted by referring physicians and required modest personnel resources.

Early hospital follow-up care within 7 to 14 days of hospital discharge for HF is a guideline recommendation and quality indicator.^2,5,6 A review by the Ontario government reported that the evidence base was of low quality, with a “lack of large, methodologically robust studies specifically focusing on the effectiveness of 7-day follow-up after discharge”⁷ and identified only 3 nonrandomized studies. A case-control study of discharged patients with HF (n = 11,985) in a large US integrated health system found that follow-up within 7 days was associated with a 19% lower odds of 30-day all-cause readmission (adjusted OR, 0.81; 95% CI, 0.70-0.94).⁸ A US Medicare study also found an association between hospital-level early follow-up within 7 days (hospital level follow-up rates were grouped into quartiles) and patient outcomes.⁹ Compared with patients in the first quartile, patients in the higher quartiles (ie, from hospitals with higher rates of follow-up) had lower rates of 30-day composite of all-cause readmission or mortality (risk-adjusted HR, 0.90; 95% CI, 0.83-0.98 for the fourth quartile). Our effectiveness findings are in line with these studies; we reported that the intervention period was associated with a lower adjusted risk of ED visit, hospitalization, or death within 30 days (HR, 0.59; 95% CI, 0.37-0.96; P = .04). One of the limitations of these observational studies is the potential bias introduced by the role of patient compliance with early follow-up; specifically, noncompliant patients are less likely to complete a follow-up visit and therefore the effectiveness of the early follow-up visit will appear greater. In contrast, our study design avoids the bias of noncompliance by defining the exposure as the intervention period rather than patient-level completion of the follow-up visit. Our study with its interventional design thereby extends the current observational literature by demonstrating that posthospitalization clinical outcomes are in fact improved with earlier and systematic cardiology follow-up visits.

Our study adds to the existing literature by also offering a qualitative evaluation of an early hospital follow-up intervention. Our findings illustrate the potential value of postdischarge cardiology specialty care, which contrasts with the traditional emphasis on follow-up in the primary care medical home in the VA. Further, we observed strong adoption of the intervention, which may have been facilitated by the “low-friction” intervention that did not require new workflows for the referring physicians. Implementation was further enabled by having a local physician champion, supportive hospital leadership, and a conducive national context. We do believe that a crucial aspect of this intervention was the creation of clinic access via dedicated clinic slots for early follow-up care. Creating these “landing spots” for recently discharged patients is a challenging issue for hospital management given the competing needs to maintain equitable and timely access for all patients. Fortunately, we did not experience a significant degree of unused clinic slots because we were also able to utilize the slots for subsequent patient follow-up visits. Perhaps fortuitously, we did not experience any major hurdles in implementing the intervention itself, probably because we had the support of departmental management. The intervention’s primary weakness was its inability to persist during change in HF provider staffing and departmental leadership. Ideally, this transitional intervention would have spread to other clinical departments and become embedded in institutional quality management efforts.

Limitations

There are several challenges and considerations in the current study. We implemented a multicomponent, multidisciplinary intervention, and the effectiveness of the individual components is difficult to ascertain. Nonetheless, we hope that by providing a detailed description of the intervention and discussing lessons learned during implementation, our experience may be of value to other institutions with similar needs for early postdischarge care. Next, the pre- vs postintervention design is a nonrandomized analysis. Patients in the preintervention period could have been sicker, given that their BNP levels were higher, and their referral for posthospital follow-up may have therefore been confounded by disease severity; however, this difference may not have been substantive because all other important HF-related predictors, including age, eGFR, BUN, hemoglobin, LVEF, systolic BP, and pulse rate, were not different. Additionally, there may have been other unknown and/or unmeasured confounders that were not accounted for, including temporal trends, comorbidities, or other cointerventions for HF. Our study was limited primarily to men due to the veterans hospital setting, which may limit translation of the study findings. Furthermore, our observed effect size should be interpreted with caution because it was significant only after adjustment for predictors; however, numerically, the effect size was similar before (HR, 0.71) and after (HR, 0.59) adjustment. One of the strengths of this analysis is that we present the incidence of scheduled appointments in addition to completed appointments, which provides greater insight into opportunities to improve early patient follow-up.

CONCLUSIONS

Our study findings, if replicated, offer the hopeful possibility that broader, universal implementation of early hospital follow-up after HF hospitalization could significantly improve quality of patients’ lives by reducing burdensome rehospitalizations. Although the VA has built the primary care medical home model (Patient Aligned Care Team), which incorporates tools of advanced clinic access,¹⁰ there is no equivalent national model in the VA for cardiology specialty care. Much work will need to be done in ambulatory care clinic redesign in order to improve early postdischarge clinic access for cardiac patients. Health systems including the VA might benefit from explicitly matching patients with complex needs to more provider time and rapid clinic access. Other authors have also suggested that HF ambulatory care is due for a redesign that improves value and care coordination and utilizes enabling technologies.^11,12 Based on our experiences, we suggest that the VA should consider designating clinicians and nurses at every hospital nationwide to be “HF champions” with training in clinical and administrative leadership. These HF champions should be trained in strategies of advanced clinic access, leveraging telemedicine, and creation of multidisciplinary care teams. Some of these individual programs already exist at individual VA medical centers, including shared medical appointments,¹³ HF guideline-directed medical therapy dashboards, and pharmacist-led clinic visits¹⁴; however, there is a need for better adoption and dissemination across the entire VA health system in order to gain maximum population-level impact.

Acknowledgments

The contents do not represent the views of the US Department of Veterans Affairs or the United States Government. The authors are grateful to the cardiology nurses, cardiology nurse practitioners, quality management staff, and Dr Charles Oh for their efforts to make this intervention a reality. They also thank the hospital medicine section and internal medicine residents for their collaboration in the care of these patients.

Author Affiliations: Southern Arizona VA Health Care System, (SD), Tucson, AZ; Phoenix VA Health Care System (JF, DS), Phoenix, AZ; Rush University Medical Center (SA), Chicago, IL; Providence VA Medical Center (WCW), Providence, RI; Northern California VA Health Care System (DS), Sacramento, CA.

Source of Funding: None.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (SD, WCW); acquisition of data (SA); analysis and interpretation of data (SD, JF, WCW, DS); drafting of the manuscript (SD, SA); critical revision of the manuscript for important intellectual content (SD, JF, SA, WCW, DS); statistical analysis (JF, DS); administrative, technical, or logistic support (SD, JF, DS); and supervision (SD, WCW).

Address Correspondence to: Sandesh Dev, MD, MS, Southern Arizona VA Health Care System, 3601 S 6th Ave, Tucson, AZ 85723. Email: Sandesh.dev@va.gov.

REFERENCES

1. Dharmarajan K, Hsieh AF, Kulkarni VT, et al. Trajectories of risk after hospitalization for heart failure, acute myocardial infarction, or pneumonia: retrospective cohort study. BMJ. 2015;350:h411. doi:10.1136/bmj.h411

2. About H2H. American College of Cardiology Quality Improvement for Institutions. Accessed October 14, 2019. https://cvquality.acc.org/initiatives/hospital-to-home/about-h2h

3. RE-AIM. Accessed October 27, 2019. https://www.RE-AIM.org

4. Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322-1327. doi:10.2105/ajph.89.9.1322

5. Yancy CW, Jessup M, Bozkurt B, et al; American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation. 2013;128(16):e240-e327. doi:10.1161/CIR.0b013e31829e8776

6. Specification Manual for Joint Commission National Quality Measures (v2018A): ACHF-02. The Joint Commission. 2018. Accessed October 15, 2019. https://manual.jointcommission.org/releases/TJC2018A/MIF0236.html

7. Health Quality Ontario. Effect of early follow-up after hospital discharge on outcomes in patients with heart failure or chronic obstructive pulmonary disease: a systematic review. Ont Health Technol Assess Ser. 2017;17(8):1-37.

8. Lee KK, Yang J, Hernandez AF, Steimle AE, Go AS. Post-discharge follow-up characteristics associated with 30-day readmission after heart failure hospitalization. Med Care. 2016;54(4):365-372. doi:10.1097/MLR.0000000000000492

9. Hernandez AF, Greiner MA, Fonarow GC, et al. Relationship between early physician follow-up and 30-day readmission among Medicare beneficiaries hospitalized for heart failure. JAMA. 2010;303(17):1716-1722. doi:10.1001/jama.2010.533

10. Patient care services: Patient Aligned Care Team. US Department of Veterans Affairs. Accessed October 15, 2019. https://www.patientcare.va.gov/primarycare/PACT.asp

11. Joynt Maddox K, Bleser WK, Crook HL, et al; American Heart Association Value-Based Models Learning Collaborative. Advancing value-based models for heart failure: a call to action from the Value in Healthcare Initiative’s Value-Based Models Learning Collaborative. Circ Cardiovasc Qual Outcomes. 2020;13(5):e006483. doi:10.1161/CIRCOUTCOMES.120.006483

12. Oliveira GH, Al-Kindi SG, Pronovost PJ. Dose titration of ambulatory care for heart failure: a new paradigm to keep patients healthy at home rather than healing in hospital. Circ Cardiovasc Qual Outcomes. 2019;12(12):e005952. doi:10.1161/CIRCOUTCOMES.119.005952

13. Wu WC, Parent M, Dev S, et al. Group medical visits after heart failure hospitalization: study protocol for a randomized-controlled trial. Contemp Clin Trials. 2018;71:140-145. doi:10.1016/j.cct.2018.06.015

14. Milfred-LaForest SK, Gee JA, Pugacz AM, et al. Heart failure transitions of care: a pharmacist-led post-discharge pilot experience. Prog Cardiovasc Dis. 2017;60(2):249-258. doi:10.1016/j.pcad.2017.08.005