Aiming to Improve Readmissions Through InteGrated Hospital Transitions

BACKGROUND

Hospital inpatient care accounts for the largest share of total health care expenses in the US. In 2012, the average cost per inpatient stay was $10,400 with total aggregated hospital costs of $377.5 billion. (Moore, 2012; and Weiss, 2012) In addition to this expense, transitions from the hospital represent a particularly vulnerable time when patients are susceptible to experiencing adverse events, which is especially true for patients with complex chronic comorbidities. For example, twenty percent of Medicare patients are readmitted to the hospital within 30 days, and 20% of all discharged patients suffer from a preventable adverse event within three weeks of discharge.(Rennke, 2013) Despite national efforts to improve transitions, the hospital discharge process largely remains chaotic, fragmented, and associated with poor outcomes. (Rennke, 2013; Hansen, 2011; Jack, 2009; Kansagara, 2015; Kansagara, 2011; & Kripalani, 2014). Currently, even the highest risk patients leave a complex, acute hospital stay to then arrive in primary care offices that are often neither equipped nor resourced to manage the challenges inherent to such transitions. Indeed, transitioning the highest risk, complex patients will require a paradigm shift in care delivery that is intensive, multidisciplinary, coordinated, and encompasses wrap-around services.(http://www.isc.hbs.edu/health-care/vbhcd/Pages/integrated-practice-units.aspx.)

RATIONALE

To better enhance transitions of care for the highest risk, complex patients, Carolinas HealthCare System (CHS) has designed an Integrated Practice Unit, called Transition Services (CHS-TS). The IPU model is based on the concept that significant improvements in outcomes will come from physician-led, team based, integrated services that are focused on a specific condition or segment of the population.(http://www.isc.hbs.edu/health-care/vbhcd/Pages/integrated-practice-units.aspx) Also integral to the IPU model is having an advanced informatics platform to continuously measure costs, processes, and patient outcomes. CHS-TS aims to improve patient outcomes through innovative approaches that leverage analytics and technology, while bridging care coordination and communication gaps. During their hospitalization, CHS-TS patients enter into a transition pathway that includes the following key services: integrated access to medical, pharmacist, and specialty providers; access to CHS disease specific management programs; dedicated care management services delivered in home and at the clinic; lab and infusion services; palliative care consultations when appropriate; and paramedicine for 24 hour support. Because this population faces many barriers to accessing traditional care, CHS-TS leverages virtual technology to provide visits with patients in their home setting when necessary and appropriate. Whether virtual or in person, each patient will receive the following seven core components (the first four of which will be used as markers for the implementation (RE-AIM) evaluation:

(i) Introduction to CHS-TS process prior to discharge (ii) Hospital follow-up evaluation within 72 hours either in home with paramedicine or CHS-TS clinic (iii) Medication reconciliation by a pharmacist within 72 hours (iv) Weekly contact with care management team (v) Entry into the Heart Success Program if appropriate (vi) Access to 24/7 phone support, 24/7 paramedicine visits, and same day clinic scheduling (vii) Coordinated transition to the next appropriate care location after 30 days from time of discharge

CHS strives to provide extraordinary care to patients and the communities it serves. This proposed evaluation is designed to guide CHS strategy and quality improvement by applying research methodology and data analytics to support the environment of a Learning Health System. Ultimately, as CHS deploys resource intensive interventions like the CHS-TS, it is important for the system and its patients to know answers to questions such as: What type of patient does the CHS-TS help?; How much benefit might be expected?; and At what cost?

AIRTIGHT (Aiming to Improve Readmissions Through InteGrated Hospital Transitions) is a pragmatic, randomized quality improvement evaluation, which seeks to evaluate the effects of the role-out of CHS-TS services for patients at high risk for a 30-day readmission. This outcomes evaluation of a quality improvement intervention is designed to be a seamless part of the routine care in a real-world setting.

INVESTIGATIONAL PLAN

Overall Study Design

AIRTIGHT is a pragmatic, randomized quality improvement evaluation. Patients who are identified by risk modeling as having a high risk (>20%) for a 30-day readmission will be eligible for CHS-TS referral (referral eligible). Each weekday, referral eligible patients will be randomly assigned to CHS-TS referrals with a total number targeted to the CHS-TS's capacity. The CHS TS capacity will be revaluated on a daily. Because variables that affect eligibility may change throughout the course of a hospital stay, exclusion criteria are applied again at the time of discharge for the purposes of analysis. For example, patients not discharged to home are excluded. A diagnosis of heart failure is also added as an additional exclusion criterion at the time of discharge because CHS Heart Success recruits patients from a similar population and there is significant overlap in the services offered between CHS TS and CHS Heart Success. All patients who remain eligible at the time of discharge (discharge eligible) will be included in the analysis. Planned enrollment is 1520 patients (n=760 per arm) with the engagement in CHS-TS services ending at the 30th day after discharge. Outcomes data will be tracked for 90 days. To mitigate the risk of selection bias, all eligibility criteria are objective and determined using data analytics within the software package, SAS. Additionally, analysis will be by intention to treat. With the use of this pragmatic design, crossovers from the CHS-TS arm to usual care are expected and may attenuate the measured effect of the CHS-TS. Since this evaluation protocol is part of a quality improvement intervention, the investigators requested that the CHS institutional review board designate the study as expedited research and grant a waiver of informed consent.

Primary Outcome Variable

30-day all cause readmission with index visit as defined by the Centers for Medicaid and Medicare Services (CMS) and with inclusion of observation patients at any CHS facility

Secondary Outcome Variable(s)

• 30-day all cause CHS readmission (as defined by CMS)
• Length of stay upon the index visit
• Length of stay upon the readmission visit
• All cause, 60 and 90-day readmission rate as defined by CMS with inclusion of observation patients at any CHS facility
• 30-day readmission rate among patients with a primary diagnosis of congestive heart failure
• 30-day readmission rate among patients with a primary diagnosis of pneumonia
• 30-day readmission rate among patients with a primary diagnosis of sepsis
• Number of Emergency Department visits within 30 days of discharge
• Intensive Care Unit visits

Independent and Dependent Variable(s)

The dependent variable under assessment is 30-day all cause readmission (yes/no). The primary independent variable of interest is the program assignment (usual care vs. CHS-TS).

Subject Selection

Selection will be through an automated process that generates a referral list at 03:00 of patients admitted during the preceding 24 hour period, called the CHS-TS Referral Eligible List (REL). At the CMC Main campus, where patient enrollment occurs Monday through Friday, the Monday REL contains admissions that meet criteria from Friday-Sunday. At the CMC Mercy campus, where patient enrollment occurs Tuesday through Friday, the Tuesday REL contains admissions that meet criteria from Friday-Monday. The REL is generated based on the study's inclusion/exclusion criteria.

Randomization

Patients that meet the eligibility criteria at the time of admission will be randomized into one of two groups on the CHS-TS Referral Eligible List (REL). A constrained randomization scheme will be utilized allocating referrals into permuted blocks each day for up to 30 total referrals (total includes CHS-TS and usual care). The total referrals are based on daily estimates of clinic capacity. The allocation will be 1:1 CHS-TS: usual care.

Inclusion criteria

• Classified as inpatient or observation as of (00:00), and
• Predixion score ≥ 0.50, and
• Carolinas Hospitalist Group is listed as the primary attending service or consulting service at CMC Main or Mercy campuses, and
• Not discharged at the time of list generation.

Exclusion criteria

• Randomized in the last 90 days into either the CHS-TS or to usual care
• Not a North Carolina resident
• Greater than 2.5 hour drive time from CMC to primary residence
• Psychiatric diagnosis codes within the last 6 months including: Schizophrenia, Suicidal Ideation, Homicidal Ideation, or Psychosis (ICD10 - R45.851, R45.850, F20.x-F29.x)
• Diagnosis of sickle cell anemia in the past year (ICD10 - D57)
• Diagnosis of drug or alcohol dependence within the last 90 days (ICD10 - F10.2x, F11.2x, F12.2x, F13.2x, F14.2x, F16.2x, F18.2x, F19.2x)
• Actively followed for a primary diagnosis of cancer (greater than 2 visits to CHS Cancer Center or on chemotherapy in last 2 months)
• Hospitalized for greater than 72 hours
• Residing in a facility prior to admission (example jail or skilled nursing facility)
• Under the care of hospice prior to admission
• Left Against Medical Advice*
• Disposition other than home (example skilled nursing facility or rehabilitation facility)*
• Disposition home with hospice*
• Heart Failure as a discharge diagnosis *

• This exclusion criterion will only be applied during the analysis.