The Discharge Companion Pilot Program

The Discharge Companion Program (DCP) pilot involved an interprofessional partnership between Tucson Medical Center, a local, nonprofit hospital, and The Medication Management Center, a university-based clinical call center. The program served patients in Tucson and the surrounding rural areas of southern Arizona.

The DCP was implemented to improve overall quality of care and reduce readmissions for patients discharged from the hospital. The Hospital Readmissions Reduction Program (HRRP), established as part of the Patient Protection and Affordable Care Act of 2010, provides financial incentives to hospitals who decrease their readmission rates for Medicare patients (1). In contrast, the Centers for Medicare & Medicaid Services (CMS) imposes penalties on hospitals with high readmission rates. The DCP addressed a need for this hospital that was directly related to these penalties.

The Medicare Payments Advisory Commission estimates that readmissions in 2017 will account for $15 billion of the projected $105 billion in total hospital-related service expenditures (2). The DCP sought to develop and implement a best-practice model to help reduce readmissions that incorporated several established quality improvement (QI) initiatives, including integration of best-practice guidelines into clinical care, reconciliation of medication lists, and an improved continuum of care for patients and health care professionals.

The hospital team, composed of four registered nurse chronic disease coordinators (CDCs), identified the need for a transition-of-care program such as the DCP from a root-cause analysis of key condition readmission rates, along with CMS HRRP data. The intent of this unique interprofessional team-based model was to provide additional pharmacovigilance to assist patients during the transition-of-care period after hospital discharge to decrease readmission rates and avoid penalties.

Pharmacists from the University of Arizona Medical Management Center reviewed patient-reported and electronic health record (EHR) discharge medication lists and health histories to assess alignment with the American Diabetes Association’s (ADA’s) best practices for diabetes management. On implementation of the program, the DCP pharmacist discovered the following issues: four patients (3%) were prescribed a statin at an inappropriate dose intensity (guideline adherence); 16 patients (10%) had barriers to appropriate medication use (adherence); 70 patients (45%) were eligible to receive one or more immunizations (preventive care deficit); and 87 patients (56%) had at least one drug utilization concern (safety).

Additionally, the medication safety evaluation, an important component of medication reconciliation, consisted of key assessment components for 1) therapeutic duplications, 2) drug-disease interactions, 3) drug-drug interactions, 4) dose appropriateness, 5) adverse drug events, and 6) high-risk medication use in the elderly. Medication adherence was also addressed. The DCP pharmacist also ensured the appropriateness of and adherence to national consensus guideline recommendations from the ADA.

This pilot program was implemented between August 2015 and July 2016. Subsequently, based on the positive findings of the pilot program, the local hospital approved full-scale implementation of the DCP.

The DCP was co-coordinated by two core QI team members: the pharmacist coordinator and the nurse coordinator. These members were selected to lead the DCP given their considerable individual and collective expertise, including respective unique scopes of practice and previous experience with innovative telehealth models.

Patients at risk for hospital readmission were identified and enrolled in the DCP pilot if they met the following criteria: 1) were ≥65 years of age and 2) had a primary discharge diagnosis of diabetes or other targeted conditions, including asthma, chronic obstructive pulmonary disorder, heart failure, myocardial infarction, pneumonia, post-coronary artery bypass graft, and renal failure. High-risk medication use in the elderly (e.g., insulin) was a consideration when patients were referred to the program; however, it was not an inclusion criterion. Initially, all qualified patients were referred to the DCP nurse by the hospital’s CDC team. Subsequently, the DCP nurse scheduled these patients for individual telephone consultations with the DCP pharmacist.

The initial patient consultation was performed via telephone by the DCP pharmacist within 1 week after discharge. Complete access to the hospital EHR was available to the DCP pharmacist for review, documentation, and preparation for the consultation. The consultation method used was dependent on where each patient was discharged to; for patients returning home, the DCP pharmacist spoke directly with patients or their caregivers (71%, n = 110), whereas for those discharged to a skilled nursing facility (SNF), the DCP pharmacist obtained the relevant medication administration record from the SNF nurse, completed the consultation based on the information provided by the SNF staff, and communicated recommendations directly to the SNF nurse (29%, n = 44).

For the direct pharmacist-patient/caregiver consultation method, teach-back education was a unique feature of this transitions-of-care model. This technique is designed to provide effective clinician-patient/caregiver communication by asking the latter to “teach back” what they have learned (3). The pharmacist assessed whether patients/caregivers understood 1) how to identify signs of a worsening condition, 2) when to seek medical attention, and 3) what lifestyle modifications to implement to prevent worsening of their condition. Additionally, the pharmacist completed a comprehensive medication safety evaluation. After the initial patient consultations, the DCP nurse coordinator contacted patients’ primary care provider, specialists, and/or dispensing pharmacy via telephone or facsimile to apprise them of the pharmacist’s recommendations and concerns and any pertinent information related to the patients’ hospitalization.

For the SNF staff consultation approach, to improve communication and avoid challenges in acquiring information by telephone (e.g., scheduling conflicts based on availability of both parties), the DCP nurse coordinator requested the medication administration record via facsimile. If this was unsuccessful, the pharmacist followed up with SNF staff within 24 hours via telephone. When the consultation was completed with SNF nursing staff, the DCP pharmacist’s notes were shared with the facility via facsimile.

For follow-up consultations, the DCP pharmacist contacted patients/caregivers or SNF staff again 3 weeks after discharge. The pharmacist revisited previously identified issues to ensure sufficient resolution and addressed any new issues regarding nonadherence, medication costs, medication changes, adverse drug events, and adherence to national consensus treatment guidelines (chronic) and Centers for Disease Control and Prevention–recommended vaccines (i.e., for influenza, pneumonia, and herpes zoster). Subsequently, the DCP nurse coordinator directly relayed any recommendations from this consultation to patients’ health care team, including specialists, primary care provider, and/or dispensing pharmacy, as needed.

Standardized, proprietary call script materials, designed for each specific discharge diagnosis based on national guidelines, were created to ensure consistency of the clinical evaluation and educational message content. The call script materials addressed three major topics: 1) how to identify signs of a worsening condition, 2) when to call for emergency assistance versus following up with a patient’s own physician, and 3) how to make lifestyle modifications to improve a patient’s overall health and well-being. Within the EHR software system, the DCP pharmacist created templates for each type of consultation (e.g., direct to patients/caregivers or via SNF staff). Additionally, the DCP nurse coordinator used a standardized call script to communicate with providers regarding the reasoning behind certain recommendations. For example, when the DCP nurse called a provider’s office, she reviewed the cardiac benefits of adding a statin cholesterol medication for patients with diabetes between 40 and 75 years of age.

A total of 482 patients from the local hospital were referred for participation in this pilot DCP during the 12-month study period. Of these, 225 patients (47%) had type 2 diabetes. Among patients with type 2 diabetes, 154 (68%) participated and received services within 30 days after discharge, including an initial telephone consultation within 1 week after discharge and follow-up telephone consultation 3 weeks after discharge. Based on the initial safety concerns identified on program implementation, respective successful interventions were made. Details are listed in Table 1.

The DCP pharmacist made a total of 166 vaccine recommendations for influenza (n = 33, 20%), shingles (n = 70, 42%), and pneumonia (n = 63, 38%). Of the 63 patients who received a recommendation for a pneumococcal vaccine, 57 (90%) accepted the recommendation, whereas 6 (10%) stated that they did not want to receive the immunization. Similarly, a majority of patients accepted the immunization recommendations for influenza (n = 26, 79%) and shingles (n = 62, 89%) vaccines.

The DCP pharmacist addressed medication adherence concerns for all patients with medication-use difficulties related to patient-reported problems (n = 16), including access barriers due to cost (n = 7, 44%) and transportation (n = 1, 6%), self-reported medication use inconsistent with prescription directions (n = 4, 25%), and recommended strategies to improve medication adherence (n = 4, 25%) for patients who had difficulty remembering to take their medications. The DCP pharmacist also contacted prescribers to recommend adding statin therapy for all patients not currently following a dose regimen, based on the current ADA guidelines. For patients with type 1 or type 2 diabetes who were 40–75 years of age and had an estimated 10-year atherosclerotic cardiovascular disease (ASCVD) risk ≥7.5%, high-intensity statin therapy was recommended. Patients with a 10-year estimated ASCVD risk <7.5% were advised to receive moderate-intensity statin therapy.

Although participation in the DCP resulted in a reduction in 30-day hospital readmissions among patients with type 2 diabetes, that decrease was not statistically significant (Table 2). The trends in readmission rates are promising, and the nonsignificance of the findings may be attributed to a limited sample size and insufficient statistical power. The QI analysis showed positive patient perceptions regarding the DCP services. Furthermore, results from a survey of providers practicing in close proximity to the hospital showed that this type of interprofessional model may positively influence their future referral practices.

Future goals for the DCP are to expand services and collaborate with accountable care organizations and other hospital entities. Telephone consultations and remote access to the EHR enable and facilitate DCP implementation regardless of patients’ proximity to hospitals, allowing for expansion of services into underserved and rural areas. Future plans include publishing other articles reporting more in-depth results, including 30-, 60-, and 90-day post-hospitalization readmission rates, return-on-investment, and Medicare beneficiary expenditures, as well as to re-examine the data to assess all hospital readmissions (i.e., not solely readmission to the same facility).

An interprofessional team collaboration with the local hospital’s nurse CDCs was crucial to the success of the pilot DCP. Additionally, the QI process provided opportunities to develop creative solutions to address issues based on receiving medication lists from the staff at the SNFs.

No potential conflicts of interest relevant to this article were reported.