A descriptive report on the impact of pharmacy workflows in the operational success of hospital at home implementation at a county academic hospital system

Abstract

Background

The hospital at home (HaH) model has become more prevalent in the American healthcare system due to its ability to decrease acute care costs and readmission risk. Recent publications have provided guidance on optimizing medication management and patient safety by leveraging clinical pharmacy services. There is limited data on pharmacoeconomic impact of HaH implementation, specifically in underinsured patients.

Objective(s)

To describe the development of HaH-related pharmacy workflows and evaluate the operational success of the program in an underinsured patient population.

Methods

This report describes HaH program implementation between August 1st, 2022, and March 19th, 2024. Patients were eligible for home treatment if they met geographic, clinical, financial, and social criteria.

Outcomes

The primary outcome measured was the quantity and cost of medication waste for patients treated at home. Secondary outcomes included HaH medication turnaround time, healthcare resource utilization, and patient safety. All study outcomes were reported using descriptive statistics.

Results

Out of 450 patients screened, 3 met criteria and provided consent for HaH enrollment. The total cost of medication waste for all 3 patients was $41.15, and 21 out of 53 dispensed doses (40 %) were wasted. The mean medication verification time was 8.1 min, and the mean medication preparation time was 50.2 min.

Conclusion

Study data provides insight into enhancing dispensing practices while establishing the benefits pharmacists bring to the HaH care team. Future research should elaborate on other measures of operational success to identify optimal performance metrics to support expanding pharmaceutical services within the HaH care model.

Highlights

• •Hospital at home programs promote health equity for underinsured patients.
• •Medication waste and turnaround time evaluate the success of pharmacy workflows.
• •Operational efficiency and pharmacy turnaround time improves with each HaH patient.
• •Optimizing prescribing patterns help reduce “as needed” medication waste.

1. Introduction

Hospitals are the standard care site for patients with acute ailments, but studies have shown that hospital wards can predispose patients to hospital-acquired infections, acute delirium, and other iatrogenic illnesses.^1,2 According to the Agency for Healthcare Research and Quality (AHRQ) in 2018, the 30-day readmission rate for Medicare and Medicaid patients was 17 % and 14 %, respectively, with an average cost of $15,200 for each readmission.³ The top diagnoses associated with 30-day readmissions include heart failure, diabetes with complications, chronic obstructive pulmonary disorder (COPD), pneumonia, acute and chronic renal failure, urinary tract infections, and skin and soft-tissue infections.³ Many of these conditions are traditionally managed in the emergency department (ED) or a general hospital ward, but issues with overcrowding have become widespread leading to an increase in patient length of stay and a reduction in quality of care.⁴

Hospital at home (HaH) is an innovative care model that provides hospital-level acute care services in the patient's home. Previous studies have shown that home-based care can provide acute care cost savings and increased patient satisfaction while maintaining quality and safety standards associated with in-hospital care for select patients with acute conditions.5, 6, 7 A meta-analysis conducted by Arsenault-Lapierre and colleagues in 2021 examined 9 randomized controlled trials of HaH patients with a combined sample size of 959 patients and found a risk ratio of 0.74 for readmission risk in patients treated at home with no difference in mortality compared to patients treated in-hospital.⁵ Another study conducted at 3 hospital sites by Leff and colleagues in 2005 found that of 214 eligible patients for HaH, 66 % were offered the choice between HaH and in-hospital care with 60 % of patients selecting to be treated at home.⁶ In 2020, Levine and colleagues published a randomized controlled trial in the United States and reported that the mean cost of acute care episodes was 38 % lower in the HaH group (n = 43) versus patients treated in-hospital (n = 48).⁷ Furthermore, there is a knowledge gap in outcomes of any kind regarding HaH programs in underinsured patients, defined as patients with insurance that does not meet their medical needs.⁸ According to a September 2024 report published by the Centers for Medicare and Medicaid Services (CMS) that reviewed 12,000 patients across 151 hospitals found that patients accepted into HaH programs were more likely to be White (83 %) and live in an urban area (93 %) while being less likely to receive Medicaid (12 %) or low-income subsidies (1.6 %).⁹ This CMS report reveals the need for additional research that focuses on outcomes from HaH implementation in underinsured patient populations to improve health equity.

While the current literature demonstrates the benefits of a specific demographic of patients being treated at home for certain acute conditions, there is limited data on the pharmacoeconomic impact of HaH implementation. Recent articles published in pharmacy-specific journals such as the American Journal of Health-System Pharmacy and the Canadian Pharmacists Journal have showcased different implementation strategies and challenges experienced in preparing and dispensing medications within the HaH setting.10, 11, 12 These studies provide guidance to hospital systems looking to optimize medication management and patient safety by leveraging clinical pharmacy services.10, 11, 12 However, given the novelty of the HaH model, there is a paucity of data that demonstrates the impact of pharmaceutical services on the efficiency of an implemented HaH program, and no data on operational outcomes that measure medication waste and total time to dispense medications from the inpatient pharmacy.

The purpose of this study was to develop an efficient pharmacy workflow for HaH implementation and evaluate the operational success of the program in an underinsured patient population at a county academic hospital system.

2. Methods

2.1. Study design

This report was an implementation-based case study conducted from August 1st, 2022, to March 19th, 2024 as this aligned with the rollout of the HaH program. Study investigators identified eligible patients upon admission to the emergency department using a tool within the electronic health record (EHR) that automatically sorted patients into a list based on the distance of their home address from the hospital. The HaH operations director, nursing manager, HaH provider (physician or nurse practitioner), and clinical case manager then manually screened patients identified by this tool via chart review to ensure clinical, social, and financial eligibility criteria were met. If deemed eligible, the HaH provider placed 2 final screening consults, one to pharmacy and one to nutrition, to determine if treating the patient at home was manageable from a medication and dietary perspective. Once screening was completed and the patient qualified, the HaH provider counseled the patient on the HaH program and offered home treatment as an option rather than in-hospital care within the brick-and-mortar pavilion. If the patient agreed to receive care at home, a consent form was signed, and the transfer process was initiated. The study included a single group of patients who were enrolled in HaH after the program went live on February 26th, 2024. Study results were designed to report findings regardless of time spent in the HaH program. Given the brief study period after program implementation, seasonal variations were expected to have minimal impact on outcomes.

The institutional review board for the participating health system deemed this study exempt from requiring review and approval. All participants were provided with information regarding the study intervention.

2.1.1. Setting and participants

The study was performed at a county, academic health system with 2 hospitals. Development and preparation of HaH implementation occurred at both county hospital pavilions, but identification, screening, and enrollment of patients into the study occurred at the hospital site designated as a level I trauma center with 402 beds. The health system conducting the study primarily serves uninsured and Medicaid patients, including those with multiple barriers to accessing healthcare.

Patients were eligible for enrollment into the HaH program if they were 18 years or older, lived within an approved zip code approximately 10 miles from either brick-and-mortar hospital site, were insured through traditional Medicare or uninsured, and admitted to the ED with a primary diagnosis of asthma exacerbation, COPD exacerbation, infection, chronic kidney disease requiring diuresis, or diabetes with related complications. Uninsured patients enrolled in the health system's financial assistance program were also considered for inclusion. Patients were excluded from HaH program enrollment if they did not meet clinical, social, and financial eligibility criteria. The specific infections included as a primary diagnosis for HaH enrollment and a detailed list of exclusion criteria can be found in the Supplemental Appendix (eMethods, eText 1–2).

2.1.2. Screening and enrollment

450 patients were screened for HaH program eligibility with 6 patients meeting enrollment criteria. Of the 444 patients who did not meet program eligibility, 12 % lived outside of the approved zip codes, 48 % had clinically complex diagnoses not conducive to treatment at home, 30 % had an insurance plan that did not cover HaH services, and 10 % did not meet social inclusion criteria. Of the 6 patients who qualified, 2 were excluded due to the hospital not having the capability to provide necessary services at home, and 1 patient did not provide consent. The remaining 3 qualified patients provided consent and were enrolled in the program to be treated at home (Fig. 1).

Fig. 1.

Study flow diagram for hospital at home screening and enrollment.

Abbreviations: HaH, hospital at home.

2.1.3. Intervention

Patients at home received daily visits from the HaH provider for evaluation and twice-daily visits from a HaH nurse for vital sign monitoring. The HaH provider and nurse were available for remote communication via a tablet device provided by the health system. Tablets were available to the patients for the full course of their admission to HaH. Patients treated at home had access to respiratory therapies, diagnostic studies, point-of-care tests, nutrition services, and lab monitoring. Ambulance and emergency medical personnel were available to pick up patients treated at home and drive them to the admitting hospital within 30 min if needed.

2.1.4. Role of the pharmacist

Pre-admission consults were completed by an inpatient pharmacist to evaluate each patient's home medication list and anticipated treatment plan. Medication therapies were reviewed and optimized for success in the home setting. This was achieved by ensuring the stability, dosing, and appropriateness of intravenous (IV) antibiotics administered via elastomeric infusion pumps, conducting renal dose adjustments, ensuring antimicrobial stewardship, completing IV to oral conversions of eligible medications, reviewing drug allergies and drug interactions, and providing drug and dosing recommendations for diabetes management and deep vein thrombosis prophylaxis (Table 1). The inpatient pharmacist also served as a clinical reference during twice-daily HaH huddles to communicate any barriers to home care from a medication perspective.

Table 1.

Hospital at home initial pharmacy screening consult used to evaluate the feasibility and safety of the anticipated medication regimen for use in the home setting.

Clinical Activity	Purpose of Clinical Activity
Optimizing Dosing Regimens	Goal is to have medications be administered: •Daily or twice-daily (if possible) •Orally (IV to PO conversion) •Minimize controlled substance use
Renal Dose Adjustment	Verify all medications are renally adjusted according to health system policy.
Antimicrobial Stewardship and IV Antibiotic Elastomeric Pump Appropriateness	Ensure appropriate antibiotic use to reduce risk of treatment failure and prevent resistant pathogens. Verify stability data and dosing information to convert IV antibiotics to elastomeric pump infusions.
Diabetes Regimen Review and DVT Prophylaxis Dosing Recommendation	Compare inpatient diabetes regimen to home regimen and ensure formulary agents are used. Recommend enoxaparin dosing for DVT prophylaxis (heparin is not used in the home setting).
Drug Allergies	Verify no drug allergies exist with current medication regimen.
Drug Interactions	Record pertinent drug interactions that are present. (i.e., drugs affecting the concentration of other drugs, etc.)

Abbreviations: IV, intravenous; PO, oral; DVT, deep vein thrombosis.

2.1.5. Medication dispensing

Before the patient transferred home, any outstanding therapies were completed in-hospital and administered by hospital nurses within the brick-and-mortar pavilion. The inpatient pharmacy prepared a 24-h supply of medications for the HaH nurse to deliver to the patient's home. A standardized medication pick-up time was established at 1:00 pm every day for all scheduled medications. Any new first doses or “as needed” (PRN) doses were ordered and streamlined for the standardized 1:00 pm pick-up time whenever possible. Any medication changes made after 1:00 pm were codified as either “urgent” or “emergent” by the HaH provider. If the situation was “urgent,” such as a minor diuretic dose decrease, the updated medication order was prepared the following day. If the situation was “emergent,” such as a dose increase for a patient reporting uncontrolled pain, the updated medication order was prepared stat and a HaH nurse would make an additional trip to the inpatient pharmacy to ensure timely delivery and administration at the patient's home.

2.1.6. Medication storage and administration

Medications that left the brick-and-mortar hospital pavilion were not allowed to be returned to the inpatient pharmacy and were either administered to the patient or wasted. Each patient was provided with a tackle box to store medications at home, and the HaH nurse refilled the tackle box during each evening visit. If medications required refrigeration during transport, the inpatient pharmacy prepared cold storage totes with freezer packs and a disposable temperature tracker to monitor for temperature excursions. All HaH nurses carried a handheld Rover® device with barcode scanning capabilities to document medication administration through Bar Code Medication Administration (BCMA) if they were physically present in the patient's home. Otherwise, the HaH nurse would video conference with the patient to witness the self-administration process and document in the Medication Administration Record (MAR).

2.1.7. Post-acute care pharmacy services

All HaH patients were automatically enrolled in the Meds-to-Beds (M2B) program. Patients were provided their post-acute outpatient prescriptions upon discharge and were formally counseled by a M2B pharmacist remotely. Advance communication between the HaH provider and the M2B pharmacy team ensured medications were in stock and available for dispensing. Discharge medications were prepared by the outpatient pharmacy at the brick-and-mortar hospital site and provided to the HaH nurse for home delivery to the patient. The M2B pharmacist provided virtual discharge medication counseling once the prescriptions were delivered to the patient. HaH patients received a follow-up telephone call from the Integrated Pharmacy Services (IPS) team 72 h post-discharge. The IPS team addressed any outstanding medication-related questions, recorded patient-reported adverse effects, and promoted medication adherence. The process map for pharmacy operations related to HaH is detailed in (Fig. 2).

Fig. 2.

Process map of the pharmacy operational workflow related to hospital at home.

Abbreviations: RPh, pharmacist; HaH, hospital at home; M2B, Meds-to-Beds; IP, inpatient; OP, outpatient; IPS, Integrated Pharmacy Services.

*Related to outpatient dispensing process (M2B and IPS).

2.1.8. Pharmacy workflow for HaH

The process started with the inpatient pharmacist screening the patient's anticipated medication regimen for potential drug-related problems and feasibility of administration in the home setting. Once accepted into the HaH program, the patient was automatically enrolled in the M2B program as all HaH patients were eligible to receive discharge prescriptions delivered to their home. The inpatient pharmacist verified all HaH medication orders to ensure the dosage form and frequency were adequate for the home setting. Once verified, the HaH medication labels were printed, and individual doses were prepared by the inpatient pharmacy. Since HaH programs provide inpatient-level care, HaH medication labels were required to adhere to the same labeling rules as medications administered in the brick-and-mortar hospital. However, HaH medication labels were also required to have outpatient label elements such as instructions for use written in layman's terms since patients may need to self-administer oral medications outside of the two nursing visits. After preparation was completed, the inpatient pharmacy used a three-bag method to organize all doses for a single HaH patient. Each dose was individually placed in a small-sized bag, which was then combined into a medium-sized bag to consolidate drugs of the same name. Finally, all medium-sized bags were placed inside one large-sized bag to make delivery easier for the HaH nurse. Completed HaH medications were secured within the inpatient pharmacy until the HaH nurse arrived for pick-up, at which point the inpatient pharmacist reviewed pertinent storage information related to drug stability and documented medication hand-off with the HaH nurse. The inpatient pharmacist conducted daily lab monitoring via electronic chart review and contacted the HaH provider if any concerns arose. Before HaH discharge, the outpatient pharmacy prepared discharge prescriptions that were delivered by the HaH nurse, and the patient received virtual counseling by a M2B pharmacist. After 72 h post-discharge, an IPS pharmacist conducted a follow-up call to answer any medication-related questions.

2.1.9. Study outcomes

The primary outcome of the study was the quantity and cost of HaH medication waste for patients treated at home. Medication waste was stratified into two different groups: frequency type (scheduled versus “as needed” doses) and dosage form (non-IV versus IV doses). Any medication that was not administered to the patient upon discharge from the HaH program was considered waste.

Medication turnaround time by the inpatient pharmacy was a secondary outcome and was defined as the total time to prepare and dispense HaH medications. Medication turnaround time was divided into two segments: verification time and preparation time. Verification time was measured from the time the HaH provider electronically signed the HaH medication order to the time the inpatient pharmacist verified the order. Preparation time was measured from when the HaH medication label was printed from the dispense queue to the completion of the dispense check (i.e., final verification) conducted by the inpatient pharmacist. Final doses prepared for administration in the brick-and-mortar hospital before the patient transferred home were excluded from data analysis.

Additional secondary outcomes related to healthcare resource utilization included medication usage based on total doses dispensed, pharmacy screening consults, documented pharmacist interventions, and discharge education provided. Medication usage was also stratified by frequency type and dosage form. Patient safety outcomes were measured via the frequency of submitted Electronic Incident Reports (EIRs) related to HaH and the number of escalations back to the brick-and-mortar hospital due to clinical deterioration while being treated at home.

Demographic data that was collected included age, gender, race, employment status, insurance status, primary diagnosis for HaH admission, house distance from the hospital in miles, and the number of prescription medications at screening (Table 2). EIRs were recorded through the internal medication error reporting system. All remaining study outcomes were collected through EHR chart review.

Table 2.

Demographic data of patients treated in the hospital at home program.

Variable	Patient Alpha	Patient Beta	Patient Gamma
Age, years	67	51	20
Gender	Male	Female	Female
Race	Hispanic or Latino	Hispanic or Latino	Hispanic or Latino
Employment Status	Unemployed	Employed	Unemployed
Insurance Status	Financial Assistance Program	Financial Assistance Program	Self-Pay
Primary Diagnosis for HaH Admission	Community-Acquired Pneumonia	Complicated Urinary Tract Infection	Pyelonephritis
House Distance from Hospital, miles⁎	9.6	11.4	10.8
Prescription Medications at Screening	5	1	0

Abbreviations: HaH, hospital at home.

^⁎Patient Beta presented at the main hospital site but lived 11.4 miles from the other brick-and-mortar hospital pavilion within the health system. Both Patient Alpha and Patient Gamma lived closer to the main hospital site.

2.2. Statistical analysis

Data on study outcomes were reported for all patients who were enrolled in the HaH program and treated at home (Table 3). If patients were escalated back to the brick-and-mortar hospital to complete their therapy, only the data collected during their time treated at home was included. All study outcomes and demographic data were reported using descriptive statistics such as percentages and means. The cost of medication waste was determined by multiplying the wholesale acquisition cost (WAC) listed by the drug wholesaler times the number of doses wasted. Medication waste was quantified as a ratio of doses wasted to doses dispensed per patient. For the stratified groups, medication waste was calculated as a proportion of the number of scheduled, PRN, non-IV, or IV doses wasted per patient over the total number of doses wasted for that same patient.”Doses Dispensed” was defined as the total doses prepared by the inpatient pharmacy. Doses that were dispensed but not administered were recorded as waste. All statistical analyses were performed using Microsoft Excel® Version 2202 (Build 16.0.14931.20648) by author JL.

Table 3.

Medication utilization among patients treated in the hospital at home program.

Patient	Alpha, n (%)	Beta, n (%)	Gamma, n (%)
Doses Dispensed By Frequency Type	22	15	16
Scheduled Doses Dispensed⁎	10 (45)	12 (80)	4 (25)
PRN Doses Dispensed⁎	12 (55)	3 (20)	12 (75)
Doses Wasted By Frequency Type†	12 (55)	5 (33)	4 (25)
Scheduled Doses Wasted⁎, ‡	0 (0)	2 (40)	0 (0)
PRN Doses Wasted⁎, ‡	12 (100)	3 (60)	4 (100)
Cost of Waste By Frequency Type, USD§	$6.36	$34.28	$0.51
Cost of Scheduled Doses Wasted	$0.00	$33.53	$0.00
Cost of PRN Doses Wasted	$6.36	$0.75	$0.51
Doses Dispensed By Dosage Form	22	15	16
Non-IV Doses Dispensed¶	18 (82)	10 (67)	12 (75)
IV Doses Dispensed¶	4 (18)	5 (33)	4 (25)
Doses Wasted By Dosage Form†	12 (55)	5 (33)	4 (25)
Non-IV Doses Wasted‡, ¶	12 (100)	4 (80)	4 (100)
IV Doses Wasted‡, ¶	0 (0)	1 (20)	0 (0)
Cost of Waste By Dosage Form, USD§	$6.36	$34.28	$0.51
Cost of Non-IV Doses Wasted	$6.36	$1.93	$0.51
Cost of IV Doses Wasted	$0.00	$32.35	$0.00
Medication Turnaround Time
Medication Orders Verified	7	3	8
Mean Verification Time, minutes	6.3	5.7	12.4
Medication Orders Prepared||	22	15	16
Mean Preparation Time, minutes	72.8	54.6	23.3
Healthcare Resource Utilization
Pharmacy Screening Consults	1	1	1
Pharmacist Interventions	2	4	2
Discharge Education Provided	Yes	No	Yes

Abbreviations: HaH, hospital at home; PRN, as needed; IV, intravenous; USD, United States Dollar.

^⁎Scheduled and PRN doses were grouped to represent 100 % of doses dispensed or wasted for each category where they are reported.

^†Doses wasted were calculated as a proportion of the total number of doses wasted per patient over the total number of doses dispensed for that same patient.

^‡Stratified groups were calculated as a proportion of the number of scheduled, PRN, non-IV, or IV doses wasted per patient over the total number of doses wasted for that same patient.

^§Cost of waste is based on wholesale acquisition cost (WAC) pricing from the drug wholesaler.

^¶Non-IV and IV doses were grouped to also represent 100 % of doses dispensed or wasted for each category where they are reported.

^||Medication orders prepared equals the number of doses dispensed for each patient.

2.2.1. Role of the funding source

The Texas Society of Health-System Pharmacists (TSHP) Research & Education Foundation awarded the study investigators grant funds to support pharmacy operations related to the implementation of the HaH program. TSHP had no role in the study design, data collection, outcomes analysis, or the decision to submit the manuscript for publication.

3. Results

450 patients were screened in the hospital's ED for potential inclusion in the program. Three patients (referred to as Patient Alpha, Patient Beta, and Patient Gamma) met the eligibility criteria for admission and provided consent for enrollment. Demographic data for the 3 patients enrolled is shown in Table 2, and detailed case reports for each patient are available in the Supplemental Appendix (eResults, eText 1–3).

3.1. Measures of operational success

Among the 3 patients treated at home, 21 out of 53 total doses dispensed had been wasted (40 %). Of the 21 wasted doses, 10 % were attributed to scheduled medications while 90 % were due to PRN medications being wasted. Patient Alpha wasted 12 out of 22 total doses (55 %) with no wasted scheduled doses and 12 wasted PRN doses. Patient Beta wasted 5 out of 15 total doses (33 %) with 2 wasted scheduled doses and 3 wasted PRN doses. Patient Gamma wasted 4 out of 16 total doses (25 %) with no wasted scheduled doses and 4 wasted PRN doses. The total cost of medication waste was $41.15, with Patient Alpha wasting 12 doses at $6.36, Patient Beta wasting 5 doses at $34.28, and Patient Gamma wasting 4 doses at $0.51. The mean verification time for HaH medications was 6.3 min for Patient Alpha, 5.7 min for Patient Beta, and 12.4 min for Patient Gamma. The mean preparation time for HaH medications was 72.8 min for Patient Alpha, 54.6 min for Patient Beta, and 23.3 min for Patient Gamma.

3.2. Healthcare resource utilization and patient safety

Medication usage by frequency type across all three patients treated at home was split between 26 scheduled doses (49 %) and 27 PRN doses (51 %). There was a total of 13 IV doses dispensed (25 %) and 40 non-IV doses dispensed (75 %), which were comprised of oral and inhaled medications. Three pharmacy screening consults were completed with a total of 18 medication orders verified by an inpatient pharmacist. Eight clinical interventions were documented by inpatient pharmacists and 2 HaH patients were provided discharge education by a M2B pharmacist. There were no reported medication errors and no escalations back to the brick-and-mortar hospital due to clinical deterioration.

4. Discussion

In this small descriptive study, the pharmacy department's role in the implementation of a HaH program was elucidated. Surrogate markers such as medication waste and medication turnaround time were reported to measure operational success. Medication waste was used to assess the degree of medication management within the HaH program while medication turnaround time evaluated the efficiency of newly implemented pharmacy workflows related to HaH.

Medication waste was selected as the primary outcome because it is unaffected by uncontrollable factors such as employee absences or unexpected surges in the volume of medication orders that need processing, whereas medication turnaround time is impacted by these variables. Metrics such as doses dispensed and doses wasted were stratified by frequency type (scheduled versus PRN) and dosage form (non-IV versus IV). This made it easier to identify any trends in medication wasting since preparing each type of dose utilized different workflows. For example, the inpatient pharmacy prepared a 24-h supply of scheduled doses every day, but only the first 24-h supply of PRN doses was prepared for home delivery. If additional PRN doses were needed, the HaH nurse had to contact the inpatient pharmacy. This workflow prevented PRN doses from accumulating in the patient's home and reduced the likelihood of medication waste occurring. Scheduled doses were rarely wasted since they were always administered at the respective due time unless they were discontinued by the HaH provider after the HaH nurse left the brick-and-mortar hospital. Any dose not administered by the time of patient discharge from the HaH program was wasted as medications were not allowed to return to the inpatient pharmacy. This workflow differed from brick-and-mortar hospital patients as any doses not administered would be returned to the inpatient pharmacy rather than wasted.

The data showed that only Patient Beta had documented waste of scheduled medications. Patient Alpha and Patient Beta both wasted 100 % of their dispensed PRN doses while Patient Gamma wasted 33 % of their dispensed PRN doses. Overall, PRN medications represented 19 of the 21 total doses wasted (90 %) across all three patients, which suggests that limiting the frequency (i.e., from “every 6 hours PRN” to “every 12 hours PRN”) of these agents during the initial 24-h transfer home may be warranted. Despite the quantity of PRN medications wasted, 79 % of reported costs due to waste were attributed to a single scheduled dose of an IV antibiotic prepared in an elastomeric pump since there were multiple components (saline bag, antibiotic, and elastomeric pump) that required wasting. This demonstrates the necessity to ensure all frontline pharmacists, nurses, and providers understand the IV antibiotic dispensing workflow for elastomeric pumps to avoid unnecessary costs due to drug and supply waste.

Verification time was measured from when the HaH provider electronically signed the HaH medication order to the time the inpatient pharmacist verified the order. Medication verification mainly occurred on the first day of HaH admission. All subsequent dispenses did not need verification and were automatically routed to the dispense queue. If the HaH provider changed the dose or medication order in any way, an inpatient pharmacist needed to reverify the medication. Preparation time was measured from when the HaH medication label was printed from the dispense queue to the completion of the dispense check (i.e., final verification) conducted by the inpatient pharmacist. If any medication had a frequency greater than once daily, then the longest time to dispense check was recorded as medications would not be considered ready for pick-up until the entire 24-h supply had been dispense checked. This recording method was utilized to prevent reporting bias since four tablets of oral acetaminophen could be prepared simultaneously and produce four similar data points versus a single data point from preparing an IV antibiotic in an elastomeric pump, which would mask the time and labor involved with compounding IV elastomeric pumps and skew the average turnaround time.

The study results demonstrated that the mean verification and preparation time for HaH medications steadily improved as frontline pharmacy staff became more familiar with HaH workflows. However, there was a steep increase in mean verification time for Patient Gamma. This was attributed to the HaH provider changing the administration frequency or order instructions, thus requiring each medication order to be reverified by the pharmacist. The mean preparation time from Patient Alpha to Patient Gamma decreased three-fold, indicating that frontline pharmacy staff were able to quickly adjust to the implemented HaH processes. It was anticipated that the first day when a patient transferred from the brick-and-mortar hospital to their home would take the longest to prepare medications since a 24-h supply of all scheduled and PRN doses must be dispensed, however, this was not the case. Days that took the longest coincided with times when HaH medication orders were entered during peak IV room hours where the IV technician had to balance compounding stat medications for critically ill brick-and-mortar hospital patients alongside elastomeric pumps for HaH patients.

The mean verification and preparation time of medication orders from three medical-surgical units at the brick-and-mortar hospital were also evaluated for comparison and found to be 17 min and 111.5 min, respectively. These times were longer than the HaH medication turnaround time, but this may be attributed to pharmacy staff being recently educated on HaH workflows causing a perceived increase in efficiency due to staff prioritizing HaH medication orders over routine brick-and-mortar hospital medication orders. It should be noted that the sample sizes used to calculate medication turnaround time for the brick-and-mortar hospital were much larger, with 224 medications being verified and 388 medications being prepared. Therefore, any comparison to medication turnaround time for HaH patients should be taken with caution.

The utilization of healthcare resources was low due to the small number of enrolled patients given the infancy of the program. Every patient that was eligible for HaH received a pharmacy consult, but not every patient required outpatient prescriptions upon discharge as seen with Patient Beta. The number of pharmacist interventions per patient enrolled in HaH was similar to the number of interventions for low-acuity patients treated in the brick-and-mortar hospital.

The inclusion of a pharmacist on the HaH team has been shown to promote patient safety through the identification of drug-related problems.^13,14 To mitigate potential drug-related problems in our study, an inpatient pharmacist participated in twice-daily huddles to review the dispensing needs and limitations for patients being treated at home. These brief, interdisciplinary meetings routinely included the HaH operations director, nursing, pharmacy, providers, clinical case management, nutrition, and transportation services to allow discussion over the clinical and operational needs of potential and currently enrolled HaH patients. The HaH huddles were critical in facilitating communication between the HaH team and the inpatient pharmacy regarding expected medication delays due to staff or drug shortages, recommendations on IV antibiotic stability in elastomeric pumps, clinical interventions such as IV to oral conversion of equally bioavailable drugs, and coordination of Meds-to-Beds services for patients approaching discharge. Inpatient pharmacists also screened every potential HaH candidate to identify feasibility concerns from a medication perspective. For example, medications requiring therapeutic drug monitoring such as vancomycin presented logistical complications due to the specific timing required to draw trough levels and the unpredictability of traffic and weather patterns.

Post-implementation feedback from ED providers revealed confusion between home infusion, home health, and HaH services. Home infusion refers to the administration of subcutaneous or IV medications that require short-term monitoring for ambulatory patients without requiring inpatient admission, resulting in reduced costs and improved patient satisfaction.¹⁵ Home health programs deliver skilled nursing and therapy services to homebound post-acute care patients to enhance transitions of care, optimize clinical outcomes, and reduce unintended rehospitalizations.¹⁶ Medicare defines a homebound patient as someone who is unable to leave their home due to illness or injury without assistance from another person.¹⁶ Both home infusion and home health services are provided in the outpatient setting while HaH programs provide acute inpatient services for admitted, hospitalized patients that must adhere to the same care standards as patients treated within the brick-and-mortar hospital. Educating providers on the differences between these services is important as it spreads awareness and promotes the growth of the HaH care model.

One unique aspect of the current study is that it was conducted at a not-for-profit, county health system that primarily serves as a safety net hospital for uninsured and Medicaid patients. Safety net hospitals aim to dismantle barriers to healthcare by ensuring services remain accessible to all patients within the county, regardless of insurance status.¹⁷ A safety net hospital with a HaH program promotes health equity by delivering acute care services to patients at home who may be unable to complete their treatment within the brick-and-mortar hospital due to dependent family members or other external factors. This is especially important for underinsured and uninsured populations because interacting with a patient in their own home allows healthcare providers and nurses to identify and address social determinants of health such as lack of social support, food insecurity, and home environment stressors that may exacerbate health conditions. This results in care and discharge planning being tailored to the individual needs of each HaH patient within their home environment, which may prevent hospital readmission due to factors within and outside of the acute care episode.⁷

The CMS Acute Hospital Care at Home (AHCAH) initiative allows health systems to be reimbursed for providing inpatient-level care in a patient's home by waiving certain Medicare conditions of participation such as the requirement for a registered nurse to be immediately available and provide continuous nursing services.¹⁸ Unfortunately, there is a limited number of private payors that will reimburse health systems for treating patients within a HaH program due to the scarcity of literature comparing patient outcomes to those seen in the brick-and-mortar setting.¹⁸ Despite the lack of supportive reimbursement models for HaH programs from private payors, underinsured and uninsured patients still benefit from health systems that offer these acute care services at home.

Due to the lack of HaH literature that focuses on reporting pharmacy-specific outcomes at the time of publication, we are unable to compare the quantity and cost of medication waste and medication turnaround time to other health systems with HaH programs. However, one study conducted by Brito and colleagues in 2020 evaluated pharmacist interventions (PIs) in HaH patients and reported 80 PIs in 53 patients treated at home, which is approximately 1.5 PIs per patient.¹⁹ In the current study, 8 PIs were recorded across 3 patients equating to approximately 2.67 PIs per patient. This increase in PIs per patient could be attributed to the novelty of the HaH program at the study site causing a shift in the behavior of pharmacists. The pharmacist may be more attentive to drug-related problems for HaH patients since drug and dosage changes have more logistical considerations regarding preparation and delivery compared to a patient treated in a brick-and-mortar hospital.

In the original pilot study for HaH conducted by Levine and colleagues in 2018, there were 9 patients (82 %) who received IV medications during their HaH admission.²⁰ In the current study, a total of 13 IV medications were administered across all 3 patients (100 %) treated at home. From a patient safety perspective, both Levine's 2018 pilot study and the current study reported that there were no adverse safety events and no escalations back to the brick-and-mortar hospital due to clinical deterioration.²⁰ Both of Levine's HaH publications from 2018 and 2020 were randomized controlled trials, however, only the HaH arm of the original pilot study in 2018 was utilized for comparison of outcomes because of the similar number of patients treated at home and the lack of a comparator arm in the current study. Demographic data reveals patients treated at home in Levine's 2018 pilot study had either Medicare or private insurance compared to patients in the current study who were either uninsured or enrolled in the health system's financial assistance program. Despite the small sample size of the current study, the results show a promising alternative to acute hospital-level care for underinsured and uninsured patients, indicating the need for more robust research focused on this vulnerable patient population. Future studies should also investigate the quantity and cost of medication waste and medication turnaround time using a larger sample size and a brick-and-mortar comparator group to provide more insight into the pharmacy-specific costs of operating a HaH program. Additional research questions that should be considered include: what are the most effective strategies for reducing medication waste and turnaround time as patient volume increases in a HaH program, how does standardization of pharmacy workflows impact medication waste and turnaround time as a HaH program expands in patient volume, which performance metrics best assess the impact of HaH program expansion on pharmacy workflow efficiency, how does an automated dispensing cabinet impact medication waste and turnaround time in a HaH program, and what is the most efficient pharmacy staffing model to reduce medication waste and turnaround time as a HaH program expands.

There were some key limitations presented in the study that should be taken into consideration. First, the small sample size, lack of a comparator group, and non-randomized design reduced the internal validity of the study, which prevented any conclusions from being drawn regarding the differences in treating patients at home compared to the brick-and-mortar hospital. The small sample size also restricted the use of inferential statistics from being applied to determine the significance of the study results and ensure outcomes were not due to random error, thus decreasing the accuracy and validity of any comparisons to the general patient population of the health system. Additionally, the paucity of data in the literature regarding the quantity and cost of medication waste and medication turnaround time made it difficult to find an established population parameter to compare home treatment with, thus limiting the generalizability of the results to other patient populations. Second, inpatient pharmacists and technicians had varying proficiency levels and rotated through different roles in the pharmacy daily, resulting in potential bias in the medication turnaround time outcomes. Third, labor costs associated with pharmacy staff preparing HaH medications that were wasted were not reported. Waste associated with PRN medications and IV elastomeric pumps would have had the greatest impact on increased labor costs. Fourth, the lack of an automated dispensing cabinet (ADC) specific for HaH patients forced the inpatient pharmacy to ensure medications were prepared by 1:00 pm for nursing pick-up every day. This time coincided with peak hours for preparing brick-and-mortar hospital medication orders which negatively impacted HaH medication turnaround time. Acquiring an ADC would allow the inpatient pharmacy an opportunity to prepare future HaH doses scheduled for subsequent days during less busy times as medications could be loaded in the ADC. Since these medications would not have left the brick-and-mortar hospital pavilion, they could be returned to the pharmacy rather than wasted if the HaH provider discontinued the medication order, or if the patient was discharged from HaH. Lastly, the inability to automatically route HaH orders to medication carousels within the inpatient pharmacy did not mimic the workflow for preparing brick-and-mortar medication orders. This barrier forced pharmacy staff to manually enter a 24-h supply of each HaH order to be retrieved from the medication carousels, contributing to a longer medication turnaround time.

Overall, HaH implementation was successful as evidenced by the low quantity of scheduled medication waste and gradual improvement in medication turnaround time. Processes that were developed or modified for HaH operations to increase efficiency such as the elastomeric pump workflow for continuous infusion of IV antibiotics and preparing a 24-h supply of medications were executed by frontline pharmacy staff without issue. Our approach of using continuous infusion IV antibiotics is supported by an article published by Peinovich and colleagues in 2024 that highlighted potential solutions to the logistical barriers of IV medication administration in HaH.²¹ Alongside continuous infusions, Peinovich also recommended utilizing IV push medications to reduce the amount of time needed to complete a patient visit since the HaH nurse would not need to stay for the entirety of an IV infusion, which could range from 30 min to 2 h.²¹ Furthermore, IV push medications require less preparation time by the inpatient pharmacy compared to elastomeric pumps, increasing staff efficiency while decreasing medication turnaround time. Another important factor to our success could be attributed to team members maintaining open lines of communication between all stakeholders and being flexible to allow accommodations that benefit the patient. All pertinent updates were communicated using a group chat to operational leaders, as well as a secure group chat within the EHR for frontline staff to provide clarity and visibility to all healthcare professionals involved with the care of HaH patients. A detailed review of the major decision points related to pharmacy workflows is available in the Supplemental Appendix (eDiscussion, eText 1).

Pharmacists are trained specialists who provide oversight into every step of the medication-use process in the acute care setting, which now expands from the walls of the hospital to the patient's own home. If other health systems are considering starting a HaH program, pharmacy leaders should be engaged as early as possible and participate in the initial planning stages given the multiple challenges to medication management in the HaH setting. Pharmacy is one of the most regulated healthcare professions with each state having different laws that inevitably impact workflows related to HaH.^22,23 Advocating for best practices early in HaH process development is critical to providing the foundation for successful program growth. There are numerous clinical and operational services pharmacists can provide in this innovative and collaborative care model that promote medication management and uphold patient safety.8, 9, 10^,13,14 However, there is insufficient data in the literature to guide pharmacy leaders in creating an appropriate and efficient pharmacy staffing model for the HaH setting. The results of the current study provide initial data as guidance, but additional research is needed to investigate the effects of HaH on pharmacy-specific outcomes related to cost, staff performance, and utilization to help justify the need for pharmacy full-time equivalents (FTE) to support the expansion of a HaH program.

5. Conclusion

Despite the limited sample size, all 3 enrolled patients were successfully treated at home with no reported medication errors, no escalations back to the brick-and-mortar hospital due to clinical deterioration, and controllable medication waste that was primarily associated with high-frequency PRN prescribing patterns. There were numerous pharmacy-specific operational challenges encountered during HaH implementation that required innovative solutions to ensure all enrolled patients received hospital-level care at home. Data from this study regarding medication waste and medication turnaround time provide insight into optimizing dispensing practices and pharmacy workflows while establishing the benefits of having a pharmacist on the HaH care team. Future research should elaborate on reporting additional performance metrics to help identify when additional pharmacy FTE would be needed to support the growth of a HaH program, as well as share lessons learned from the experiences gained from overcoming unique barriers to treating patients at home.

Funding

This study was supported by a grant (ADH-550) provided by the Texas Society of Health-System Pharmacists Research & Education Foundation. The funding source was not involved in study design, data collection, data analysis, data interpretation, report writing, and did not take part in the decision to submit this manuscript for publication. The insights expressed in this study are those of the authors and do not necessarily reflect the views of the Texas Society of Health-System Pharmacists.

CRediT authorship contribution statement

Jason N. Levy: Writing – review & editing, Writing – original draft, Visualization, Project administration, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Joshua Wollen: Writing – review & editing, Visualization, Methodology, Formal analysis, Conceptualization. Phuoc Anne Nguyen: Project administration, Methodology, Funding acquisition, Conceptualization. Catina Brimmer: Project administration, Methodology, Funding acquisition, Conceptualization. Rohan Dwivedi: Writing – review & editing, Resources. Shane Tolleson: Writing – review & editing, Visualization, Supervision, Methodology, Conceptualization.

Declaration of competing interest

Jason Levy reports financial support was provided by Texas Society of Health-System Pharmacists Research & Education Foundation. The other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The supplemental appendix contains additional information such as the specific clinical exclusion factors to hospital at home program enrollment, detailed case reports for the enrolled patients included in the study, the major decision points related to pharmacy operational workflows, and an explanation of challenges encountered during patient screening.