Précis
This single-center, descriptive, retrospective chart review identified barriers to outpatient chemotherapy use, revealing avoidable inpatient stays and highlighting targets for stewardship interventions.
Abstract
Background
The primary site of chemotherapy administration has transitioned from the inpatient to outpatient setting for reasons such as optimized resource utilization, increased patient satisfaction, decreased overall costs, and enhanced reimbursement. Chemotherapy regimens are considered for appropriateness for outpatient administration based on the adverse event profile, risk for tumor lysis syndrome, need for intensive monitoring, and indication for treatment (eg, acute leukemia). At the University of California, San Diego, the following order question—which must be addressed at the time an oncologist signs the treatment plan—was incorporated into all inpatient chemotherapy treatment plans that have been designated as appropriate for outpatient administration: “This treatment plan has been identified as being appropriate for outpatient administration. Please provide a justification for giving this inpatient.” Options include “housing,” “transportation,” and “other: fill in the blank.” Outpatient-appropriate chemotherapy administered during an inpatient admission may not yield favorable patient outcomes and can impose a financial burden on the health care system through direct and indirect costs. This was the catalyst for our review.
Objective
The primary objective was to identify prescriber-specified reasons for the inpatient administration of chemotherapy regimens of interest and to determine interventions that may decrease inpatient administration. The secondary objective was to describe health care utilization (length of stay, intensive-care-unit level of care, and medication cost) associated with inpatient use of regimens of interest
Methods
This single-center, descriptive, retrospective chart review included hospitalized patients 18 years and older who had received at least 1 cycle of an outpatient-appropriate regimen between January 1, 2017, and September 14, 2022. Answers to the order question were reviewed and categorized to guide future interventions. Additional admission and demographic data were collected to further describe health care utilization.
Results
A total of 925 patients were included, with 987 corresponding provider-identified, non–mutually exclusive justifications (housing: n = 45 of 925 [4.9%]; transportation: n = 98 [10.6%]; and other: n = 844 [91.2%]). The most common other justification was “comorbidities or symptoms requiring hospitalization” (n = 258 [26.1%]). Eighty-three (9.0%) regimens were justified by prescribers solely because of housing or transportation needs, accounting for 1580 potentially avoidable hospital days and an estimated inpatient drug cost of $478,265 based on cost data from 2023. More than 25% of patients were discharged within 24 hours of their last dose, and nearly 21% were readmitted at least once to receive another cycle of an outpatient-appropriate regimen.
Conclusions
The variety of justifications for inpatient use of outpatient-appropriate cancer-directed therapy (CDT) highlights that there isn’t a one-size-fits-all approach to this issue. The frequent use of “other: acutely or critically ill” and “other: poor performance status” to justify inpatient CDT administration could indicate that some patients may not have been fit for treatment. The in-hospital mortality rate for the patient cohort was 6.4%. The instances of discharge shortly after CDT administration, coupled with the fact that more than 1 in 10 admissions were due to lack of housing or transportation, indicate that a significant portion of chemotherapy administrations could have been deferred to the outpatient setting if appropriate patient support had been in place. The data provide an assessment of factors that lead to inpatient administration of outpatient-appropriate regimens and highlight targetable areas for chemotherapy stewardship programs. Interventions were identified and developed to target increasing outpatient access, improving patient-centered care, and maximizing reimbursement.
Background
Over the past 3 decades, the primary site of chemotherapy administration has transitioned from the inpatient to the outpatient setting. As early as 1996, guidelines were published outlining appropriate reasons to hospitalize patients for chemotherapy, acknowledging the value in limiting inpatient administration to specific circumstances.1 Today, the justifications for inpatient chemotherapy have become even more narrow as the ability to safely deliver care in the outpatient setting continues to improve.2 Several institutions have implemented some form of an inpatient chemotherapy stewardship committee or algorithm.3-8 Others have investigated transitioning regimens traditionally considered logistically challenging to the outpatient setting, such as hyper-CVAD-A (cyclophosphamide, vincristine sulfate [Oncovin], doxorubicin hydrochloride [Adriamycin], and dexamethasone—part A) due to the administration of cyclophosphamide every 12 hours, or EPOCH-R (etoposide phosphate, prednisone, vincristine sulfate, cyclophosphamide, doxorubicin hydrochloride, and rituximab [Rituxan]), which requires portable infusion pump administration and, therefore, appropriately selected patients.2,9-14
This transition was driven by myriad factors that can be generally grouped into patient safety, patient satisfaction, resource utilization, cost containment, and reimbursement.2,15 The outpatient setting offers access to skilled nursing, supervising oncologists, and pharmacist support, mimicking the inpatient setting but with the advantage of reduced nosocomial infection risk and access to new agents and regimens with fewer toxicities and highly effective antiemetic regimens.2,15 These advances in safety in the outpatient setting are a portent of outpatient administration playing a more dominant role.2,15 Patient satisfaction also is improved by avoiding multiday admissions and potential treatment delays while awaiting an inpatient bed and through increased physical and mental comfort in the clinic vs an inpatient admission.2 Additionally, outpatient chemotherapy administration helps alleviate strain on inpatient resources, address capacity issues, and avoid expenses associated with utilization of these resources.2,16
Rapidly rising drug costs for cancer-directed therapy (CDT) and decreasing payer reimbursement are also prominent motives for driving chemotherapy administration to the ambulatory space.2-4,14,16 Advancements in therapeutics are associated with ever-increasing acquisition costs, with 69% of novel medication launches in the past 5 years carrying an annual price tag of more than $100,000 per medication.17 In addition, use of the diagnosis-related group (DRG) payment system in the acute care setting often does not provide adequate reimbursement for utilization of high-cost medications, leading to revenue losses.2,3 Transitioning site of care to the outpatient setting allows for reimbursement tied directly to administration and acquisition costs, the ability to bill for drug waste, and the potential for increased revenue for institutions eligible to purchase medications under the 340B Drug Pricing Program.3 In addition, patients avoid costly hospital stays and have increased access to patient assistance programs, helping to reduce their financial burden.2
The University of California, San Diego (UCSD) comprises 2 large academic medical centers with 5 infusion centers and a robust formulary system. This includes a multidisciplinary pharmacy and therapeutics (P&T) committee and a P&T oncology subcommittee, an inpatient request for outpatient restricted medication program, and a high-cost nonformulary request system. The P&T committee is tasked with maintaining the health-system formulary to provide the safest, most efficacious, and most economical selection of available agents and with several additional responsibilities regarding appropriate medication utilization. Regarding CDT treatment plans, the electronic health record (EHR) does not permit treatment plan conversion between the inpatient and outpatient settings. This functionality is intentionally disabled, limiting inpatient prescribers’ ability to order a nonformulary or outpatient-only regimen without pharmacist intervention. Given these restrictions, institutional guidance on appropriate inpatient use of CDT regimens with both available inpatient and outpatient treatment plans has yet to be developed. There are several reasons why an outpatient regimen may need to be administered in the inpatient setting, such as high tumor burden, comorbid conditions requiring admission, or the inability to arrange outpatient treatment within a reasonable time frame.
The aim of this study team was to understand the rationale for and assess the appropriateness of inpatient administration of CDT. Prior to designing interventions to optimize the use of these regimens, we aimed to elucidate provider-specified reasons for administering CDT to hospitalized patients. In 2017, an order question was incorporated into all UCSD inpatient treatment plans that were deemed appropriate for outpatient administration—ie, treatment plans that (1) were built for both inpatient and outpatient administration, (2) conferred a low risk of tumor lysis syndrome, (3) had few monitoring requirements, and (4) were logistically appropriate for an infusion center that was open (at that time) for less than 12 hours each day—to better describe provider justification for inpatient use and make providers aware that the regimen is typically given in the outpatient setting. The order question states: “This treatment plan has been identified as being appropriate for outpatient administration. Please provide a justification for giving this inpatient.” The prescriber is required to answer in order to sign the orders. Options are “housing,” “transportation,” or “other: fill in the blank.”
Although some of the regimens deemed appropriate for outpatient administration could pose a risk of tumor lysis syndrome during cycle 1 (eg, those that contain rituximab), that risk typically lessens with ongoing treatment and was deemed low enough to prompt the order question. On the other hand, induction regimens for acute leukemia, such as hypomethylating agents and venetoclax (Venclexta; AbbVie)—which are eventually transitioned to outpatient administration but are standardly initiated in the inpatient setting—were deemed not appropriate for outpatient administration (ie, failed to meet No. 3: had few monitoring requirements) and did not prompt the order question. Notably, outpatient-restricted regimens and most high-cost regimens do not have an inpatient treatment plan available in the EHR and are manually entered by pharmacists; these regimens do not include the justification order question and were not included in this review, but their use is evaluated monthly by the P&T oncology subcommittee.
The treatment plans that met the 4 criteria and prompted the order question are hereafter referred to as regimens of interest. The purpose of this study was to investigate and describe prescriber-identified reasons for the use of regimens of interest in the acute care setting and to identify (and later design) interventions that may decrease inpatient use when appropriate.
Methods
This was a single-center, descriptive, retrospective chart review that included adult patients (≥ 18 years) admitted to UCSD between January 1, 2017, and September 14, 2022, who received at least 1 cycle of a regimen of interest for an oncology or hematology condition during an inpatient admission. Patients were excluded if they received treatment as part of a clinical trial or through a patient assistance program. We identified patients through EHR queries. The study was designated by the UCSD Institutional Review Board as not human subjects research and exempted from continuing review.
All data were collected from the first admission during which a regimen of interest was administered (ie, the index admission). The following data were collected from the EHR at the time of admission: age, sex, race, home zip code, admission diagnosis, and ECOG status. If no ECOG status was documented at the time of admission, the earliest documented ECOG status during the admission was used. Data collected regarding the regimen of interest included medication doses and schedules, date and time of administration of each dose, cycle number, goal of treatment, and whether treatment was continued outpatient. Other data collected included admission and discharge dates and times, vital status at discharge, vital status at 30 days post discharge, and whether the patient was transitioned to comfort care or hospice at discharge. An electronic data pull was utilized for all data except for ECOG status, continuation of therapy to the outpatient setting, and enrollment into hospice or comfort care, which were collected manually via chart review.
Answers to the order question were collected and categorized into housing, transportation, and/or other. These answers were not mutually exclusive, so multiple answers may have been selected. Justifications listed as other were further categorized by study investigators as comorbidities or symptoms requiring hospitalization, acutely or critically ill, intensive monitoring requirements, emergent treatment required, acute leukemia treatment, progressive or relapsed disease, modified to no longer be appropriate for outpatient use, poor performance status, social or logistical constraints, unclear rationale, or other. Housing and transportation were separately analyzed further because these factors were deemed modifiable compared with some of the other patient- and disease-specific factors. Drug cost data were sourced from Micromedex RED BOOK (Merative) and based on the lowest reported wholesale acquisition costs as of May 2023.18 Hospital-adjusted expenses were derived from data from the American Hospital Association Annual Survey, which lists the hospital-adjusted expenses per hospital day for nonprofit hospitals in California at $4719 in 2022.19
The primary outcome was to identify and categorize prescriber-specified reasons for the use of regimens of interest in the hospital and to determine interventions that may decrease inpatient use. Secondary outcomes included a description of health care utilization and patient characteristics, including in-hospital mortality and 30-day mortality post discharge, transition to comfort care or enrollment in hospice, time from administration of the last dose to patient discharge, continuation of the regimen to the outpatient setting, distance from the patient’s home zip code to the nearest health system–affiliated infusion center, and ECOG performance status at admission. Exploratory outcomes included cost and reimbursement data. Descriptive statistics were used throughout and performed using Microsoft Excel (Microsoft Corporation).
Results
Table 1 lists the 10 most frequently utilized outpatient-appropriate regimens included in this analysis. A total of 925 unique patients were included in this study after the exclusion of 9 patients who were treated as part of a clinical trial. The median patient age was 60 years, slightly less than half of patients were women, approximately half were White, and more than half had a baseline ECOG performance status of 0 to 2. Other baseline characteristics are listed in Table 2. Information about the diagnosis, including whether it was a new diagnosis, was not collected as part of this review.
A total of 987 provider-identified non–mutually exclusive reasons for inpatient administration were provided (Figure [A]) across 1277 unique admissions. Non–mutually exclusive justifications for inpatient administration were as follows: 45 (4.9%) for housing, 98 (10.6%) for transportation, and 844 (91.2%) for other. The 844 other justifications were further classified as follows (Figure [B]): comorbidities or symptoms requiring hospitalization (n = 258 [30.6%; 26.1% of total]); acutely or critically ill (n = 143 [16.9%; 14.5% of total]); intensive monitoring requirements (n = 115 [13.6%; 11.7% of total]); unclear rationale (n = 68 [8.1%; 6.9% of total]); emergent treatment required (n = 64 [7.6%; 6.5% of total]); poor performance status (n = 62 [7.4%; 6.3% of total]); progressive or relapsed disease (n = 47 [5.6%; 4.8% of total]); modified to no longer be appropriate for outpatient use (n = 42 [5.0%; 4.3% of total]); social or logistical constraints (n = 25 [3.0%; 2.5% of total]); acute leukemia treatment (n = 16 [1.9%; 1.6% of total]); or other (n = 4 [0.5%; 0.4% of total]).
Eighty-three (9.0%) index inpatient administrations were documented as entirely related to either housing or transportation (ie, other was not also selected as a reason), accounting for a total of 1580 potentially avoidable hospital days based on the combined number of days to administer the 83 regimens. Approximately 24% of patients resided in counties without an institution-affiliated infusion center. Among those with transportation issues that warranted their admission, 69% lived within 40 miles of an affiliated infusion center, with an average distance of 61 miles.
Of the index admissions, admission status was coded as emergent (via the emergency department) for 722 patients (78.1%), elective for 127 (13.7%), urgent (a planned admission that occurs within 48 hours) for 71 (7.7%), and other/unknown for 5 (0.5%). The baseline ECOG performance status was greater than 2 for 156 patients (16.9%), and 255 patients (27.6%) did not have an ECOG performance status documented during their admission (Table 2). A total of 1277 admissions associated with the use of a regimen of interest were identified (Table 3). The median time from admission to initiation of CDT was 4.9 days (IQR, 2-9), and the median time from last dose to discharge was 2.6 days (IQR, 1-7). Median length of stay was 12.9 days (IQR, 6-23), with 251 patients (27.1%) discharged within 24 hours of their last dose of a treatment plan–based CDT. Nearly 21% of patients had at least 1 readmission for administration of a regimen of interest, with 70 patients (7.6%) readmitted more than once. Of the 897 patients with regimens of interest that were eligible to be continued in the outpatient setting—excluding, for example, autologous hematopoietic stem transplant regimens or induction regimens that are only 1 cycle—458 patients (51.1%) remained on the same therapy as an outpatient. In-hospital mortality during the index admission was 6.4%; at 30 days post discharge, 85.9% of the patients who survived to discharge remained alive.
The bone marrow transplant service ordered nearly half of the treatment plans (48.0%), and the hematology/oncology consult team ordered the remainder. The most-used base regimens were single-agent rituximab (11.7%), R-CHOP (10.4%; rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, and prednisone), CyBorD (7.5%; cyclophosphamide, bortezomib, and dexamethasone), modified FOLFOX-6 (6.5%; leucovorin calcium [folinic acid], fluorouracil, and oxaliplatin), and azacitidine (6.1%) (Table 1).
The total cost of parenteral CDT medications administered in the inpatient setting solely because of a lack of housing or transportation was approximately $478,265. Given the total of 1580 potentially avoidable hospital days and an estimated cost of $4719 per hospital day based on 2023 American Hospital Association data,19 hospital-adjusted expenses may exceed $7 million over the 5 study years.
Discussion
Transitioning CDT medication administration to the outpatient setting, when appropriate, offers several advantages for both health systems and patients; however, this transition may be hindered by multiple factors. Social barriers are often significant and include unstable housing, unreliable transportation or untenable transportation times, insurance concerns, insufficient support systems at home, and low health literacy.2
In this study, we sought to determine the underlying factors driving the use of outpatient CDT regimens in the inpatient setting at a large academic medical center. Overall, most treatment plans were justified for inpatient use by the prescriber due to comorbidities or symptoms requiring hospitalization or to an acutely or critically ill patient status. However, 25% of all patients with these justifications also received at least 1 dose of a treatment plan–based medication within 24 hours of discharge, indicating that some—if not all—of these patients may have been medically ready for discharge and that the CDT potentially could have been given post discharge.
The next most common rationale for inpatient administration, at 14%, was the need for intensive monitoring requirements. This category was largely made up of reasonable concern for tumor lysis syndrome and patients with previous infusion reactions or intolerance—but also responses indicating a lack of knowledge regarding regimens that could safely be given in an outpatient setting, namely ifosfamide (Ifex)–based regimens (due to the need for concurrent [Mesnex]) and other regimens requiring continuous infusions (eg, EPOCH). This lack of knowledge represents an opportunity to educate providers as well as outpatient nurse case managers who sometimes facilitate planned admissions.
According to the prescribing oncologist, approximately 9% of administrations were due entirely to lack of housing or transportation support and were associated with significant costs. A notable observation from the data is that although 24% of patients lived in counties without an institution-associated infusion center, most patients with transportation issues (69%) resided relatively close to an affiliated infusion center (ie, within 40 miles). However, the average distance of 61 miles suggests that a significant portion of patients still face considerable travel burdens. This highlights the potential impact of transportation as a barrier to accessing care, even in areas with nearby infusion centers, and underscores the need for targeted solutions to address these challenges, such as improving transportation options or expanding services to underserved areas. Transportation insecurity as a barrier to care is a complex issue that has been well described.20 Briefly, a combination of patient inability to drive themselves, lack of timely public transport directly to cancer centers, weekday appointments that may preclude a working family member or friend from providing transportation, and the financial cost of gas or fare, to name a few, can make transportation difficult.20 Moreover, patients with a cancer history are more susceptible to transportation barriers and their downstream effects, with higher rates of transportation insecurity and more frequent appointments than patients with no cancer history.21 Transportation insecurity may lead to patients with cancer missing, delaying, and/or prematurely terminating necessary care and is associated with a higher risk of emergency department use and all-cause mortality.20,21 These detrimental consequences may influence oncologists to admit these patients for treatment.
Improved social support for accessing outpatient infusion centers could have helped avoid more than $7 million in hospital-adjusted expenses. As with many health care organizations, transportation and housing resources are limited at UCSD. No-cost transportation is often available only through very specific insurance plans or foundations and, when available, is commonly unreliable. Other transportation options may be available through social work support at a lower cost than a ride-sharing app, but these are still associated with out-of-pocket costs that may be prohibitive for some patients. Additional difficulties are encountered by patients who use wheelchairs or are bedbound. Housing support through a social worker is individualized; however, exceptionally few options are available for patients experiencing homelessness, and health system–provided family housing is very selective and frequently at capacity. When housing or transportation support is available, social workers and case management nurses typically require at least a week’s notice to arrange appropriate assistance. Limited grant or foundation funding is occasionally available based on the cancer diagnosis, but it also takes weeks to apply for and organize. While researching our institutional support options for this project, we discovered that there is no standardized onboarding for newly hired oncologists about outpatient social work resources at UCSD, which may leave some relying on inpatient admissions for treatment. Lastly, although telehealth appointments, which became prevalent during the COVID-19 pandemic, can alleviate some of the burden of travel, telehealth is limited in its nature and may not work for some disease states that require frequent physical exams.
Acknowledging where UCSD was falling short in providing patient support for transport and housing (based on the data from this project), pharmacy department leadership developed a project proposal that included a request for a donor-sponsored, health system–owned space for patients and their caregiver, if applicable, to stay locally as often as needed during treatment, with transportation support to get back and forth. The executive suite selected the proposal to become a formal initiative, and efforts are ongoing. This intervention may be transformative for UCSD patients, but many health systems do not have the funding or the operational support to create and sustain such a program. We recognize that housing and transportation challenges are typically persistent and require sustained efforts, and that addressing these issues through a short-term strategy will not be sufficient to achieve lasting solutions. As a result, one of our objectives is to transition oncology care closer to patients’ homes, which will necessitate dedicated personnel to coordinate communication among providers, manage appointments, and ensure alignment of treatment goals.
For patients requiring emergent workup and treatment (8% in this study), the outpatient setting presents several barriers that may compel providers to admit their patients for these services. Timely appointments for imaging and biopsies are often difficult to schedule in the ambulatory space and can lead to extended delays in diagnosis and treatment initiation. Scheduling infusion appointments poses similar challenges due to schedule limitations when treatment must start imminently. Some of this strain might be alleviated through institutional efforts to extend infusion center hours and spread appointments across consistently slower hours of the day and slower days of the week. Additional barriers that could cause undue delays in treatment initiation may include the need for insurance authorization prior to initiation (dependent on treatment center policies), lack of insurance coverage, and care coordination for more complex regimens.
The use of CDT in patients with advanced cancer and/or poor ECOG performance status is an increasingly recognized concern, with a high proportion of these patients receiving aggressive or futile therapy in their last month of life.22 Aggressive medical care in this setting is associated with significantly greater health care expenditure without associated overall survival or quality-of-life improvements for patients.22 In our patient population, approximately 7% of the CDT regimens were administered with a justification of poor performance status, and nearly 17% of all patients had a baseline ECOG performance status greater than 2. Overall, in-hospital mortality was slightly higher than 6%, and 30-day postdischarge mortality was approximately 14%. Thirty-day mortality was 13.6% in patients with a prescriber justification of poor performance status and 18.6% in patients with an ECOG status greater than 2. Providers should be encouraged to have up-front discussions with patients and families regarding goals of care and expected benefits of aggressive treatment at patients’ end of life.
Medication costs continue to rise at a rapid rate as health systems contend with often-inadequate DRG payment for associated inpatient stays. Increasing drug costs and reduced reimbursement are major incentives pushing CDT administration to ambulatory spaces. A plethora of literature demonstrates the cost savings of transitioning specific chemotherapy regimens from the hospital to infusion centers.2-4,7-12,14 At UCSD, our outpatient infusion centers are qualified registered entities under the federal 340B Drug Pricing Program, allowing drug acquisition from manufacturers at significantly reduced costs; inpatient settings are not eligible to purchase drugs at 340B pricing. Reimbursement in the hospital setting is based on DRG, which does not reliably provide adequate reimbursement for utilization of high-cost medications. On the other hand, authorized outpatient use of these medications provides set, reliable reimbursement of the medication purchase price and associated resource utilization costs. Additionally, when a parenteral compounded product does not utilize the entire contents of the vial, the residual medication cannot be billed for in the inpatient setting; infusion centers, however, can bill payers for drug waste.
Patients also may reap significant benefit from outpatient CDT administration. A randomized, crossover study of patients receiving both an inpatient and outpatient cycle of high-dose cisplatin found that most patients preferred the outpatient setting.23 Another study of patients receiving either inpatient or outpatient (ie, home-based) FOLFOX regimens in Korea reported that patients receiving outpatient treatment had significantly higher satisfaction with their treatment than those with hospital-based therapy.24
At UCSD, unnecessary admissions for CDT administration contribute to persistently high overall and team-based censuses; these admissions also may utilize a room that a patient with more urgent medical needs may require, leading to boarding for days in the emergency department until a hospital bed becomes available. Additionally, safely transitioning continuous-infusion regimens to the outpatient setting has been associated with drastic decreases in resource utilization.25 Several institutions have implemented strategies on the inpatient side to reduce the unnecessary or inappropriate use of outpatient-appropriate oncology regimens as well as high-cost CDT and nononcology medications. Strategies generally involve a designated criterion for use (ie, policy/algorithm) or a regularly maintained inpatient CDT regimen formulary in conjunction with a multidisciplinary stewardship peer-review process.3-8 Programs such as these are crucial for shifting institutional culture away from avoidable inpatient administration of CDT.
Clinical pharmacists in both the inpatient and outpatient settings can prove valuable in transitioning appropriate CDT regimens to infusion centers. Inpatient pharmacists should be involved in institutional chemotherapy stewardship initiatives and in establishing guidelines for appropriate use. Additionally, pharmacist expertise can be vital in engaging frontline and senior medical leadership to intervene and promote appropriate use. Outpatient pharmacists can assist in addressing denials from payers and accelerating authorization; providing patient, caregiver, and nursing staff education on toxicities and supportive care; and offering additional necessary support for oncologists, patients, and their families. While transitioning numerous CDT regimens to the outpatient setting, the University of Arizona utilized pharmacists to educate providers about treatment safety in the outpatient setting and to address patient eligibility for outpatient chemotherapy and supportive care services.2 The university also established accessible after-hours care for its outpatient infusion patients, utilizing oncology fellows, nurses, and pharmacists to address urgent issues regarding chemotherapy, infusion pumps, or toxicities.
The results of this descriptive study have been used to develop interventions targeted to increase outpatient access, improve patient-centered care, reduce costs, and maximize reimbursement. The interventions derived from this study include an institution-level project to increase the transportation and housing options available to patients in need. Currently, the UCSD hospital system is looking to rent an entire floor of a local housing facility to give patients and caregivers a place to stay locally while receiving treatment or during recovery from a stem cell transplant. In addition, a group of pharmacist specialists, pharmacy managers, and oncologists has formed a committee to evaluate inpatient chemotherapy usage when regimens exceed 1 cycle. The idea here is that the group evaluates such things as clinical appropriateness of an inpatient level of care or continuing chemotherapy if nearing end of life and helps facilitate outpatient administration when appropriate. Plans include incorporating nurse case managers and social workers into the group.
Additionally, adjustments to the order question were approved to increase its value as a tool. This included making answers mutually exclusive, combining housing and transportation into a single option, and adding 2 new discrete options: admitted to the intensive care unit and comorbidities/symptoms requiring hospitalization. The inpatient oncology pharmacists have also been educated on assessing inpatient treatment plans as they are modified for individual patients to determine whether the order question should be included in the modified plan. This will allow future data collection to assess our interventions post implementation. Finally, one-on-one, just-in-time provider and support staff education is provided to increase awareness of outpatient resources and the impact of avoidable inpatient days for patients receiving regimens of interest.
Limitations
This study had limitations. This study’s single-center, retrospective nature limited the generalizability of our findings. Also, patient diagnoses were not examined because the focus of this study was the treatment plans themselves, and manual chart review to determine the clinical necessity of inpatient administration was not feasible because of the high number of patients included. The use of cost data from 2023 also likely overestimated the actual costs because prices likely fluctuated over the 5-year study period, with lower prices in the earlier years. In addition, answers to the order question were dependent on the ordering physician and may not provide the best representation of the true reason for inpatient use, and the subcategorization of fill-in-the-blank justifications by study investigators was subjective. Moreover, the ECOG performance status reported herein was the first charted per admission and was not reliably documented in all oncology progress notes; thus, it may not fully represent the performance status at the time of chemotherapy administration. Also, mortality data may be incomplete because deaths in the community are not consistently documented in the EHR.
Future directions include repeating the evaluation of the inpatient use of outpatient-appropriate regimens after updates to the order question, implementing initiatives developed by the newly formed review committee, and formalizing housing and transportation support. Evaluations should also include a comprehensive cost evaluation that includes antiemetics and ancillary medications and an outpatient cost comparison.
About the Authors
Janine G. Martino, PharmD, BCOP, APh, is a pharmacist specialist in bone marrow transplant and immune effector cell therapy at the University of California, San Diego.
Kaitlyn Wells, PharmD, is an oncology pharmacist at Oregon Health & Science University in Portland.
Sam Martinez, PharmD, BCOP, is an outpatient infusion pharmacy manager at the University of California, San Diego.
Craig A. Stevens, PharmD, BCPS, is the pharmacist specialist in medication use policy and outcomes at the University of California, San Diego.
Conclusions
The inpatient administration of chemotherapy regimens that could be given as an outpatient is a multifaceted issue. If institutions try to collect the rationale for these administrations, it may shed light on areas that can be addressed. A portion of chemotherapy administrations examined in this study could have been deferred to the outpatient setting and were associated with significant medication-related costs. A noteworthy number of administrations of outpatient-appropriate chemotherapy were prescriber justified due solely to a lack of housing or transportation support, and the associated admissions represent a cost burden to the health system via direct and indirect costs. Implementation of a support program for housing and transportation needs at UCSD aims to address an overlooked area in need of improvement. Additionally, a multidisciplinary team is beginning to work on realtime assessments of the need for inpatient chemotherapy and to provide support for outpatient transitions. Further development of appropriate guidance on the use of CDT regimens in the inpatient setting and improved support for outpatient administration may offer significant benefits to both patients and the health system.
REFERENCES
Dollinger M. Guidelines for hospitalization for chemotherapy. Oncologist. 1996;1(1 & 2):107-111.
McBride A, Persky D. Transitioning select chemotherapeutics to the outpatient setting improves care and reduces costs. Oncol Issues. 2021;36(4):56-64. doi:10.1080/10463356.2021.1927629
Foster AE, Reeves DJ. Inpatient antineoplastic medication administration and associated drug costs: institution of a hospital policy limiting inpatient administration. P T. 2017;42(6):388-393.
Johnson S, Smith S. Development of inpatient chemotherapy and supportive care therapy guidelines to promote appropriate inpatient and outpatient utilization of high-cost medications. J Am Coll Clin Pharm. 2019;3(1):318-319. Accessed May 1, 2025.
Hennes ER, Reed M, Mably M, et al. Implementation of a chemotherapy stewardship process. Am J Health Syst Pharm. 2020;77(15):1243-1248. doi:10.1093/ajhp/zxaa157
Li E, Schleif R, Edelen B. Hospital management of outpatient oncology treatment decisions: a survey to identify strategies and concerns. J Oncol Pract. 2013;9(5):e248-e254. doi:10.1200/JOP.2012.000814
McBride A, Campen CJ, Camamo J, et al. Implementation of a pharmacy-managed program for the transition of chemotherapy to the outpatient setting. Am J Health Syst Pharm. 2018;75(9):e246-e258. doi:10.2146/ajhp170138
Randle H, Smith LV, Padilla-Tolentino E, Goodgame BW. Evaluation of the use of non-formulary oncology medications restricted to outpatient use in hospitalized patients after implementation of a criteria-for-use algorithm. J Oncol Pharm Pract. 2020;26(4):882-890. doi:10.1177/1078155219877920
Corsi MP, Shea K, W Knoebel RW. Impact of transitioning inpatient chemotherapy regimens to the outpatient setting. J Oncol Pharm Pract. 2020;26(6):1324-1330. doi:10.1177/1078155219891566
Keshvani N, Hon M, Gupta A, et al. Reducing hospitalizations: institution of outpatient infusional EPOCH-based chemotherapy at a safety net hospital. J Oncol Pract. 2019;15(8):e644-e651. doi:10.1200/JOP.18.00738
Moore DC, Arnall JR, Plesca DC. Transitioning historically inpatient chemotherapy regimens for hematologic malignancies to the ambulatory care setting. Am J Health Syst Pharm. 2018;75(22):1824-1830. doi:10.2146/ajhp180060
Martin AL, Frank JP, Waggoner ML. Making moves: transitioning R-EPOCH to the ambulatory setting. J Oncol Pharm Pract. 2018;24(8):617-622. doi:10.1177/1078155217722408
Mabrey FL, Gardner KM, Shannon Dorcy K, et al. Outpatient intensive induction chemotherapy for acute myeloid leukemia and high-risk myelodysplastic syndrome. Blood Adv. 2020;4(4):611-616. doi:10.1182/bloodadvances.2019000707
Banh C, Valsvik K, Arredondo A, et al. Transitioning ifosfamide chemotherapy regimens to the ambulatory setting: reviewing cost savings and safety profile. Support Care Cancer. 2022;30(3):2755-2766. doi:10.1007/s00520-021-06653-4
Jacobson JO, Mulvey TM. Time to focus on inpatient safety: revision of the American Society of Clinical Oncology/Oncology Nursing Society chemotherapy administration safety standards. J Clin Oncol. 2012;30(10):1021-1022. doi:10.1200/JCO.2011.40.9409
Use of chemotherapy/high-cost drugs in hospitalized patients. IPD Analytics Rx Brief: Drug Management. September 2022. Accessed May 6, 2025. https://secure.ipdanalytics.com/User/Login?ReturnUrl=%2fUser%2f
Graboyes EM, Chaiyachati KH, Sisto Gall J, et al. Addressing transportation insecurity among patients with cancer. J Natl Cancer Inst. 2022;114(12):1593-1600. doi:10.1093/jnci/djac134
Jiang C, Yabroff KR, Deng L, et al. Transportation barriers, emergency room use, and mortality risk among US adults by cancer history. J Natl Cancer Inst. 2023;115(7):815-821. doi:10.1093/jnci/djad050
Waller A, Turon H, Bryant J, Zucca A, Evans TJ, Sanson-Fisher R. Medical oncology outpatients’ preferences and experiences with advanced care planning: a cross-sectional study. BMC Cancer. 2019;19(1):63. doi:10.1186/s12885-019-5272-6
Cox KM, Goel S, O’Connell RL, et al. Randomized cross-over trial comparing inpatient and outpatient administration of high-dose cisplatin. Intern Med J. 2011;41(2):172-178. doi:10.1111/j.1445-5994.2010.02201.x
Joo EH, Rha SY, Ahn JB, Kang HY. Economic and patient-reported outcomes of outpatient home-based versus inpatient hospital-based chemotherapy for patients with colorectal cancer. Support Care Cancer. 2011;19(7):971-978. doi:10.1007/s00520-010-0917-7
Svirskis D, Behera S, Naidoo N, et al. Stability of vincristine sulfate, doxorubicin hydrochloride and etoposide phosphate admixtures in polyisoprene elastomeric pump supporting transition of the EPOCH regimen to outpatient care. J Oncol Pharm Pract. 2019;25(4):831-840. doi:10.1177/1078155218764285
