Outcomes of Inpatient Administration of Restricted Antineoplastic Medications at a Large Academic Medical Institution

INTRODUCTION

The administration of medications to treat oncology conditions, including chemotherapy and immunotherapy, has increasingly shifted from the inpatient setting to the outpatient setting as a result of new regimens with minimal toxicities, rising drug costs, and a movement toward value-based health care.^1,2 As payer reimbursements decrease, healthcare organizations must identify opportunities for reducing costs while continuing to provide high-quality health services.² A survey by Li and Schleif showed that a majority of hospitals currently utilize some management strategy to ensure the appropriate use of outpatient oncology medications. The most common approaches to restricting such medications include using order sets, tumor-specific protocols, or product formularies, or involving an oncology pharmacist to facilitate appropriate medication usage.³

Although no standard exists among institutions for maintaining the active control of medication use, pharmacy and therapeutics (P&T) committees themselves can effectively restrict medication usage to specific areas of administration, provider specialties, or patient populations.^4,5 The Joint Commission requires the committees to maintain an annually reviewed formulary and to consider innovative treatment regimens while balancing clinical benefit with cost-effectiveness.^4,5

At our institution, the P&T committee created a new formulary status––“formulary, outpatient-restricted use only”––to account for medications that are not typically expected to resolve clinical symptoms in the acute setting, do not require inpatient monitoring, and have a low risk of serious adverse effects. When a prescriber deems that inpatient administration of a formulary, outpatient-restricted medication is warranted, a loosely defined review process consisting of a conversation between the requesting prescriber, the patient’s pharmacist, and the chair of the oncology P&T committee occurs (Figure 1). Often, peer-reviewed literature or ongoing clinical trial data are cited to justify the use of the particular oncology or hematology medication. Prescribers can contact the oncology P&T committee chair in a variety of ways, such as by making a direct phone call, sending a secure email, or meeting face-to-face. But the information relayed during these discussions is inconsistent and difficult to keep track of. Typically, inpatient and outpatient pharmacy staff inform the prescribers and committee chair about the cost of the medication, non-restricted medication alternatives, and the feasibility of continuing the medication in the outpatient setting depending on the patient’s insurance coverage. Owing to the informal nature of these requests, it is difficult to review the safety and efficacy of formulary, outpatient-restricted medications. Therefore, the purpose of this study was to describe our institution’s use of formulary, outpatient restricted oncology and hematology medications in the inpatient setting.

METHODS

RESULTS

Between January 1, 2015 and May 1, 2017, 24 medication requests were approved and 23 patients met the study criteria (Table 1). Of the 23 patients, nine (39.1%) were diagnosed with a hematologic malignancy; six (26.1%) were diagnosed with a solid tumor; and eight (34.8%) were receiving a medication secondary to their malignancy (Table 2). One patient received two restricted medications, brentuximab and nivolumab, during the same hospital admission. Of the 24 medications administered, 11 (45.8%) were antineoplastics (including monoclonal antibodies and immunotherapy agents); five (20.8%) were hormonal agents; and eight (33.3%) were supportive care agents (Table 3).

Thirteen (54%) formulary, outpatient-restricted medications were continued to the outpatient setting after discharge; eight (33%) of those medications were not continued. Denosumab for hypercalcemia of malignancy accounted for three (13%) of the formulary, outpatient-restricted medication approvals and was administered as a single inpatient dose without the intention of its being continued in the outpatient setting. Of the eight medications that were not continued to the outpatient setting, five discontinuations were a result of patient death and one was a result of disease progression. One patient transferred care to another institution (it is unknown if the medication was continued), and one patient was lost to follow up.

At the time of chart review, 10 (43.5%) patients had died. Five (21.7%) patients died within 30 days of hospital discharge; all were either full code or “do not attempt resuscitation (DNAR)/full care” prior to medication administration but were transitioned to either comfort care or hospice prior to death. In four of these five patients, death was related to their cancer.

Eight patients (34.8%) ultimately transitioned to comfort care and another four (17.4%) patients transitioned to outpatient hospice. The median hospital LOS was 16 days (interquartile range [IQR], 12–37.8 days) and the median hospital LOS after medication administration was 7.5 days (IQR, 5–13.8 days). An average of 1.68 outpatient-restricted medication doses were administered in the acute care setting. At the time of chart review, six (26.1%) patients were still alive and seven (30.4%) patients were eventually lost to follow up. Moreover, eight (33.3%) patients showed disease improvement and seven (29.2%) patients showed disease progression after the administration of a formulary, outpatient-restricted medication. We were unable to determine the progression status in 10 (41.7%) patients.

We further described the level of evidence supporting the use of formulary, outpatient-restricted medications. Fourteen (58.3%) medications were supported by phase 3 clinical trial data and 10 (41.7%) medications were supported by phase 1 or phase 2 clinical trial data (Table 4). When we compared the primary and secondary outcomes between medications supported by phase 3 data and those supported by phase 1 or phase 2 data, respectively, eight versus five medications (57.1% vs. 50%) were not continued to the outpatient setting; ten versus six patients (71.4% vs. 66.7%) reached 30-day survival; and five versus three patients (35.7% vs. 30.0%) showed disease progression (Table 5). Given the small sample size of this cohort, we did not test for statistical significance.

The total AWP of therapies during this study time period was $217,335.96, with an average cost of $9,055.67 per patient. The AWP was $10,662.54 in 2015, $92,420.03 in 2016, and $101,314.87 in 2017.

DISCUSSION

Our study describes the frequency with which medications designated as formulary, outpatient-restricted were given to patients who were admitted to an NCI-academic medical center and continued to the outpatient setting. Approximately one-third of these medications were not continued after discharge, and five were discontinued due to the patient’s death. Furthermore, all five patients who did not meet overall 30-day survival were either full code or DNAR/full care prior to receiving an outpatient-restricted medication, but they were transitioned into comfort care or hospice shortly after medication administration.

To decide whether to pursue treatment in the case of terminal illness is extremely difficult, especially when it concerns patients who are approaching the end of their life.^6–8 Our results showed that some patients expired shortly after having received restricted medications, and it is not clear whether they benefited from those treatments. Moreover, our patient population had a variable baseline ECOG performance score, which suggests that some of them may not have had optimal functional status at the time the medication was administered. Some formulary, outpatient-restricted medications were used as a last-ditch effort when the patient had few treatment options and, in some cases, was nearing the end of life. The literature supporting the use of these medications in such circumstances was limited at times. Finally, the cost for using these formulary, outpatient-restricted medications was not reimbursed, which had a significant financial impact on the inpatient pharmacy.

Our study findings highlight the difficulties regarding the appropriate use of oncology or hematology medications for advanced disease in the acute setting. These results suggest that our institution might benefit from improving its current formulary, outpatient-restricted medication approval process to better evaluate whether a patient warrants treatment in the acute setting. Therefore, developing a standardized review process with a request form for the use of a “formulary, outpatient-restricted” oncology or hematology medication could enable more transparency for appropriate treatment decisions to be made in these clinical scenarios. We plan to require the following information from prescribers when a formulary, outpatient-restricted medication is requested: reason for inpatient administration; previous failed lines of therapy; supporting published literature; anticipated benefit to patient; expected safety issues; outpatient insurance coverage; cost; and number of doses anticipated in the inpatient setting. The form would also allow us to collect data on the use of specific formulary, outpatient-restricted medications more easily, and report outcomes to the P&T committee.

To our knowledge, this is one of the first studies to look at outcomes associated with the use of outpatient-restricted oncology and hematology medications in the inpatient setting. Several other studies have analyzed cost savings associated with policies that were implemented regarding the use of expensive medications. Recently, Foster and Reeves described such savings after implementing a policy that defined appropriate criteria for inpatient antineoplastic administration at a community hospital.⁹ The authors analyzed the cost savings and appropriate medication utilization before and after policy implementation. The hospital achieved an estimated annual savings of $163,602 and reduced the number of patients receiving medications out of convenience after policy implementation.⁹

Durvasula and colleagues developed a High-Cost Medication Review Committee at a large academic medical center to standardize the process of approving inpatient requests for high-cost medications,² with a threshold of $5,000 for a single medication dose or $10,000 for a course of therapy. Specific drug classes included biologics, immunomodulators, hematologic drugs, antineoplastics, metabolic drugs, and antimicrobials. Through this standard review process, the institution reduced its non–evidence-based use of high-cost medications, saving the pharmacy $491,000.²

These studies provide insight into a valuable cost-containment strategy, especially regarding the treatment of oncologic or hematologic conditions, for which costs continue to rise.

Lastly, although we sought to compare the outcomes of medication continuation and overall survival based on medication administration supported by phase 3 clinical studies compared with phase 1 or phase 2 clinical studies, our limited study population made such a comparison difficult. Our findings suggest that providers are using oncology medications with varying levels of evidence at our institution.

The study has several other limitations. Having data from one single center limits generalizability. Our study was a retrospective review that relied on EHRs containing complete and accurate documentation. Moreover, the population was predominantly Caucasian, male, and small in number, and the study took place within the short timeframe of 2.5 years. Another limitation was the large proportion of patients whose EHR had missing information, such as their baseline ECOG score, disease progression assessment, and overall vital status (alive or deceased).

Although we grouped the restricted medications based on clinical data supporting their use, this information was not documented in the EHR. Instead, the groupings were based on independent searches of available literature. As a result, it is possible that our categorizations on the clinical data supporting medication use differ from those of the prescribing provider, which highlights another limitation of the study.

CONCLUSION

At UCSD, formulary, outpatient-restricted oncology and hematology medications were requested for patients with a poor ECOG performance status or who were near the end of life. Approximately one-third of the patients did not continue treatment with these medications to the outpatient setting. Based on the study results, we plan to incorporate an outpatient-restricted medication request form at our institution, and complete a post-implementation study with an expanded data set and collection timeframe. Having a formalized process for requesting formulary, outpatient-restricted oncology and hematology medications could optimize drug utilization in the inpatient setting.