You are here

Inpatient Antineoplastic Medication Administration And Associated Drug Costs: Institution of a Hospital Policy Limiting Inpatient Administration

Alexandra E. Foster PharmD
David J. Reeves PharmD, BCOP


The cost of cancer care is increasing rapidly. According to the Agency for Healthcare Research and Quality, the costs associated with cancer treatment in the United States in 2012 totaled $87.2 billion.1 An integral portion of this amount can be attributed to antineoplastic medications, as the cost of new anticancer agents has risen to more than $10,000 per month.2 The global cost of antineoplastic medications has grown an average of 9.8% annually since 2010, with 2015 experiencing a 14.2% increase in the cost of antineoplastic medications to $83.7 billion globally.3

Given the continually increasing costs of antineoplastic medications, finding and implementing cost-saving strategies, as well as maximizing reimbursement, have become an aspect of health care that requires great deliberation. One such strategy is administering antineoplastic medications to patients in the outpatient clinic setting. While antineoplastic medications were historically administered in the hospital setting, there has been a shift toward providing these medications in outpatient clinics to promote cost-savings and to maximize reimbursement through utilization of drug discounts (such as federal 340B outpatient drug discount pricing available to eligible health care organizations) and avoidance of the common method of inpatient reimbursement based on diagnosis-related group.4,5 It should be noted, however, that there are instances in which inpatient administration of antineoplastic medications is necessary. Justifications for hospitalization to receive antineoplastic medications have been established and, overall, reflect clinical parameters, such as the need for prolonged observation, prevention or management of side effects, or the minimization of certain treatment risks that cannot be effectively managed in an outpatient clinic setting.5

A policy defining the appropriate criteria necessitating inpatient administration of antineoplastic medication was implemented in a community teaching hospital in December 2013 in an attempt to shift administration to the outpatient clinic setting when possible (i.e., administration in the outpatient clinic setting after hospital discharge) (Table 1). The policy also outlines a peer-review process for regimens that do not meet designated criteria (Figure 1). The objectives of this study were to evaluate the necessity and drug costs of administering antineoplastic medications in the inpatient setting prior to and after implementation of the policy that defined the appropriate criteria necessitating inpatient administration of antineoplastic medications and to explore potential associated hospital inpatient drug cost-savings.


Study Design

A retrospective chart review was conducted of patients 18 years of age and older receiving inpatient antineoplastic medications during January, April, July, and October of 2010, 2012, 2014, and 2015 at a community teaching hospital. The time periods for data collection prior to the implementation of the policy (2010 and 2012) were selected to demonstrate the baseline utilization of inpatient chemotherapy. Patients receiving chemotherapy in 2011 were not included because data could not be accessed as a result of an electronic medical record change at the hospital. Patients younger than 18 years of age and pregnant women were excluded. The study was approved by the local institutional review board.

Data collected from electronic medical records included demographic information, diagnosis, reasons for hospital admission, antineoplastic regimen (cycle number, schedule, agents, routes of administration, and dosing), documented prior and current adverse effects, and reasons for inpatient administration. Average wholesale price (AWP) in 2015 dollars as listed in Lexicomp Online (Pediatric and Neonatal Lexi-Drugs, Hudson, Ohio: Lexi-Comp, Inc., accessed in January 2015) was utilized to calculate the drug costs of antineoplastic regimens specific to the doses received by each patient. In the case of multiple pricing options, the AWP was calculated based on the strength and package size utilized by the institution or the lowest AWP available if multiple package sizes were used. Two agents, L-asparaginase and alemtuzumab, were excluded from cost analyses due to lack of availability of drug cost data in 2015. The costs assigned to each regimen included only the drug itself and did not take into consideration supportive-care medications such as antiemetics, mesna, etc.

The necessity for inpatient administration of antineoplastic medication was determined based on adherence to the hospital policy. Both authors of this study reviewed each regimen to make this determination. Because the policy was not in effect in 2010 and 2012, the criteria were retrospectively applied to patients receiving antineoplastic medications prior to 2014. Taking into consideration the diagnosis, clinical condition, cycle number, antineoplastic regimen, documented adverse effects, and number of days the regimen was administered prior to discharge, patients were separated into two groups—those meeting one or more criteria and those not meeting criteria for inpatient administration. Patients meeting at least one of the 10 criteria outlined in the policy were considered appropriate for inpatient administration, and patients not meeting criteria were considered inappropriate for inpatient administration (i.e., chemotherapy could have been given in the outpatient clinic setting after hospital discharge). Patients admitted to the inpatient unit for a 23-hour observation to receive antineoplastic medications are billed/reimbursed as outpatients and were removed from multiple analyses (including those for cost) to reflect the population truly reimbursed as inpatients. Due to the cost of rituximab (Rituxan, Genentech), much interest has focused on the administration of this agent on the day prior to or following admission for an antineoplastic regimen that otherwise meets criteria for hospital admission. For this reason, additional designations of appropriateness for inpatient rituximab were used to determine if the drug could have been administered in the outpatient clinic setting (i.e., the chemotherapy portion of the regimen met criteria for inpatient administration; however, rituximab could have been administered prior to admission).


The primary outcome was the proportion of patients appropriately receiving antineoplastic medications in the inpatient setting (i.e., those meeting institutional criteria for inpatient drug administration). The four years were compared to determine the presence of any trends associated with policy implementation. Secondary outcomes included a comparison of drug costs of regimens meeting and not meeting policy criteria for inpatient antineoplastic drug administration throughout the four years, trends in cancer diagnoses and drug regimens throughout the four years, and a comparison of the use and drug cost of rituximab in the inpatient setting when outpatient clinic administration was feasible throughout the four years.

Statistical Analysis

Descriptive statistics were used to represent the data in the two groups of patients (those meeting criteria and those not meeting criteria for inpatient antineoplastic medication administration). Chi-square and Kruskal-Wallis ANOVA tests were utilized to analyze data and assess for trends in the primary and secondary outcomes between the four years. A P value of less than 0.05 was considered statistically significant.


Inpatient antineoplastic regimens were administered to 648 patients in January, April, July, and October of 2010, 2012, 2014, and 2015 (179 in 2010; 206 in 2012; 123 in 2014; and 140 in 2015). Based on these numbers, the annualized number of regimens administered in the inpatient setting in 2010, 2012, 2014, and 2015 was estimated to be 537, 618, 369, and 420, respectively. Baseline characteristics of patients receiving inpatient antineoplastic regimens were generally similar across the four years, with the majority of patients being women and admitted specifically for antineoplastic medication administration (Table 2). However, a difference in the number of patients admitted for 23-hour observation was observed (28% of patients in 2010; 29% of patients in 2012; 15% of patients in 2014; and 22% in 2015; P = 0.03). Of all regimens administered in the inpatient setting, 80% in 2010, 78% in 2012, 83% in 2014, and 91% in 2015 met criteria for inpatient administration (P = 0.005) (Table 3). When the years were compared, the significant differences were present between the years 2010 and 2015 (P = 0.003) and 2012 and 2015 (P < 0.001), after Bonferroni correction. The same trends were apparent when patients admitted for 23-hour observation were excluded (Table 3).

Drug Cost Analysis

The total annualized AWPs of antineoplastic medications administered in the inpatient setting are listed in Table 4. The annualized AWP of antineoplastic medications for patients not meeting criteria for inpatient administration decreased from $269,049 in 2010 to $105,218 in 2014 and $105,447 in 2015—an estimated annual inpatient drug cost-savings of at least $163,602. The annualized inpatient drug cost of patients meeting criteria for inpatient administration of antineoplastic regimens averaged $1.27 million per year during these four years, with the cost increasing from $1.06 million in 2010 to $1.32 million in 2015.

Trends in Diagnoses, Antineoplastic Regimens, and Reason for Administration

Excluding patients admitted for 23-hour observation, top diagnoses and regimens were similar through the years studied; leukemia and lymphoma were the most common, followed by sarcoma and non–small-cell lung cancer in 2010 and 2012, respectively, and gynecological malignancies in 2014 and 2015 (Table 5). Top regimens correlated with the top diagnoses, with regimens often used for leukemia and lymphoma being most commonly administered in all years studied. In patients not meeting criteria for administration of antineoplastic medications in the inpatient setting, there was a trend toward a decreasing variety of regimens, with only one (carboplatin/paclitaxel—a relatively inexpensive regimen) administered to more than 5% in 2015. In addition, all monoclonal antibodies administered in the inpatient setting in 2015 met criteria in the institutional policy.

Reasons for inpatient chemotherapy administration are illustrated in Figure 2. The number of patients receiving these medications for convenience decreased and was the lowest in 2015. After initiation of the policy regarding inpatient anti-neoplastic medication use in 2013, there were four instances in which the department chair approved the inpatient use of these medications (three in 2014 and one in 2015).


There was a statistically significant difference in the number of patients receiving rituximab in the years studied (P = 0.049) (Table 6); however, no difference was present between individual years after Bonferroni correction for multiple comparisons. The same was true for rituximab that could have been administered in the outpatient clinic setting either the day before or after hospitalization. The potential annualized inpatient drug cost-savings associated with shifting rituximab to the outpatient clinic setting, when appropriate, ranged from $111,004 in 2010 to $234,548 in 2014 (Table 6).


Implementation of a policy limiting inpatient administration of antineoplastic medication to those patients who meet the criteria listed in Table 1 significantly increased the proportion of regimens that met those criteria and decreased the inpatient drug costs associated with administration of regimens in patients who did not meet those criteria. Moreover, a decrease in convenience as the motivating factor for inpatient administration was observed after initiation of the policy. Though the existence of the policy was disseminated to all stakeholders (physicians, pharmacists, nurses, etc.), there were still patients receiving inpatient antineoplastic medications who did not meet criteria. Many of these were patients receiving relatively inexpensive regimens (i.e., generic, non-monoclonal antibody-based regimens) that were not subjected to the approval process in the policy or that were given based on orders during holidays or weekends with limited availability and accountability of those enforcing the policy.

Despite a decrease in the cost of drugs administered to those not meeting criteria, there was an increase in drug cost from 2010 to 2015 of those regimens that did meet criteria. This could be attributed to higher costs of new agents approved after 2010 or the increased number of patients receiving chemotherapy appropriately in the inpatient setting. The increased drug cost observed in 2012 is most likely due to chance; one patient received ipilimumab (Yervoy, Bristol-Myers Squibb) during one of the months in which data was collected, increasing annualized drug costs for that year. Although the drug cost may be skewed by this, the number of patients receiving antineoplastic medications that did not meet criteria remained similar to 2010 and was significantly greater than in 2015.

Areas of opportunity identified through this analysis include rituximab utilization in the inpatient setting and enforcement of the protocol. There was an increase in the number of rituximab administrations that could have been shifted to the out-patient clinic setting over time. Multiple regimens allow for the administration of rituximab the day prior to admission for chemotherapy that otherwise would meet criteria for inpatient administration.68 Administration in this manner could lead to further inpatient drug cost-savings. In addition, though there was a decrease in the number of inpatient antineoplastic medication administrations not meeting criteria, more could be done, such as education or increased weekend/holiday accountability, to increase enforcement so that closer to 100% of the patients meet the criteria outlined in Table 1.

With the changes in reimbursements for hospitals, exploring potential methods for cost-savings is more imperative than ever. This study has helped delineate the benefits associated with implementation of a policy that defines the appropriate criteria for the use of inpatient chemotherapy. This could be used to further encourage health care professionals to administer chemotherapy in the outpatient clinic setting and serve as a starting point for other institutions interested in implementing similar initiatives to help decrease inpatient drug costs.

Limitations of this study include its retrospective chart review design and study population. The study may be subject to selection bias, as the study population included patients receiving inpatient chemotherapy during select months of four select years. While variations in the regimens and associated costs of chemotherapy administered would differ throughout each calendar year, limiting the study population to those specific time periods allowed for data to be collected consistently over different years. Additionally, due to the use of AWP instead of actual institutional cost of antineoplastic medications, the drug cost-savings identified may not be reflective of the full cost-saving potential available to the institution. However, because medication costs vary at different institutions, use of the AWP for each antineoplastic medication allowed for calculation of drug cost-saving estimates with generalizability to institutions outside of this community teaching hospital. It is important to note that potential drug cost-savings calculated in this study could increase even further by maximizing reimbursement in the outpatient setting and taking advantage of programs such as 340B outpatient drug pricing.


Implementation of a policy establishing criteria for inpatient antineoplastic drug administration led to a decrease in both the number of patients receiving antineoplastic medications in the inpatient setting who did not meet the criteria and the associated drug costs. More education and implementation of further policies, specifically directed toward appropriate inpatient administration of rituximab, may increase the impact of this policy on the institution’s oncology medication costs. Further studies regarding the impact of maximizing administration of antineoplastics in the outpatient clinic setting on reimbursement are necessary to fully elucidate the drug–cost benefit of shifting antineoplastic medication administration to the outpatient clinic setting.

Figures and Tables

Hospital Antineoplastic Medication Approval Process Algorithm

Reasons for Inpatient Antineoplastic Medication Administration

Justification for Hospitalization for Antineoplastic Medication Administration

  • Emergent chemotherapya
  • Higher-dosage cisplatin, ≥ 75 mg/m2
  • Intra-arterial chemotherapy/chemoembolization/heated intraperitoneal chemotherapy
  • Induction for acute leukemia
  • Ifosfamide
  • Complex chemotherapy programsb
  • High-dose methotrexate, ≥ 500 mg/m2
  • 23-hour observation (billed and reimbursed as an outpatient visit)
  • Other special circumstances upon approval of department chairs
  • Rituximab when clinically necessary (i.e., emergency, monitoring required)

aDefined as antineoplastic medications for initial treatment of leukemia or lymphoma using standard regimens and highly chemosensitive solid tumors requiring an immediate reduction in tumor size.

bDefined as regimens requiring more than six hours of continuous observation and drug administration.

Baseline Characteristics

2010 2012 2014 2015 P Value
All patients (N = 648; N [annualized] = 1,944)
Patients, n 179 206 123 140
Annualized patients, n 537 618 369 420
Male, n (%) 75 (42) 79 (38) 54 (44) 57 (41) 0.780
Age, median (IQR) 60 (41–79) 59 (39–79) 58 (33–83) 61 (44–78) 0.941
23-hour observation, n (%) 50 (28) 59 (29) 19 (15) 31 (22) 0.031
Admitted for anticancer drug administration, n (%) 117 (65) 118 (57) 71 (58) 82 (59) 0.368
Subgroup—excluding 23-hour observation (N = 489; N [annualized] = 1,467)
Patients, n 129 147 104 109
Annualized patients, n 387 441 312 327
Male, n (%) 69 (53) 70 (48) 53 (51) 51 (47) 0.698
Age, median (IQR) 58 (38–78) 57 (37–77) 59 (34–84) 60 (40–80) 0.852
Admitted for anticancer drug administration, n (%) 68 (53) 60 (41) 52 (50) 55 (50) 0.202

IQR = interquartile range.

Compliance With Criteria for Inpatient Antineoplastic Medication Administration

2010 2012 2014 2015 P Value
All patients (N = 648, N [annualized] = 1,944)
Met criteria for inpatient administration, n (%)a 144 (80) 160 (78) 102 (83) 128 (91) 0.005
Met criteria for inpatient administration, n [annualized] (%) 432 (80) 480 (78) 306 (83) 384 (91)
Excluding 23-hour observation (N [annualized] = 1,140)
Met criteria for inpatient administration, n (%) 94 (73) 101 (69) 83 (80) 97 (89) 0.001
Met criteria for inpatient administration, n [annualized] (%) 282 (73) 303 (69) 249 (80) 291 (89)

aDifference between 2010 and 2015 (P = 0.003) and 2012 and 2015 (P < 0.001) after Bonferroni correction for multiple comparisons.

Annualized Average Wholesale Price (AWP)* (Excluding Patients Admitted for 23-Hour Observation)

2010 2012 2014 2015

Total cost of drugs for patients receiving inpatient anticancer drugs $1,330,391 $ 2,310,154 $ 1,251,497 $1,429,267

Average cost of regimen per patient $3,438 $5,238 $4,011 $4,371

Met institutional criteria for inpatient anticancer drug administration
 No $269,049 $758,664 $105,218 $105,447
 Yes $1,061,343 $ 1,551,490 $ 1,146,278 $1,321,034

*AWP as of January 2015 as listed in Lexicomp Online (Pediatric and Neonatal Lexi-Drugs, Hudson, Ohio: Lexi-Comp, Inc.).

Trends in Diagnoses and Antineoplastic Regimens Administered, Excluding Patients Admitted for 23-Hour Observation (N [Annualized] = 1,140)

2010 2012 2014 2015

Diagnosis, n (%) AML 21 (16) AML 26 (18) AML 21 (20) AML 27 (25)
NHL 21 (16) NHL 23 (16) NHL 18 (17) NHL 17(16)
ALL/BL 10 (8) Ovarian 14 (10) ALL/BL 8 (8) ALL/BL 16 (15)
Sarcoma 9 (7) NSCLC 8 (5) Ovarian 7 (7) Endometrial 38 (7)

Top five inpatient regimens, n (%) 7 + 3 11 (9) HIDAC 12 (8) 7 + 3 9 (9) HIDAC 13 (12)
Hyper-CVAD 11 (9) 7 + 3 10 (7) PC 8 (8) 7 + 3 11 (10)
PC 10 (8) Rituximab 10 (7) HD MTX 7 (7) HD MTX 8 (7)
HIDAC 8 (6) PC 9 (6) HIDAC 6 (6) Rituximab 8 (7)
HD MTX 7 (5) AIM 8 (5) Hyper-CVAD 6 (6) PC 7 (6)

Top inpatient regimens not meeting criteria (> 5%) PC 7 (20) PC 9 (20) PC 8 (38) PC 5 (42)
Gemcitabine 5 (14) Bortezomib 4 (9) Paclitaxel 2 (10)
Cisplatin 3 (9) BR 3 (6) RCHOP 2 (10)
Avastin/FOLFIRI 2 (6) Docetaxel 3 (6)
Doxil 3 (6)
Bortezomib 2 (6) Gemcitabine 3 (6)
Rituximab 3 (6)
RCHOP 3 (6)

7 + 3 = seven days of standard-dose cytarabine + three days of daunorubicin or idarubicin; AIM = doxorubicin, ifosfamide, mesna; ALL = acute lymphoblastic leukemia; AML = acute myeloid leukemia; BL = Burkitt’s lymphoma; BR = bendamustine, rituximab; HD MTX = high-dose methotrexate; HIDAC = high-dose intermittent cytarabine; CVAD = cyclophosphamide, vincristine, doxorubicin, dexamethasone; FOLFIRI = fluorouracil, leucovorin, irinotecan; NHL = non-Hodgkin’s lymphoma; NSCLC = non–small-cell lung cancer; PC = paclitaxel, carboplatin; RCHOP = rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone.

Inpatient Rituximab Use (Excluding 23-Hour Observation)

2010 2012 2014 2015 P Value

Received rituximab, n* 13 30 22 24 0.049

Received rituximab, n [annualized] 45 93 54 72

Rituximab could have been outpatient, n (%)* 1 (7) 10 (32) 11 (61) 9 (38) 0.014

Rituximab could have been outpatient, n [annualized] (%) 3 (7) 30 (32) 33 (61) 27 (38)

Total cost of rituximab $435,054 $945,6713 $387,854 $737,469
Could have been outpatient $111,004 $168,591 $234,548 $211,728

*No difference between years after Bonferroni correction for multiple comparisons.

Author bio: 
At the time of writing, Dr. Foster was a student in the Department of Pharmacy Practice in the College of Pharmacy and Health Sciences at Butler University in Indianapolis, Indiana. Dr. Reeves is an Associate Professor in Butler’s College of Pharmacy and Health Sciences and a Clinical Oncology Pharmacist in the Department of Pharmacy at St. Vincent Indianapolis Hospital in Indianapolis.


  1. Soni A. Top five most costly conditions among adults age 18 and older, 2012: estimates for the US civilian non-institutionalized population Statistical Brief #471 Agency for Healthcare Research and Quality. April 2015; Available at: Accessed March 3, 2017
  2. Kantarjian HM, Fojo T, Mathisen M, Zwelling LA. Cancer drugs in the United States: justum pretium—the just price. J Clin Oncol 2013;31:3600–3604.
  3. IMS Institute for Healthcare Informatics. Global oncology trend report: A review of 2015 and outlook to 2020 June 2016; Available at: Accessed March 3, 2017
  4. Warren A, Shankar A. Oncology transactions and the 340B drug pricing program. J Oncol Pract 2013;9;(2):89–91.
  5. Dollinger M. Guidelines for hospitalization for chemotherapy. Oncologist 1996;1:107–111.
  6. Evans AM, Carson KR, Kolesar J, et al. A multicenter phase II study incorporating high-dose rituximab and liposomal doxorubicin into the CODOX-M/IVAC regimen for untreated Burkitt’s lymphoma. Ann Oncol 2013;24:3076–3081.
  7. Jerman M, Jost LM, Taverna C, et al. Rituximab-EPOCH, an effective salvage therapy for relapsed, refractory or transformed B-cell lymphomas: results of a phase II study. Ann Oncol 2004;15:511–516.
  8. Ohmachi K, Niitsu N, Uchida T, et al. Multicenter phase II study of bendamustine plus rituximab in patients with relapsed or refractory diffuse large B-cell lymphoma. J Clin Oncol 2013;31:2103–2109.