¹Department of Pediatric Oncology, Dana-Farber Cancer Institute, USA

²Division of Hematology/Oncology, USA

³Harvard Medical School, Boston, MA, USA

⁴Department of Pharmacy, Children's Hospital, USA

Correspondence to: Dr S K Parsons, Department of Pediatric Oncology, Dana Farber Cancer Institute, 44 Binney Street, Boston, MA 02115, USA

We studied the practice patterns regarding intravenous (i.v.) ondansetron in children receiving stem cell transplants (SCT) at The Children's Hospital, Boston to identify cost efficiencies. The pharmacy provided information on material and preparation costs on 36 patients who received i.v. ondansetron during 41 SCT in 1995. We examined the effects of frequency, duration, and route of administration on costs. There were 498 days of ondansetron administration costing $49 083 (95$). Tremendous variation existed in frequency and duration with one third receiving i.v. ondansetron once daily, despite published evidence of equivalence of once a day and divided dosing. A switch to once daily i.v. dosing for all patients would have resulted in 28% savings. The median duration of use was 11 days (range 1-48); placing a cap for 7-10 days based on the length of SCT conditioning regimens, would produce savings of 48-60% over current use. By shifting administration route from i.v. to oral, a savings of 67% over current use, without a cap on duration, would be realized. Identifying areas for cost savings can be achieved after thorough analysis of all the component costs. We demonstrated that significant cost reductions could be realized by simple changes in prescribing practices without jeopardizing efficacy. These savings are achieved by standardizing dosing interval, route of administration and duration of treatment without altering daily dosage or access to an effective antiemetic. Bone Marrow Transplantation (2000) 25, 553-557.

antiemetics; drug costs; serotonin antagonists; ondansetron; pediatric transplantation

While the introduction of the selective serotonin receptor antagonists has greatly enhanced the control of acute chemotherapy-induced nausea and vomiting, the drug acquisition cost of these agents is considerably higher than conventional antiemetic therapy.^1,2,3,4 There is general consensus about the need to balance clinical efficacy and cost.^5,6,7 However, identifying and understanding the individual contributing factors to all components of costs and the actual steps required to accomplish that balance are less well defined.

The 5-hydroxytryptamine (5-HT₃) receptor antagonists have been extensively tested in clinical trials, initially in general oncology and more recently, in stem cell transplantation (SCT) and have been shown to be more effective than conventional agents in emesis control following moderately and highly emetogenic chemotherapy in children and adults.^8,9,10 These agents have demonstrated similar clinical efficacy when administered parenterally as either once daily or divided dosing (ie the first 24 h after initiation of treatment) in the acute phase of chemotherapy-related nausea and vomiting.^11,12 Recent studies have shown that the oral formulation of these agents is as effective as the parenteral route both for moderate and highly emetogenic chemotherapy.^{13,14,15,16,17,18,19}

Obtaining effective control of emesis in the SCT patient can be challenging because of the highly emetogenic nature of the conditioning regimens, often administered over several days, the added emetogenicity of radiation, and the combined acute and delayed patterns of emesis experienced. Nevertheless, several studies in adults and children have demonstrated that the 5-HT₃ receptor antagonists are effective in controlling conditioning-related emesis, both during high dose chemotherapy^8,14,15 and total body irradiation.^{10,20,21,22,23,24,25} Recent studies in SCT demonstrate that once daily dosing is equivalent to divided dosing and that equivalence is demonstrated with either the parenteral or oral formulation, mirroring findings in the general oncology population.^8,12,24,25

While the efficacy of the 5-HT₃ receptor antagonists is superior in the acute phase of chemotherapy or radiotherapy-induced nausea and vomiting, this efficacy diminishes dramatically in the delayed phase (ie >24 h). This delayed phase of therapy-related emesis is experienced more often in chemotherapy regimens containing platinum-containing agents, high-dose cyclophosphamide or doxorubicin.^26,27 Therefore, the adoption of alternative therapy for delayed nausea and vomiting has been useful.^28,29,30

Faced with a steadily rising annual budget, our institution's Pharmacy and Therapeutics Committee directed each clinical program to evaluate utilization patterns and maximize the cost effectiveness of those drugs with the highest level of annual expenditure. Antiemetics and colony-stimulating factors were two drug categories identified by our program for an initial evaluation. In response to requests to be more cost conscious, our clinical program formally studied the practice patterns regarding parenterally administered (i.v.) ondansetron (Zofran; Cerenex, Research Triangle Park, NC, USA), the 5-HT₃ antagonist on formulary, in children receiving SCT at the Children's Hospital, Boston to identify areas for potential cost efficiencies.

We have developed a paradigm for analyzing drug costs in a health care environment which is increasingly cost-conscious. Specifically, this paradigm characterizes the cost relationships that exist between clinician prescribing practices, pharmacy stocking and distribution practices, and manufacturer's decisions regarding drug production, available dosing units, and pricing structure (Figure 1). In this paper we report our findings regarding the use of antiemetics in the SCT population; the results of the colony-stimulating factors evaluation utilizing this model are reported elsewhere.³¹ The goal of this analysis was to identify possible cost-containment strategies in the use of highly potent, yet expensive, 5-HT₃ antagonist antiemetics while maintaining clinical efficacy and access to these clinically important drugs.

Methods

Patient sample

The sample included 36 patients who received 41 transplants and parenterally administered ondansetron. This sample was derived from the total of 55 transplants in 50 patients performed in FY95. Patients were excluded from this analysis if complete cost information was not available or if they received a combination of parenteral and oral ondansetron. In addition, one patient was excluded from the modeling as a weight outlier (eg >twice the weight of the next heaviest patient). Information from the hospital's pharmacy records and financial records was linked to each SCT hospitalization via unique identifiers including name, medical record number, date of birth, and dates of hospitalization. The dates of intravenous ondansetron use, daily dosage, dosing schedule, and the total number of doses dispensed per patient for each SCT hospitalization were collected.

Costs

Pharmacy provided all information on both the material cost and the cost of dispensing (preparing) one dose of ondansetron, depending on dosage form (ie dose fee code). For this study, the total direct cost for one dose of ondansetron is the sum of material and preparation cost, exclusive of drug administration costs (eg nursing labor or supplies) or institutional overhead. Material costs are based on the institutionally negotiated discounts of the wholesale acquisition cost (WAC), available to health care facilities via group purchasing agreements. The WAC is approximately 80% of the average wholesale price (AWP). When combined with institutional discounts from group purchasing agreements, the resultant cost is approximately 75-80% of the AWP. Material costs in our institution were $165/40 mg for the parenteral formulation, $112/40 mg for the oral solution, and $8/4 mg for the oral tablet. The preparation cost is the fee associated with dispensing the drug and is dependent upon the formulation (eg parenteral, oral tablets or solution). This fee, set by the Pharmacy Department was $1.51 for a tablet, $7.53 for the oral solution, and $20.07 for the i.v. preparation. The cost information was not inflation-adjusted to current year dollars and thus reflects 1995 dollars (95$).

In this model, we calculated the effects of dosing schedule, duration, and route of administration on drug costs using weight-based dosing (0.45 mg/kg/day) for the i.v. formulation. For the oral formulations (tablets and solution), we used the manufacturers' age-based recommendations.³² These recommendations for oral administration are: 4 mg three times daily for children 4-11 years and 8 mg twice daily for children 12 years and older. Since no recommendations are available for children less than 4 years of age, the oral solution was modeled using weight-based dosing (0.45 mg/kg/day) on a three times daily divided dosing schedule. Information on cost of these oral forms of ondansetron, in addition to the various charges for preparing the different forms of the same drug, were obtained from our pharmacy. The duration was modeled for either 7 or 10 days based on an additional 24 h of anti-emetic coverage after completion of the preparative regimen typically of 6-9 days duration.

Results

Patient and treatment characteristics of the 41 SCT are depicted in Table 1. Four hundred and ninety-eight days of i.v. ondansetron were administered, accounting for an annual expenditure of $49 083 (95$). Initial review of pharmacy's dispensing records of intravenous ondansetron revealed considerable variability with respect to dosing schedule and duration of treatment. Despite convincing data of the utility of a daily dosing regimen, ondansetron was administered on a daily basis in only eight (20%) of the transplants, on a divided schedule in 28 transplants (68%), and a combination of daily and divided dosing in the remaining five (12%). The median duration of treatment was 11 days (range, 1-48 days) at an average dose of 0.45 mg/kg/day, independent of dosing schedule. Therefore, the prescribing practices of clinicians became the focus of this analysis given the dramatic variability of use, and the implications of these decisions.

We first modeled cost variations related to frequency of i.v. ondansetron administration. The preparation costs were $20.07/day for the daily dose and three times more, or $60.21/day, for three times daily dosing. Therefore, an overall saving of 28% for the total population and 35% in the subset of patients who previously received only divided dosing would be achieved by switching to once daily dosing. The impact of this change would vary by patient weight, reflecting the different ratios of material to preparation costs (Figure 2).

Next, we examined cost variations related to duration of ondansetron administration. This model was designed to limit or cap the duration of proposed daily treatment at 24 h after completion of the SCT conditioning regimen given the published maximal efficacy of serotonin antagonists for acute phase regimen-related emesis. The median length of the conditioning regimen was 6 days (range, 6-9 days). A 10-day cap would result in additional savings of 28%, while a 7-day cap would translate to savings of 45% compared to daily i.v. dosing (Figure 3). Overall, the combined strategy of daily i.v. dosing for a fixed 7-day duration would result in up to 60% savings (estimated $29 680/year) over current practice.

Finally, we evaluated cost variations related to route of ondansetron administration, parenteral vs oral. This model was based on growing evidence of the equivalence of the two routes in adults as previously cited. By shifting from an intravenous to oral route of administration, either solution or tablets, a cost saving (over current use) of 67% would be realized without a cap on drug duration (Figure 3). The magnitude of these savings is a function of both decreases in material cost as well as a dramatic decrease in the preparative costs. Specifically, the preparative cost for tablets and the solution is significantly less than the i.v. formulation (92% and 63% less, respectively). If this approach was combined with a cap on the duration of treatment, the savings would be even greater.

Discussion

The highly selective 5-HT₃ receptor antagonists offer significant therapeutic advantages in the prevention of acute chemotherapy- or radiotherapy-induced nausea and vomiting in both general oncology and SCT populations. These advantages include superior emesis control, ease of administration, and a more favorable toxicity profile than conventional antiemetic therapy. Despite mounting evidence of their therapeutic benefit, the considerably higher material costs of these agents necessitate careful control of their use. Although drug costs in our institution and elsewhere account for a relatively small percentage of total SCT-related cost (estimated 11%), they represent a significant portion of the variable costs and, therefore, warrant close scrutiny. The physical and psychological costs of poorly controlled nausea and vomiting can also be significant and should be taken into account when evaluating the balance between efficacy and costs.^33,34

We demonstrated in this analysis that considerable cost savings can be achieved by standardizing dosing schedule and duration, the latter related to the length of the conditioning regimen. In addition, further savings could be achieved by switching from i.v. to oral formulations, based on published data in the SCT population.^21,24,25 These savings are achieved without changing practitioner-prescribing access to this important class of drugs. Other cost reduction strategies, such as decreased daily dosage or substitution with less expensive alternative agents, have been proposed,^{35,36,37,38,39} but were not considered in our analysis.

To realize projected cost savings, standardized guidelines addressing each of these variables (schedule of administration, duration of treatment and route), must be developed and implemented. The use of standardized guidelines for these agents has been shown in several studies to enhance cost control of their use.^5,39,40,41 In addition, education of clinical staff, pharmacy, and families is necessary, including formal recommendations of alternate treatment strategies in case of poor antiemetic control. Post-implementation monitoring of emetic control and side-effects is also imperative to ensure no adverse clinical impact. These strategies are currently underway at our institution.

Even with the use of standardized guidelines, some individualization of treatment may be required. For example, using ondansetron solution, although more expensive than tablets, may be most appropriate for a younger child to ensure maximum patient tolerance. In addition, patients receiving highly emetogenic conditioning regimens should be targeted for the most aggressive anti-emetic therapy that may require use of multiple agents and/or sequential administration of these agents. In these instances, standardized guidelines must be balanced with formal clinical reevaluation.

Utilizing the model described in this study, we were able to identify potential areas of cost savings, namely modification of the prescribing practices of clinicians. These modifications were based on published studies of equal clinical efficacy of various regimens. In a similar analysis of the use of colony-stimulating factors in the SCT population, we demonstrated cost efficiencies by altering stocking and distributing practices of the pharmacy and had implications for the manufacturer related to vial sizes needed for the pediatric population.³¹ In sum, these analyses highlight the need to identify and understand all of the factors contributing to drug costs when making decisions about appropriate access to 'expensive' drugs to maintain the balance between optimal efficacy and reasonable costs.

Acknowledgements

We gratefully acknowledge the contributions of Peter J Blanding, RPh, and Frank Federico, RPh in the preparation of this manuscript. SKP is a Dyson Scholar in Clinical Research at the Dana-Farber Cancer Institute. LEH participated in the International Medical Student Exchange Program between the University of Groningen, the Netherlands, and Harvard Medical School, Boston. KJL was awarded a fellowship from the Harvard Medical School Academic Societies' Committee on Student Research.

1 Nightengale BS, Mauch RP. A pharmacoeconomic review of granisetron and ondansetron in the prophylaxis of chemotherapy-induced nausea and vomiting. P & T 1998; 23: 119-130,

2 Johnson N, Bosanquet N. Cost effectiveness of 5-hydroxytryptamine₃ receptor antagonists: a retrospective comparison of ondansetron and granisetron. Anti-Cancer Drugs 1995; 6: 243-249, MEDLINE

3 Grunberg SM. Cost-effective use of antiemetics. Oncology (Huntingt) 1998; 12: (3 Suppl.4) 38-42, MEDLINE

4 Stewart DJ, Dahrouge S, Coyle D, Evans WK. Costs of treating and preventing nausea and vomiting in patients receiving chemotherapy. J Clin Oncol 1999; 17: 344-351, MEDLINE

5 Goldsmith M, Wodinsky H. Controlling costs using evidence-based guidelines: 5-HT₃ antagonists as a model. Managed Care Cancer 1999; 1: 20-24,

6 Grunberg SM. Innovative approaches in the treatment of emesis. Support Care Cancer 1997; 5: 9-11, MEDLINE

7 Jones AL, Lee GJ, Bosanquet N. The budgetary impact of 5-HT₃-receptor antagonists in the management of chemotherapy-induced emesis. Eur J Cancer 1993; 29A: 51-56,

8 Fox-Geiman M, Fisher SG, Kiley K et al. Double-blind randomized comparison of oral granisetron, oral ondansetron and IV ondansetron for regimen-related nausea and vomiting (n/v) in patients undergoing stem cell transplants. Proc ASCO 1999; 18: 592a,

9 Hainsworth JD, Harvey W, Pendergrass K et al. A single-blind comparison of intravenous ondansetron, a selective serotonin antagonist, with intravenous metoclopramide in the prevention of nausea and vomiting associated with high-dose cisplatin chemotherapy. J Clin Oncol 1991; 9: 721-728, MEDLINE

10 Hewitt M, Cornish J, Pamphilon D, Oakhill A. Effective emetic control during conditioning of children for bone marrow transplant using ondansetron, a 5-HT3 antagonist. Bone Marrow Transplant 1991; 7: 431-433, MEDLINE

11 Hainsworth JD, Hesketh PJ. Single-dose ondansetron for the prevention of cisplatin-induced emesis: efficacy results. Semin Oncol 1992; 19: (6 Suppl.15) 14-19,

12 Perez EA. Comparative efficacy of oral and intravenous granisetron for the prevention of acute chemotherapy-induced emesis. Clin Ther 1996; 18: 578-590, MEDLINE

13 Perez EA, Hesketh PJ, Sandbach J et al. Comparison of single-dose oral granisetron versus intravenous ondansetron in the prevention of nausea and vomiting induced by moderately emetogenic chemotherapy: a multicenter, double-blind, randomized parallel study. J Clin Oncol 1998; 16: 754-760, MEDLINE

14 Gralla RJ, Navari RM, Hesketh PJ et al. Single-dose oral granisetron has equivalent antiemetic efficacy to intravenous ondansetron for highly emetogenic cisplatin-based chemotherapy. J Clin Oncol 1998; 16: 1568-1573, MEDLINE

15 Frakes LA, Brehm TL, Kosty MP et al. An all oral antiemetic regimen for patients undergoing high dose chemotherapy with peripheral blood stem cell transplant. Bone Marrow Transplant 1997; 20: 473-478, MEDLINE

16 Gandara DR, Roila F, Warr D et al. Consensus proposal for 5HT3 antagonists in the prevention of acute emesis related to highly emetogenic chemotherapy. Dose, schedule, and route of administration. Support Care Cancer 1998; 6: (3) 237-243, Article MEDLINE

17 Beck TM, Ciociola AA, Jones SE et al. Efficacy of oral ondansetron in the prevention of emesis in outpatients receiving cyclophosphamide-based chemotherapy. Ann Intern Med 1993; 118: 407-413, MEDLINE

18 Spector JI, Lester EP, Chevlen EM et al. A comparison of oral ondansetron and intravenous granisetron for the prevention of nausea and emesis associated with cisplatin-based chemotherapy. Oncologist 1998; 3: 432-438, MEDLINE

19 Krzakowski M, Graham E, Goedhals L et al. A multicenter, double-blind comparison of i.v. and oral administration of ondansetron plus dexamethasone for acute cisplatin-induced emesis. Ondansetron Acute Emesis Study Group. Anticancer Drugs 1998; 9: 593-598, MEDLINE

20 Mehta NH, Reed CM, Kuhlman C et al. Controlling conditioning-related emesis in children undergoing bone marrow transplantation. Oncol Nurs Forum 1997; 24: 1539-1544, MEDLINE

21 Abbott B, Ippoliti C, Bruton J et al. Antiemetic efficacy of granisetron plus dexamethasone in bone marrow transplant patients receiving chemotherapy and total body irradiation. Bone Marrow Transplant 1999; 23: 265-269, MEDLINE

22 Orchard PJ, Burns L, Rogosheske T et al. A prospective, randomized trial of the anti-emetic efficacy of ondansetron and granisetron during bone marrow transplant. Blood 1998; 92: (Suppl.1) 279a,

23 Jurgens H, McQuade B. Ondansetron as prophylaxis for chemotherapy and radiotherapy-induced emesis in children. Oncology 1992; 49: 279-285, MEDLINE

24 Spitzer TR, Friedman CJ, Bushnell W et al. Oral granisetron (Kytril) and ondansetron (Zofran) in the prevention of hyperfractionated total body irradiation induced emesis: the results of a double-blind, randomized parallel group study. Blood 1998; 92: (Suppl.1) 278a,

25 Spitzer TR, Bryson JC, Cirenza E et al. Randomized double-blind, placebo-controlled evaluation of oral ondansetron in the prevention of nausea and vomiting associated with fractionated total body irradiation. J Clin Oncol 1994; 12: 2432-2438, MEDLINE

26 Latreille J, Pater J, Johnston D et al. Use of dexamethasone and granisetron in the control of delayed emesis for patients who receive highly emetogenic chemotherapy. J Clin Oncol 1998; 16: 1174-1178, MEDLINE

27 Kris MG, Gralla RJ, Clark RA et al. Incidence, course, and severity of delayed nausea and vomiting following the administration of high-dose cisplatin. J Clin Oncol 1985; 3: 1379-1384, MEDLINE

28 Italian Group for Antiemetic Research. Ondansetron vs. metoclopramide, both combined with dexamethasone in the prevention of cisplatin-induced delayed emesis. J Clin Oncol 1997; 15: 124-130, MEDLINE

29 Kris MG, Gralla RJ, Tyson LB et al. Controlling delayed vomiting: double-blind, randomized trial comparing placebo, dexamethasone alone, and metoclopramide plus dexamethasone in patients receiving cisplatin. J Clin Oncol 1989; 7: 108-114, MEDLINE

30 Roila F, Ballatori E. The efficacy and cost-effectiveness of various antiemetic regimens. Curr Opin Oncol 1998; 10: 310-315, MEDLINE

31 Parsons SK, Levine KJ, Hoorntje LE et al. A fiscal review of colony stimulating factor use in pediatric stem cell transplantation: an institution-based analysis (submitted for publication).

32 Zofran Product Information. Glaxo Wellcome, October. 1996;

33 O'Brien BJ, Rusthoven J, Rocchi A et al. Impact of chemotherapy-associated nausea and vomiting on patients' functional status and on costs: survey of five Canadian centres. CMAJ 1993; 149: 296-302, MEDLINE

34 Soukop M, McQuade B, Hunter E et al. Ondansetron compared with metoclopramide in the control of emesis and quality of life during repeated chemotherapy for breast cancer. Oncology 1992; 49: 295-304, MEDLINE

35 Pai VB, Nahata MC, Koepke J et al. Efficacy and safety of oral ondansetron in the control of acute chemotherapy induced nausea and vomiting (ACINV). Proc ASCO 1999; 18: 603a,

36 Hesketh PJ, Beck T, Uhlenhopp M et al. Adjusting the dose of intravenous ondansetron plus dexamethasone to the emetogenic potential of the chemotherapy regimen. J Clin Oncol 1995; 13: 2117-2122, MEDLINE

37 Holdsworth MT, Raisch DW, Duncan MH et al. Assessment of chemotherapy-induced emesis and evaluation of a reduced-dose intravenous ondansetron regimen in pediatric outpatients with leukemia. Ann Pharmacother 1995; 29: 16-21, MEDLINE

38 Beck T, York M, Chang A et al. Oral ondansetron 8 mg twice daily is as effective as 8 mg three times daily in the prevention of nausea and vomiting associated with moderately emetogenic cancer chemotherapy. Cancer Invest 1997; 15: 297-303, MEDLINE

39 Berard CM, Mahoney CD. Cost-reducing treatment algorithms for antineoplastic drug-induced nausea and vomiting. Am J Health-Syst Pharm 1995; 52: 1879-1885, MEDLINE

40 Ninane J, Ozkaynak MF, Kurtin P, Siegel SE. Variation in the use of ondansetron as an antiemetic drug in children treated with chemotherapy. Med Pediatr Oncol 1995; 25: 33-37, MEDLINE

41 Vermeulen LC, Matuszewski KA, Ratko TA et al. Evaluation of ondansetron prescribing in US academic medical centers. Arch Intern Med 1994; 154: 1733-1740, MEDLINE

Figure 1 Factors influencing drug cost.

Figure 2 Fiscal impact of dosing schedule and route. Daily cost for 20 kg patient for three alternate routes of ondansetron: i.v. preparation, oral solution and oral tablets, each on a once daily or divided schedule assuming a constant daily dose (0.45 mg/kg). The material cost for each category appears in the lower box; the preparation cost contributes to the remainder of the total cost.

Figure 3 Impact of alternate drug routes and schedule on projected total cost of daily ondansetron. Depicts the total cost of modeled strategies as a % of total cost from actual i.v. use, based on individualized dosing schedule. The i.v. (no cap) reflects the change to uniform once-daily dosing, the 10 and 7 day i.v. caps reflect limits in usage to those days respectively based on conditioning regimen. Once daily dosing is assumed for all i.v. models. The p.o. data reflect manufacturer's age-based recommendations (4 mg tablets three times daily for 4-11 years; 8 mg twice daily for 12 years). Since there are no recommendations for children <4 we modeled 0.15 mg/kg three times daily solution. There is no cap on the p.o. data.

Table 1  Patient and treatment characteristics of study sample (n = 36)