Economic Models of Prophylactic CSF 721 Cost Minimization Models

The initial economic models of prophylactic CSF administration were based on the results of previously discussed pivotal RCTs and crude estimates of direct medical costs. These cost minimization models using baseline probabilities and resource utilization from the RCT of Crawford et al. (65) and local direct institutional cost information generated average risk threshold estimates for cost savings with prophylactic rHuG-CSF use in the range of 40% (19,20). These models assumed that all patients experiencing FN will be hospitalized for empiric parenteral antibiotic therapy (Fig. 16). No impact of CSF on the duration of hospitalization or infection-related mortality was assumed. The expected cost of a specific choice was calculated from the sum of the products of the terminal value or cost and the probability of each branch. The total cost of hospitaliza-

Fig. 16. Decision tree for a cost minimization model of the use of prophylactic granulocyte colony-stimulating factor (rHuG-CSF) or no rHuG-CSF in chemotherapy patients at risk for FN. Model assumes that all patients with FN are hospitalized, that the risk of hospitalization of FN in those receiving prophylactic G-CSF is reduced by 50%, and that there is an equal duration of hospitaliza-tion and mortality among those hospitalized for FN in both groups. Cost considerations are limited to those related to hospitalization for FN (average length of stay X average cost/day [$]) and G-CSF (average duration of administration X average cost/day for drug and administration). Patients not hospitalized and not receiving G-CSF have no associated costs. (Data from ref. 19.)

Fig. 16. Decision tree for a cost minimization model of the use of prophylactic granulocyte colony-stimulating factor (rHuG-CSF) or no rHuG-CSF in chemotherapy patients at risk for FN. Model assumes that all patients with FN are hospitalized, that the risk of hospitalization of FN in those receiving prophylactic G-CSF is reduced by 50%, and that there is an equal duration of hospitaliza-tion and mortality among those hospitalized for FN in both groups. Cost considerations are limited to those related to hospitalization for FN (average length of stay X average cost/day [$]) and G-CSF (average duration of administration X average cost/day for drug and administration). Patients not hospitalized and not receiving G-CSF have no associated costs. (Data from ref. 19.)

tion was the product of the average daily cost of hospitalization and the average hospital LOS. The expected cost per treatment cycle represents the marginal or net cost associated with hospitalization for FN and treatment with CSF. The expected cost for each management strategy was calculated as the sum of the products of the costs and probabilities of each outcome.

7.2.2. Sensitivity Analyses

A series of sensitivity analyses were undertaken estimating the expected cost associated with each strategy while varying the assumptions concerning each variable over the range of reasonable values. In a one-way sensitivity analysis, the expected cost was calculated for each value of the variable, forming a curve of cost values for each decision choice. Thresholds were generated for each variable at which the expected costs were equal for management with and without CSFs. The relationship between the threshold risk of hospitalization for FN while varying the cost associated with hospitalization was studied in a two-way sensitivity analysis. A family of threshold curves was generated on the basis of multivariate sensitivity analyses for a combination of two or more variables.

Under baseline assumptions, a threshold risk for FN of 40% was estimated where the added cost of CSF was offset by the reduction in cost associated with hospitalization for FN (Fig. 17). As the daily cost increases, the threshold risk of hospitalization favoring the use of CSF decreases. Likewise, the longer the expected LOS, the lower the threshold favoring the use of CSF. Above this risk threshold, the overall costs of treatment are less when a CSF is used, whereas below it the use of a CSF actually

Risk of Hospitalization for FN

Favors No CSF

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Cost of Hospitalization I Day ($)

Fig. 17. Two-way sensitivity analysis of the threshold for cost associated with FN hospitalization and granulocyte colony-stimulating factor (rHuG-CSF) use based on the decision model shown in Fig. 16. The horizontal axis varies the cost/day ($) for hospitalization, and the vertical axis varies the risk of FN hospitalization. The threshold curve demonstrates lower thresholds for cost saving use of rHuG-CSF with increasing cost of hospitalization. Combinations of risk and cost above the threshold curve are associated with a reduction in cost with the use of prophylactic G-CSF, whereas those below the threshold curve are associated with greater cost with the use of rHuG-CSF. Illustrated are three sequential cost estimates and the accompanying risk threshold based on the original study using crude direct medical costs, * (Threshold = 40% [19]), the updated total hospital cost estimate, + (Threshold = 23% [16]), and the recent addition of indirect and out-of-pocket cost estimates, # (Threshold = 18% [81]).

increases the costs of treatment, although the added cost of the CSFs is partially offset by the reduction in cost of hospitalization. As the daily cost of hospitalization or the LOS increase, the FN risk threshold favoring the use of these agents decreases. The American Society of Clinical Oncology (ASCO) used the early results to support its clinical practice guidelines for the use of CSFs. The original risk threshold estimates were subsequently updated using the single-institution estimates of total costs for FN hospitalization described above (16). Incorporation of total hospital costs into the new FN cost estimates and their use in the previously reported cost-minimization models, generated FN risk threshold estimates for CSF use of 20-25% (Fig. 17).

7.2.3. Model Extensions

Additional efforts to refine risk threshold estimates further for the prophylactic use of the CSF have included the addition of indirect and out-of-pocket cost estimates, a study of the impact of incorporating an option to treat selected low-risk patients with FN in the ambulatory setting, consideration of special high-risk populations such as the elderly, and the use of predictive models permitting individualization of risk estimation and the study of other practice changes that might influence threshold estimates.

Was this article helpful?

0 0
10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment