Imputing missing patient-level data and propensity score matching in cost-effectiveness analysis

Author(s):  
Naazish S. Bashir ◽  
Thomas D. Walters ◽  
Anne M. Griffiths ◽  
Shinya Ito ◽  
Wendy J. Ungar
2019 ◽  
Vol 41 (11) ◽  
pp. 2297-2307.e2 ◽  
Author(s):  
Bereket Molla Tigabu ◽  
Majid Davari ◽  
Abbas Kebriaeezadeh ◽  
Mojtaba Mojtahedzadeh ◽  
Kourosh Sadeghi ◽  
...  

2021 ◽  
pp. 112972982199136
Author(s):  
Kairong Wang ◽  
Yingfeng Zhou ◽  
Na Huang ◽  
Zhenqi Lu ◽  
Xiaoju Zhang

Background: Peripherally inserted central catheters (PICCs) and totally implanted vascular access ports (PORTs) have been widely used for medium- to long-term chemotherapy. PICCs are associated with lower insertion cost, but higher complication rates than PORTs. However, there is a paucity of cost-effectiveness comparisons between the devices. We aimed to compare the cost-effectiveness of PICCs and PORTs for medium- to long-term chemotherapy from catheter insertion to removal. Methods: A cost-effectiveness analysis was conducted based on propensity score matching (PSM) from the hospital perspective. Data were collected from a retrospective cohort study. The total cost outcome comprised insertion, maintenance, removal and complication costs. The effectiveness outcome was the complication-free rate. The primary and supplemental outcomes were cost-effectiveness ratios (CERs) and incremental cost-effectiveness ratios (ICERs). Results: A total of 1050 patients (after PSM for 417 patients) were included. The average total cost for 3–6 month ($603.55 ± 78.68 vs $1270.21 ± 128.84), 6–9 month ($731.40 ± 42.97 vs $1414.48 ± 155.43), and 9–12 month ($966.83 ± 53.78 vs $1587.76 ± 160.56) dwell times were all significantly lower for PICCs than for PORTs ( p < 0.001). PICCs were associated with significantly lower complication-free rates than PORTs during the 3–6 month (65.22% vs 90.58%, p < 0.001), 6–9 month (53.33% vs 91.80%, p < 0.001), and 9–12 month (44.44% vs 88.46%, p = 0.015) dwell times. Ultimately, PICCs were associated with lower CERs than PORTs for the 3–6 month (928.54 vs 1395.84) and 6–9 month (1380.00 vs 1537.48) but higher CER for the 9–12 month (2197.34 vs 1804.27) dwell times. ICERs were 2564.08 and 1751.49 with dwell times of 3–6 months and 6–9 months, respectively. Conclusion: This study provided economic evidence that informs vascular access device choice for medium- to long-term chemotherapy. For 3–9 month dwell times, PICCs were more cost-effective than PORTs. Furthermore, ICERs were applied and the choice was determined by willingness-to-pay. For 9–12 month dwell times, PORTs might be more cost-effective than PICCs, and studies with larger sample size would be needed to verify this finding in the future.


2007 ◽  
Vol 25 (18_suppl) ◽  
pp. 1081-1081
Author(s):  
S. E. Jones ◽  
A. Benedict ◽  
D. Cameron ◽  
S. Jourdan

1081 Background: Docetaxel was shown to be superior to paclitaxel in OS and TTP (median OS: 1.28 vs 1.06 yr, HR=1.41; median TTP: 0.47 vs 0.30 yr, HR= 1.64) for the treatment of patients with metastatic breast cancer progressing after an anthracycline-based regimen (Jones et al. J Clin Oncol. 2005;23:5542). A cost-effectiveness analysis based on this head-to-head comparison was performed considering clinical effectiveness, quality-adjusted life-years, and direct medical costs in the UK. Methods: A probabilistic Markov model was developed to examine results over a 10-yr long time-horizon. Patient level data were available on PFS and OS, treatment cycles and doses, number of cycles affected by adverse events, G-CSF use and post-failure treatment. Generalized gamma regression, fitting patient-level data best were used to model baseline PFS and OS in the paclitaxel arm. HRs adjusting for all covariates were applied to the baseline hazards to generate the docetaxel arm. Resource use and utility values for health states were obtained from published literature and practicing UK oncologists. Unit costs came from 2005 NHS reference costs; drug costs from the British National Formulary 2006 without hospital discount. Costs and benefits were discounted at 3.5%. A Monte-Carlo simulation and extensive 1-way sensitivity analyses were conducted. Results: Docetaxel is more costly (£13,500 vs £10,600), but yields higher health benefits than paclitaxel (2.01 vs 1.48 LYs and 1.18 vs 0.85 QALYs for docetaxel and paclitaxel, respectively) over a 10-yr time horizon. The discounted Incremental Cost-Effectivness Ratio (ICER) of docetaxel vs paclitaxel 3-weekly was estimated to be £5,532/LY gained (95% CI 2,250–12,700) and £8,741/QALY gained (95% CI 3,400–17,300). The ICER was most sensitive to the HR for PFS, OS and the cost of docetaxel and paclitaxel. However, ICERs remained below £20,000/QALY at extreme values of the parameters. Conclusions: Compared with paclitaxel 3-weekly, over a 10-yr time horizon docetaxel provides survival and quality-adjusted survival benefit to metastatic breast cancer patients failing anthracycline regimens at an acceptable cost in a UK setting. Docetaxel is cost-effective compared to paclitaxel. No significant financial relationships to disclose.


2012 ◽  
Vol 15 (7) ◽  
pp. A469-A470
Author(s):  
P. McEwan ◽  
V. Foos ◽  
A. Lloyd ◽  
J.L. Palmer ◽  
M. Lamotte ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document