| Quality assessment | Summary of findings | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Study | Limitations | Applicability | Population | Intervention | Comparator | Incremental cost (2010 £) | Incremental effects | ICER | Uncertainty |
| Borget, et al., 2009 | Very serious limitations1 | Partially applicable2 | A theoretical cohort of women with breast cancer. The base case is a 45-year-old woman with stage II breast cancer receiving four cycles of chemotherapy with a ≥20% risk of febrile neutropenia (FN). | Primary filgrastim (11-day) | Primary PEG-G-CSF | £1282.783 | <0 QALYs | Dominated | Results were also robust to changes in model inputs. |
| Primary filgrastim (6-day) | Primary PEG-G-CSF | - £506.693 | -0.106 QALYs | £4770.00 per QALY gained3 | |||||
| Danova, et al., 2008 | Very serious limitations4 | Partially applicable5 | A hypothetical cohort of 45-year-old women with stage II breast cancer receiving 4 cycles of chemotherapy associated with a ≥20% risk of FN. | Primary PEG-G-CSF | Primary filgrastim (6-day) | £36.706 | 0.10 QALYs | £349.86 per QALY gained6 | One-way and two-way sensitivity analysis was conducted but range of ICER was not reported. The paper only reported when the highest PEG-G-CSF and the lowest filgrastim price were used, ICER is still below per £43,5226 QALY. |
| Lathia, et al., 2009 | Potentially serious limitations7 | Partially applicable8 | Patients with diffuse large B-cell lymphoma (the most common subtype of non-Hodgkin Lymphoma) receiving induction chemotherapy. Base-case analysis considered a cohort of 64-year-old men and women | Primary filgrastim (did not report if it is 6 or 11 days) | Nothing | £1992.489 | 0.002 QALYs | £0.99 million per QALY gained9 | All one-way sensitivity analysis yielded ICERs of greater than £0.58 million9 per QALY gained. |
| Primary PEG-G-CSF | Nothing | £5765.089 | 0.004 QALYs | £2.52million per QALY gained | |||||
| Liu, et al., 2009 | Very serious limitations10 | Partially applicable11 | Women aged 30-80 years with early stage (I-III) breast cancer | Primary PEG-G-CSF | Primary filgrastim (6-day) | £505.5412 | 0.052 QALYs depends on scenarios | £ 9773.87 per QALY gained12 | When the relative risk of FN was ≤1.3 for 6-day filgrastim versus pegfilgastim, the ICER exceeded |
| Primary filgrastim (11-day) | Primary PEG-G-CSF | £ 1046.6312 | -0.028 QALYs depends on scenarios | Dominated | |||||
| Lyman, 2009 (a) | Very serious limitations 13 | Partially applicable 14 | A hypothetical cohort of patients with intermediate- or high-grade non-Hodgkin lymphoma receiving myelosuppressive chemotherapy (e.g, CHOP-21) with an FN risk of approximately ≥20%. A 65-year-old was chosen as base line. | Primary PEG-G-CSF | Primary filgrastim (6-day) | £192.9615 | Range: 0.042-0.155 QALYs (depends on scenarios) | Range: £1244.61-4594.00 15 per QALY gained (depends on scenarios) | The probability for PEG-G-CSF to become more cost-effective over filgrastim was 50% with the threshold of £11132.4715 per QALY gained, 80% for £22,264.9415 per QALY gained, and 91% for £37,108.2315 per QALY gained. |
| Lyman, 2009 (b) | Very serious limitations 16 | Partially applicable 17 | Women 30-80 years with early stage (I to III) breast cancers who were receiving adjuvant myelosuppressive chemotherapy and had an FN risk of ≥20%. | Primary filgrastim (6-day) | Primary PEG-G-CSF | -£ 1005.6318 | Range: -(0.043-0.094) QALYs depends on scenarios | Range: -£(10698.30-23386.35) 18 per QALY gained | Probabilistic sensitivity analysis show that the probability that strategy A is cost-effective compared with B was 50% for a threshold value of £14,843.2918 per QALY gained, 80% for a threshold value of £22,264.9418 per QALY gained, and 90% for a threshold value of £29,686.5818 per QALY gained. |
| Primary filgrastim (11-day) | Primary PEG-G-CSF | -£ 4899.7718 | -(0.022-0.050) QALYs depends on scenarios | Dominated | |||||
| Ramsey, 2009 | Very serious limitations19 | Partially applicable20 | Women aged 30 to 80 years with early stage (I to III) breast cancer receiving myelosuppressive chemotherapy with an FN risk of approximately 20%. The reference patient was 49 years old with stage II breast cancer receiving six cycles of chemotherapy. | Primary PEG-G-CSF | Secondary PEG-G-CSF | £6459.0621 | 0.076 QALYs | £86091.0921 per QALY gained | One-way: when FN case fatality was less than 2%, the ICER exceeded £148,432.921 per QALY gained. The probability that pegfilgastim primary prophylaxis would be considered cost-effective at the threshold value compared with secondary prophylaxis was 12% for a WTP of £37,108.2321 per QALY gained, 40% of a WTP of £74,216.46 21 per QALY gained, and 75% for a WTP of £148,432.9221 per QALY gained. |
| Timmer-Bonte, et al., 2008 | Potentially serious limitations22 | Partially applicable23 | Patients with small cell lung cancer at risk of FN defined as 60 years of age or older, extensive disease, a Karnofsky performance stats of 40% to 70%, and/or having received prior chemotherapy. Patients have received primary prophylaxis with antibiotics or with antibiotics plus G(M)-CSF. | Secondary antibiotics + G(M)-CSF | Secondary antibiotics | £4970.0324 | 0.02 FN-free cycle | £0.29 million 24 per FN free cycle | Result is robust to probability of FN and treatment cost of FN (although when using higher FN-related costs, the strategies are less distinct in their monetary effects, but still favour antibiotics). |
| Secondary sequential approach (Antibiotics after the first episode of FN and antibiotics plus G(M)-CSF after another episode of FN.) | Secondary antibiotics | £1839.8724 | -0.11 FN-free cycle | Dominated | |||||
| Timmer-Bonte, et al., 2006 | Potentially serious limitations 25 | Partially applicable 26 | Small-cell lung cancer patients receiving standard dose chemotherapy. | Primary antibiotics + G(M)-CSF | Primary antibiotics | First cycle: £611.7827 Entire treatment period: £4609.0427 | First cycle: 14% decrease of the probability of FN Entire treatment period: 23% decrease of the probability of FN | First cycle: £44.9827 per percent decrease of the probability of FN Entire treatment: £329.2827per percent decrease of the probability of FN | Sensitivity analysis has only been conducted for cycle 1. G(M)-CSF is cost saving if the probability of FN is more than 84%, the price of prophylactic G(M)-CSF is less than £421.9527 per patient, or the cost of an episode of FN amount to greater than £10,366.0727. The acceptability for the willingness to pay was approximately 50%. |
| Whyte, et al., 2011 | Very serious limitations28 | Partially applicable29 | The base case consisted of a cohort of 52-year-old female patients diagnosed with stage II breast cancer in line with data on presenting characteristics. | Secondary lenograstim (11 days) | Nothing | £968 | 0.023 QALYs | Dominated | Results are highly sensitive to baseline FN risk. When willingness to pay is £20,000 per QALY, for a patient with a FN risk level of 11% -37%, secondary PEG-G-CSF is most cost-effective; for patients with a higher risk level, primary PEG-G-CSF is the most cost-effective. Using a WTP threshold of £30,000, primary prophylaxis with PEG-G-CSF was cost-effective for baseline FN risks greater than 29%. |
| Secondary lenograstim (6 days) | Nothing | £462 | 0.023 QALYs | Dominated | |||||
| Secondary filgrastim (11 days) | Nothing | £852 | 0.024 QALYs | Dominated | |||||
| Secondary filgrastim (6 days) | Nothing | £397 | 0.024 QALYs | Dominated | |||||
| Secondary PEG-G-CSF | Nothing | If baseline risk =24%: £274 If baseline risk =31%:£253 | If baseline risk =24%: 0.042 QALYs If baseline risk =31%: 0.069 QALYs | If baseline risk =24%: £6,500 per QALY gained If baseline risk =31%: £3,651 per QALY gained | |||||
| Primary lenograstim (11 days) | Nothing | £8326 | 0.075 QALYs | Dominated | |||||
| Primary lenograstim (6 days) | Nothing | £4355 | 0.075 QALYs | Dominated | |||||
| Primary filgrastim (11 days) | Nothing | £7434 | 0.077 QALYs | Dominated | |||||
| Primary filgrastim (6 days) | Nothing | £3865 | 0.077 QALYs | Dominated | |||||
| Primary PEG-G-CSF | Nothing | If baseline risk =24%: £3559 If baseline risk =31%:£3252 | If baseline risk =24%: 0.128 QALYs If baseline risk =31%:0.181 QALYs | If baseline risk =24%: £38,482 per QALY gained If baseline risk =31%: £26,824 per QALY gained | |||||
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF; however it only counts the cost of G(M)-CSF without counting cost of chemotherapy. Not all estimates of input data come from the best available source (systematic review). Have conflicts of interest.
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. This study doesn't look at all interventions of interest. Health effects are not discounted at an annual rate of 3.5%.
Uprated from 2006 British Pounds using inflation factor of 115% (http://eppi
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF; however it only counts the cost of G(M)-CSF without counting cost of chemotherapy. Have conflicts of interest.
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. This study is conducted in Italy, not in the UK. Doesn't look at all interventions of interest.
Converted from 2008 Italian Euros using a PPP exchange rate of 0.78 then uprated by inflation factor of 105% (http://eppi
Only the abstract of this study has been published at the moment, so it is unclear whether all input data of this study come from the best available source.
This study is conducted in Canada, not in the UK. Doesn't look at all interventions of interest.
Converted from 2009 Canadian dollars using a PPP exchange rate of 0.55 then uprated by inflation factor of 106% (http://eppi
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF; however it only counts the cost of G(M)-CSF without counting cost of chemotherapy. No costs were modelled beyond 1 year; while on the other hand, the effectiveness was modelled for lifetime. Have conflicts of interest.
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. This study doesn't look at all interventions of interest.
Uprated from 2006 British Pounds using inflation factor of 115% (http://eppi
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF; however it only counts the cost of G(M)-CSF without counting cost of chemotherapy. Not all estimates of input data come from the best available source (systematic review). Have conflicts of interest.
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. This study is conducted in the U.S.A, not in the UK. Doesn't look at all interventions of interest.
Converted from 2006 U.S.A dollars using a PPP exchange rate of 0.69 then uprated by inflation factor of 108% (http://eppi
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF; however it only counts the cost of G(M)-CSF without counting cost of chemotherapy. Not all estimates of input data come from the best available source (systematic review). Have conflicts of interest.
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. This study is conducted in the U.S.A, not in the UK. Doesn't look at all interventions of interest.
Converted from 2006 U.S.A dollars using a PPP exchange rate of 0.69 then uprated by inflation factor of 108% (http://eppi
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF; however it only counts the cost of G(M)-CSF without counting cost of chemotherapy. Not all estimates of input data come from the best available source (systematic review). Have conflicts of interest.
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. This study is conducted in the U.S.A, not in the UK. Doesn't look at all interventions of interest.
Converted from 2006 U.S.A dollars using a PPP exchange rate of 0.69 then uprated by inflation factor of 108% (http://eppi
Not all estimates of input data come from the best available source (systematic review).
This study is conducted in the Netherlands, not in the UK. Doesn't look at all interventions of interest. The value of health effects is not expressed in terms of quality-adjusted life years (QALYs).
Converted from 2005 Netherlandish Euros using a PPP exchange rate of 0.78 then uprated by inflation factor of 109% (http://eppi
Not all estimates of input data come from the best available source (systematic review).
This study is conducted in the Netherlands, not in the UK. Doesn't look at all interventions of interest. The value of health effects is not expressed in terms of quality-adjusted life years (QALYs).
Converted from 2002 Netherlandish Euros using a PPP exchange rate of 0.78 then uprated by inflation factor of 115% (http://eppi
Whyte et al 2011 modelled three functions of G-CSF: 1). Reducing incidence of febrile neutropenia. 2). Reducing short-term mortality (by preventing febrile neutropenia). 3). Reducing long-term mortality (by maintaining chemotherapy dose). Only the efficacy data for the first function of G-CSF (reducing incidence of febrile neutropenia) was obtained from a systematic review. Efficacy data for the other two functions of G-CSF were estimated based on assumptions, and are in contrast with more direct evidence: Sung et al (2007), Papaldo et al (2005), Pettengell (2008), and Shitara, et al., 2011)
This study is looking at a combined effectiveness of chemotherapy and G(M)-CSF, not just G(M)-CSF. Doesn't look at all interventions of interest.
From: 5, Reducing the risk of septic complications of anticancer treatment
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.