顆粒性白血球生長激素於化學治療引發嗜中性白血球低下之成本效果分析

Abstract

研究背景：顆粒性白血球生長激素 (granulocyte colony-stimulating factor, G-CSF)在預防對癌症患者的嗜中性白血球減少症併發燒 (febrile neutropenia, FN)上之功效已被證實。因此目前臨床治療指引皆建議於FN的高危險群，可預防性使用G-CSF。然而關於預防性使用G-CSF其經濟效益上的影響的研究結果卻多有衝突。而與臨床指引相左，台灣於G-CSF的健保給付規定僅針對血液腫瘤病人給予初級預防性G-CSF之給付。 研究目的：本研究之目的為藉由使用醫院資料庫與全民健保資料庫，分析於健保體制下，於非何杰金氏淋巴癌 (non-Hodgkin’s lymphoma, NHL) 與乳癌病人預防性使用G-CSF之成本效益 (cost-effectiveness)。 研究方法：本研究納入於2001至2010於健保資料庫中新診斷為乳癌或NHL，並於觀察期間開始使用新的化學治療處方者。於醫院資料庫則納入診斷為乳癌或NHL之病人，乳癌病人於2010年，NHL病人則於2010至2011年開始新化學治療處方者。所有病人皆以開始化學治療之第一天當成「進入研究世代日期 (index date)」，而後追蹤至化學治療完成，或「進入研究世代日期」後一年為止。根據使用G-CSF的方式，病人進一步被分為「初級預防」、「次級預防」、「無預防」三組。使用G-CSF的方式、病人有無發生FN和其相關費用皆以化療週期為單位進行資料收集。以廣益估計方程式 (generalized estimating equations, GEE)分析 G-CSF使用方式對FN發生之影響；成本效益分析方面，計算「初級預防」與「次級預防」之增量成本效益，並進行敏感度分析。 研究結果：由健保資料庫之乳癌病患中可看出預防性使用G-CSF於下降嗜中性白血球低下之風險有顯著效果 (初級預防：OR=0.18，95%CI=0.13-0.25；次級預防：OR=0.54，95%CI=0.36-0.81)。於降低FN的風險上則未觀察出這樣的效果。對於NHL的病人而言，在健保資料庫中G-CSF的初級預防顯著的下降發生嗜中性白血球低下與FN的風險 (OR=0.17，95%CI=0.11-0.27)，然而使用G-CSF進行次級預防則沒有相同的效果 (OR=0.79，95%CI=0.56-1.12)。針對醫院病人進行分析時，預防性使用G-CSF亦未顯示出減少嗜中性白血球與FN風險的效果。 使用健保資料庫的數據進行分析時，相較於無預防，初級預防為下降一次FN的發生所需增加的花費在乳癌病人與NHL病人分別為NT$30,732與NT$1,704。相反的，當使用醫院資料庫進行分析時，相較於無預防，初級預防為下降一次FN所需增加的花費於乳癌病人為NT$87,115；而於NHL病人間，初級預防並不合乎成本效益。次級預防在兩種癌症與兩個資料庫裡皆不合乎成本效益。 研究結論：研究中顯示無論是初級抑或次級預防，兩者皆有利於降低嗜中性白球下降之風險；相較於次級預防，初級預防更具成本效益。此外，在台灣健保制度下，初級預防性使用G-CSF在NHL病人上，相較於乳癌病人更具成本效益。

Background: The beneficial effect of granulocyte colony-stimulating factor (G-CSF) prophylaxis on reducing neutropenia and febrile neutropenia (FN) in cancer patients is well-established. Therefore, prophylactic G-CSF is recommended by recent clinical guidelines to manage patients with high risk of FN. However, the economic impact of G-CSF prophylaxis is conflicting between studies. Most studies were based on claim database. The validity of database in detecting neutropenic events still remained uncertain. Furthermore, against to recommendation of guidelines, regulations in Taiwan’s national health insurance (NHI) limited the reimbursement of primary prophylactic G-CSF among patients with hematologic malignancy. Objectives: The objectives of this study were to assess the cost-effectiveness of G-CSF prophylaxis among breast cancer and non-Hodgkin’s lymphoma (NHL) patients under Taiwan’s NHI system by using electronic medical record from one cancer center and national health insurance research database (NHIRD). Methods: Eligible patients were those who newly diagnosed with breast cancer or NHL only, and initiated a new chemotherapy regimen between 2001 and 2010 in NHIRD. For patients in the hospital, those who diagnosed with breast cancer only and initiated a chemotherapy regimen in 2010 were included. Patients diagnosed with NHL only and initiated a new chemotherapy regimen between 2010 and 2011 were also included. All patients were followed until the planned chemotherapy courses were completed or one year after the first date of chemotherapy regimen. Patients were further categorized into three subgroups as “primary prophylaxis”, “secondary prophylaxis” and “no prophylaxis”. The purpose of G-CSF use, occurrence of neutropenia and FN and related costs were evaluated by cycles. Generalized estimating equations (GEEs) were used to examine the impact of G-CSF use on risk of neutropenia and FN. The incremental cost-effectiveness of primary and secondary prophylactic G-CSF were calculated and sensitivity analyses were performed. Results: Significant beneficial effects of prophylactic G-CSF were shown in reducing neutropenia risk among breast cancer patients retrieved from NHIRD (primary prophylaxis: OR=0.18, 95%CI=0.13-0.25; secondary prophylaxis: OR=0.54, 95%CI=0.36-0.81). No such effect were shown in reducing FN risk. For NHL patients, primary prophylactic G-CSF reduced risk of neutropenia and FN significantly in NHIRD (OR=0.17, 95%CI=0.11-0.27) while secondary prophylactic G-CSF did not show the same result (OR=0.79, 95%CI=0.56-1.12). Beneficial effects were not shown when analysis performed in hospital patients. When using data from NHIRD, the incremental cost per FN reduced of primary prophylaxis is NT$30,732 and NT$1,704 among patients with breast cancer and NHL, respectively, compared to no prophylaxis. In contrast, the incremental cost-effectiveness ratio of primary prophylaxis in the hospital is NT$87,115 among patients with breast cancer, and is more costly in NHL patients when compared with no prophylaxis. Secondary prophylaxis is more costly compared with no prophylaxis and primary prophylaxis in both cancer and data sources. Conclusion: We found that both primary prophylactic G-CSF and secondary prophylactic G-CSF have beneficial effect on reducing neutropenia risk, while primary but not secondary prophylactic G-CSF was more cost-effective. In addition, primary prophylactic G-CSF is more cost-effective for NHL patients than for breast cancer patients under Taiwan’s NHI system.