探討臺灣乳癌患者存活期照護之型態與結果

Abstract

研究背景：存活者照護(survivorship care)階段可謂為癌症照護連續體的典範轉移。然而，為乳癌存活者提供適切照護對於健康照護體系而言仍是挑戰。照護連續性(continuity of care)與照護協調性(coordination of care)可促進照護的公平與效率，進而提升照護品質與健康結果。本研究使用具全國代表性的長期資料，以照護連續性與照護協調性量化指標分析乳癌存活者照護現況。臺灣目前對於存活者後續追蹤與照護尚無明確的政策規範或臨床指引，本研究結果無論在政策制定規劃資源配置或專業照護團隊進行醫療決策時皆可做為重要參考，對於建構與發展癌症存活者的理想照護模式亦有所助益。 研究方法：本研究首先檢視病患完成主要癌症治療、進入存活者照護階段時，照護過程裡的場域轉換是否有間斷性照護(care fragmentation)的問題。接著分別以醫師和醫療院所為測量基礎，分析追蹤期間的癌症專科門診照護連續性，以及一般醫療的門診照護連續性。另一方面，本研究亦評估乳癌存活者是否有照護協調性風險的可能，並且以照護密度(care density)測量追蹤期間的照護協調性。研究目的為(1)瞭解乳癌存活者照護連續性與照護協調性的測量結果，採用趨勢卡方檢定(Cochran–Armitage trend test)與多項式迴歸(Polynomial regression analysis)分析歷年趨勢；(2)使用廣義估計模型(Generalized estimating equations, GEEs)檢視乳癌存活者照護連續性與照護協調性的相關因素；(3)以傾向分數配對(Propensity score matching, PSM)以及廣義估計模型(GEEs)，探究乳癌存活者照護連續性與照護協調性對於追蹤服務、健康結果與醫療費用的影響。資料來源係使用癌症登記檔長表串聯全民健康保險資料庫與死亡登記檔，選取2002至2007年間確診為I-III期新發乳癌的成年女性，主要觀察期間為癌症診斷後第二年至診斷後第五年。 研究結果：本研究發現，(1)在照護連續性方面，間斷性照護(i.e.完成治療後的主要照護地點與先前治療醫院不同)的比例隨著追蹤期間上升，但隨著診斷年度增加而下降。以醫院為分析單位時，照護連續性指標數值(continuity of care indicators, COCI)在不同診斷年度或追蹤期間未有顯著趨勢。以醫師為分析單位時，COCI在不同診斷年度之間為持平狀態；癌症專科醫師COCI隨著追蹤期間而下降，一般醫療醫師COCI則隨著追蹤期間而上升。在照護協調性方面，病患有照護協調性風險的比例隨著診斷年度增加而上升；照護密度隨著追蹤期間下降，但隨著診斷年度增加而提升。(2)病患特性的年齡、期別、賀爾蒙受器狀態、手術術式、共病症指數分數、診斷前是否曾住院、追蹤期間門診次數、月投保金額、職業類型，地區特性的都市化程度與居住地每平方公里醫師人數，醫療提供者特性的醫師年齡、醫師乳癌照護服務量，以及醫院評鑑層級、權屬別、醫院乳癌照護服務量等變項為照護連續性與照護協調性相關因素。(3)有間斷性照護的乳癌存活者使用乳房超音波的可能性較低(OR=0.88, 95%CI: 0.81-0.96)，出現住院、可避免急診等負向健康結果的可能性較高(OR=1.13, 95%CI: 1.05-1.22; OR=1.55, 95%CI: 1.16-2.07)。以醫師為分析單位時，癌症專科COCI高者較可能使用乳房X光攝影、乳房超音波(OR=1.26, 95%CI: 1.19-1.32; OR=1.12, 95%CI: 1.06-1.18)，發生住院、急診或有延遲性症狀的可能性較低(OR=0.78, 95%CI: 0.71-0.85; OR=0.88, 95%CI: 0.82-0.95; OR=0.84, 95%CI: 0.77-0.91)。一般醫療COCI高者，發生住院、可避免住院、急診或有延遲性症狀的可能性較低(OR=0.77, 95%CI: 0.70-0.85; OR=0.77, 95%CI: 0.61-0.97; OR=0.75, 95%CI: 0.68-0.82; OR=0.91, 95%CI: 0.84-0.98)。以醫院為分析單位時，研究結果亦有一致的方向。無論是癌症專科或一般醫療，COCI較高的乳癌存活者，總醫療費用較低(p<.0001)。(4)有照護協調性風險的乳癌存活者使用乳房X光攝影、乳房超音波的可能性較高(OR=1.25, 95%CI: 1.19-1.31; OR=1.33, 95%CI: 1.25-1.40)，而發生住院、可避免住院、急診、可避免急診或延遲性症狀的可能性也較高(OR=2.16, 95%CI: 1.95-2.38; OR=2.33, 95%CI: 1.69-3.23; OR=2.78, 95%CI: 2.53-3.05; OR=3.91, 95%CI: 2.77-5.51; OR=1.51, 95%CI: 1.42-1.59)，總醫療費用也較高(p<.0001)。全部門診照護密度高的乳癌存活者，使用乳房X光攝影、乳房超音波的可能性較高(OR=1.62, 95%CI: 1.55-1.69; OR=1.53, 95%CI: 1.46-1.60)，發生住院、可避免住院、急診、可避免急診的可能性較低(OR=0.78, 95%CI: 0.74-0.81; OR=0.76, 95%CI: 0.71-0.81; OR=0.88, 95%CI: 0.82-0.94; OR=0.81, 95%CI: 0.76-0.87)，總醫療費用也較低(p=0.01)。僅將癌症專科與一般醫療醫師門診納入照護密度的分析時，亦有相同方向的結果。 結論：照護連續性與照護協調性對於乳癌存活者的監測性照護利用、健康結果與醫療費用皆有顯著影響。政策制定者與健康照護提供者需重視並提升追蹤期間的照護連續性與照護協調性，以期提供適切的存活者照護。

Background: The survivorship care phase is a paradigm shift in the cancer care continuum. However, it remains a challenge for the health care systems to provide appropriate care for breast cancer survivors. The continuity and coordination of care promote fairness and efficiency, which in turn consolidates the care quality and health outcomes. In this study, nationally representative long-term data were utilized to analyze the current care situation of breast cancer survivors with indicators such as care continuity and care coordination. So far, there are no specific policy criteria or clinical guidelines for follow-up visits and care of survivors in Taiwan. This study will not only serve as an important reference for policy formulation, planning, resource allocation, and medical decision-making by the professional care teams, but also help develop an ideal care model for cancer survivors. Methods: This study first investigated whether there was care fragmentation during the transition from the primary cancer treatment to survivorship care. Subsequently, the continuity of outpatient care in oncology and the primary care physician (PCP) was analyzed by physician or institution based measures, respectively. On the other hand, this study evaluated the possibility that breast cancer survivors were at risk for poor care coordination, and we measured care coordination during follow-up using quantitative indicators (i.e. care density). The research objectives were as follows: (1) Explore the value of the care continuity and care coordination for breast cancer survivors and analyze the annual trend by the Cochran–Armitage trend test and polynomial regression analysis; (2) Examine the factors related to the continuity and coordination of care in breast cancer survivors with the generalized estimating equations (GEEs); (3) Evaluate the impact of care continuity and care coordination for breast cancer survivors on the follow-up care, health outcomes, and medical costs using models based on propensity score matching and GEEs. The data were retrieved from the National Health Insurance Research Database and the death registry databases. The adult women who were newly diagnosed with stage I–III breast cancer between 2002 and 2007 were selected. The main observation period was from the second to the fifth year after the cancer diagnosis. Results: This study found that (1) In terms of care continuity, the proportion of care fragmentation (i.e., the location of primary care after the completion of treatment is different from the hospital providing previous treatments) rose in the follow-up period but declined with the increase of diagnostic year. With institution-based measures, no significant trends in the value of continuity of care indicators (COCI) in different diagnostic years or follow-up periods were discovered. With physician-based measures, COCI remained stable among different diagnostic years. However, oncology COCI decreased during the follow-up period, while PCP COCI increased during the follow-up period. Regarding care coordination, the proportion of patients at risk of poor care coordination increased with the increase of diagnosis year; care density declined during the follow-up period but increased with the increase of diagnostic years. (2) The following variables can be used as the related factors responsible for care continuity and coordination: age, stage, hormone receptor status, surgical procedure, comorbidity index score, hospitalization before diagnosis, the number of outpatient visits during the follow-up period, monthly insurance premium, occupation type, the degree of regional urbanization, the number of physicians per square kilometer residence, the age of medical providers, the volume of breast cancer care service of the physician, and the accreditation level/ownership/the volume of breast cancer care services of the hospital. (3) Breast cancer survivors with care fragmentation were less likely to use breast ultrasound (OR=0.88, 95% CI: 0.81–0.96), and more likely to have negative health outcomes such as hospitalization and avoidable emergency department visits (OR=1.13, 95% CI: 1.05–1.22; OR=1.55, 95% CI: 1.16–2.07). With physician-based measures, patients with high oncology COCI had a higher tendency to use mammography and breast ultrasound (OR=1.26, 95% CI: 1.19–1.32; OR=1.12, 95% CI: 1.06–1.18; respectively), and a lower tendency for hospitalization, emergency department visits, and delayed symptoms (OR=0.78, 95% CI: 0.71–0.85; OR=0.88, 95% CI: 0.82–0.95; OR=0.84, 95% CI: 0.77–0.91; respectively). Patients with high PCP COCI were less likely to experience hospitalization, avoidable hospitalization, emergency department visits, and delayed symptoms (OR=0.77, 95% CI: 0.70–0.85; OR=0.77, 95% CI: 0.61–0.97; OR=0.75, 95%CI: 0.68–0.82; OR=0.91, 95% CI: 0.84–0.98; respectively). Similar results are observed with physician-based measures. In addition, the breast cancer survivors spent lower total medical expenses (p<0.0001) when they received a high continuity of care either in oncology or in PCP. (4) Breast cancer survivors at risk for poor care coordination were more likely to receive mammography and breast ultrasound (OR=1.25, 95% CI: 1.19–1.31; OR=1.33, 95% CI: 1.95–1.40; respectively), and also more likely to have negative health outcomes such as hospitalization, avoidable hospitalization, emergency department visits, avoidable emergency department visits, and delayed symptoms (OR=2.16, 95% CI: 1.95–2.38; OR=2.33, 95% CI: 1.69–3.23; OR=2.78, 95% CI: 2.53-3.05; OR=3.91, 95% CI: 2.77–5.51; OR=1.51, 95% CI: 1.42–1.59; respectively). The total medical expenses for them were also higher (p<0.0001). The breast cancer survivors with a high care density in overall outpatient settings were more likely to receive mammography and breast ultrasound (OR=1.62, 95% CI: 1.55–1.69; OR=1.53, 95% CI: 1.46–1.60; respectively), but less likely to experience hospitalization, avoidable hospitalization, emergency department visits, and avoidable emergency department visits (OR=0.78, 95% CI: 0.74–0.81; OR=0.76, 95% CI: 0.71–0.81; OR=0.88, 95% CI: 1.88; OR=0.81, 95% CI: 0.76–0.87; respectively). The overall medical expenses were also lower (p=0.01). The results were consistent with those based on analysis only including oncology and PCP outpatient visits. Conclusion: Care continuity and care coordination have a significant impact on the surveillance care utilization, health outcomes, and medical costs for breast cancer survivors. Policymakers and health care providers need to address and improve care continuity and care coordination during the follow-up period to provide appropriate care for survivors.