衰弱及多重用藥對於臺灣老年族群之臨床不良影響

Abstract

研究背景：近來在老年醫學領域，學者們逐漸開始探討衰弱與多重用藥之複合效應對臨床不良事件發生風險的影響。然而綜觀過去文獻，不僅追蹤時間較短，而且並未考量衰弱與多重用藥隨時間變動的特質，可能對分析結果造成影響。此外，過去文獻受限於前瞻性收案，導致研究樣本數少，無法進一步依年齡分層探討複合效應的影響。 研究目的：分析不同年齡層之老年族群，追蹤其十年間的衰弱與多重用藥之變化趨勢，再合併兩者以建立複合效應(combined effect)，並在考量複合效應隨著時間變動的影響下，分析其與不良事件的相關性。 研究方法：本研究為一回溯性世代研究，利用台灣全民健康保險資料庫，篩選於2007年1月1日年齡介於65至100歲的門診患者，並按照年齡分成65-74歲、75-84歲和85歲以上等三組。首先會在2006年收集研究對象之衰弱和多重用藥狀態作為背景資料，隨後自2007年起以季為單位，記錄研究對象在2007至2015年間每季衰弱、多重用藥和不良事件狀態，總共追蹤36季。衰弱乃根據研究對象之多重共病衰弱指數(multimorbidity frailty index, mFI)定義，分為健康、輕度衰弱、中度衰弱與重度衰弱共4組；多重用藥定義為使用5-9個藥品，而過度多重用藥則定義為使用藥品數超過10個以上；隨後根據研究對象之衰弱與多重用藥狀態建立複合效應，總共12組；本研究所分析之不良事件包含全死因死亡事件、全部住院事件和未計畫性住院，並以廣義估計方程式(generalized estimating equations, GEEs)估計複合效應與不良事件的相關性。 研究結果：本研究納入100,000位研究對象，研究族群之平均年齡為74.1歲，其中65-74歲者占56.8%，75-84歲者占34.9%，而85歲以上者則占8.3%。在各組複合效應中，以健康/無多重用藥者人數最多，共計61,008 位，占研究族群之61.0%，其次為健康/多重用藥者，共16,155位，占16.2%，重度衰弱/過度多重用藥組人數則最少，共359位，僅占0.4%。 在衰弱方面，研究族群在追蹤起點時的mFI平均值為0.046，其中輕度衰弱之盛行率為14.57 %、中度衰弱為4.48 %、而重度衰弱則為1.99 %。比較不同年齡層之衰弱長期變化趨勢，65-74歲和75-84歲族群之mFI在追蹤期間明顯上升，反觀85歲以上者之mFI則較無明顯變化。 而在多重用藥方面，研究族群在追蹤起點時平均使用3.2個藥品，多重用藥的盛行率為26.08 %；而過度多重用藥盛行率則為5.46%。在長期追蹤下，不同年齡層的使用藥品數變化趨勢有所不同，65-74歲和75-84歲族群呈現相似的上升趨勢，而85歲以上者用藥數的增加幅度相對較小。 衰弱與多重用藥組成之各組複合效應皆與死亡顯著相關。若比較衰弱與多重用藥對死亡風險的相互影響，可發現在相同多重用藥程度下，衰弱與死亡風險正相關，例如：在多重用藥者中，健康者的死亡風險1.59 (95% CI: 1.53-1.64)，當多重用藥者愈衰弱，其死亡風險愈高，最高者為重度衰弱/多重用藥者，風險為7.18 (95% CI: 6.84-7.54)。另一方面，多重用藥對死亡風險的影響在不同衰弱程度中有所差異。在健康與輕度衰弱者中，使用較多藥品者有較高的死亡風險，例如：在健康者中，多重用藥與過度多重用藥者的死亡風險分別為1.59 (95% CI: 1.53-1.64)和2.65 (95% CI: 2.47-2.85)；而在中度與重度衰弱者中，多重用藥程度與死亡風險則為負向相關。 而在全部住院事件方面，各複合效應皆和住院風險顯著相關。而在衰弱與多重用藥對住院風險的相互影響方面，各複合效應之間的風險分布趨勢與死亡大抵相似，不過在中度與重度衰弱者有所不同，多重用藥對住院風險的影響較不明顯。此外，未計畫性住院與各複合效應顯著相關，且整體風險分布情形與全部住院事件相似。 研究結論：在不同年齡層的老年族群中，衰弱與多重用藥之長期變化趨勢有所差異，不宜將各年齡層概括討論。而在不良事件發生風險方面，衰弱與多重用藥會互相影響彼此的不良事件發生風險，因此在照護高齡者時應密切監測其衰弱與多重用藥情形，並力求處方簡化，以期提升整體照護品質。

Background: The impact of combined effect of frailty and polypharmacy on the adverse outcomes in the elderly has raised significant concerns. However, existing studies were limited to short follow-up period and one-time measurement of frailty and polypharmacy. In addition, little is known whether these combined status varied or resulted in different consequences among different age groups in the elderly. Objectives: We aimed to measure the longitudinal trends of frailty and polypharmacy over a 10-year follow-up period, and to further investigate the association between different status of frailty and polypharmacy on risks of adverse outcomes (mortality, all-cause hospitalizations and unplanned hospitalizations). Furthermore, we tested whether the impacts of these combined status varied across different age groups. Methods: People who aged between 65 to 100 years old in 2007 were identified from the National Health Insurance Reimbursement Database (NHIRD), and were further divided into 3 age groups (aged 65-74, 75-84 and 85+). The status of frailty and polypharmacy was collected in 2006 as baseline, and then quarterly measured from 2007 to 2015. Frailty was categorized into fit, mild frailty, moderate frailty and severe frailty according to the multimorbidity frailty index. The chronic use of 5 to 9 medications was considered as polypharmacy, and 10 medications or more as excessive polypharmacy. Subjects were further classified into 12 groups taking into account both their status of frailty and polypharmacy. The adverse outcomes of interest were all-cause mortality, all-cause hospitalization and unplanned hospitalization. Generalized estimating equation model was used to examine the association between these combined effects and risk of adverse outcomes. Results: We identified 100,000 older adults as our study subject. Mean age of our subjects were 74.1 years (aged 65-74: 56.8%; aged 75-84: 34.9%; aged 85+: 8.3%). Most of the subjects were categorize as fit / non-polypharmacy (61.01%), followed by fit / polypharmacy (16.16%), and the least of the subjects were categorized as severe frailty/excessive polypharmacy (0.4%). The mean value of mFI was 0.046 at baseline. Among the study subjects, 14.57% were mild frailty, 4.48% were mild frailty and 1.99% were severe frailty. The longitudinal trends of mFI varied across different age groups. Those who aged 65-74 and 75-84 had higher increase in mFI while those who aged more than 85 didn’t show significant change of mFI during the follow-up period. The mean number of medications used was 3.2 at baseline. Among the study subjects, the prevalence of polypharmacy and excessive polypharmacy were 26.08% and 5.46%, respectively. During the follow-up period, higher increases in drug use was observed in elderly who aged 65-74 and 75-84. Compared to fit / non-polypharmacy group, all other combined effect of frailty and polypharmacy were associated with the risk of mortality. Noteworthy, frailty and polypharmacy would both modify the risk of mortality. In the subjects with polypharmacy, the risk of death increased as subjects became frailer. For example, within the polypharmacy group, the lowest aOR of mortality were 1.59 (95% CI: 1.53-1.64) for the fit group, and the highest were 7.18 (95% CI: 6.84-7.54) for the severe frailty group. Besides, the risk of mortality varied among different polypharmacy status within each frailty groups. In fit and mild frailty groups, the risk of mortality increased as subjects used more medications. For example, within the fit frailty group, the aORs of mortality were 1.59 (95% CI: 1.53-1.64) and 2.65 (95% CI: 2.47-2.85) for the polypharmacy and excessive polypharmacy groups. This phenomenon is the opposite within moderate and severe frailty groups. When it comes to the risk of all-cause hospitalization, the overall risk pattern of different combined effects was similar to that of the all-cause mortality, except for the impact of polypharmacy within moderate and severe frailty groups. In subjects categorized as moderate and severe frailty, polypharmacy didn’t show obvious impact on the risk of hospital admission. Furthermore, the overall risk pattern of unplanned hospitalization was similar to that of the all-cause hospitalization. Conclusions: The longitudinal trends of frailty and polypharmacy continued to increase over the last 10 years, and varied across different age groups in the elderly. Furthermore, by demonstrating that frailty and polypharmacy would both modify the risk of adverse outcome in the elderly, this study indicates that these factors should be monitored to optimize the quality of care of older people.