社區長者預防跌倒信念與行動可能性預測模式驗證之研究

目的：本研究運用健康信念模式來探討社區長者之預防跌倒信念，並進一步發展適用於國內社區長者預防跌倒行動可能性之預測模式，以可靠且有效地測量臺灣社區長者採取預防跌倒行動之可能性。方法：本研究共分三階段進行，第一階段採用焦點團體訪談，以健康信念模式為架構設計訪談大綱，訪談逐字稿以直接內容分析法進行分析；第二階段採用Q方法，Q分類排序所得之資料以主成份分析（Principle Components Analysis）及最大變異轉軸法（Varimax Rotation）進行因素抽取及類型區分；第三階段採用量性問卷調查，以焦點團體訪談資料為基礎及參考相關文獻，建構量化測量工具及進行資料收集，再以結構方程模式（Structure Equation Modeling, SEM）進行預測模式驗證分析。結果：第一階段研究共辦理六場焦點團體訪談，每場各招募8位社區長者參與，逐字稿內容共整理出54個濃縮意義單位，25個編碼，上述編碼再依據健康信念模式之六個類別進行歸類結果如下：共有紀大反應差、慢性疾病頭暈、生理功能退化、情緒心情不佳及環境不安全等五個編碼歸類於自覺罹患性；共有身體不舒服、身體受傷、心理害怕及生活無法自理等四個編碼歸類於自覺嚴重性；共有個人身心健康、減少家人負擔及周遭人快樂等三個編碼歸類於自覺行動利益；共有個人身心因素、外在人為因素、生活環境因素及缺乏知識技能等四個編碼歸類於自覺行動障礙；共有自我提醒、個人經驗、他人經驗、知識分享及衛教宣導等五個編碼歸類於行動線索；共有經常提醒自己、成為生活習慣、使用輔具及注意安全等四個編碼歸類於自我效能。第二階段研究共招募42位社區長者參與Q-分類排序，根據Q分類結果之分析共歸納出四種因素類型社區長者，各因素命名如下：因素類型Ⅰ Considerate perspective、因素類型Ⅱ Promising perspective、因素類型Ⅲ Adaptable perspective、因素類型Ⅳ Ignorant perspective。第三階段研究共收回704份有效問卷，根據SEM分析整體模式適配性良好，路徑分析顯示自覺嚴重性、自覺行動利益、行動線索及自我效能對預防跌倒行動可能性均具有正向之顯著影響力，對整體模式的預測力為39.0%；自我效能是直接影響社區長者採取預防跌倒行動可能性之最強因子，同時在自覺行動利益與行動可能性之間具有部份中介作用，在自覺行動障礙與行動可能性之間則具有完全中介作用。結論：本研究同時結合理論模式與被介入者（社區長者）的觀點來探討老人預防跌倒信念，並以Q方法成功的將社區長者區分為四種不同信念類型，也發展出適合於社區長者預防跌倒行動可能性之測量工具，可提供未來在設計預防跌倒介入方案之應用，使方案內容更能符合社區長者的心理與實際需求，以促進社區長者對預防跌倒活動的參與率與遵從度。

關鍵字

社區長者；預防跌倒信念；行動可能性；模式驗證

並列摘要

Purpose: The aim of the present study was to explore fall-prevention beliefs in community-dwelling older adults with the application of health belief model (HBM). The prediction model was further developed for domestic community-dwelling older adults in order to reliably and effectively measures the likelihood of taking fall-prevention actions for community-dwelling older adults in Taiwan. Method: The study was conducted in three phases. In the first phase, focus group interviews were used. The interview outlines were constructed on HBM. The interview transcripts were analyzed using direct content analysis. In the second phase, Q method was adopted to investigate the pattern of perception on fall-prevention beliefs. Principal component analysis with Varimax rotation was conducted on the data retrieved from Q-sorting to extract relevant perspectives. In the third phase, quantitative questionnaire survey was adopted. The questionnaire was constructed on the results from focus group interviews and reference literature, and the collected data were used to verify the prediction model with the application of structure equation modeling (SEM). Results: Six focus group sessions were held in the first phase, with 8 community-dwelling older adults recruited in each. Transcripts were classified into 54 condensed meaning units and 25 codes. Referring to six dimensions in HBM, codes were distributed as follows. Perceived susceptibility consisted of five codes which were slow response due to aging, dizziness caused by chronic diseases, physical degeneration, poor emotion condition, and unsafe environment condition; perceived severity consisted of four codes which were physical wellness, physical injuries, mental fear, and inability of taking care of themselves; perceived benefits consisted of three codes which were personal physical wellness, burden relief for family, and happiness creation to surroundings; perceived barriers consisted of four codes which were personal physical and metal factors, external interpersonal factors, living environmental factors, and insufficiency of knowledge and skills; cues to action consisted of five codes which were self reminder, personal experiences, other’s experiences, knowledge sharing and promotion of health education; self-efficacy consisted of four codes which were frequent self-reminding, formation of habit, usage of assistive devices, and awareness of safety. In the second phase, Q-sorting was performed by the 42 recruited community-dwelling older adults. Four statistically independent perspectives were derived from the analysis and reflected distinct viewpoints on beliefs related to fall prevention. The correspondent names were: Considerate perspective (factor I), Promising perspective (factor II), Adaptable perspective (factor III), and Ignorant perspective (factor IV). In the third phase, 704 valid samples were collected. The new HBM demonstrated a good fit among the prediction model verified by SEM. Path analysis indicated that perceived severity, perceived benefits, clue to action and self-efficacy have significant positive effect on the likelihood of fall-prevention action, and the prediction model explained 39.0% of the variance in the likelihood of action. Self-efficacy was found to be the most effective factor which directly affected the likelihood of taking fall-prevention action in community-dwelling older adults. It had partial mediation effect on perceived benefits and the likelihood of action, and had complete mediation effect on perceived barriers and the likelihood of action. Conclusion: Both the theoretical concepts and the perspectives from the intervened (community-dwelling older adults) were incorporated to discover fall-prevention belief in elderly. Community-dwelling older adults were successfully classified into four distinct perspectives by Q method. A feasible measurement was developed for predicting the likelihood of fall-prevention action in community-dwelling older adults. It is recommended that future design of fall-prevention intervention programs adopt the developed measurement. Future interventions are expected to satisfy the mental and practical needs from community-dwelling older adults to improve their participation in and adherence to program activities.