探討臺灣地區人類乳突狀病毒疫苗施打之推廣方式－－利用擴散模型分析

An Exploration of Ways to Promote Inoculation with Human Papilloma Virus Vaccine in Taiwan

目的：研究分析哪些推廣因素能有效促進台灣市場上，人類乳突狀病毒疫苗接種數量之提升。希望能推動更多人實際接種人類乳突狀病毒疫苗，嘉惠於個人與整個社會，達到疫苗開發之目的。方法：本研究分別採用三種不同的擴散模型，來分析人類乳突狀病毒疫苗的接種資料。資料來源為IMS資料庫所估計，台灣地區自 2007年 1月至 2010年 6月，每半年之人類乳突狀病毒疫苗銷售數量；衡量方法採用線性最小平方法，檢視外部擴散模型、內部擴散模型以及混合擴散模型之資料適配度與預測能力。結果：混合影響模型預估之外部影響係數為0.0203，內部影響係數為0.4555，解釋能力達到99.47%，且具有非常高之預測準確度 (MAPE=6.7%)；其次為內部影響模型，其預估內部影響係數為0.6284，對原始資料的解釋能力達到99%，預測誤差則為13.19%，具有好的預測準確度；外部影響模型預估之外部影響係數為0.1420，但其預測誤差為73.38%，參數解釋能力也僅有76%，因此模型解釋力較為不佳。結論：本研究的實證結果指出混合影響模型最能解釋在台灣地區，人類乳突狀病毒疫苗接種之擴散情形，其中內部影響係數 (q=0.4555) 大於外部影響係數 (p=0.0203)，這些數據意含著，民眾比較會參考之前已接種過疫苗個案的經驗及意見，甚於大眾媒體的推廣資訊。因此，雖然政府目前強力推動人類乳突狀病毒疫苗接種之電視廣告、車廂廣告及醫院海報之公共宣傳策略，仍需投入更多比例之資源在推動人際間的宣傳與口碑上，才能確實讓更多民眾有實際接種人類乳突狀病毒疫苗之意願。

Purpose: The aim of this study was to determine the factors that might increase the number of Human Papilloma Virus (HPV) vaccine inoculations in Taiwan. Promoting more people to be inoculated with HPV vaccine would benefit both those individuals and the society. Methods: This study applied three different diffusion models to analyze the data about HPV vaccine inoculation. These data were derived from the IMS (Intercontinental Medical Statistics company, IMS) health database of sales of HPV vaccine every six months from January, 2007 to June, 2010 in Taiwan. After adopting NLS (Nonlinear Least Squares, NLS)as the estimation method, we examined the predictive ability of external, internal and mixed diffusion models. Results: The mixed diffusion model’s coefficient of external influence was 0.0203, the coefficient of internal influence was 0.455, the explanation ability was 99.47%, and it fit the data best (MAPE=6.7%). This was followed by the internal diffusion model, whose coefficient of internal influence was 0.6284, explanation ability was 99%, and MAPE was13.19%. The external diffusion model’s coefficient of external influence was 0.1420, but the MAPE was 73.38%, and the explanation ability was only 76%. Conclusion: The results of this study showed that the mixed diffusion model best predicted the diffusion of HPV vaccine inoculation in Taiwan, and that internal influence (p =0.4555) was more important than external influence (p=0.0203). This suggested that the reason the public was motivated to receive inoculations of HPV vaccine was the experience and opinions of people who were already vaccinated and not TV commercials. The Government should therefore provide more resources for increasing the internal effect, for example 'word-of mouth' promotion rather than the public promotion they focus on now, to convince more people to receive HPV vaccine inoculations.