Abstract: Forecasting is an important aspect of health services research,and it is crucial to making health planning and health policy.At present,methods such as regression analysis,factor analysis,time series analysis,trend extrapolation,gray model,auto regressive integrated moving average（ARIMA） model and artificial neural network model are widely used in prediction analysis,however,with strict requirements on index values of predicting future trends,the conventional forecasting methods are lack of flexibility in the adjustment of indexes,and their research objective is too simplex.In order to overcome the shortcomings of the conventional forecasting methods,the paper,from the perspective of methodology,explores the application of system dynamics in health services research combined with the characteristics of health services.

Key words: System dynamics, Health services research, Forecasting

摘要： 预测是卫生服务研究中的重要方面，对于制定卫生规划与卫生政策至关重要。目前，回归分析、因子分析、时间序列分析、趋势外推法、灰色模型、自回归积分滑动平均（ARIMA）模型及人工神经网络模型等方法被广泛应用于预测分析中，但常规预测方法对未来变化趋势进行预测的指标数值要求严格、指标调整不够灵活、研究目标过于单一。为克服常规预测方法的不足，本文从方法学角度出发，结合卫生服务的特点，探讨系统动力学方法在卫生服务研究中的应用。

关键词: 系统动力学, 卫生服务研究, 预测