標準人口的選用，其目的是校正差異性人口結構所造成的影響，但是當所選用的標準人口不一樣時，可能會產生不同的結果。本研究以台灣區域23縣市為實例，探討以不同標準人口對直接標準化死亡率排續的影響。依台灣縣市地區23縣市19個年齡層別，計算各縣市各年齡層別死亡率，並與各縣市直接標準化死亡率排序做比較，各縣市標準化死亡率係以台灣地區及23縣市年齡別人口為標準人口調整各縣市死亡率而得。我們發現兩地區或多地區中若每一年齡別死亡率均高的地區，則標準人口不論如何選擇，在直接標準化死亡率排序上，此一地區都會較高，例如台北市每一年齡別死亡率均低於花蓮縣，則標準人口不論是何種組成其排序台北市都低於花蓮縣。若兩地區或多地區年齡別死亡率不規則，選擇不同人口為標準人口，在直接標準化死亡率排序上，會因標準人口的不同而有所不同，例如苗栗縣及台南市年別死亡率分佈有交錯的情況，此2縣市以不同人口為標準人口，其直接標準化率排序因標準人口的不同而有變動。對於由後者得到的直接標準化死亡率再使用上我們宜應謹慎。利用直接標準化死亡率，標準人口的比較上，我們最好能計算直接標準化率95%的信賴區間。
We compared the effects of different standard populations to the rank of the direct standardized mortality rate of 23 countries and cities in Taiwan. The result of the ranking is based on the comparison of direct standardized mortality rate, which is the adjustment of mortality of 19 age groups used in standard populations in 23 countries and cities. When comparing mortality rates for 2 or more areas in all age groups, the area having the higher age specific mortality rates consistently showed a higher direct standardized mortality rate, even when different standard populations in these areas were chosen. For instance, the mortality rates of all age groups in Hwalien Country are higher than Taipei City, and the direct standardized mortality either, even adopting the different standard populations. In some cases, the mortality rates in all age group are very variable when comparing one area to others. When we compared the direct standardized mortality rates, the result was very different, depending upon the choices of the different standard populations. For instance, the comparison of mortality rates over all age groups was variable like Miaoli Country and Tainan City, so by choosing different standard populations, the result of direct standardized mortality rates showed differences. Therefore, it would be very careful to use the direct standardized mortality rate. To use the direct standardized mortality, we had better adopt 95%confidence interval in the comparison of the mortality among areas.
