Czaderny Krzysztof
Warsaw School of Economics, Szkoła Główna Handlowa w Warszawie
Maria Skłodowska Curie Institute – Oncology Centre, Warsaw, Centrum Onkologii – Instytut im. Marii Skłodowskiej-Curie
Przegl Epidemiol. 2018;72(2):235-246.
This study investigates spatial distribution and future trends in prostate cancer mortality in Poland. Since detection of prostate cancer is based on patients’ requests (not an invitation system), hypothesis of self-selection of patients was also verified
Age-period-cohort model was estimated to assess mortality trends. A spatial analysis was performed by Kulldorff’s and Moran’s tests. Structural equation model was fitted to identify factors influencing decision on participation in prostate cancer screening. Statistical analysis is based on data from the National Cancer Registry for 1980-2014 and a 2014 questionnaire survey of 1,700 men aged 45+.
The annual number of registered deaths due to prostate cancer is expected to increase from 4,440 in 2014 to 6,550 in 2030. Prostate cancer mortality in Poland is to a large extent spatially uniform (global Moran’s Ig=0.066) and was not found to be related to spatial patterns in oral contraceptive use in women which was recently hypothesised. Nevertheless, a Kulldorff’s cluster can be identified over the Lubusz and Greater Poland areas for both the 1980-1984 and the 2010-2014 periods (p<0.001). The structural equation model showed that the early detection of prostate cancer is influenced mainly by physician’s activeness and local availability of healthcare (p<0.050). The direct effect of socio-economic status on early detection of prostate cancer is insignificant.
When considering prediction results, one should request more actions aimed at facing the upward trend of deaths, including making research funding in prostate cancer a priority, improving understanding of prostate cancer among males, and offering to patients individualised risk-adapted strategy modelled on the guidelines of the European Association of Urology. Currently, uptake of prostate cancer screening is largely affected by external factors.
本研究调查了波兰前列腺癌死亡率的空间分布及未来趋势。由于前列腺癌的检测基于患者的请求（而非邀请系统），因此也验证了患者自我选择的假设。
采用年龄-时期-队列模型评估死亡率趋势。通过 Kulldorff 检验和 Moran 检验进行空间分析。拟合结构方程模型以确定影响参与前列腺癌筛查决策的因素。统计分析基于 1980 - 2014 年国家癌症登记处的数据以及 2014 年对 1700 名 45 岁以上男性的问卷调查。
预计因前列腺癌登记的年死亡人数将从 2014 年的 4440 人增加到 2030 年的 6550 人。波兰的前列腺癌死亡率在很大程度上在空间上是均匀的（全局 Moran's I = 0.066），并且未发现与最近假设的女性口服避孕药使用的空间模式相关。然而，在 1980 - 1984 年和 2010 - 2014 年期间，在卢布斯和大波兰地区均可识别出 Kulldorff 聚类（p < 0.001）。结构方程模型表明，前列腺癌的早期检测主要受医生的积极性和当地医疗保健的可及性影响（p < 0.050）。社会经济地位对前列腺癌早期检测的直接影响不显著。
考虑到预测结果，应要求采取更多行动以应对死亡人数上升的趋势，包括将前列腺癌研究资金列为优先事项、提高男性对前列腺癌的认识，以及根据欧洲泌尿外科学会的指南为患者提供个性化的风险适应策略。目前，前列腺癌筛查的接受程度在很大程度上受外部因素影响。
