Knox E G, Lancashire R J
Department of Social Medicine, University of Birmingham.
J Public Health Med. 1991 Aug;13(3):142-50. doi: 10.1093/oxfordjournals.pubmed.a042610.
The spatial and temporal distributions of cot deaths and of infant respiratory deaths in Birmingham between 1964 and 1984 were examined. Respiratory deaths declined and cot deaths increased, and the combined rate showed a downward trend. Seasonal variations were identical in the two groups and in each case were more powerfully related to date of death than to date of birth. Cumulative mortalities were parallel by age, depended crucially upon the quarter of the year in which the child was born, and were identical for the two classes of death. The separation of the temporal trends probably owed more to changes in diagnostic practice than to biological or epidemiological processes. Powerful space-time interactions were detected for both groups and for cross-pairs belonging to different groups. There was a four-yearly oscillation in the combined death rate, of doubtful statistical significance, but synchronous with a similar oscillation reported by other investigators using national data. The four-yearly oscillation was also synchronous with the four-yearly pattern for national isolations of Mycoplasma pneumoniae.
对1964年至1984年间伯明翰市婴儿猝死和婴儿呼吸道死亡的时空分布进行了研究。呼吸道死亡人数下降,婴儿猝死人数增加,两者合并发生率呈下降趋势。两组的季节性变化相同,在每种情况下,季节性变化与死亡日期的相关性都比与出生日期的相关性更强。累积死亡率按年龄呈平行关系,关键取决于孩子出生的年份季度,且两类死亡的累积死亡率相同。时间趋势的分离可能更多是由于诊断实践的变化,而非生物学或流行病学过程。在两组以及属于不同组的交叉配对中均检测到了强烈的时空相互作用。合并死亡率存在每四年一次的波动,其统计学意义存疑,但与其他使用全国数据的研究者报告的类似波动同步。这种每四年一次的波动也与全国肺炎支原体分离的四年模式同步。