Doll R, Peto R
Br Med J. 1977 Jun 4;1(6074):1433-6. doi: 10.1136/bmj.1.6074.1433.
A total of 20540 male doctors who replied to a questionnaire on their smoking habits that was sent to them on 1 November 1951, and who were aged 35 years and over, were classified according to their occupation as listed in the Medical Directory for 1952 and followed up until 1 November 1971. Examination of the mortality rates in 11 occupational groups showed gross heterogeneity for smoking-related diseases but not for all other diseases grouped together. On average, general practitioners smoked 37% more cigarettes than did hospital physicians and surgeons and the overall death rates among general practitioners were about 23% higher than among physicians and surgeons of similar ages. This excess death rate was chiefly accounted for by a 38% excess mortality from smoking-related diseases such as lung cancer, chronic bronchitis, and ischaemic and pulmonary heart disease. The few other statistically significant associations between occupation and disease were thought to be due either to chance or to the effect of the disease on the choice of specialty.
1951年11月1日，共有20540名年龄在35岁及以上的男性医生回复了一份关于他们吸烟习惯的调查问卷，这些医生根据1952年《医学名录》中列出的职业进行分类，并随访至1971年11月1日。对11个职业组的死亡率进行检查发现，与吸烟相关的疾病存在明显的异质性，但其他所有疾病合并在一起则不存在。平均而言，全科医生吸烟量比医院内科医生和外科医生多37%，全科医生的总体死亡率比年龄相仿的内科医生和外科医生高出约23%。这种额外的死亡率主要是由吸烟相关疾病（如肺癌、慢性支气管炎、缺血性和肺源性心脏病）导致的38%的额外死亡率造成的。职业与疾病之间其他一些具有统计学意义的关联被认为要么是偶然因素，要么是疾病对专业选择的影响。
