Individual variations in sickness absence

Records of personal sickness absence, including all spells of one day's duration or more, have been kept at this refinery for more than 20 years. The distributions of sickness spells and also calendar days of absence have been analysed for single years and also for periods of up tO 20 years' continuous service for the 1,350 hourly paid male employees. It is shown that both these measures of sickness absence are distributed among the men in an unequal fashion (negative binomial) rather than at random (Poisson) and thus resemble the distribution of industrial accidents first described almost 50 years ago. This pattern of distribution is not related to occupation or to length of service. Analysis of the distribution of lateness and absenteeism for reasons other than sickness or holidays shows that these also follow this pattern. It is postulated that this could be a principle applicable to all forms of industrial absenteeism as well as to accidents. The personal records of I87 men with continuous service from 1946 to I965 have been studied to investigate the trends in their sickness absence over this 20-year period. In contrast to the well-recognized pattern that in any one period of time young men have more spells of absence than their older fellows, this secular study shows that sickness spells have not decreased with age. This apparent paradox is explicable by the rising national trend in sickness absence and by a high labour turnover in young men with frequent sickness spells.