BACKGROUND: Undertaking research on the role of regular GPs with regard to rates of sickness absence is methodologically challenging, and existing results show a wide divergence. We investigated how long-term sickness absence is affected by the characteristics of doctors and their patient lists. MATERIAL AND METHODS: The study encompassed all those vocationally active residents of Oslo and Bergen in 2005-2006 who had the same regular GP throughout 2006 (N = 298,039). Encrypted data on sickness absence for each individual in 2006, as well their age, gender and level of education were merged with data on the regular GPs (N = 568) and their patient lists, and subsequently analysed with the aid of logistic regression. The outcome variable was at least one period of sickness absence which had been paid for by the Norwegian Labour and Welfare Administration (NLWA). The explanatory variables included the age, gender, list length and list status (open/closed) of the regular GPs, as well as variables that characterised the composition of the patient lists. The analyses were stratified by gender and controlled for individual age and education. RESULTS: The age, gender and list length of the regular GPs were not associated with sickness absence paid for by the NLWA. The odds ratio for sickness absence > 16 days was reduced for both women and men when the list contained many highly educated patients, a high proportion of elderly people and few disability pensioners. Men on lists with a high proportion of men and lists with a high proportion of vocationally active patients also had lower odds rates for sickness absence > 16 days. Among women, the rate of sickness absence was lower for those on open lists than for those on closed lists. INTERPRETATION: In addition to well-known individual factors, the study shows that the likelihood of sickness absence is affected by the socio-demographic composition of the patient list to which one belongs.
BACKGROUND: Undertaking research on the role of regular GPs with regard to rates of sickness absence is methodologically challenging, and existing results show a wide divergence. We investigated how long-term sickness absence is affected by the characteristics of doctors and their patient lists. MATERIAL AND METHODS: The study encompassed all those vocationally active residents of Oslo and Bergen in 2005-2006 who had the same regular GP throughout 2006 (N = 298,039). Encrypted data on sickness absence for each individual in 2006, as well their age, gender and level of education were merged with data on the regular GPs (N = 568) and their patient lists, and subsequently analysed with the aid of logistic regression. The outcome variable was at least one period of sickness absence which had been paid for by the Norwegian Labour and Welfare Administration (NLWA). The explanatory variables included the age, gender, list length and list status (open/closed) of the regular GPs, as well as variables that characterised the composition of the patient lists. The analyses were stratified by gender and controlled for individual age and education. RESULTS: The age, gender and list length of the regular GPs were not associated with sickness absence paid for by the NLWA. The odds ratio for sickness absence > 16 days was reduced for both women and men when the list contained many highly educated patients, a high proportion of elderly people and few disability pensioners. Men on lists with a high proportion of men and lists with a high proportion of vocationally active patients also had lower odds rates for sickness absence > 16 days. Among women, the rate of sickness absence was lower for those on open lists than for those on closed lists. INTERPRETATION: In addition to well-known individual factors, the study shows that the likelihood of sickness absence is affected by the socio-demographic composition of the patient list to which one belongs.