Quality assessment in sickness certificates – changes over an eight-year period in Sweden and associated factors

This study investigates changes in the quality of information in sickness certificates over time, before and after the 2011 national introduction of a revised sickness certificate, and identified factors associated with certificate quality. Four experts independently assessed the quality of a total of 783 sickness certificates issued in primary care in 2004, 2009, and 2012. A Global Quality Score (GQS) was constructed by the research group, rating quality on a 10-point scale. The cut-off for high quality was set to GQS &gt;5. The inter-rater reliability of the GQS was tested using Intra-Class Correlation and internal consistency by Pearson correlation coefficients (r). Sickness certificates issued in 2004 and 2009 were merged into one group and compared to sickness certificates in 2012. Logistic binomial regression analyses examined associations between patient-, sick leave-, and physician-related variables and GQS. The GQS demonstrated moderate to good inter-rater reliability (ICC= 0.79, 95% CI: 0.6–0.9) and good internal consistency. Certificate quality improved significantly (<i>p</i> &lt; 0.001) from 3.6 (SD 1.5) in 2004/2009 to 4.7 (SD 1.2) in 2012. In 2004/2009, mental disorder or an extended length of sick leave was negatively associated with high quality. In 2012, only female sex of the patient was negatively associated with high certificate quality. While sickness certificate quality was overall low, it improved significantly after the 2011 revision of the sickness certificate. Also, diagnosis, sick leave duration, and patient’s sex were factors associated with the quality and varied over time, highlighting the need for further research regarding potential remaining differences in quality of sickness certificates.