Patient characteristics and degrees of discontinuity of care in Danish general practice: a cohort study

Continuity of care is crucial for effective health care, but patients often experience discontinuity of care. Understanding the patterns and associated patient characteristics can guide interventions to improve continuity of care.

To investigate longitudinal provider discontinuity of care to identify disparities in patient characteristics across degrees of discontinuity of care in Danish general practice.

Cohort study using data from all Danish general practice patients aged ≥12 years, who were alive between 2007 and 2018.

Logistic regression was employed to estimate discontinuity of care at level 0 versus: 1-2 shifts, 3-4 shifts, 5-6 shifts, 7-8 shifts, and ≥9 shifts. Five regression models were used to analyse the odds of different levels of discontinuity of care relative to demographic, regional, municipal, socioeconomic, and morbidity factors.

The majority of males were likely to have lower levels of discontinuity of care than females; however, males in the subgroup with the most shifts had a higher likelihood of experiencing the highest level of discontinuity. Younger adults aged 25-44 years had a higher likelihood of moderate-to-high discontinuity of care compared to those aged 12-24 years, whereas older adults had a lower probability. Discontinuity of care was higher among residents of the Capital and Zealand Regions, and varied by municipality type - being lowest in intermediate areas and highest in peripheral and rural municipalities at the most severe levels. Individuals who were unemployed, those in the lower income quartiles, and those classified as being of 'other ethnicities' and single status had increased probability of discontinuity of care. Patients with lower or moderate-to-high morbidity were also more likely to experience discontinuity of care than patients with no chronic diseases.

This study revealed disparities in discontinuity of care linked to lower socioeconomic status and higher morbidity, with varying odds by region, municipality type, age, and gender.