BMC Health Services Research (Jan 2021)

The effect of the integration of health services on health care usage among patients with type 2 diabetes in North Karelia, Finland

  • Katja Wikström,
  • Marja-Leena Lamidi,
  • Päivi Rautiainen,
  • Hilkka Tirkkonen,
  • Petri Kivinen,
  • Tiina Laatikainen

Journal volume & issue
Vol. 21, no. 1
pp. 1 – 9


Read online

Abstract Background The need to improve the care of people with complex care requirements has been driving the reforms integrating care processes. This study examines the effect of the integration of health services on health care usage and the processes and outcomes of care among type 2 diabetes patients. Methods Data include all type 2 diabetes patients who lived in North Karelia, Finland, between 2014 and 2018. Health care contacts and glycated haemoglobin (HbA1c) measurements were obtained from the electronic health records. Logistic, Poisson and linear models with generalised estimating equations and the Friedman test were used to study the differences between years. Results The health care usage was highest in 2017, the first year of a new organisation, and smallest in the following year. Before the new organisation, the health care usage was lowest in 2014, being slightly higher compared with 2018. Between the last two years, the mean number of contacts per person declined from 3.25 to 2.88 (-0.37, p < 0.001). The decreasing pattern seen in total health care usage was most obvious among contacts with primary health care nurses. The number of contacts increased only among specialised care nurses between the last two years. The number of HbA1c measurements was also in its lowest in 2018 but in its highest in 2015. Between the years 2014 and 2018, the difference in the mean number of contacts was − 0.05 (p = 0.011) for those not measured, -0.02 (p = 0.225) for those measured and within the target level of HbA1c, and 0.12 (p = 0.001) for those measured and not at the target level of HbA1c. Conclusions Health care integration first increased the health care usage but then brought it to a slightly lower level than before. The changes were most obvious in primary health care nurses’ appointments, and no decline was observed in secondary-level care. Even though the numbers of HbA1c measurements and the proportion measured declined, measurements increased among those with poor glycaemic control. The observed changes might reflect the better targeting and more concordant services in different service units.