Can casemix-systems be applied in Danish primary care?

Background:
New technology in terms of IT systems, better data infrastructure and improved registrations of health data provide new opportunities for health care systems to improve the care experience of individual patients, improve public health and reduce healthcare costs. Application of "Big Data", which covers the collection, storage, analysis, processing and interpretation of large amounts of data can via a casemix system provide new and insightful information about the morbidity burden of populations in terms of co-morbidity in addition to index conditions/multi-morbidity and related resource consumption.
.
Aim:
The objective of this study was to conduct a preliminary analysis of the usefulness of the ACG casemix system in Denmark. This includes presenting the results for a region of Denmark and assessment of the usefulness and quality of the results.


Methods and Data:
This cross-sectional study uses individual data on gender, age and diagnoses for the year 2013 from the Danish General Practice Database (DAMD) to make ACG groupings of a sample of 700,443 citizens of Zealand in Denmark. The Johns Hopkins Adjusted Clinical Groups ACG software, Version 10 December 2011 was applied to make the ACG groupings. Individuals were divided into patients linked to sentinel clinics, which diagnose code at least 70% of their patients, and patients associate to clinics without sentinel status.
Through comparisons with results for other countries and sentinel versus non-sentinel status, the results were used to assess whether the ACG system can be used to estimate the morbidity profile and related resource consumption.


Results:
Via the establishment of the Danish General Practice Quality Database (DAMD) and a political desire to reform the Danish general practice sector preconditions for an implementations of a casemix system has been implemented in Denmark. By comparison with similar results in Sweden and other countries, the results of the ACG grouping indicate, that the ACG grouping provides useful and meaningful results. However, there are significant differences between diagnose coding and ACG-groupings, when comparing the distribution of ACG values for sentinel clinics with non-sentinel clinics. ACG groupings based on all clinics in Region Zealand rather than sentinel clinics leads to a inappropriate an non-useful ACG groupings when compared to results from Sweden and other countries where the ACG system is implemented. There can be significant differences in diagnostic coding systems and coding practices across countries, which should be taken into account in country comparisons.


Conclusions:
In Danish sentinel clinics coding of diagnoses seems to be relatively good. This means that it is likely that the corresponding ACG / RUB-grouping provide useful estimates of morbidity burden (casemix) and related resource need. Before a casemix system is implemented in Denmark, however, further research should be undertaken of the applicability of the DAMD data and the corresponding ACG groupings.