Keeping patients out of hospital by improving patient trajectories in primary care - methodological development and effect

It is possible to study patient trajectories using data from different types of electronic registers and health records. If these register data are linked to measures of effect and quality, it is possible to identify patient trajectories which is most beneficial both from a patient and system perspective. However, this is seldom done due to challenges getting permission to link data, lack of appropriate methods for compiling such data and lack of appropriate measures of quality of integrated services from a patient perspective. The aim with this project was to 1) develop tools to harvest and combine data from different registers, 2) visualise patient trajectories across services, 3) identify variables which reflects quality from a patient perspective and 4) investigate the relationship between patient experienced quality and different patient trajectories. A register based longitudinal cohort study with data on contacts with somatic health services during a two-year period for 170000 person. A survey with a random sample of 12500 among those with chronic complaints were invited to answer questions about health, function and perceived quality four months later. It is developed software to combine data from different sources and visualise individual patient trajectories. Those taking part in the survey where shown their own trajectory and answered questions about their experiences based on this. It was found that type or frequency of use of somatic health and social care services were not related to patient?s experience of cooperation among health care workers. There were, however, a strong linear relationship with increasing age. In another investigation on predictors for unplanned hospital admissions, it was found that in addition to previous unplanned hospital admissions, having been an inpatient in 3 or 4 different wards and having a chronic disease in 4 or more organ system where the most influential variables. This project has generated a data set which also in the future will give unique possibilities to study patient trajectories across services and importantly also include the patient?s own judgement linked directly to their individual trajectory.

There is a lack of knowledge about the trajectories patients with chronic diseases go through within the primary health care service. However, there is a lot of data gathered about these patients in different data systems like the electronic medical recor d (EMR). The problem is that the EMR systems do not generate information about trajectories. There is thus a need for the development of methods that can be used to extract and analyse relevant information from EMR systems. When such methods are developed , the information extracted can be used to identify and investigate the effect of different trajectories. There are two main areas of use for information about trajectories within the primary health care service in Norway today. - To investigate the qual ity and thus the effect of the services delivered to the best for the patients. This can be done by finding indicators for quality of care by identifying trajectories that are close to the chronic disease management models (CDM). - To investigate the coor dination and integration of health care services by e.g. describing the trajectories that actually are used and identify the ones that lead to reduced use of specialist care. Using information gathered in the data systems in primary care can thus be used to improve care and fulfil political aims by providing knowledge about what is happening in everyday practice and the effect of the different ways of providing integrated practice in primary care. To achieve this, the project will make use of the recen t change in Norwegian legislation for research on routine medical data and the advent of a new way of extracting, organizing and analyzing data called Event Stream Analysis (ESA). ESAs gives the opportunity to look at the effectiveness of the existing tra jectories using data from the electronic medical record. We will also establish measures of CDM to discover whether these, in line with existing theory, keep patients with chronic conditions out of hospital.