The Added Value of Using Primary Care Data in Population Health Management

James Barrett ^a, Paul Molyneux ^a, Stephen Sutch ^b

Introduction
Population Health Management (PHM) is a key data driven methodology aimed at improving the health of the whole population. PHM techniques include population profiling to understand health needs, followed by targeted interventions to improve the physical and mental health of specific segments of the population. This can include allocating scarce resources based on measured need. In many countries the most easily available national datasets come from secondary care (Hospital). However, there are limitations with the coverage and completeness of secondary care datasets which may impact their suitability for use in PHM. In this paper we will assess the suitability of secondary care datasets for PHM and look at the added value a rich primary care dataset can bring to PHM.

Methods
A study was made of the differences in disease prevalence for the same UK population when they are based on primary care data only compared with secondary care data only. From this we were able to establish the extent of morbidity information that is present in one data set and not in the other. The impact of these differences on typical PHM analyses was also assessed.

Results
Results will be shared showing some of the differences in the comprehensiveness, completeness and richness of datasets that do and do not include full primary care data. Significant morbidity such as angina, eczema and depression are often not present in the secondary care record as they can be managed in primary care. A large proportion of the population were found to have primary care data but no secondary care data. Examples of analyses relevant to PHM that benefit from using the full primary care data set will be shown, including some real-world case studies utilising such an integrated care data set.

Discussion
In general, secondary care data is more available than primary care data and many PHM projects are based on secondary care data alone. In this study we have shown that the increased coverage and completeness seen in primary care data allows a range of PHM analyses that would not be possible with just secondary care data. PHM analysts need to be aware of the limitations of the data set when drawing conclusions based on secondary care data only.


^a Johns Hopkins HealthCare Solutions, United Kingdom
^b Johns Hopkins University, United States

Original Version in PDF