“Gazing into the crystal ball”: applying outcome prediction tools for early identification of patients who may require palliative care in a general practice database

Almost two thirds of patients who die expected deaths in Australia do not receive palliative care support, which is known to improve quality-of-life. General practitioners, while ideally placed to identify these patients, have low rates of referral to palliative care services. Barriers include difficulty in predicting poor prognosis and time constraints. Running a computerised search tool in electronic GP databases may help by identifying target patients. There are a number of patient morbidity and mortality prediction tools which have been validated in primary care (e.g. Charlson Co-morbidity Index, SPICT Tool). This study aims to apply these tools to a GP database to determine their feasibility as a screen for patients who require palliative care interventions.
 
Methods
The database that will be used collates data from 10 general practices in metropolitan Sydney. Each outcome tool will be converted into a search algorithm and applied to the database retrospectively to obtain data on the number of patients identified as being at risk of morbidity or mortality. This will then be compared to actual outcomes.
 
Findings
Conclusions will be made on the feasibility of using outcome tools in an electronic GP database to identify patients in need of palliative care and recommendations made on which tools have the best predictive value.
 
This study aims to demonstrate the use of information readily available in GP electronic medical records to screen for patients at risk of deterioration, thereby enhancing practice and improving patient outcomes.

“Gazing into the crystal ball” (350kb pdf)