|
CLINICAL DATA ANALYTICS LANGUAGE |
CLINICAL DATA ANALYTICS LANGUAGE
Brief Description:
General purpose access to data from clinical information systems, beyond retrieval for point of care work, is needed for many aspects of the hospital’s work particularly for clinical research, logistics & operational planning, and auditing patient safety.
Current systems, with rare exceptions, only provide access to data identified in standard reports with no flexibility to make ad hoc enquiries or to pursue new directions of enquiry. The clinical data analytics language the Laboratory has developed has two objectives: to enable the expression of any question that can be answered from the data in the database; and, to compute and display that answer.
A prototype of the language has been developed for the CareVue information system used in the ICU at the Royal Prince Alfred Hospital. It provides for the use of local medical dialects, SNOMED CT terminology, identification of any physical databases to be used in a query, specification of patient groups, a variety of statistical functions, and constraints over any medical variable, Time, and Location.
Project Aim:
The aim of this project is to make it easy to ask any question, using a restricted natural language, about the data stored in a clinical information system including questions based on the semantic contents of the clinical notes.
Project Objectives:
The objectives of this project are to enable clinicians:
1. at the point-of-care to retrieve case materials based on their contents and not the identity of the patient.
2. acting as researchers to obtain aggregated information about groups of patients and statistically test hypotheses about the differences between groups.
3. acting as administrators to likewise readily obtain questions about aggregations of the patient data appropriate for the managing the logistics of the ward.
4. acting as auditors to investigate patient cases based on semantic contents of records and compare and contrast aggregations of various patient groups with each other or with individual patients.
Brief Project Methodology:
The methodology is to attach CDAL to a clinical information system which then acts as a host for it and whose data contents and architecture it is trained to understand. The steps in the methodology involve:
1. Study the existing clinical information system and establish its internal structure and the methods it uses to store, organise and retrieve data.
2. Interview the staff to establish the local terminology they use to talk about the data items collected by the CIS.
3. Map the local terminology and the host CIS terminology (both at the user and internal level) to SNOMED or any other TTOC (thesaurus, terminology, ontology or classification, e.g. UMLS, ICD-10AM, NIC, NOC, etc.).
4. Design and implement the generic methods for retrieving all data from the host clinical information system.
5. Compute the TTOC indices for historical records held in the CIS.
6. Test the system with trusted staff and develop methods for operationalising the system.
7. Train all staff to use the system.
8. Set up a specific research question for investigation and demonstrate that it can be answered.
9. Collect feedback on the use of the system and ways to improve it.
10. Revise data collection methods in the ward, if considered necessary.
Example Uses of CDAL:
1. Recently in the ICU ward we were able to answer the question: do men have a significantly different blood pressure to women for patients admitted in the last 24 hours? The counterintuitive answer was produced within 2 minutes showing that women have the higher MAP of 80 vs 75 with a probability of this occurring by chance at .023. The surprising result was explained as being due to the fact that the men were sicker than the women.
2. The RPAH Intensive Care Unit and the MayoEpidemiology and Translational Research in Intensive Care (METRIC) of the Mayo Clinic wish to collaborate on testing out a model for predicting patients at risk of Acute Respiratory Distress Syndrome (ARDS). This requires CDAL to monitor patient respirational rates, X-ray pathology reports, and computing severity of illness scores (APACHE IV). CDAL will be used to collect all these variables from the patient records and share the at-risk cases between the two institutions in real-time for mutual advice and discussions.
How will this project improve patient care:
This project will improve patient care by enabling staff to investigate their practices with the evidence in their own CIS and hence strengthen their evidence based medicine practices.
Staff will do more research without loss of time for current activities. They will effectively be able to do their research while they do their ward rounds. With greater access to analysis of their data staff will be able to promote the research dimension of their practices and so improve their professional standing.
Staff will also be able to plan more ambitious research knowing that they have the analytical tools readily available to evaluate its progress. The increase in knowledge gleaned from their research will in turn be used to improve clinical practice and consequently the patient care.
Trainee staff will be better inculcated into an evidenced based medicine approach to their work and learn more quickly the essential information needed for particular types of decision making. It will save staff time having to search through voluminous collections of data to get the data items relevant for not only immediate care tasks but long term care improvement projects.
It will help train staff to use a stable terminology consistently and hence reduce the confusion in communication over care tasks.
It will provide fast semantic retrievals from the clinical notes, and so increase the value of those notes to staff and improve the quality of what they record in them as well as their extent for all categories of clinical staff.
It will improve patient safety by increasing the amount of knowledge used for clinical decision making.
In the long term CDAL is a technology on the pathway to creating more advanced technologies such as real-time audit of patient care.
|
