Q: We usually search our EMR, how is the CTI process any better?
A: The CTInformatics ACTRIS automated pre-screening records review process has several advantages over typical 'EMR Code Search' approaches:
Cast the net wider
Because of the efficiency of the automated search process we are able to cast the net wider than by selecting ICD9 codes alone. In the heart failure domain this is a particular advantage as prior research has shown that ~15% of patients are likely to have HF but are not diagnosed. Additional patients may be diagnosed in the clinical narrative, but not coded, or mis-coded. By looking at all patients receiving cardiac care services the automated system is able to include some of these eligible patients into the selection.
Find 'invisible' data
In many clinical settings there are conditions, symptoms, procedures, and lab results that are documented in the clinical narrative text, but not included in a structured database field. This is common when outside labs are cited into the narrative, and when for mentions of certain symptoms. By using a combination of rules and statistical natural language processing tools we are able to annotate otherwise invisible (without manual review) values of interest.
Make final selections easier
Automated exclusions can make final selections easier and more targeted. For example, in a diastolic heart failure setting, using a code search to find HF patients with EF>40 and BP>130 is often easy to do. It is much harder to find which of those patients meet the broader range of inclusion/exclusion criteria for a study. For example, CTInformatics ACTRIS system can Auto-Exclude on:
- ulcerative carotid arterial plaques,
- serum creatinine > 2.5 mg/dL,
- being treated with dialysis,
- not hospitalized related to heart failure in the past year,
- late-stage kidney disease,
- known or suspected baroreflex failure,
- autonomic neuropathy,
- prior surgery, radiation, or endovascular stent placement in the carotid sinus region,
- significant uncontrolled symptomatic bradyarrhythmias,
- infiltrative cardiomyopathy or a systemic illness known to cause infiltrative cardiomyopathy,
- pericardial constriction,
- solid organ or hematologic transplant.
Automate Pharma Expertise:
Keeping in mind all the alternative medications available to treat a given disorder is a daunting task, even for domain experts. For one of our studies a review of antihypertensive medications was required. Our system automatically evaluated for over 150 antihypertensive medication formulations to identify distinct medications given to patient. The system has awareness of combo meds and treats each component as a distinct medication. This allows for rapid and comprehensive screening for patients receiving 3+ antihypertensive medications.
Minimize manual chart review:
By selecting a highly targeted set of patients the amount of manual chart review time is reduced. Because the screening is automated it produces consistent results across large data sets. It reduces the amount of training, review, and Q/C time needed. It allows reviews of very large patient populations without overwhelming clinical research staff.
