Lost Souls®: Detect & Track Radiology Recommendations

Summary:

Automated detection and tracking of patients who did not obtain the indicated, overdue follow-ups which were recommended in radiology reports.

Objectives of Lost Souls® Recommendation Tracker:

Problem:

Solution:

In 51% of cases, physicians fail to obtain the indicated follow-up exams per Annals of Emergency Medicine, Dutta, Sayon et al. "Automated Detection Using Natural Language Processing of Radiologists Recommendations for Additional Imaging of Incidental Finding", Aug 2013, Volume 62, Issue 2, 162 - 169.
Numerous lawsuits against physicians and hospitals arise from missed follow-ups, for example:
- $1.5 million settlement, surprisingly despite the patient failing to get their follow-up:
  - D. Collins, JD "Missed follow-up leads to stage IV renal cancer, death", Urology Times, 1/14/2014.
- $2.5 million settlement:
  - R. Kreisman, JD "$2.5 Million Settlement for Doctor’s Failure to Follow Up on Signs and Symptoms of Lung Cancer", Chicago Malpractice Attorney Blog, 2/28/2014.
- $850,000 settlement:
  - R. Kreisman, JD "Doctors’ Failure to Timely Diagnose Lung Cancer Leads to $850,000 Settlement – Sahagun v. Aven", Chicago Malpractice Attorney Blog, 8/30/2014.

Detects past radiology reports that made recommendations for follow-up or biopsy, while understanding various phrasing patterns within reports, to find which patients did not get the appropriate follow-up which may be now overdue.
Sorts and Filters the recommendations by relevance and discards the recommendations which were already fulfilled, and optionally discards recommendations which are unlikely to be relevant.
Tools to efficiently track and manage the recommendations.
Secure Messenger component facilitates communication with clinicians or patients, tracking message delivery, and clinician responses pertaining to whether the test will be performed.
Designed and written by a Radiologist, with thorough understanding of all pertinent aspects such as filtering out recommendations of low relevance, estimating the importance of findings, zero-effort usage (no need for radiologists to click any buttons to begin tracking a recommendation, and no need to change reporting style), and streamlined efficient work-flow that makes sense.
Scientific Abstract & Presentation at The Society of Imaging Informatics in Medicine 2016.

Software application accessible from within institutional Local Area Network via most web-browsers (available as a PC tower, 1U rack server, virtual-machine image).
Automatically obtains reports from existing dictation software.
Optional Secure Messenger component that is accessible to clinicians and patients from the public internet.
For residency and fellowship programs, optionally bundled with Boomerang™, the application for providing effortless feedback to residents by highlighting attending's edits to trainee's reports.