Burke W. Mamlin

A simple error classification system for understanding sources of error in automatic speech recognition and human transcription.

OBJECTIVESnTo (1) discover the types of errors most commonly found in clinical notes that are generated either using automatic speech recognition (ASR) or via human transcription and (2) to develop efficient rules for classifying these errors based on the categories found in (1). The purpose of classifying errors into categories is to understand the underlying processes that generate these errors, so that measures can be taken to improve these processes.nnnMETHODSnWe integrated the Dragon NaturallySpeaking v4.0 speech recognition engine into the Regenstrief Medical Record System. We captured the text output of the speech engine prior to error correction by the speaker. We also acquired a set of human transcribed but uncorrected notes for comparison. We then attempted to error correct these notes based on looking at the context alone. Initially, three domain experts independently examined 104 ASR notes (containing 29,144 words) generated by a single speaker and 44 human transcribed notes (containing 14,199 words) generated by multiple speakers for errors. Collaborative group sessions were subsequently held where error categorizes were determined and rules developed and incrementally refined for systematically examining the notes and classifying errors.nnnRESULTSnWe found that the errors could be classified into nine categories: (1) announciation errors occurring due to speaker mispronounciation, (2) dictionary errors resulting from missing terms, (3) suffix errors caused by misrecognition of appropriate tenses of a word, (4) added words, (5) deleted words, (6) homonym errors resulting from substitution of a phonetically identical word, (7) spelling errors, (8) nonsense errors, words/phrases whose meaning could not be appreciated by examining just the context, and (9) critical errors, words/phrases where a reader of a note could potentially misunderstand the concept that was related by the speaker.nnnCONCLUSIONSnA simple method is presented for examining errors in transcribed documents and classifying these errors into meaningful and useful categories. Such a classification can potentially help pinpoint sources of such errors so that measures (such as better training of the speaker and improved dictionary and language modeling) can be taken to optimize the error rates.

Chapter 5 – Regenstrief Medical Informatics: Experiences with Clinical Decision Support Systems

The discipline of clinical informatics endeavors to improve the process and outcomes of health care by enabling efficient access to information. Care providers can then use this information, both in the form of medical knowledge and in the form of patient data collected during clinical practice, to make decisions and comply with appropriate standards of care. The Regenstrief Institute began work on clinical information systems in 1972, when Dr. Clement McDonald and colleagues conceptualized and began construction of a computerized patient management system for outpatient diabetes care, developed to meet three primary goals: first, it was built to eliminate the problems inherent in paper records by making clinical data available to authorized users “just-in-time” as medical decisions are made; second, it was designed to aid in the recognition of diagnoses and adoption of pertinent care practices by assisting clinicians during their record-keeping activities; third, the system was designed to aggregate and analyze clinical information to be used in health care support systems, such as those for public health, health services research, and quality improvement. The first installation of the Regenstrief Medical Record System (RMRS) at Wishard Memorial Hospital occurred in 1974 and, over the next few years, the use of this system expanded outside of the diabetic clinic into a few of the hospital’s many general medicine clinics. From early in its history, the Regenstrief system has included mechanisms for tailoring rules based on the data, to generate reminders and alerts to care providers. This chapter provides a history of the development and growth of the RMRS into a region-wide source of clinical data, the Indiana Network for Patient Care (INPC), and a summary of the research on the decision support interventions themselves, made possible by this infrastructure. Additionally, lessons learned throughout the more than 30 years of experience in both building and maintaining this system are detailed, alongside some reflections that may be useful for future system builders.

Towards Standardized Patient Data Exchange: Integrating a FHIR Based API for the Open Medical Record System.

Interoperability is essential to address limitations caused by the ad hoc implementation of clinical information systems and the distributed nature of modern medical care. The HL7 V2 and V3 standards have played a significant role in ensuring interoperability for healthcare. FHIR is a next generation standard created to address fundamental limitations in HL7 V2 and V3. FHIR is particularly relevant to OpenMRS, an Open Source Medical Record System widely used across emerging economies. FHIR has the potential to allow OpenMRS to move away from a bespoke, application specific API to a standards based API. We describe efforts to design and implement a FHIR based API for the OpenMRS platform. Lessons learned from this effort were used to define long term plans to transition from the legacy OpenMRS API to a FHIR based API that greatly reduces the learning curve for developers and helps enhance adhernce to standards.