Spelling performance on the web and in the lab.

Several dictionary websites are available on the web to access semantic, synonymous, or spelling information about a given word. During nine years, we systematically recorded all the entered letter sequences from a French web dictionary. A total of 200 million orthographic forms were obtained allowing us to create a large-scale database of spelling errors that could inform psychological theories about spelling processes. To check the reliability of this big data methodology, we selected from this database a sample of 100 frequently misspelled words. A group of 100 French university students had to perform a spelling-to-dictation test on this list of words. The results showed a strong correlation between the two data sets on the frequencies of produced spellings (r = 0.82). Although the distributions of spelling errors were relatively consistent across the two databases, the proportion of correct responses revealed significant differences. Regression analyses allowed us to generate possible explanations for these differences in terms of task-dependent factors. We argue that comparing the results of these large-scale databases with those of standard and controlled experimental paradigms is certainly a good way to determine the conditions under which this big data methodology can be adequately used for informing psychological theories.

Tool-supported Interactive Correction and Semantic Annotation of Narrative Clinical Reports.

OBJECTIVES: Our main objective is to design a method of, and supporting software for, interactive correction and semantic annotation of narrative clinical reports, which would allow for their easier and less erroneous processing outside their original context: first, by physicians unfamiliar with the original language (and possibly also the source specialty), and second, by tools requiring structured information, such as decision-support systems. Our additional goal is to gain insights into the process of narrative report creation, including the errors and ambiguities arising therein, and also into the process of report annotation by clinical terms. Finally, we also aim to provide a dataset of ground-truth transformations (specific for Czech as the source language), set up by expert physicians, which can be reused in the future for subsequent analytical studies and for training automated transformation procedures. METHODS: A three-phase preprocessing method has been developed to support secondary use of narrative clinical reports in electronic health record. Narrative clinical reports are narrative texts of healthcare documentation often stored in electronic health records. In the first phase a narrative clinical report is tokenized. In the second phase the tokenized clinical report is normalized. The normalized clinical report is easily readable for health professionals with the knowledge of the language used in the narrative clinical report. In the third phase the normalized clinical report is enriched with extracted structured information. The final result of the third phase is a semi-structured normalized clinical report where the extracted clinical terms are matched to codebook terms. Software tools for interactive correction, expansion and semantic annotation of narrative clinical reports has been developed and the three-phase preprocessing method validated in the cardiology area. RESULTS: The three-phase preprocessing method was validated on 49 anonymous Czech narrative clinical reports in the field of cardiology. Descriptive statistics from the database of accomplished transformations has been calculated. Two cardiologists participated in the annotation phase. The first cardiologist annotated 1500 clinical terms found in 49 narrative clinical reports to codebook terms using the classification systems ICD 10, SNOMED CT, LOINC and LEKY. The second cardiologist validated annotations of the first cardiologist. The correct clinical terms and the codebook terms have been stored in a database. CONCLUSIONS: We extracted structured information from Czech narrative clinical reports by the proposed three-phase preprocessing method and linked it to electronic health records. The software tool, although generic, is tailored for Czech as the specific language of electronic health record pool under study. This will provide a potential etalon for porting this approach to dozens of other less-spoken languages. Structured information can support medical decision making, quality assurance tasks and further medical research.

Automated misspelling detection and correction in clinical free-text records.

Accurate electronic health records are important for clinical care and research as well as ensuring patient safety. It is crucial for misspelled words to be corrected in order to ensure that medical records are interpreted correctly. This paper describes the development of a spelling correction system for medical text. Our spell checker is based on Shannon's noisy channel model, and uses an extensive dictionary compiled from many sources. We also use named entity recognition, so that names are not wrongly corrected as misspellings. We apply our spell checker to three different types of free-text data: clinical notes, allergy entries, and medication orders; and evaluate its performance on both misspelling detection and correction. Our spell checker achieves detection performance of up to 94.4% and correction accuracy of up to 88.2%. We show that high-performance spelling correction is possible on a variety of clinical documents.

Internal benchmarking paves way for medical transcription improvements.

Accurate, detailed data collection is essential to improving productivity in medical transcription departments. Transcription staff are best positioned to identify nonvalue-added duties and processes and should be involved in improvement planning. Even a small percentage of improvement in productivity can translate into huge dollar and time savings.

Home-based radiology transcription and a productivity pay plan.

Shands Hospital in Gainesville, Fla., decided to evaluate the way it provided transcription services in its radiology department. It identified four goals: increased productivity, decreased operating expense, finding much needed space in the radiology department and increasing employee morale. The department performs 165,000 procedures annually, with 66 radiologists, 29 faculty, and 37 residents and fellows on staff. Six FTEs comprised the transcription pool in the radiology department, with transcription their only duty. Transcriptionists were paid an hourly rate based on their years of service, not their productivity. Evaluation and measurement studies were undertaken by the hospital's management systems engineering department. The transcriptionists' hours were then changed to provide coverage during the periods of heaviest dictation. The productivity level of the transcription staff was also measured and various methods of measurement reviewed. The goal was a pure incentive pay plan that would reward employees for every increase in productivity. The incentive pay plan was phased in over a three-month period. Transcriptionists were paid for work performed, with no base pay beyond minimum wage. The move to home-based transcription was planned. The necessary equipment was identified and various issues specific to working at home were addressed. Approximately six months later, the transcriptionists were set up to work at home. The astounding results achieved are presented: 28% increase in productivity, operational cost savings exceeding $25,000 and a space savings of 238 square feet.

The quality of operation notes: can simple word processors help?

In a retrospective study the quality of operation notes on specifically designed proformas was compared to those produced by word processors and predesigned templates and notes in patients undergoing operations for colorectal cancer. The operations in the two groups were similar. Computer notes scored higher on all criteria, were all legible and took the same time to generate. The results were given to staff in an attempt to improve the quality of future notes. Prospective comparison demonstrated that the quality of computer notes improved even further, while that of manual forms did not change. Word processor computer notes are easy to set up, amend and produce but staff should be trained in the use of personal computers and machines must be readily available in operating theatres. Word processor databases will, in addition, enable the data collected to be analysed automatically to yield information for audit and research.