Method for production of medical records and other technical documents

Abstract

A computer implemented system for the production of medical records, legal documents and other frequently produced semi-technical documents. This is accomplished by generating an intelligent computer-guided interview and the use of serialized scriptable objects. Major program elements include a knowledge base text file, a parse engine, and an execution module. The knowledge base uses a unique rule syntax. The parse engine converts the textual knowledge base file to a compiled binary representation which can then be interpreted by the execution module. The execution module leads the user through the interview by generating a series of questions and presents possible answers in the form of pick lists. The data is recorded with a computer pen and collated into a document file. This file is coded in binary format and can be written to and recalled from disk. If the file is recalled from disk the user can continue to answer questions or change answers previously given. The result is a mobile computing system whereby data is input in a structured format. When the program is executed the user is prompted for answers to questions and, based upon the user's response, the final document can change considerably. Depending upon each answer, the program may change the subsequent questions being asked, change the list associated with a question, change the text being generated, or change the entire structure of the document. The data collected may be also be stored in a database for analysis at a later date. Finally, the data collected is output in a narrative text format which can be tailored according to the traditions and expectations of the user's profession. The program will output the text via a printer or it can be transmitted via electronic means.

Claims

We claim:

1. A document generation system for the purpose of creating technical or professional documents comprising:

a. a mobile computing device incorporating a user interface device for entering data in various locations, said data entered pursuant to a menu driven user interface wherein choices can be selected quickly so as to make use of said mobile computing device;

b. wherein said menu driven user interface is a guided interview wherein: the user is asked a series of question posed by said system and the subsequent questions asked vary in response to answers previously given;

c. a programmable and reusable knowledge base which is distinct from any execution program accessible by said mobile computing device such that, said programmable knowledge base is defined by a script language such that the user can perform modification of the knowledge base resulting in alteration of the final document; and

further comprises a navigation means for routing or guiding the user through the interview.

2. The system of claim 1, further comprising:

a bi-directional inference means for producing medical records, the bi-directional inference means containing the following elements:

multiple variable elements which may contain single or multiple variable values;

variable states which track user information;

rule elements which execute various actions such as screen display, text output, user prompts and pick lists depending upon the values of the variables; and

inference tables which reset or infer the values of multiple variables based upon the input value of a single variable;

the bi-directional inference means modifying the knowledge base such that the final document may be altered in a forward direction by entering answers to the questions asked in the guided interview and in a reverse direction by entering a medical diagnosis.

3. A document generation system for the purpose of creating technical or professional documents comprising:

a. a mobile computing device incorporating a user interface device for entering data in various locations, said data entered pursuant to a menu driven user interface wherein choices can be selected quickly so as to make use of said mobile computing device;

b. wherein said menu driven user interface is guided interview wherein: the user is asked a series of questions posed by said system and the subsequent questions asked vary in response to answers previously given;

c. a programmable and reusable knowledge base which is distinct from any execution program accessible by said mobile computing device further comprising a drawing means whereby additions in the form of bit mapped images may be:

displayed on the screen;

edited by the user; and

incorporated into the final document produced.

Description

FIELD OF INVENTION

This invention relates generally to the production of technical documents and more particularly to the production of electronic medical records by mobile health care workers.

BACKGROUND OF INVENTION

As the entire world struggles through the transition from the industrial to the information age, the need for information and document management is felt throughout the workplace. Medicine can be considered a prototypical field where that need is acutely felt. In medicine, as in many other fields, the increasing risk of litigation, the presence of third party payers, and the increasing complexity of the work have lead to a mountainous burden of paperwork.

Modern patients lament the fact that physicians don't spend as much time with them as Marcus Welby would have. Few realize that most doctors spend at least as much time on paperwork as they spend directly with a patient for any given encounter. Superficial, cursory physician/patient encounters are not acceptable but unfortunately the work of documentation must be done and we cannot turn back the clock. Our only hope of improving this situation is to make the documentation process more efficient.

Structured Documentation

Through the years, individuals have sought to improve the process of documentation in several different ways. Kempster (U.S. Pat. No. 3,960,634, 1976) sought to simplify the process by adding adhesive sheets to an existing record, thus adding small amounts of information to a preexisting record in an incremental fashion. Sonsteby (U.S. Pat. No. 4,865,549, 1989) attempted to add structure to the documentation process. She utilized color coded adhesive labels to organize the document and eliminate omissions of crucial data elements.

The concept of structured documentation is not new. It should be considered simply a way to practice good medicine. It is often estimated that in medicine the correct diagnosis is revealed in the patient's history as often as 90% of the time. A structured interview is simply a way of enhancing the quality of care by reminding the physician not to omit important details. An example of structured documentation is noted in one of the earliest textbooks still in widespread use today (The Early Diagnosis of the Acute Abdomen, Sir Zachary Cope, 1921, 15th ed., pg 21). The concept is still being refined and improved today (Wrenn, et al.: The use of structured, complaint-specific patient encounter forms in the emergency department. Annals Emergency Medicine, May 1993; 22:805-812). Unfortunately these attempts all comprise written manual methods. Attempts to incorporate information age technologies to bear in the medical profession have thus far been less than successful.

Visual Images

Others sought to improve upon the documentation process by adding visual representations to their documents. Assuming "a picture is worth a thousand words" these individuals felt that they could improve the documentation process and avoid long winded descriptions. Abramsom (U.S. Pat. No. 3,951,062, 1976) accomplished this by using a rubber stamp which consisted of an outline of a particular anatomic structure. Rees (U.S. Pat. No. 4,869,531, 1989) accomplished the same goal by using adhesive backed stickers with drawings of anatomical structures that could be mounted on the chart. In either case the physician could then add the particular elements of the individual case to the outline or scaffold drawing. This could be accomplished in a relatively accurate fashion without the need for great artistic talent on the part of the user.

Mobile Workers

Fuller (U.S. Pat. No. 4,038,666, 1977) attacked a different problem. How could documentation be performed by a mobile worker? His device, known as a "portable medical data recorder" consisted of a roll of paper on two rollers encased in a device with a window for writing. When the paper in the window was filled, the paper could be advanced and a fresh writing surface exposed.

Dictation/Transcription

All of these were manual methods that required written input on the part of the user. Documentation was still very time consuming and produced only a physical paper record. Eventually some modern technological methods were applied in order to increase the efficiency of the professional user who needed to produce documentation. The most important of these was dictation and subsequent transcription. We are not aware of any patents relating to standard transcription methods even though they are in widespread use. The method is quite obvious i.e. a tape recorder and a typist. Jachmann and Sweet (U.S. Pat. No. 5,146,439, 1992) developed a more advanced form which they described as "digital dictation" but this was a technical innovation that had little impact on the individual performing the dictation or on the final document produced.

Dictation/transcription systems are in widespread use today. These systems have several drawbacks:

1) Compared to writing or typing, there is a clear time saving for the individual doing the dictating. However, this is simply represents a cost shifting mechanism. Transcribing is still very labor intensive and skilled transcriptionists are still in short supply in most areas.

2) Similar scenarios require repetitive duplicate dictation on the part of the dictating user. Some have attempted to overcome this drawback by combining transcription with the use of "boiler plate" methods. A boiler plate is where a variable is inserted in the context of some preexisting text. This is seem in the mail merge functions of many word processors.

3) Transcription does not even come close to the dream of the "Electronic Medical Record" so often discussed among health care planners. There are three reasons for this.

a) Transcription still produces a paper document in most cases. Although most modern transcription centers use computerized word processors the transcribed documents are almost always printed to paper before there are reviewed or distributed. The electronic file produced by the transcription software is rarely transmitted or archived in its electronic state. The paper document is often reproduced by photocopying or transmitted by facsimile machine and this requires additional time and effort. If one then wishes to transmit or store the paper document by electronic means, it must first be converted, usually by means of an optical scanner. This adds a redundant step to the work.

b) Even if they are scanned into electronic documents, most documents are simply scanned as a graphical image not as text files. Optical character recognition is an additional step that still requires an operator to proofread and correct the document.

c) Even if the document is scanned, converted and stored in a text based format, it is still very difficult to retrieve data for outcome studies (very important for health planning). This is because the data is not input in any sort of systematic format. The same physical symptom could be known as chest pain, chest pains, substernal pain or pains in the chest. Complex search engines must be utilized to extract any meaningful data from free text documents. This can be very labor intensive as well.

Computerized Medical Records

As computers have become common place over the past decade it is only logical that they would be utilized to address these problems. Buchanan and Dowdle (U.S. Pat. No. 5,148,366, 1992) developed "computer-assisted documentation system for enhancing or replacing the process of dictating and transcribing". They themselves described their technique as a "boiler plate" methodology. It is much like the process commonly described as "mail merge" in word processing systems. Buchanan and Dowdle (U.S. Pat. No. 5,267,155, 1993) later improved upon their system by adding a relational database.

Computerized Speech recognition

One of the leaders in the field of document production over the past few years has certainly been Kurzweil AI, Inc. Kurzweil Al has been a dominant market player in the field of medical record production. Investigators at Kurzweil hold at least ten patents. Nine of these relate specifically to voice recognition apparatus which is the main strength and emphasis at Kurzweil. Only one patent by Goldhar (U.S. Pat. No. 5,101,375, 1992) relates to the actual structure or production of the document itself and this only in a limited way. His patent related to the use of a database table lookup scheme for providing binding and capitalization of text strings in a report. This is a function that is accomplished in different ways by other programs that provides narrative output, e.g. grammar checking programs. The methods developed at Kurzweil have certainly been very innovative and useful for those who can utilize speech recognition in their work. However there remain several shortcomings to Kurzweil's and other speech recognition systems.

1) Speech recognition is suitable for radiologists, pathologists, and attorneys who do their work at a fixed desk but is not suitable for mobile workers. Thus for these workers who move about constantly, speech recognition introduces the same redundancy and duplication into the flow of work as standard dictation. Mobile workers must take notes and then return to a central dictating station to dictate a formal record from their notes.

2) The advantage of speech recognition is that the document is produced immediately but a disadvantage is that the speaker must speak much slower than with standard dictation. This time consuming drawback can often offset the other advantages of the system.

3) The currently available systems available from Kurzweil and others offer speech recognition as their only advantage. These systems actually utilize the same basic boiler plate approach advanced by Buchanan and Dowdle. The voice recognition systems are used to simply fill in the blank holes in preprinted text. Many documentation needs in the world today are simply far too complex for such an approach.

Expert Systems

As computers became more powerful and sophisticated, expert systems were developed and applied to a variety of problems. DeLano (US DEFPUB T998008, 1980) developed a method for decision tree analysis. This involved the calculation of probabilities and weighted decisions. Gangarosa, Patrick, Futtu James M., and Green Andrew S. (U.S. Pat. No. 4,835,690, 1989) developed an integrated expert system for medical imaging, set-up, and scheduling. This system was used for organizing x-rays. It had nothing to do with producing medical charts. Skeirik (U.S. Pat. No. 4,920,499, 1990) developed an expert system with natural-language rule updating. Spiece (U.S. Pat. No. 4,798,543, 1989) used an expert system to perform interactive training in the military. Bodick, et al. (U.S. Pat. No. 4,945,476, 1990) developed an interactive expert system and method for creating and editing a knowledge base for use as a computerized aid to the cognitive process of diagnosis. This program was used to assist in making diagnoses in a pathology depart by means of cataloging and accessing case features and photographs. Little, et al. (U.S. Pat. No. 5,359,509, 1994) applied an expert system to the process of health care payment adjudication of claims and payments.

More recent inventions include Loftin, Wang, Baffes and Hua's (U.S. Pat. No. 5,311,422, 1994) development of general purpose architecture for intelligent computer-aided training and Mahajan (U.S. Pat. No. 5,404,528, 1995) who developed a scripting system. Although there is a great deal of interest in the application of expert systems in the medical field, most of this interest is directed toward using such systems to arrive at a diagnosis. We are aware of no case where an expert system has been used as an aid to the rapid production of technical documents.

Automated Data Entry

Sokolski (U.S. Pat. Nos. 3,800,439, 1974 and 3,900,961, 1975) developed a document reading, data input technology known as the SCAN-TRON.TM. method of data entry. This method comprises a specially designed paper form used for the purpose of collecting data for computerized analysis. Scan-Tron.TM. sheets are filled in by the individual being surveyed. The user must fill in a small box or circle with a pencil mark. A common use for Scan-Tron.TM. sheets is in the grading of standardized examinations. These data entry sheets can then be read by a special electronic device and the data fed directly into a computer. This bypasses the expensive step of a human data entry clerk. The disadvantage to these scannable sheets is that they are not at all fast or expeditious for the original user. There are two reasons for this. One, the Scan-Tron.TM. answer sheet must be a distinct second sheet of paper. The user must look back and forth between the two sheets and be very cautious not to lose his place. The other reason the use of these sheets is very slow is that the small boxes or circles must be filled in very precisely. Generations of students have been warned that if the are not cautions and diligent in the filling out of their answer sheets their grade will suffer. This degree of attention to detail is time consuming.

Patented Pen Computer Processes

One other patent is of some interest. Roach and Hollander (U.S. Pat. No. 5,310,997, 1994) secured a patent for an automated order and delivery system. Although this patent comes from a totally different field, it is an example of a computerized process utilizing a pen based computer to improve the flow of work in a given industry.

Emergency Medicine--State of the Art

The state of the art of documentation in the emergency department was recently reviewed in a monthly newsletter directed toward emergency department managers and directors (ED Management, American Health Consultants, Nov. 1995). This article begins by pointing out that "producing medical charts quickly and thoroughly has always been a challenge for emergency physicians". They proceed to review the currently available technology. Standard dictation/transcription is in widespread use. However chart turnaround time is often too slow and tracking and editing charts is cumbersome. These authors note that voice recognition is improving, but remains too slow for widespread acceptance. Voice recognition systems in use today are uniformly combined with template systems. This same ED Management article points out that a Smithsonian Technology award was recently granted to a company called TRANSQUICK.TM. (Atlanta, Ga.) for developing a methodology to smoothly toggle from computerized speech recognition to a standard medical transcriptionist.

The authors of this review then turned their attention to what they felt was the best currently available computerized documentation system, EMstationm.TM. (Datamedic Clinical Systems, Waltham, Mass.). EMstation.TM. is a template driven system that incorporates many useful reminders for billing and reimbursement. It also manifests a very creative use of the Windows GUI with regard to the use of color and placement of items on the screen. However, at its core this product remains essentially a template driven system. Any template system is still limited to the initial document structure provided by the template. The user simply fills in the blanks. These systems can be customized but this still requires someone; the user, designer, or domain expert, to envision each and every potential scenario in advance. Each and every potential scenario or combinations of scenarios must be programmed as a separate template. Thus, to use medical examples it would be logical to create a template for nausea and vomiting as these two complaints are logically related. However, the complaints of ankle sprain and shortness of breath have no logical connection although they are both very common. Creating templates for combinations of even two or three common complaints would be essentially impossible.

Shortcomings of Currently Available Systems

Other drawbacks of currently available template computerized charting systems are noted.

1. These devices are designed to enhance reimbursement. They are structured for physician data entry. They make no allowance for shared data or data entry by less experienced and/or less highly paid individuals.

2. Whatever inference might be present in these systems is a function of the structure of the template. As such it is unidirectional, that is to say, it begins with the patient's complaint. We are aware of no other software system which allows for bi-directional inference, i.e., working backward from a diagnosis. This corresponds to the that way doctors often think when they produce charts, especially if the charts are produced after the fact.

3. These systems are desktop which have been modified and customized for different types of input such as mouse and pen. The result is a system that often requires such hybrid input in order to function effectively. These programs are not designed for use by the mobile worker with data input from a pen only.

4. A corollary to number 3 above is that the currently available systems too often require keyboard entry to complete or modify the data. Many systems depict the user typing on a laptop computer. Again this is not suitable for the mobile worker.

5. Those emergency department, medical record systems which do a better job of facilitating data entry, such as CLINPLEX.TM., produce documents in the form of an outline or telegraphic speech. These documents are similar to the short hand notes people make to themselves in the course of their work. Handwritten doctor's notes have long been considered unintelligible. This is true for two reasons. One is the tradition of sloppy handwriting. The other is the shorthand and technical abbreviations. Because of these two problems, dictated notes are now considered the standard of practice for insurance reimbursement and legal documentation. Computer systems that produce abbreviated text output only address half of the problem.

6. Currently available systems generally do not incorporate any reminder function to improve the quality of care provided. The authors of the ED management review point out that a computer system should be able to help remind a doctor not to prescribe penicillin if the patient is allergic. Some systems do incorporate some reminders but they are still template based. For this reason these reminder functions when they do exist don't really respond to the actual unique data input such as a patient's drug allergy. They simply respond to the current template.

7. In the systems currently available the number of user definable fields is usually quite limited and not programmable by the end user. Consequently there is no means for the user to collect much personally useful data for the purpose of outcomes analysis.

Background Summary

The authors of this recent review of currently available technology (ED Management) conclude, "there is no cutting-edge technology in transcription . . . we're not even close to that." Furthermore in regard to template systems such as EM STATION.TM. they conclude "The one deficit I see with all of these systems is they are motivated by improving reimbursement . . . ED's will need something to make them more efficient, improve outcomes and keep them out of trouble". These authors seem to view currently available computer systems as less than successful. They then turn their attention to a review of other methodologies for improving the documentation process. These include the use of court stenographers and dictation assistants.

Thus the reader will see in the following sections that the present invention combines some features that have been known. These features are part of the process of producing documents, managing information and practicing medicine. The reader will also see that the present invention, Computer-assisted Documentation (CaD), succeeds in computerizing these processes, combining a vast array of useful functionality into one smooth process. CaD furthermore adds new functionality to produce a unique and powerful tool to create technical documents.

SUMMARY AND OBJECTS

Accordingly, the overall object and advantage of our invention is to assist the professional or technical expert in producing documentation or records on a daily basis. There is a great need for a more efficient means of managing information and producing records. Our invention is a means of automating the production of such documents. This is accomplished by several means as outlined below.

Speed and Efficiency

It is, therefore, the first major object of this invention to provide a method to produce documents more efficiently.

Rapid Documentation

One advantage of our invention is that mechanically it is simply a faster way to record information. It is faster because data is recorded by quick taps of a pen on a computer screen. These pen strokes are translated to words or sentences in the completed document. Older methods such as writing or dictating are much slower because they require the operator to generate each and every word individually.

Decreased Number of Steps

It is a further object of the invention to increase the efficiency of the process by reducing the number of steps required by eliminating unnecessary and redundant procedures. The means by which this is accomplished is outlined in more detail in the operation of the preferred embodiment.

Shorthand Documentation

It is yet a further object of the invention to increase efficiency of the documentation process by providing the expert user a means of recording data in a shorthand form. When menu selections are made, this can produce various forms of text output. Knowledge base rules can be used to generate grammatically correct phrases and sentences from a single pen click. The inference function of the program can be programmed by the domain expert so that single choices produce longer text fragments. These text fragments can contain variables which are inferred or defaulted according to the user's personal preference. An example might be the description of some surgical procedure that is frequently performed in the same manner or a document such as a will which is frequently produced in a similar fashion.

Expert Extender

It is still a further object of the invention to provide increased efficiency by enabling an expert extender (i.e., nonexpert) to perform some expert functions in a more accurate way. The knowledge base rules can be written so as to incorporate the knowledge of the domain expert into the computer program. The knowledge base can be programmed at the user level in much the same way that a spread sheet program can be written for a specific application. This knowledge base can then be parsed into a reusable program. The nonexpert individual is then guided through the interview process in a structured and controlled fashion simply by answering the questions as they appear on the computer screen. Thus our invention has the advantage of increasing efficiency by freeing the expert to spend more time in decision making, i.e., actually using his/her expertise. The expert user is relieved of redundant documentation tasks and assistants can be utilized in a more cost effective manner. One example of the application of this feature might be to allow less technically proficient individuals to staff a computer help desk.

Multiple Simultaneous Documents

It is still a further object of the invention to provide increased efficiency by allowing the user to produce multiple documents concurrently. Although a user can actually only work on one document at a time, incomplete documents may be saved to disk and later restored. Once a document is restored, the user can move forward and continue to answer questions or go back and change answers previously given to questions. This can be accomplished with written records but is very difficult to do with speech recognition. This requires extra specific effort to communicate to the transcriptionist what changes are to be made. Thus our invention has the advantage of better harmonizing with the way some individuals really work. This feature is essential for certain professions such as emergency medicine where physicians always care for multiple patients simultaneously.

Multiple Contributors to the Same Document

A corollary object and advantage of the invention is to allow a means for multiple workers to contribute to different parts of the same document. Examples of this function include medical records where technicians might record the vital signs and doctors and nurses might each make their own contributions to the record. In the production of standardized accounting report different individuals might each make their own contributions to the final document.

Bi-directional Inference

It is still a further object of the invention to provide increased efficiency by the use of a concept known as "bi-directional inference". Some of the efficiency of CaD is due to the process of inference, i.e., the program allows the value of a variable to be set or determined by the previous selection of another variable. An important and unique feature of the CaD software is that inference can operate in a reverse direction. That is, the reverse of the normal flow of the document and reverse of the apparent thinking process of the user. This is an essential component of the CaD program. It is particularly useful in the expert mode of the preferred embodiment.

In the medical setting, particularly in the emergency department, a physician is able to make a correct diagnosis much more quickly he or she might be able to document the reasoning processes. Bi-directional Inference allow the user to enter some "bottom line" information, in this case a diagnosis, and the CaD system will infer appropriate information throughout the document. This inferred information can then be edited by the user. This process dramatically increases the efficiency of the CaD system. This function is absent in template based systems. As a simple example of how this works consider the fact that a diagnosis of a "heart attack" almost always implies that an EKG was obtained and demonstrated a "heart attack". A template which starts from the beginning complaint of "chest pain" can make no such assumption.

Furthermore complaint based templates can never be specific enough. A chief complaint of "shortness of breath" may mean asthma in a child, anxiety in a young adult but is very likely to signify a heart attack or other heart disease in an elderly person. The advantage of bi-directional inference according to the preferred embodiment is that it corresponds to the way physicians often actually perform their documentation. A physician may arrive at a diagnosis very quickly. The medical record should reflect how the physician arrived at that diagnosis.

Electronic Format

It is still a further object of the invention to provide increased efficiency by producing documents in an electronic format. Handwritten, checklist and most dictation systems used today produce paper documents. There is a great deal of talk and interest in the "electronic medical record". Although most transcription systems use a word processor and produce an electronic file, the far more common practice is to produce a paper document first. Then extra steps are taken to change the document back to an electronic format by scanning or faxing. This is usually necessary because with transcription the user cannot the edit the document in real time as it is being produced (typed). One more advantage of this invention is that because the document can be edited by the original producer in real time, it can immediately be saved or transmitted in electronic format. Thus this invention will allow users to move much closer to the dream of the "paperless office".

Accuracy and Professionalism

It is, therefore, the second major object of this invention to provide a method to produce documents which are more accurate, complete and professional.

Professional Appearance

An object and advantage of this invention is to provide a means of producing charts which are more professional in appearance than those produced by most other methods. High quality documents can be produced with transcription systems but this requires the extra steps of proofreading, editing and retyping. By the appropriate use of the rule syntax in the knowledge base, this invention allows the incorporation of the user's expertise and grammar rules to influence the structure and content of the document. The user's expertise can include the traditional way to organize a document according to the user's profession or according to local custom. Correct spelling and grammar can be built into the knowledge base in advance by the user. The result is a high quality textual report in narrative or prose form. Accurate, concise and readable documents are very important and useful in today's world with increased risk of litigation. This type of document is especially useful in the medical profession where reimbursement by third party payers is based upon the written medical record. Also, according to the preferred embodiment, the CaD program is used to produce accurate medical records. This is advantageous because doctors are particularly known for illegible hand writing. According to the preferred embodiment these documents are created in a narrative or prose format. This feature is useful in a wide variety of fields where the documents once produced, will be read by nontechnical persons.

Individualized Documents

It is another object of the invention to provide a method to provide documents which are more professional in terms of their variety and individuality. Unlike boiler plate methods which can be fast and efficient but result in a document which is very limited or even amateurish, this invention has the advantage of producing documents which are unlimited in their variety. This can be accomplished because the knowledge base syntax rules allow the users input to change the structure and content of the final document in several ways. A simple syntax for the nodes or potential branches in the knowledge base structure allows for essentially unlimited (billions) numbers of possible variations in the final document. In addition to this variability there are the dynamic changes that can occur at runtime. Depending upon the user's input several changes can occur. The subsequent questions can change. The list of potential answers to the questions can change. The text output can change or the actual structure of the document could change.

It is still a further object and advantage of the invention to provide a professionalism and variety in document production due to the fact that the knowledge base script language is not bound to the language or customs of a particular profession. It can be reused across many different fields.

Graphic Images

It is another object of the invention to provide increased accuracy by means of incorporating pictures and diagrams into documents. The advantage of this function is illustrated by the well known saying "a picture is worth a thousand words". The incorporation of visual or graphic images is difficult with any currently available methods. It can easily be accomplished with hand written records but hand writing is otherwise the most cumbersome and inefficient method available. Drawings cannot be easily incorporated into documents produced by checklists, transcription or newer speech recognition systems.

Increased Accuracy

It is another object of the invention to provide a method to produce more accurate documents. This is accomplished because the knowledge base can be programmed in such a way that the user is intelligently guided through an interview. The final document thus produced is more accurate and complete because the user is reminded not to omit certain questions. The user could also be reminded to take a certain action. At first glance it would seem that this function would be most useful to the nonexpert user. However, this ability would actually be of tremendous value to the expert user as well. In a review of the world's literature on the value of continuing medical education for physicians, Davis (JAMA, 1995) found that reminder systems were by far the most effective means of positively changing physician behavior. This invention has the advantage of enabling the expert user or supervisor to incorporate such reminder systems into the day to day work of the regular users.

Mobile Documentation

It is, therefore, the third major object of this invention to provide a means of documentation for the mobile professional. One advantage of this invention is that because it is portable, the user would not be tied down to a desk, Dictaphone or fixed computer but would be free to move about unencumbered. User input is collected via a hand held computer with pen input device and this computer can be connected to a printer through a wireless network.

Concurrent Documentation

It is, therefore, also an object of the invention to provide a means whereby this documentation can occur in real time while the user is actually performing his job rather than after the fact. If the user is forced to postpone documentation until some later time then there is increased risk of forgetting crucial bits of information before the documentation is completed. Real time documentation has the advantage of increased accuracy because the user has the pertinent facts readily available.

It is, therefore, the forth and final major object of this invention to provide a method of data collection and management. This is useful in two ways.

Data Collection for Outcome Analysis

It is, therefore, an object of this invention to provide a method of data collection for use in outcome's analysis. In our world today there is a great deal of pressure to measure outcomes and evaluate the effectiveness of what we do. The most difficult aspect of this in most situations is the actual collection of adequate data. Usually this requires hiring someone to retrospectively review the appropriate documents. An alternative method is to take a dictated record and scan it into electronic format using optical character recognition. A complex search engine is then utilized to extract any meaningful data. Both of these methods are expensive and time consuming. Many times the records in question do not even contain the information being sought. None of the currently available methods of documentation are very good at organizing and retrieving data. Hand written and free text dictation offer no control over the data elements. These methods depend solely upon the operator to include the proper information. Check list systems with preprinted forms are the best currently available attempts to address this problem. These systems are fast, efficient and potentially complete at collecting data. Retrieving the data is another matter. These systems still require a human reviewer/data entry clerk. The alternative is to use a scannable answer sheet which requires very careful attention to detail. In this case the advantage of speed and efficiency is lost because these sheets cannot be filled in quickly.

The advantage of this invention is that each data element is structured in the form of a pick list and stored as a variable. The appropriate data elements are collected in real time as the work is being performed. The user is prompted for data elements he might otherwise have forgotten. Although this program is not based upon a database or internally structured as a database, an interface allows the user to specify that the contents of any variable be translated to a database field. Thus the user can decide which data elements are to be collected and saved for future analysis. This information can then be analyzed retrospectively using commercially available database programs.

A further advantage if the present invention is that the number of user definable fields is unlimited. Variable (field) names can be created by the domain expert/knowledge base programmer. The data collection and analysis process can be individualized.

Data Collection for Scientific Research

It is, therefore, an object of this invention to provide a method of data collection for use in performing scientific research. Once a user decides what questions should be asked, these questions can be programmed into the knowledge base and asked repeatedly in the appropriate context. Thus computer assisted documentation has rather obvious application to field research in sociology and politics. Perhaps less obvious is the advantage of this methodology in such fields as clinical medicine where the research is extraneous to the ongoing task of patient care. In this situation the program could prompt the user to ask the appropriate question or remind the user that a given patient is a suitable candidate for a research protocol. In either case the necessary data elements would be recorded during the course of the patient's care and would be retrievable, if necessary, for research purposes.

Ultimate Object

The overall object and advantage of this invention is to liberate the expert to greater utilization of his or her expertise. Freed from the drudgery of documentation these individuals could use their time more effectively. Perhaps the clearest advantages can be seen in health care which has so preoccupied the public eye in the past few years. Patients idealize the physicians of the past and their caricatures such as Marcus Welby and the Norman Rockwell paintings. Patients lament the fact that today's physicians seem too rushed and busy to spend enough time with them. However, these physicians of the past were never depicted spending hours on paperwork. This is the crucial difference. By utilizing this invention physicians could spend more time with patients. Thus physician's time could be utilized more efficiently and health care costs should decrease.

Up until now people have only talked and dreamed of the "paperless office". Others have utilized complex, redundant technologies in an attempt to convert paper documents to an electronic format. CaD represents a quantum leap forward in the technology of record keeping and document management. The advantages of this invention are by no means limited to health care applications. This invention can be effectively utilized wherever an expert individual is called upon to record and interpret data which is then recorded in the form of individual documents. This invention is of particular advantage in any situation which requires a relatively structured interview.

DRAWING FIGURES

FIG. 1 is a flow chart illustrating an overview of the process of computer-assisted documentation.

FIG. 2 shows a schematic of wireless LAN transmission between the two computing devices.

FIG. 3 depicts a typical system configuration for a desktop computing device.

FIG. 4 depicts typical system configuration for a hand-held computing device.

FIG. 5 is a flow chart which illustrates the parse routine whereby a binary knowledge base file is produced from an ASCII knowledge base file.

FIG. 6 is a modular flow chart depicting the components of the run-time execution routine.

FIG. 7 is a schematic representation of the various types of persistent objects.

FIG. 8 illustrates the six divisions of the knowledge base text file.

FIG. 9 is a schematic representation of the LIST object structure.

FIG. 10 is a schematic representation of the TEMPLATE object structure.

FIG. 11 is a schematic representation of the VAR object structure.

FIG. 12 is a screen snapshot demonstrating the numeric data entry grid.

FIG. 13 is a schematic representation of the NODE object structure.

FIG. 14 is a flow chart which illustrates various node relationships, including looping, branching, and return of control.

FIG. 15 is an example of a naming convention used in the preferred embodiment.

FIG. 16 is a schematic representation of the inference object structure.

FIG. 17a is the first of six pages an abbreviated knowledge base (KB) file illustrating many aspects of the knowledge base syntax.

FIG. 17b is the second page of the abbreviated knowledge base file.

FIG. 17c is the third page of the abbreviated knowledge base file.

FIG. 17d is the fourth page of the abbreviated knowledge base file.

FIG. 17e is the fifth page of the .abbreviated knowledge base file.

FIG. 17f is the sixth page of the abbreviated knowledge base file.

FIG. 18 is a screen snapshot which demonstrates features of the user interface.

FIG. 19 is a screen snapshot which shows features of the CaD specific File menu.

FIG. 20 is a screen snapshot which shows features of the CaD specific Chart menu.

FIG. 21 is a screen snapshot which illustrates a pick list and text window with a menu bar and tool bar.

FIG. 22 illustrates the edit buttons for the text edit window and text mode.

FIG. 23 is a screen snapshot which shows the login dialog box for name and password entry.

FIG. 24 is a screen snapshot which shows the login/logout buttons as well as other features.

FIG. 25 is a screen snapshot which illustrates the most basic use of the input and output of a variable and the selection of the chief complaint in the preferred embodiment.

FIG. 26 is a screen snapshot which depicts example of input selection from a pick list for more complex manipulation of lists.

FIG. 27 is a screen snapshot which depicts another example of input selection for complex list handling as illustrated in FIG. 26.

FIG. 28 is a sample of a document produced which demonstrates complex variable manipulation using the .backslash.SETVARFROMSUM and VARCONVERSE commands.

FIG. 29 is a flow chart which illustrates the potential for dynamic changes based upon user input.

FIG. 30 is a screen snapshot which depicts input selection of a chief complaint of neck injury.

FIG. 31 is a screen snapshot which depicts context specific mechanism of injury choices presented to a user after selecting a complaint of neck injury.

FIG. 32 is a screen snapshot which depicts input selection of a chief complaint of laceration.

FIG. 33 is a screen snapshot which shows a context specific question, "Which seat in the car?", presented to a user after selecting auto accident as a mechanism.

FIG. 34 is a screen snapshot which shows a context specific question, "Where is the laceration located?", presented to a user after selecting laceration as a complaint.

FIG. 35 is a screen snapshot which depicts context specific mechanism of injury choices presented to a user after selecting a complaint of laceration.

FIG. 36 is a series of screen snapshots which depict altered text output demonstrating context specific grammar changes.

FIG. 37 is a sample document demonstrating context specific changes in the structure of the document.

FIG. 38 is a sample document demonstrating expert mode with personal preferences selected by inference.

FIG. 39 is a screen snapshot which depicts two alternate means of text input, a free text annotation and a pen driven input keyboard on the screen.

FIG. 40 is a screen snapshot shows an example of a drawing template.

FIG. 41 is an example of a simple but complete document according to the preferred embodiment.

FIG. 42 is a screen snapshot which depicts an example of data recalled in database for analysis.

FIG. 43 is a flow chart which illustrates how multiple factors contribute to flexible document structure.

FIG. 44 is a flow chart which illustrates steps to generate a medical record document in a hospital emergency department according to the preferred embodiment.

FIG. 45 is a flow chart which illustrates conventional steps to produce an electronic medical record.

                        REFERENCE NUMBERS
     Ref.
       #    Description                             FIG. #
      100   complete CaD program                    1, 44
      102   User, e.g. physician                    1, 6, 43, 44,
                                                    45
      104   Interviewee, e.g. patient               1, 44, 45
      106   structured interview process            1
      108   user input                              1
      110   document produced - end result          1, 6, 44, 45
      112   binary knowledge base file              1, 5, 6, 44
      114   real time retrieval (SCT files)         1
      118   user editing process                    1
      120   text file storage - TXT files           1
      122   archival storage - TXT, SCT, DB files   1, 6
      124   data analysis                           1
      130   hand-held computing device              1, 2
      132   desktop computing device                1, 2
      134   wireless LAN transmitter/receiver       2, 3, 4
      140   central processing unit - CPU           3, 4
      142   read only memory - ROM                  3, 4
      144   random access memory - RAM              3, 4
      146   mass storage device - hard disk drive   3, 4
      148   keyboard input device                   3
      150   mouse input device                      3
      152   computer monitor - display screen       3
      154   printer device                          3
      156   LCD screen/digitizeddigitalized input device 4
      158   pen input device                        4
      200   knowledge from domain expert            5
      202   syntax language definitions             5
      204   ASCII knowledge base file               5
      208   parse engine                            5
      210   persistent objects - POs                5, 6
      212   binary code format definitions          5
      214   parse module of CaD editor program      5, 6
      216   execution module of CaD editor program  6
      218   knowledge base manager                  6
      220   view manager                            6
      222   display manager                         6
      224   object manager                          5, 6
      226   inference manager                       6
      228   reusable document file - SCT file       6
      231   chart (document) persistent object      7
      232   section persistent object               7
      233   node persistent object                  7
      234   text persistent object                  7
      235   rule persistent object                  7, 43
      235   node rules                              43
      236   list persistent object                  7
      237   variable persistent object              7
      238   variable set persistent object          7
      240   Kbknowledge base syntax - [LISTS] division 8, 17a, 43
      242   Kbknowledge base syntax - [TEMPLATES]   8, 17b
            division
      244   Kbknowledge base syntax - [VARS] - variable 8, 17c
            division
      246   Kbknowledge base syntax - [NODES] division 8, 17d
      248   Kbknowledge base syntax - [INFERENCE]   8, 17e, 43
            division
      248   Inferences                              43
      250   Kbknowledge base syntax - [SECTIONS]    8, 17e, 43
            division
      252   list name                               9, 17a
      254   list sort flag                          9, 17a
      256   sample list/list contents               9, 17a
      262   template name                           10
      264   template subvariables                   10
      268   VAR name                                11, 42
      270   variable attribute - value              11, 42, 43
      272   variable attribute - state              11
      274   variable attribute - longname           11, 17c
      276   variable attribute - associated list    11, 17c
      278   variable attribute - default value      11, 17c
      280   variable attribute - help text          11, 17c
      282   variable attribute - type               11
      284   modal dialog data entry form - MR #     12
      286   numerical entry buttons                 12
      288   numerical display box                   12
      290   "OK" - enter button - MR # modal dialog entry 12
      292   node - name                             13
      294   node execution - line by line           13, 43
      300   nodal program control - straight transfer 14
      302   nodal program control - conditional branch 14
      304   nodal program control - return of control 14
      306   nodal program control - transfer to     14
      310   typical 3 letter abbreviation           15
      312   compound name syntax                    17a, 43
      312   node name syntax                        43
      314   commonly accepted abbreviation          15
      316   inference table - input variable        16, 17e
      318   inference table - output variable       16, 17e
      334   [NEW] example of a list keyword         17a
      338   menu text.backslash.output text         17a
      340   subvariable name line                   17b
      342   example of a simple rule in a node      17d
      344   template definition                     17c
      354   variable - TYPE = MULTSELECTABLE        17c
      356   variable - TYPE = OMITTABLE             17c
      358   variable - TYPE = INFER_ONLY            17c
      360   variable - TYPE = RATCHET               17c
      362   variable - TYPE = NUMERIC_ENTRY         17c
      363   variable - TYPE = DRAWING
      364   variable - TYPE = FILENAME              17c
      366   list of nodes to executed as a section  17e
      370   simple text output                      17d, 43
      372   .backslash.n = insert carriage return   17d
      374   %VAR% variable substitution in text fragment 17d
      376   VAR = insert variable value             17d
      378   routine node redirection                17d
      380   IF rule                                 17d
      382   ELSE rule                               17d
      384   SETTEMPLATE variable set reassignment   17d
      386   VAR .backslash.NODISP variable option   17d
      387   .backslash.ARTICLE                      17d
      388   .backslash.ABORT action                 17d
      390   .backslash.TEXT action                  17d
      392   .backslash.SETVARIABLE command          17d
      393   .backslash.SETVARFROMSUM command        17d
      394   DO .backslash.action syntax             17d
      398   section - display flag                  17e
    400399  document title definition               17f
      404   title bar with file ID                  18, 21
      406   menu bar                                18, 21
      408   tool bar                                18, 21, 24
      410   prompt window                           18, 21, 24,
                                                    25, 36
      412   parse from text button                  18
      414   File menu - parse from text menu selection 19
      416   execute Chart menu - interview mode     20
      417   execute Chart menu - edit mode          20
      419   execute Chart menu - expert mode        20
      420   execute Chart menu - document sections  20
      422   execute (run) program (currently) loaded KB 21
      424   toggle between interview mode &and edit mode 21
      426   omit current question button            21
      428   open drawing window button              21
      430   list window with active pick list       18, 21, 24
      432   "OK" button for quick pen input         18, 21, 24,
                                                    25
      434   text edit window                        18, 21, 24
      436   edit button - text edit mode            18, 21
      438   standard windows scroll bar             21
      440   print file button                       21
      442   text edit button - green check = question 22
            answered
      444   text edit button - green question, optional 22
      446   text edit button - omitted question     22
      448   text edit button - red question, mandatory answer 22
      450   User menu - use to log in, open dialog  23, 24
      452   log in dialog box                       23
      454   log in button                           24
      456   log out button                          24
      458   d/c interview and return to edit mode   24
      502   pick list with chest pain selected      25
      504   text output of "chest pain" in chief complaint 25
      506   text output of "chest pain" later in HPI 25
      512   gastrointestinal symptoms include nausea 26
      514   respiratory symptoms include cough      27
      516   Chart section = "Review of Systems"     28
      518   example of .backslash.SETVARFROMVAR command 28
      520   example of VAR .backslash.CONVERSE command 28
      522   flow chart question - chief complaint   29
      524   flow chart question - mechanism of injury? 29
      526   flow chart question - which seat?       29
      528   flow chart question - location of laceration? 29
      530   flow chart question - laceration mechanism? 29
      532   chief complaint answer = neck injury    30
      534   mechanism answer = motor vehicle accident 31
      536   chief complaint answer = laceration     323
      538   which seat answer = driver's            332
      540   location answer = hand                  34
      542   mechanism answer = broken glass         35
      552   "occupation" pick list                  36
      554   variable longname prompt window         36
      556   selection = clerical worker             36
      558   selection = disabled                    36
      562   complaint #1 = twisted left ankle       37
      564   complaint #2 = right knee pain          37
      566   left knee exam - minimal                37
      568   right knee exam - detailed              37
      570   left ankle exam - detailed              37
      572   right ankle exam - minimal              37
      574   Dr. Lewis - younger doctor              38
      576   Dr. Lewis' preferences                  38
      580   Dr. Welby - older doctor                38
      582   Dr. Welby's preferences                 38
      584   text annotation, drawing window         39, 40
      586   keyboard overlay, pen input device      39
      588   drawing template                        40
      590   clear button                            40
      592   cancel button                           40
      594   "OK" (enter) button                     40
      598   record definitions (filenames)          42
      610   impact of user selections               43
      620   document structure                      43, 44
      632   medical history                         44, 45
      634   physical examination findings           44, 45
      636   diagnosis                               44
      638   clinical judgment of physician          44
      654   medical traditions                      44
      656   physician's past experience &and training 44
      658   local hospital rules                    44
      672   secondary user - RN                     44
      674   secondary user - technician             44
      676   minor historical details                44
      702   take notes                              45
      704   sit down &and dictate                   45
      706   transcription                           45
      708   dictated note                           45
      710   repeat similar information in each note 45
      712   information gathered by staff members   45
      714   fax transmission                        45
      716   optical scanner                         45