WO2000028459A2 - Method for automated medical outcomes data feedback, medical records integration, and healthcare provider response - Google Patents

Abstract

A method and computer apparatus for automated medical outcomes data feedback, medical records integration, and healthcare provider response. The process includes generating a medical record entry regarding a medical intervention, generating at least one question concerning the medical intervention, generating a message to the patient, generating a feedback to the questions, generating a feedback to the healthcare provider, or the healthcare provider staff, containing the generated feedback, matching an appropriate response action with the generated feedback, generating a new medical record entry, and generating a suggested response action in a compiled list for the provider or appropriate provider staff.

Description

METHOD FOR AUTOMATED MEDICAL OUTCOMES DATA FEEDBACK,

MEDICAL RECORDS INTEGRATION, AND HEALTHCARE PROVIDER

RESPONSE

Technical Field

This invention pertains to the art of methods and apparatuses for outcomes data collection, and more particularly to methods and apparatuses for automated medical outcomes data feedback. This application claims priority to a U.S. Provisional Patent Application Serial No. 60/107,894, entitled METHOD OF AUTOMATED MEDICAL OUTCOMES DATA FEEDBACK, MEDICAL RECORDS INTEGRATION, AND HEALTHCARE PROVIDER RESPONSE, filed on November 10, 1998. Background Art One known method for remote data monitoring is described in U.S. Patent No.

5,715,393 to Naugle. Naugle's method uses two computer systems that operate independently, but are linked in such a way that they can exchange electronic mail with each other. The computer systems automatically send mail back and forth between each other, which activates a monitoring program. The monitoring program generates an e-mail, which stimulates a particular medical action. One drawback of the Naugle patent is that there is no capability of interaction with human patients. Naugle requires two computers responding back and forth to generate status reports.

Another known type of automated reporting is described in U.S. Patent No. 5,704,366 to Tacklind et al. Tacklind discloses a system for monitoring and reporting medical information including a stand alone monitor for storing data records

(comprising measured values and time stamps) and for transmitting the records to a remote reporting unit over a communication system. The remote recording unit processes the raw data into graphs, generating chronological graphs. An important disadvantage is that although the remote recording unit generates reports, it does not take action, nor does it dispatch to appropriate receivers, depending upon the nature of the data or the processed information that it has received.

One other type of automated medical retrieval system is described in U.S. Patent No. 5,776,056 to Bu et al. Bu discloses a health self-checking system using a remote controller and a television and a method thereof, capable of easily checking a user's health using a remote controller and a television by checking user's heart rate. Bu discloses a monitor strapped to a patient's chest, thereby monitoring the various physical functions of the patient. Bu's data retrieval system merely relates the numerical physical responses from the patient to a computer at the hospital. The receiver of the information does not create any type of response.

One more type of medical automated data retrieval is described in U.S. Patent No. 5,633,910 to Cohen. Cohen discloses a patient monitoring system. An automated phone call system calls the patient at home and plays a list of questions that the patient has to respond to using the numerical key pad on their telephone. The Cohen patent is very similar to a telephone survey using a telephone keypad. Again, there is no response generated automatically on the other end after the reception of this data. Still another known type of computer based system for retrieving data is described in U.S. Patent No. 5,803,906 to Pratt et al. Pratt discloses a computer- based system for providing up-to-date health histories of animals, for example, in a feed lot. The Pratt patent uses an identification means around the animals' necks to identify which animal is at the feed trough. When the animal is identified the computer tells the feed lot which food and which medicine to release into the food trough for that specific animal. However, in the Pratt patent there is no response generated from the other end.

One final known type of monitoring system for outcome data retrieval is described in U.S. Patent No. 5,778,882 to Raymond et al. Raymond discloses a health monitoring system which tracks the state of health of a patient and compiles a chronological health history of the patient using a multi-parametric monitor which periodically and automatically measures and records a plurality of physiological data from sensors in contact with the patient's body. The Raymond patent does not allow for any generated response from the patient. Also, the data received from the patient does not automatically generate a response from the healthcare provider, or the healthcare provider staff. The present invention contemplates a new and improved method for automated medical outcomes data feedback. Thus, this invention is simple in design, effective in use, and overcomes the foregoing difficulties and others while providing better and more advantageous overall results. Disclosure of the Invention

In accordance with one aspect of the present invention a new and improved process for automated medical outcomes data feedback, medical records integration, and healthcare provider response, includes generating the following: a medical record entry regarding a medical intervention, at least one question concerning the results of a medical intervention, a message to a patient, an answer to the questions (intervention feedback), a reply to the healthcare provider, or the healthcare provider staff, containing the generated feedback, an appropriate matching response action to the generated feedback, generating a new medical record entry, and generating a suggested response action in a compiled list for the provider or appropriate provider staff.

In accordance with another aspect of the current invention, the knowledge base consists of the appropriate candidate questions for given interventions, the candidate provider response protocols for given feedback, and the rules of inference for reasoning by the automated generator of electronic mail messages and task lists in electromagnetic form as elements of at least one expert system computer program or document based equivalent.

In accordance with another aspect of the current invention, the message to the patient is either an electronic mail message or a hard copy mail message, including a plurality of questions, a plurality of choices for possible answers to each of the questions, a free text area for the patient to enter a brief message, and all necessary symbols, text or graphics necessary to successfully return the contents (outcomes feedback) to the point of message origin.

In accordance with another aspect of the current invention, the response action consists of one of the following: a phone call from the healthcare provider, or the healthcare provider staff, a follow up appointment, an electronic mail message, a pager message, a new or modified medication prescription, a letter, a postcard, or an end to the transaction. In accordance with yet another aspect of the current invention, the current invention is not limited to physiological parameters measured by machines.

According to still another aspect of the current invention, the process can be repeated multiple times. In accordance with yet another aspect of the current invention, a new and improved computer apparatus for generating automated feedback includes means for generating questions, means for sending a message to the patient, means for matching the generated feedback with an appropriate response action , and means for generating a list of response actions suggested for each provider staff member. One advantage of the present invention is that the speed of correspondence between a healthcare provider, or the healthcare provider staff, and the patient is significantly increased.

Another advantage of the current invention is that the patient has almost immediate access to the healthcare provider, or the healthcare provider staff, with any questions they may have.

Yet another advantage of the current invention is that the process, in dispatching the patient feedback, is completely automated. Even if the patient feedback cannot successfully be matched to a single recommended response action, but instead to multiple candidate response actions, the invention produces a list of candidate responses for the healthcare provider to choose from.

Still another advantage of the current invention is that the healthcare provider, or the healthcare provider staff, can now track medical outcomes that occur outside of the hospital or office. This advantage is both in regard to the collection of what occurred consequent to the intervention, and when it occurred. Still another advantage of the current invention is that the structured query and timely request for feedback provides more reliable data than that collected only after the patient returns to the provider's hospital or office, or with freeform questions and answers.

Still another advantage of the current invention is that the structure of the query provides a systematic approach to outcomes measurement that provides more consistency in measurement and thus increases the ability to compare outcomes across treatments and patients with more rigor. Still other benefits and advantages of the invention will become apparent to those skilled in the art to which it pertains upon a reading and understanding of the following detailed specification. Brief Descripition of the Drawings

The invention may take physical form in certain parts and arrangement of parts. A preferred embodiment of these parts will be described in detail in the specification and illustrated in the accompanying drawings, which form a part hereof and wherein: FIGURE 1 is a flow chart showing the inventive automated process.

The full process can be described as composed of two sub-processes, each generated automatically by components of the invented technology — the solicitation of feedback with a structured question and the mediation of provider response with a generated task list. Description of the Preferred Embodiment

Referring now to the drawing, which is for the purpose of illustrating a preferred embodiment of this invention only, not for purposes of limiting the same, FIGURE 1 shows a flow chart of the inventive process. The top of the flow chart represents the fact that the input for the invention and method is an entry into the medical record. A computer program of the present invention is preferably stored or loaded into a computer with access to the medical record data entry process. Those skilled in the art will appreciate that this input may be the result of a redundant fork of data input streams or of accumulated data gathered for updating records in batch mode. The direct entry of medical record data on input forms of this invention is not precluded; however, the inefficiency of entering data for this invention as a duplicate of other medical records systems makes that undesirable. The data input stream may be filtered or unfiltered. Unfiltered input would consist of all data entered in the medical record regardless of relevance to the invention. Such data would require an additional component of the invention in the form of program elements that would recognize data tagged as relevant to the invention. Which elements are of relevance will be explained shortly. Filtered input consists of data that has already had irrelevant data removed. In either case, the data is required to be of a syntax that delineates individual data elements. This syntax can be of a proprietary convention (comma- delineated character strings, quoted strings, etc.) or, preferably, a standardized convention such as Arden Syntax or Health Level 7 (recognized standards for those skilled in the art of medical informatics). Those skilled in the art of medical record keeping will recognize that there are at least four types of medical record entries: "subjective data" reported but not confirmed, "objective findings" in examination that presumably would be recognized and described similarly by any trained colleague, "assessments" made as a result of the findings and "plans" or treatments which are justified interventions. The last type may be regarded as different types of medical interventions, which may be further subdivided in classification. One type of intervention is a prescription of medication; a second is any type of physical therapy; a third is surgery; and there remain other types of intervention (i.e. wound debridment, counseling, psychological therapy, etc.).

Such medical record entries are often called clinical notes. These notes may include observations, assessments, and reports by the patient. The surgery, medical therapy, and other intervention entries are found as a description of the intervention, be it an action performed or a recommendation made. The medication entries are composed of drug, quantity and frequency, i.e., this drug for this length of time, this size pill, etc. Surgical records are typically accompanied by the anesthesiologist's report, including drugs and time intervals, as well as a log of monitor parameters, like oxygen levels in the blood, respiratory rate, heart rate, etc. Surgeons may simply give a description of the intervention, without including very many variables. It may be a detailed description of the procedure or a brief account. They might describe the intervention by reference to site and objective accomplished. For example, a certain intervention may be applied to the right collateral ligament of the knee, i.e., "repair of the anterior cruciate ligament in the right stifle joint. " Entries may be reduced to standard codes from recognizable vocabularies (e.g. ICD9CM, CPT4, or Read codes, etc. , which are recognized standardized vocabularies in the health care industry). Entries of relevance to this invention are those for which there is an expected result or outcome at some anticipated time interval. These will be largely, but not limited to, treatment and intervention. This invention will increase the number of cases for which outcomes may be measured by enabling the report of outpatient events that occur as a consequence of medical intervention that have been impracticable to acquire in the past with any systematic rigor. Furthermore, the invention will track and facilitate the workflow of provider responses to feedback for each and every medical intervention. Those skilled in the art will acknowledge the impracticality of doing this with a manual system of operation. Currently feedback in medical practice is largely limited to the provider notification of exceptional patient events only and precludes the capture of details regarding the outcome of the provider's interventions. Examples of details not captured include (1) precisely when the symptoms have abated, (2) when results are delayed, limited, partial or (3) when other non-threatening events occur, etc. The computer program of the present invention receives these medical record entries as input, represented by arrow 1 of FIGURE 1. This input is translated into a question to be asked of the patient. That translation is dependent upon the programs ability to recognize the interventions of the medical record data entry. All four types of medical entries noted above, would, in an ideal circumstance, have a code that coincides with third party payment. The code system that has been found to be most functional in current practice is the ICD9-CM code (The International Classification of Diseases: 9th revision, Clinical Modification (ICD-9-CM); 6th ed. Washington (DC): Health Care Financing Administration, July, 1998). Currently, a healthcare provider, or healthcare provider staff, has to submit bills to Medicare/Medicaid with ICD9-CM codes in order to get paid. The ICD9-CM code does not, however, fully cover everything the healthcare provider, or the healthcare provider staff, might want to write in a medical record. There are other standardized lists such as the "Physicians' Current Procedural Terminology" (4th ed. Chicago, IL: American Medical Association, 1999) (CPT4), which is tailored to label procedures in contrast to the diagnosis based ICD9-CM codes. There are also "Clinical Terms Version 3"

(England: National Health Service Centre for Coding and Classification, March, 1998) (Read Codes) and many others for specialty domains. In addition, there is a Thesaurus produced by the Unified Medical Language System^® (UMLS^®) project that tries to unify the standardized lists. This invention can use standardized terminology, as specified in these coding systems, or non-standard terminology. Where medical records employ standard terminology to record interventions, these codes can be used by the invention to map interventions to potential outcomes and feedback soliciting queries. Outcome data collected in terms of standard vocabularies will be preferable for comparability and wide scale analysis, but the invention is not limited to standard coding systems. Non-standard vocabulary systems can be supported at the expense of comparability, but have the advantage of flexibility. If medical records systems do not provide a parsimonious coded output suitable for mapping to expected outcomes and feedback soliciting queries then one of two alternatives are available. First, additional software can be used to translate free text record entries into standardized coding schemes. Second, rules for the expert system within the invention may be written to recognize non-standard, but frequently, used terms in the medical record. Ideally, the medical record system will provide data entry users with pick lists of recognized terminology to facilitate standardized intervention entry. In any case, the expert system of the present invention, takes coded input representing the medical interventions (arrow 1 of FIGURE 1) and generates queries to be e-mailed or otherwise sent to the patient to elicit outcome reports; one or more queries per intervention.

The healthcare provider, or the healthcare provider staff, must record a great deal of information into the medical record. At a minimum, the healthcare provider, or the healthcare provider staff, must record enough clinical history to justify the recommendation, the diagnostic choices, and the prescribed treatments. When the entry goes into the medical record system, the invention captures it. So, the current invention is kept aware of all entries to the actual medical record and is guaranteed sufficient information to drive the process of soliciting outcomes.

In the mind of the healthcare provider, associated with each medical intervention is a list of potential outcomes-some good, some bad-that can be expected to occur at some anticipated interval. For example, for a drug, the symptoms are expected to be resolved within a certain number of days. For a surgical procedure, recovery should occur in a certain number of days. For lacerations and incisions, the skin should be healed and the sutures, or staples, should no longer be required after a certain number of days. The whole practice of medicine is intended to resolve symptoms, and there are expected intervals for when that should occur. Naturally, the desired outcome is not always what happens. Prior to the eventual realization, all possibilities may be anticipated. For instance, after sewing up a laceration, the healthcare provider, or the healthcare provider staff, expects any one of the following: 1) the wound is healed and dry and ready for suture removal, 2) the incision has festered and remains unhealed, because the patient did not take care of the incision sight, or 3) the wound has swollen and the sutures are tearing out, allowing the laceration to reopen. So, there are a number of outcomes to expect, with some expected interval for each of these listed treatments, that correspond with regimented, or tabulated, entries to the medical record. The invention capitalizes on this relationship between interventions and anticipated outcomes to structure a request for feedback on the outcome of medical treatment (to fully describe the invention, the response to this feedback must be included as part of the full feedback loop supported, but discussion of the provider response will be left for later in the document).

The present invention embodies the association of intervention with expected outcomes in what those skilled in the art refer to as a knowledge base. This knowledge base can consist of a simple look-up table in which interventions are matched to outcomes or it can be more complex, e.g., a group of rules used by an expert system for deducing a list of possible outcomes. This knowledge base is stored or loaded into the computer housing the present invention (represented in FIGURES 1 by the "Expert System" block).

Since there is no code list that covers all possibilities, there is a need for the entry of free text in the medical record. However, such free text entries are infrequent due to the prolific use of codes for third party payment claims. Free text will become increasingly rare as omissions and inadequacies are discovered and added to the standardized vocabularies. Meanwhile, when free text intervention record entries are necessary, a brief collection of expected outcomes and intervals can be entered at the same time as any novel interventions are recorded. For example, the most common interventions can be listed, whether they are surgical or medical or otherwise, and a look-up table can be created, which would have the expected responses. If the invention does not find any listing of the medical intervention in the table or the expert system reports no recommendation, the program displays an input form to ask the user for a list of possible outcomes and when to expect them. These are then transformed into the syntax of suitable form to be added to the knowledge base and reviewed offline by human domain experts for possible permanent addition. Hence, the invention will either be capable of offering a recommendation for the feedback soliciting query based on it's knowledge of outcomes or ask for the expected outcomes explicitly, including the time interval. With the accumulation of more entries in the knowledge base over time, the program is enriched with more interventions for which it can offer a recommendation.

In the preferred embodiment, the invention can take two forms. One is electronic and the other is a paper hard copy. In FIGURE 1, the flow chart shows the feedback loop progression, which is the same for both e-mail and a paper hard copy. Functionally they are the same, and if a batch e-mail can be generated, then a string of postcards can also be generated, which the office personnel can stamp with bulk postage, drop in the mail (arrow 2), and thereby operate substantially the same as the e-mail. A written postcard would contain the same content, a parsimonious list of potential outcomes, one or more of which will be checked off and immediately dropped into the mail. Self-address features of the postcard fulfill all of the requirements for return postage and labeling (arrow 5). The outcome indicated on the returned postcard can easily be entered manually into data entry forms included in the invention computer program or read by a scanner capable of distinguishing which outcome was checked off by the patient.

In the preferred embodiment, the current invention is adaptable to any form of medical intervention, in any medical field. For example, surgical treatments in one kind of practice could make templates with the question structures all the same, or the same set of answers, but the question might look dramatically different between surgery of a podiatrist and a fertility practice. But the current invention sends out different questions depending on the medical intervention. The inventive program has a look-up table or expert system that associates appropriate questions with specific medical record entries as well as the set of potential outcomes for the intervention. This library, or knowledge base of associated interventions, questions, and outcomes, is constructed offline from such resources as published literature, actuarial data, outcomes analysis, and local clinical experience. It is a dynamic list of associations which is enlarged and identified by the use of the invention to collect feedback and outcomes. It may be substantially constructed from interviews of domain experts, either universally recognized specialists or merely local authorities. At this point in time, the elicitation of rules and protocol can be performed by one skilled in the art who understands the technology of rule-based expert systems and the details of this specific invention. As knowledge eliciting technology is developed to automate this capture of domain expertise and its transformation into rules, it may be employed to construct and maintain the knowledge base for this invention. For example, application of the Quinlan C4.5 algorithm and its derivatives for machine learning can be applied to accumulated data regarding medical interventions, recorded outcomes and provider responses to deduce rules for both query generation and response recommendations. Those familiar with the art recognize this algorithm as a systematic program for inducing rules suitable for expert systems from such data sets.

As stated earlier, these rules, or knowledge base, are stored or loaded on the computer housing the present invention. The computer either also stores an e-mail program or is connected to another computer that does. Those skilled in the art are familiar with e-mail services delivered over the Internet as requiring only messages (character strings) as input that conform to a specified syntax. Any such input received by such programs is delivered to the recipient of the e-mail account at the address of the addressee. Hence the delivery of the feedback query is provided by such a service outside the present invention. The present invention merely delivers a query packaged in a form compliant with the requirements for such e-mail services (arrow 2 in FIGURES 1). For example, the embodiment of the invention programmed in the Java language delivers the feedback query packaged in compliance with the Sun Microsystems' JavaMail API. This packaging involves the assignment of the patient as the addressee and the provider as the sender. The e-mail process time-stamps the message both at transmission and at receipt by the recipient's e-mail reading program. Similarly the reply will be time-stamped at the point of reply and provider receipt. This time-stamp is the simple result of the invention's computer program requesting the time and date from the operating system of the computer when the event occurs. This provides a record of the sending (arrow 2) and receiving (arrow 5) of the request for feedback and the patient reply that is included in the medical record (recording both the sending of the feedback request and the receipt of a reply).

Each request for feedback may refer to a single intervention or, alternatively, a compilation of all questions appropriate to ask on a given day from multiple interventions whose appropriate interval for feedback happen to fall upon the same day. That is, a separate e-mail or postcard may be used for each set of expected outcomes with a distinct time interval between intervention and expected outcomes or several questions to be asked on the same day may be combined into a single e-mail or postcard.

Henceforth in this document, to simplify descriptions, e-mail will be used to illustrate the technology without loss of generality regarding both electronic and paper mail. All statements regarding e-mail are fully intended to imply that either the same thing can be said of paper mail or that the same can be accomplished with a trivial extension as exemplified thus far. Recall that it was said that the full process can be described as composed of two sub-processes, each generated automatically by components of the invented technology — the solicitation of feedback with a structured question and the mediation of provider response with a generated task list.

The solicitation of feedback can be accomplished by a computer program that accepts as input the interventions described in standard or non-standard terms, which are matched in electronically stored tables or are patterns that trigger the rules of an expert system. This computer program is stored or loaded on a computer with access to the electronic medical record data entry and e-mail services. Tables may explicitly contain paired interventions and outcomes or, more efficiently, a database may store intervention/outcome associations in a relational database schema. An object-oriented programming approach would store these relationships as instances of a intervention- outcome relationship class in a context supporting efficient storage, retrieval, indexing and sorting, etc. An expert system can be used for more sophisticated reasoning in deriving the list of outcomes expected for a given intervention. In any case, the input to the program segment would be a representation of interventions and the output would be an exhaustive list of the expected outcomes. For instance, in an expert system, the interventions are represented as clauses stating necessary conditions of if- then rules to be fired in the automated process of reasoning from input to recommendations for feedback questions. In their simplest form, they might look like the following example, although more sophisticated rules embedding more complicated logic are possible and to be utilized.

(protocol

(intervention "Recommendation to watch for redness, swelling or other systemic symptoms") (question "Which of the following have occurred''' )

(answers "Nothing, no evidence remains of the tick bite I am still waiting to hear results of the blood work

"Nothing, no evidence remains of the tick bite I was informed that the results of the blood work were negative

"Redness remains where the tick was removed "

"The area of redness has grown larger than a dime where the tick was removed "

"Swelling has occurred or remains where the tick was removed "

"Redness and swelling both occurred or remain where the tick was removed "

"Other skin lesions have arisen not located at the site where the tick was removed "

"Any of the following symptoms have occurred malaise, fever, fatigue, headache, or other flu like symptoms

"Other symptoms that I did not expect have occurred "))

(protocol

(intervention "Recommendation to watch for three more days for redness, swelling or other systemic symptoms")

(question "What has happened over the last three days⁹' )

(answers "Nothing, no evidence remains of the tick bite I am still waiting to hear results of the blood work "

"Nothing, no evidence remains of the tick bite I was informed that the results of the blood work were negative

"Redness remains where the tick was removed "

"The area of redness has grown larger than a dime where the tick was removed "

"Swelling has occurred or remains where the tick was removed '

"Redness and swelling both occurred or remain where the tick was removed "

"Other skm lesions have arisen not located at the site where the tick was removed "

"Any of the following symptoms have occurred malaise, fever, fatigue, headache, or other flu- like symptoms

"Other symptoms that I did not expect have occurred "))

(protocol

(intervention "doxycyc ne lOOmg BID PO 10 days")

(question "Please select one or more responses concerning your condition ")

(answers "I have taken my medication twice daily every day for one week"

"I have taken my medication almost regularly for one week (missed < two doses)" "I have missed more than two doses but am still taking medication after 7 days" "I have stopped taking my medication within seven days of starting" "The redness/swelling under my arm has improved or disappeared after one week on doxycychne"

"The redness/swelling under my arm has worsened after one week on doxycychne" "One or more of the following symptoms malaise, fever, fatigue, headache, etc , occurred while taking doxycychne for one week"

"No change after one week of doxycychne" "Other - please explain ")) (protocol

(intervention "Recommendation to stay on doxycychne and we will send another request for feedback in 3 days") (question "Please select one or more responses concerning your condition ")

(answers "I have taken my doxycychne medication twice daily every day for 10 days"

"I have taken my doxycychne medication almost regularly for 10 days (missed < two doses)" "I have missed more than two doses but am still taking medication after 10 days" "I stopped taking medication before 10 days was up" "The redness/swelling under my arm has improved or disappeared (and there are no other symptoms) after 10 days on doxycychne"

"The redness/swelling under my arm has worsened after 10 days on doxycychne" "One or more of the following symptoms malaise, fever, fatigue, headache, etc , occurred while taking doxycychne for 10 days" "No change after ten days of doxycychne"

"Other - please explain "))

(protocol

(intervention "E-mail Congratulations") (question "We are happy to hear that you are doing so well Please call if there is anything else we can do for you We would be interested in knowing which of the following you think are true statements (indicate by marking with an X ")

(answers "I found this to be an easy way to give my doctor some feedback" "I think this communication helped me to get healthier faster"))

(protocol

(intervention "extend doxycychne lOOmg BID PO 10 days more") (question "Please select one or more responses concerning your condition ") (answers "I have taken my medication twice daily every day for 17 days" "I have taken my medication almost regularly for 17 days (missed < two doses)"

"I have missed more than two doses but am still taking medication after 17 days" "I have stopped taking my medication before 17 days was up'

"The redness/swelling under my arm has improved or disappeared (no other symptoms) in the last 17 days on doxycychne" "The redness/swelling under my arm has worsened in the last 7 days on doxycychne"

"One or more of the following symptoms malaise, fever, fatigue, headache, etc , occurred while taking doxycychne for an additional 10 days"

"No change during the last 7 days of doxycychne" "Other - please explain ")) Rules illustrated above are stored in a distinct file accessed by an expert system shell that resolves all the implications generated by the assertion of facts represented in the input (interventions). The action clauses of these rules represent the question and answers to be offered to the patient in requesting feedback. A simple parsing program can transform the recommended actions into the friendly content of an e-mail message.

Another group of rules provides the infrastructure of reasoning specific to the task of maintaining the proper recipient, email address, and generating e-mail compatible syntax for content. These rules are exemplified below. They are not necessarily kept in a distinct rule base, but it is advantageous to do so. This allows medical domain experts an unencumbered view of the protocols embedded (as exemplified above) in a separate file. This provides an inviting ease of maintenance, debugging, refinement and extension.

(deftemplate protocol "A query template for Jess."

(slot intervention) (slot question) (multislot answers) (slot free Text) (slot reply Index) )

(deftemplate recipient (slot name) (slot eaddress) )

(defclass afeedbackRequest IntegratedOutcomes .Query) create query and store

(defrule QueryCreateNStore

?ιnterv <- (intervention ?ι) (protocol (intervention ? )

(question ?qstn) (answers S'answrs)

) =>

(facts)

(bind ?fbRequest (new IntegratedOutcomes .Query ?ι)) (?fbRequest setQuestion ?qstn)

PfbRequest setAnswers S'answrs) (store afeedbackRequest ?fbRequest) (retract ?mterv) )

(load-facts protocols.txt) alter store

(defrule alter-store

?query <- (protocol (intervention ?n)

(question ?questn) (answers ?answ) ) =>

(printout t "about to fire alter-store rule" crlf) (modify ?*qry*

(intervention ?ιι) (question ?questn) (answers ?answ) )

(retract ?query)

(printout t crlf "alter-store rule firing" crlf) ) The expert system output is an e-mail message that can be provided as input to a program that packages and sends actual e-mail via the Internet (arrow 2 of FIGURE 1). A program that supports the Javamail API from Sun Microsystems would enable this operation. Thereby, any device that reads email (arrow 3) and supports replies (arrow 4) can be used by the patient to reply with feedback (traditional mail would be replied to by marking the appropriate box and dropping it back in the mailbox relying on pre-printed return postage). An extension of the method is possible with pagers that allow responses and other two-way wireless communications devices. The transmission of e-mail, coming and going, is mediated by standard Internet technology not unique to this invention. Other forms of two-way wired or wireless communication, as well as paper forms via postage or courier, are not excluded from implied feasibility.

This completes the description of the first sub-process, that of timely soliciting feedback from the patient regarding a specific intervention. What follows is a description of the technology implementing the second sub-process - that of mediating the provider's response to the feedback received.

The patient's reply is indicated in such a way that the indicated outcome is recognized by a parsing component of the invention's computer program. This can be as simple as recognizing a pair of characters between which the patient is asked to place another character (e.g., "...please put an 'x' in the space between the bars, " | I ", next to the most appropriate answer to the question..."). More sophisticated graphics and user interface devices (e.g., applets) are supported in e-mail with a growing list of MIME types supported by email technology. This provides additional means to accomplish the same function or similar steps in the invention. The structured responses are key to the reliability and validity of the feedback. More complicated response structures may potentially be used. However, diversity in user interface devices will detract from the comparability of feedback data and should be minimized. It is not the intent of this description to limit the scope of technology employed for this role in the invention.

The computer program of the current invention, stored or loaded in a computer, then parses the e-mail reply identifying the actual response of the patient (arrow 5 of FIGURE 1). If the program is confused or the content garbled in transit, the program can generate another copy of the request, perhaps with an explanation for the redundant request. The logic of the response can be evaluated programmatically with an extended rule base to assure the quality of the response. The output of this expert system consult would be a recommended step of action to be taken by a provider or provider staff member - a response action appropriate for the feedback regarding the medical intervention. The degree of certainty about the quality of the response as being a coherent reply from the intended person may be in question. Any specified level of uncertainty in the quality of the response (e.g., missing pieces of the e-mail, multiple outcomes indicated though mutually exclusive, erroneous patient information or erroneous date, etc.) can be used as a trigger for a regenerated request in the same fashion as a garbled reply. This process of qualifying the feedback as valid is not further described. However, it is seen as a generalization of the same content exemplified in the previous and ensuing description. Confirmation of a previous reply is simply analogous to the original feedback request.

A validated reply (arrow 5) then serves as the input to another consultation with either a flow-chart, look-up table, database, expert system or any other knowledge-based artificial intelligence system capable of hashing the intervention/actual-outcome pair into an appropriate response recommendation for the provider and his or her staff. Using an expert system as an example, the identified outcome is asserted in the working memory of the expert system shell. Any possible chain of inference leading to a response action is evaluated and recommended response actions form the output of the program. These actions are parsed into elements of a task list with appropriate staff identified as suggested actors. Thus, the ultimate output (arrow 6) is one or more task lists that can be viewed collectively or individually by role (e.g., nurse's list, receptionist's list, physician's list). This list can be used to guide and manage provider-side responses. For instance, the receptionist can sit down and use the receptionist's list to perform each item checking it off as it is aiding workflow management (arrow 7). Similarly the nurse can use the nurse specific list to follow up on all patients found in the nurse's list. The physician's list will consist of all those items specific for the professional as well as any items for which the expert system could not resolve a single recommended response (either there were multiple candidates of equal merit or no candidates). Any member of the provider staff can delegate a task on their list to another role as would commonly be done in natural practice (e.g., a nurse decides the doctor should actually make this call, the doctor decides the receptionist should simply call and make the patient another appointment, etc.). The invention facilitates workflow (by design) in any and all cases of intervention. This differs from technologies intended to simply alert the staff in cases of aberrant events or status that exceeds a specified threshold. Although such one way "notification" technologies could be accomplished on the same infrastructure, the design of this invention goes far beyond notification to complete the feedback loop to facilitate the provider's response, enabling outcomes to be recorded for all interventions normal and abnormal.

Examples of rules of protocol for provider response actions:

(respProtocol (feedback "Nothing; no evidence remains of the tick bite. I am still waiting to hear results of the blood work.") (actions "Ask again in three days"))

(respProtocol (feedback "Nothing; no evidence remains of the tick bite. I was informed that the results of the blood work were negative.") (actions "Email congratulations"))

(respProtocol (feedback "Redness remains where the tick was removed.")

(actions "Make follow-up appointment."))

(respProtocol

(feedback "The area of redness has grown larger than a dime where the tick was removed.")

(actions "Make follow-up appointment."))

The response action taken by the provider or provider staff (arrow 7) is in actuality another medical intervention. Whether it is a simple phone conversation, a change in medication, or a new appointment, it is an intervention worthy of a medical record entry. As such, this action serves as the next intervention (unlabeled arrow of FIGURE 1), triggering another cycle of anticipated outcome at an expected interval. Hence the invention is invoked to cycle around the feedback loop again and again until the feedback indicates a cessation of illness or symptom (or the patient dies despite expert care).

It is easy to gloss over the fact that the reply to the feedback request is naturally of a form easily added to the medical record, automatically (arrow 5a). Output of the expert system is stored or loaded into the computer program containing the medical records. Those skilled in the art will recognize the capacity to put both the feedback request and the feedback reply in the record. Furthermore, those skilled in the art will recognize that the contents of the feedback reply can be used as input to the medical record as text. Thus each reply is parsimoniously added to the record without human effort or time consumption. This is accomplished at the same time that the response action recommendation is dispatched to a task list. Therefore, it is present in the record when the provider or staff member makes contact with the patient and, in the medical record chronology, it precedes the ensuing intervention entry that starts the next feedback cycle.

Similar to the first sub-process, the protocols (rules in the case of an expert system technology) for inferring the response appropriate for a given intervention- feedback-request reply can be stored in separate files. These can be kept distinct from the other rules of domain specific administration for formatting response descriptions, assigning duties and dispatching the implicit response tasks in respective lists. As was the case in query formation, the separation is not required but facilitates refinement, debugging, maintenance, and extension. It abstracts the protocols that are of interest to the clinicians from the nuts and bolts of the knowledge-based system.

Examples of rules of non-protocol inference for provider response actions:

,, Response Protocol Rulebase

(deftemplate respProtocol

"A response protocol template for Jess ' (slot feedback) (multislot actions))

(deftemplate respAction

"A responseAction template for Jess " (slot feedback) (multislot actions)

(slot responselndex))

(defclass afeedbackResponse ResponseManager ResponseAction)

,, create ResponseAction and store

(defrule QueryResponseNStor

"if there is a feedback reply that matches a protocol, then create a feedback response, store it for java code, and delete the triggering fact" (enable by commenting out)

'response <- (feedbackreply 'patient 'intervn 'provider 'feedback) (respProtocol (feedback 'feedback) (actions $'acts)

) =>

(printout t "Firing QueryResponseNStor rule" crlf)

(bind 'fbResponse (new ResponseManager ResponseAction 'feedback)) ('fbResponse setActions $'acts)

('fbResponse setPatient 'patient)

('fbResponse setlntervention 'intervn)

('fbResponse setActor 'provider)

(store feedbackResponse 'fbResponse) (retract 'response)

(load-facts respProtocol txt) .these are removed by a reset ,, assign all appointment making to receptionist

(defrule AssignAppMakingToRecept

"if there is a feedbackResponse that involves making an appointment, then assign the Actor role to the receptionist"

(receptionist 'receptn)

'response <- (feedbackresponse (actions $'acts (contains $'acts "make appointment")))

) =>

(printout t "Firing AssignAppMakingToRecept rule" crlf)

('response setActor 'receptn)

(store feedbackResponse 'response)

Using the task list as a worksheet, the provider staff member is able to follow up on the receipt of feedback with the listed response action. If this transaction is performed in the presence of medical record access, medical record entries may conveniently be made (subjective and objective entries as well as assessments and plans) similar to entries made during or near a traditional office visit or exam. For example, the results of a phone conversation between nurse and patient may be entered into the medical record during or after the phone call. As the staff member marches through the list of tasks recommended, medical entries are made including the ensuing plan of action or recommendation. The reply to the feedback query is already present in the record and can inform the conversation with the patient. The response action results in a plan of action, recommendation or some adjustment in the medical case management, which implicitly is another intervention. As such, this medical record entry will spawn another cycle in the invention for generating a feedback request at the appropriate interval. Thus, each intervention initiates a feedback loop in which the inventive method first generates a feedback request and, second, dispatches the reply to the medical record as well as transforms it into a response action on a role specific task list. The cycle is perpetuated until either the patient reports complete resolve of the problem or dies. Example 1

This example shows a representative form that could be sent to a patient. E-mail or Postcard: Mr. Smith,

On [this date ] you were given a [name of treatment] treatment or prescription. It is common for the [name of treatment] treatment or prescription to result in the

[relief of the symptom] at about this time. To help us serve your medical needs better, would you please indicate how things actually turned out? the symptoms are now gone. the symptoms are going away. I am better, but not completely. the symptoms are unchanged. the symptoms have worsened.

Please let us know if the [medication, procedure, surgery, or other therapy] was successfully completed per doctor's instructions. successfully completed per doctor's instructions, but delayed. unsuccessfully completed per doctor's instructions. started, but not completed. not attempted.

Please indicate your desire for our response from this point nothing. I am now fine. I would like to talk to the nursing or PA staff. I would like to have a follow-up appointment. I specifically need to speak with the doctor.

If necessary, please use the following space for a brief message to the receptionist nurse PA doctor

Example 2

Consider a case of a laceration that was debrided and sutured and the patient is sent home with an antibiotic and one pain reliever. These three interventions generate three expectations, hence three questions, each with a set of possible results. There are three expected outcomes from a sutured laceration: 1) the wound heals, 2) it has not healed together completely, and is still red on the edges, or 3) the sutures have come out, and the wound is gaping open. The patient would check which of the three outcomes has happened. The two drugs would generate two more questions. Mr. Jones, on this date you were given this antibiotic. Can you tell me, 1) are the symptoms now gone, 2) are the symptoms going away, but not completely, 3) the symptoms are unchanged, or 4) the symptoms have worsened. Then a third question: you were given this pain reliever, and 1) are you still feeling pain, 2) feeling some pain, but it is tolerable, 3) all the pain is gone, or 4) the pain has been gone, and I stopped taking the pills before they were used up.

Example 2 represents three entries to the record, which generate three questions, in a multiple-choice format, with the range of possible outcomes expected. All the patient has to do is check the one that applies to them. It is possible, because of the timing, that the procedure described in Example 2, will generate three separate communications to the patient. The wound is expected to heal, and the sutures ready to come out, at one interval, which is different than the length of time to take the antibiotics, which is a little bit shorter, and both of those are different from the time to take the pain reliever, which would be shorter still. The patient will receive one e- mail or postcard for each point in time that the healthcare system would like to know the outcome.

If the means of communication with the patient is via hard copy mail, the healthcare provider, or the healthcare provider staff, will need to read the written reply and enter it into the computer. If it is a multiple-choice question, with one of the choices having been checked off, then all the healthcare provider, or the healthcare provider staff, has to do is pull up the medical record, and enter the information. The medical record could be primed with an entry form that would only require the click of a button. The current invention could also include a hybrid between sending an e-mail and a postcard. The same medical, practice could send some patients an e-mail and others a postcard. As is shown in FIGURE 1, arrow 5a, the response e-mail and the written, keyed-in response lead to the same new medical record entry. Either the response is being keyed in, or the e-mail is being translated 100% through automation, and the transaction will negotiate with the medical record system.

In the most preferred embodiment, the current invention contemplates using only e-mail to send the communication to the patient. Once the patient receives the e- mail, the patient can click on a button and return it to the healthcare provider, or the healthcare provider staff. The computer could even set up an appointment for the patient and have the computer send an e-mail saying, "You have an appointment on such and such a date." The program could automatically interact either with an appointment-book-aiding program, where the healthcare provider, or healthcare provider staff, actually makes the appointment, or it could interact with a fully automated appointment system. Some of the feedback that generate medical record entries will cause another iteration of expectation and feedback, for example, as in Example 1, if "symptoms are going away. I am better, but not completely," has been checked, it is duly noted in the medical record the appropriate healthcare provider staff member is alerted, and a subsequent e-mail or postcard is automatically initiated as a follow up, all automatically by the invention.

The current invention allows for an automated medical entry that is of an objective nature. The current invention allows direct medical record entry of patient feedback. Thus, the medical record entry can be of a subjective nature. In currently known systems, most of what the patient tells his physician is not written into the medical record. Thus, an immense amount of information is lost from the subjective clinical history. With more of the subjective clinical history captured by the automated system of the current invention, the healthcare provider, or the healthcare provider staff, has more information at their disposal to diagnose the medical conditions of their patients. The invention allows this information to be captured at minimal expense and burden to the provider work flow, and can be directly processed by programmatic aids or filtered in any way that proves to be clinically effective in patient health care. In addition, with known systems it is difficult to translate what the patient said, and/or what the patient meant. Also, it is difficult for the healthcare provider, or the healthcare provider staff, to make objective use of the information the patient provided in freeform. The current invention instead provides the patient with a structured question to which they give the healthcare provider, or the healthcare provider staff, a fixed answer. In addition to the multiple-choice questions the current invention includes an opportunity for the patient to add something to each answer. At the bottom of the communication, as shown in Example 1 , the communication asks if there are any other things the patient would like to add. The current invention could write the freeform text entered by the patient into the medical record automatically. Then, it is information that the healthcare provider, or the healthcare provider staff, could use in his defense, his arguments, his recommendations, or his comparisons in the future. Example 3

If the patient comes back in six months and says they do not feel very good, then doctor asks them if they have been limping. If the patient responds, "No, I haven't been limping, but I am not doing any better than I was six months ago," the doctor could go back to the record and see where the patient wrote, "I have been limping a lot this week, is there anything I should do about that. " The doctor could point that out to the patient and say, "Look, you responded six months ago that you were limping and now you say your not limping, so you are getting better." The automatic system allows the free text to be entered for free in the medical record as long as there is space. The patient could theoretically reply to the feedback query and that reply could be associated with more than one response action. For example, the appropriate response action for what they checked off could be a nurse phone call or a provider phone call. It could be either or both. This ambiguity of more than one recommendation will occur in one of two contexts. One type would be that it does not matter which choice is made, because either one will work. This could be resolved by, 1) putting the task on the list of the lowest paid employee so that the resources are maximized, or 2) putting the task on the list of the person who has the least to do. So, in essence the inventive system does resolve this case down to a single response action. If this is not the case and ambiguity remains, the response can be added to the provider's list of replies requiring professional judgement and mediation. The provider may, in turn, choose an action delegating the response to another staff member's task list.

The alternative is the situation where, in fact, more than one response action is appropriate. In that case, the program makes a list of unresolved items, and when the appropriate decision maker logs on to this portion of the program, the program can tell the decision maker, "You need to decide whether a nurse or provider calls back Mrs. Jones on this one problem. " In other words, what the program cannot figure out, it presents as questions for the human user in the office to resolve. Even if there are multiple responses that are sufficient, the current invention can turn the inventive process into a human aided process. Both elected decisions, from human intervention or from automation, will generate a response action, will make a work list. The work list could be the nurse's list of phone calls, or it could be the doctor's list of phone calls, or it could be a list of appointments to make for the office staff. And at that point, the response action could be automated.

These are examples of rules that can embody policy chosen by the health care staff. Unresolvable scenarios are presented to appropriate staff for final decisions for which a response action is to be taken. For instance, the program may not be able to resolve where a patient reporting equivocal progress should receive a phone call from the nurse, a phone call from the physician, or a follow-up office visit. The decision concerning the appropriate response action is placed on the appropriate staff members task list (triage nurse or physician) and processed along with other human interventions. Once the decision-maker resolves the choice of an appropriate response action the task is dispatched to the actor. The invention incorporates programmed rules and those rules could be customized to a particular practice. For example, the program would always pick a certain person, and could even pick a certain person depending on the day of the week. For example, on Monday, Wednesday, Friday, one person, and on Tuesday and Thursday, a different person. It would be dictated by office policy, and implemented in the program.

The end result of the response action is one or more of the following: 1) a new clinical note medical record entry regarding patient's status, accompanied by a modified expectation for future outcomes 2) a new or modified prescription 3) communication or instruction with or without modified treatment plans, or 4) a referral to other healthcare specialists, or 5) a scheduled re-exam. Each of these involves a medical record entry and represents an intervention with an associated expectation or outcome. The only event that does not spawn a response action is a patient feedback indicating that all is well. Such patient feedback is informative, because it brings closure to the interactive cycle of the expectation-feedback-response. The current invention conveniently documents the resolution of medical problems with minimal effort. In the absence of the current invention, medical records fail to capture such information economically and with timely precision. Example 4

Imagine a patient who experiences a delay in resolution of symptoms. A feedback request is received and the patient indicates the outcome describing a partial resolution of symptoms at the interval for which symptoms are usually gone. The response action usually taken in the practice of medicine and embedded in the knowledge base of the current invention is to continue therapy and request more feedback at a later time. With the current invention in place, the feedback reply indicating a partial or delay of symptom resolution might suggest simply sending email to the effect of continuation of therapy. Showing up on a provider staff member's task list, the recommendation of sending email would be reviewed and approved. The program of the current invention can generate an e-mail message conveying this instruction and inform the patient to expect another feedback request at an appropriate interval. Example 5

In another case, imagine the expert system of the current invention cannot find any question or outcomes to build a query recommendation for the medical intervention. The program of the invention that is stored or loaded on a computer gives notice of this inability at the time of medical record entry and presents an input form to the provider that asks the provider to enter an appropriate question and the outcomes he knows of that are possible. This input is then transformed into the proper format and added to the knowledge base of the current invention. Example 6

Another possibility is that the appropriate action is to do nothing, the transaction is ended. So, it is the obvious termination of this loop of constant regeneration. It would be natural, if the postcard or the e-mail gave the opportunity for the patient to check off 1) everything is just fine now, or 2) symptoms are gone, or 3) symptoms are gone, that resolves the problem, thank you for your help, to end the transaction. In the most preferred embodiment, the e-mail reply by the patient will be almost instantaneous. The incentive to respond is greatly enhanced by quick action on the part of the health care system. Similarly, the health care system will find the current invention equally useful in quickly communicating between the healthcare provider, or healthcare provider staff, and the patient, and better meeting the medical needs of their patients. Closer monitoring of expected outcomes provided by this invention is anticipated to result in urgent intervention if things do not go as planned, which in turn will prevent more severe outcomes. These benefits will increase the incentive of both parties to participate. Quick action on the part of the healthcare system is made possible by timely feedback. The systematic nature and structure of the patient feedback improve the utility and credibility of the feedback.

The current invention is going to help practitioners fulfill outcomes reporting regulations in areas that they currently cannot. Presently, healthcare providers, or healthcare provider staff, can only measure outcomes that happen inside the hospital (mortality is a lone exception). There is no effective systematic widely used mechanism for measuring outcomes of events that happen outside the hospital. The hospital can measure if the patient came back in to get the sutures removed, but if the sutures are absorbed and removal is unnecessary, there is no effective means to find out if healing went as expected. If the patient is discharged with medication, there is no efficient mechanism to get answers for the following questions: 1) when did the patient stop taking the medication, 2) what day did the symptoms go away? With the current invention, these various events that occur outside of the hospital can now be measured expeditiously and with less ambiguity than the recollection reported at the patient's next visit.

The current inventive process generates the following: a medical record entry regarding a medical intervention, at least one question concerning the medical intervention, a message to a patient, a feedback from the patient to the questions, a message to healthcare provider containing the generated feedback, then matching an appropriate response action with the generated feedback, generating a new medical record entry containing the feedback, and generating a response action list.

In the descriptions thus far, there has been only one response action by the provider for every feedback query sent to, and received from, the patient. It is also possible for the feedback to spawn more than one provider response. For instance, the outcome may warrant a phone call from the provider and the resulting conversation warrant that a follow-up appointment be made with no significant interval or email interposed. The response action can include one or more of the following: a provider phone call, a follow-up appointment, an e-mail or postcard, or an end to the transaction. The provider phone call can generate a second response action, which can include one or more of the following: a follow-up appointment, an e-mail or postcard, writing on a prescription, or an end to the transaction. The message to the patient, in the preferred embodiment, includes a free text area for the patient to enter free text regarding the medical intervention, or any other questions they may have. This free text area can generate a third response action, which can include one or more of the following: writing a new prescription, contacting the patient, sending another e-mail, setting up a follow-up appointment, or end to the transaction. The three sections of the query in Example 1 correspond to three possible responses. The iterative process of intervention, feedback and response can be repeated as many times as necessary, and each time include as many response actions as fitting.

The current invention allows the healthcare provider, or the healthcare provider staff, to capture the decrease, increase, or cessation of a symptom, for example, pain, vomiting, diarrhea, cough, etc. These particular symptoms are not capable of being measured by other means when they occur outside of a hospital.

The current invention is not limited to the medical field. It could be used in a wide number of applications. Any situation where an automatically generated request for structured feedback is required, this invention is relevant. This invention could be used in any field where an inquiry is generated and the feedback initiates a response action specified by articulated rules that are capable of being embodied in a production rule or flow chart, or any other form of artificially intelligent programming. The current invention is not limited to a current computer system, but could be in included in any artificial intelligence programming.

The current invention could also be coupled with current, or future technology involving writing and refilling prescriptions via the Internet. The current invention could also be coupled with an automatic phone call, where the phone call is generated automatically and continues to call until the patient responds. The current invention could also acknowledge the receipt of the phone call and the response of the patient. It would be a matter of routine skill to select an appropriate conventional computer system and implement the claimed process on that computer system. The invention has been described with reference to a preferred embodiment. Although the invention has been described with reference to the preferred embodiment illustrated in the attached drawing figures and examples, it is noted that equivalents may be employed and substitutions made herein without departing from the scope of the invention as recited in the claims.

Having thus described the invention it is now claimed:

Claims

CLAIM

1. A process for automated medical outcomes data feedback, medical records integration, and healthcare provider response, the process characterized by: a) generating a medical record entry regarding a medical intervention; b) generating at least one question concerning the medical intervention; c) generating a message to a patient; d) generating a feedback to the at least one question; e) generating a message to a healthcare provider containing the generated feedback; f) matching an appropriate response action with the generated feedback; g) generating a new medical record entry documenting the feedback; h) generating a new medical record entry documenting a response action; and, i) generating a first response action.

2. The process of claim 1, wherein generating a message to a patient further characterized by: generating an electronic mail message.

3. The process of claim 1 , wherein generating a message to a patient further characterized by: generating a hard copy mail message.

4. The process of claim 1 wherein generating a message to a patient further characterized by: providing a plurality of selections for possible answers to each of the questions; and, generating a free text area for entering free text.

5. The process of claim 2 wherein generating a message to a patient further characterized by: providing a plurality of selections for possible answers to each of the questions; and, generating a free text area for entering free text.

6. The process of claim 3 wherein generating a message to a patient further characterized by: providing a plurality of selections for possible answers to each of the questions; and, generating a free text area for entering free text.

7. The process of claim 1, wherein generating a first response action further characterized by: generating a provider phone call.

8. The process of claim 1, wherein generating a first response action further characterized by: generating a follow-up appointment.

9. The process of claim 1, wherein generating a first response action further characterized by: ending the transaction.

10. The process of claim 7 wherein, after generating a first response actions, the process further characterized by: responding to the questions of the patient regarding the medical intervention; and, generating a second response action based on the patient's feedback.

11. The process of claim 10 wherein generating a second response action based on the patient's feedback further characterized by: generating a follow-up appointment.

12. The process of claim 10, wherein generating a second response action based on the patient's feedback further characterized by: writing a prescription.

13. The process of claim 10, wherein generating a second response action based on the patient's feedback further characterized by: ending the transaction.

14. The process of claim 11 wherein generating a follow-up appointment further characterized by: automatically generating a follow-up appointment.

15. The process of claim 4, wherein, after generating a free text area for entering free text, the process further characterized by: analyzing the free text; and, generating a third response action based upon the free text.

16. The process of claim 5, wherein, after generating a free text area for entering free text, the process further characterized by: analyzing the free text; and, generating a third response action based upon the free text.

17. The process of claim 6, wherein, after generating a free text area for entering free text, the process further characterized by: analyzing the free text; and, generating a third response action based upon the free text.

18. The process of claim 15, wherein generating a third response action based upon the free text further characterized by: writing a prescription.

19. The process of claim 16, wherein generating a third response action based upon the free text further characterized by: writing a prescription.

20. The process of claim 17, wherein generating a third response action based upon the free text further characterized by: writing a prescription.

21. The process of claim 15, wherein generating a third response action based upon the free text further characterized by: contacting the patient.

22. The process of claim 16, wherein generating a third response action based upon the free text further characterized by: contacting the patient.

23. The process of claim 17, wherein generating a third response action based upon the free text further characterized by: contacting the patient.

24. The process of claim 15, wherein generating a third response action based upon the free text further characterized by: sending an electronic message.

25. The process of claim 16, wherein generating a third response action based upon the free text further characterized by: sending an electronic message.

26. The process of claim 17, wherein generating a third response action based upon the free text further characterized by: sending an electronic message.

27. The process of claim 15, wherein generating a third response action based upon the free text further characterized by: making a follow-up appointment.

28. The process of claim 16, wherein generating a third response action based upon the free text further characterized by: making a follow-up appointment.

29. The process of claim 17, wherein generating a third response action based upon the free text further characterized by: making a follow-up appointment.

30. The process of claim 15, wherein generating a third response action based upon the free text further characterized by: ending the transaction.

31. The process of claim 16, wherein generating a third response action based upon the free text further characterized by: ending the transaction.

32. The process of claim 17, wherein generating a third response action based upon the free text further characterized by: ending the transaction.

33. The process of claim 1, wherein the process further characterized by: repeating the process multiple times.

34. The process of claim 10, wherein the process further characterized by: repeating the process multiple times.

35. The process of claim 15, wherein the process further characterized by: repeating the process multiple times.

36. The process of claim 16, wherein the process further characterized by: repeating the process multiple times.

37. The process of claim 17, wherein the process further characterized by: repeating the process multiple times.

38. A computer apparatus for generating automated data feedback, the apparatus characterized by: a) means for generating at least one question; b) means for generating a message; c) means for matching an appropriate response action with a generated feedback; and, d) means for generating a first response action.

39. The apparatus of claim 38, wherein the message is an electronic mail message.

40. The apparatus of claim 38, wherein the message is a hard copy mail message.

41. The apparatus of claim 38, wherein the message characterized by: a plurality of selections for possible answers to each of the questions; and, a free text area for entering free text.

42. The apparatus of claim 39, wherein the message characterized by: a plurality of selections for possible answers to each of the questions; and, a free text area for entering free text.

43. The apparatus of claim 40, wherein the message characterized by: a plurality of selections for possible answers to each of the questions; and, a free text area for entering free text.

44. The apparatus of claim 38, wherein the apparatus is directed to a process for automated medical outcomes data feedback, medical records integration, and healthcare provider response, and the first response action consists of one of the following: a) a provider phone call; b) a follow-up appointment; c) an e-mail to the patient; or d) an end to a transaction.

45. The apparatus of claim 44, wherein the apparatus further characterized by: means for generating a second response action based on the patient's feedback to the first response actions; and, means for generating a third response action based on the free text.

46. The apparatus of claim 38, wherein the apparatus further characterized by: means for repeating a process for generating automated data feedback.

47. A computer apparatus for automated medical outcomes data feedback, medical records integration, and healthcare provider response, the apparatus characterized by: a) means for generating a medical record entry regarding a medical intervention; b) means for generating a plurality of questions concerning the medical intervention; c) means for providing a plurality of selections for possible answers to the questions; d) means for generating a free text area; e) means for generating an electronic mail message to the patient; f) means for receiving a feedback to the questions; g) means for matching the appropriate response action with the generated feedback; h) means for generating a new medical record entry; i) means for generating a first response action, the first response action consisting of one of the following: 1) a healthcare provider, or healthcare provider staff, phone call;

2) a follow-up appointment; and,

3) an end to a transaction; j) means for generating a second response action, the second response transaction consisting of one of the following:

1) a follow-up appointment;

2) writing a prescription; and, 3) an end to the transaction; k) means for generating a third response action, the third response actions consisting of one of the following:

1) writing a prescription;

2) a phone call from the healthcare provider, or the healthcare provider staff;

3) an electronic mail message;

4) a follow-up appointment; and,

5) an end to the transaction; and,

1) means for repeating a process, for generating automated medical outcomes data feedback, medical records integration, and healthcare provider response, multiple times.