SHARED KNOWLEDGE BASES FOR CLINICIANS

Prof. (Dr.) Rao G. Nanduri

Guest Lecture

Abstract:

The practical experience gained by clinicians in their day-to-day practice of medicine becomes their real primary knowledge after finishing their education and starting their clinical practice. This knowledge grows rich day by day. This knowledge combined with the secondary knowledge on medicine derived from books, journals etc (this was their primary knowledge during medical education) make them wiser and more efficient in their profession. The more practical experience a physician has, the more the number of patients do prefer to consult him/her for diagnosis and treatment. The professional efficiency of a clinician is therefore based primarily on his/her own practical clinical experience supported by knowledge gained from various other sources. Every day is a separate experience by itself for all the doctors in a large teaching hospital as they come across many types and varieties of patients almost every day. Such cases, if they belong to extraordinary type, atypical nature, unusual and rare disorders, complex and complicated ailments which are difficult to diagnose and treat, are usually discussed among all the doctors working in that hospital in monthly clinical meetings in the hospital; but unfortunately, that experience, a very rich source of knowledge, stays limited to the clinicians of that hospital only. Not even 5 percent of those cases are published in domestic or international journals. Imagine the situation when all the doctors of one hospital learn about all the interesting patients treated in all other hospitals in their city, also those from the hospitals of the whole state and even those from the whole country if not the whole world. It is certain that those doctors exposed to such sources, would gain more knowledge, become more efficient in their profession and do earn more laurels to themselves, to their organization and society than otherwise. Suppose, the doctors from all hospitals in a country share such experiences of theirs with the doctors of all other hospitals through an advanced high-tech medium on a topic of common interest, which can be shared at frequent intervals with more such past experiences, like any professional journal published once in a quarter or twice in an year, how does it influence and effect the working conditions of those hospitals and their quality standards in healthcare delivery ? A very novel method of distributing such sharable knowledge bases of different specialties in an uniquely designed format of presentation loaded on a CD or Flash Drive with an embedded interactive software to read the presentations and query them to obtain intelligent answers has been designed by the author. This paper describes in detail the necessity to prevent wastage of the said vast sources of practical knowledge, the shared knowledge bases which are separate for each specialty, their practical utility in daily practice of medicine and the problems that are likely to be experienced while implementing this novel concept.

Introduction to Shared Knowledge:

Sharing of knowledge and experience between any two or more people is a well known accepted method to enhance knowledge. The people in medical practice are on constant urge to improve their clinical knowledge and professional skills in diagnosis and treatment through various ways. If they do not do so, they know very well that their career, income, name and moral standing are at stake; and they should wake up in time to avoid the perils of inadequate knowledge. The usual methods for updating their knowledge and enhancing their professional efficiencies are reading the latest books, journals, proceedings of the conferences of respective subjects, attending and partaking in clinical meetings, group discussions, tele conferences, postmortem examinations, critical analysis of all undiagnosed, mismanaged, atypical, complex and complicated patients and those with diseases of unknown origin etc. Out of these, the periodic clinical discussions and critical analysis that take place in each hospital on all cases of clinical interest are the most useful to an upcoming clinician as those are based on actual happenings in the hospital. This material is a rich source of knowledge and it grows richer every day as more and more cases are added to the list of discussed cases. 'Learning from Past Experiences and Becoming Wise from Past Mistakes' are the well known methods to acquire knowledge since ancient times.

When the data collected from the cases discussed in various clinical forums is stored in an organized manner in the form of database files, which are called in the computer jargon, 'A Knowledge Base' and its retrieval by the user in different formats if made possible through an artificial intelligence software, it makes the foundation for a vast building to come up above which houses many more such Knowledge Bases. These Knowledge Bases are multipurpose knowledge banks and serve the needs of multiple users i.e. clinical researchers, undergraduate and postgraduate students, teaching faculty, practicing clinicians, statisticians, hospital administrators, state level and country level healthcare planners, public health staff, clinical auditors, world health bodies etc.

Most of the large teaching hospitals, medical institutes, other hospitals recognized for postgraduate training, hospitals affiliated to medical research labs and hospitals of Armed Forces, are in the habit of regularly conducting clinical meetings. Some of the hospitals do conduct such meetings separately for each specialty. Unfortunately, not much importance is given to these meetings by the clinicians and administrators; as a result, all the clinicians working in the same hospital too are not in a position to utilize such a good opportunity available to them free of cost. Even otherwise, the contents of these discussions remain with the staff of those hospitals as very few hospitals do have the habit of inviting the clinicians of other nearby hospitals to attend these clinical meetings and take part in the discussions. Not even 5% of these discussed cases are sent for publication in medical journals. Thus, a very rich source of knowledge remains obscure to majority who needs it and in the strict sense of the facts, it goes waste.

For an undergraduate student in the medical sciences, the primary knowledge is the text books and journals and one acquires secondary knowledge from the patients in the hospital where he/she learns the practical lessons. The status of these two types of knowledge are reversed when a medical student becomes a fully fledged clinician; the practical experience one gains in the clinical practice becomes the primary source of knowledge and his/her text book cum journal knowledge stays as a back-up support in the form of secondary knowledge. This is the very reason for people preferring an experienced doctor to the highly qualified but little experience. This is true for all specialties in medicine. The clinician's primary knowledge grows with the years of his/her experience and the professional efficiency increases along with the primary knowledge supported by his/her secondary knowledge. The primary knowledge of one clinician becomes the secondary knowledge of the other clinician if the latter makes an effort to learn from the former. This is the basis for enhancement of secondary knowledge of the clinicians through sharing of knowledge by one another. The' Shared Knowledge Bases' is a new concept for medical profession in the present era of 'High Tech', though, the sharing of knowledge and exchange of ideas and opinions have been in practice since ages through conferences, discussions and personal meetings. These knowledge bases also form the foundation for creation and development of an 'Evidence Based System' for the benefit of hospital and health administrators as well as the periodical peer reviews by experts in quality healthcare. They add to the collections of a medical library and widens the scope of knowledge to the users of library.

Methodology:

The collection of material, its review by experts, proper representation of the data in the Shared PEKBs and development of a suitable software to suit the multi-user environment are the key issues. Every patient a clinician sees or admitted in a hospital does not qualify to be included in the Knowledge Base. That is the very reason for the foremost decision to restrict the recommended Knowledge Bases and the data entry to only those cases that were already selected by different specialists in various hospitals for discussion in the clinical meetings. The criteria for selecting the clinical record of the patients to qualify for such an honour to be a part of Knowledge Base for future use, have to be drawn by a team of experienced clinicians for each specialty separately and the same team may be asked to identify the actual data elements to form the different fields for the data base file so that representation of knowledge in the form of data is easy and retrieval of knowledge is accurate and fast. The reviews and editorial work involved are no less than those in selecting articles for publication in a professional journal of international repute.

The software designed and developed exclusively for these Knowledge Bases does aim at (i) Quick data entry in a previously designed fixed format (ii) Accurate, error free and complete data with auto-validation technique (iii) Simple reading, downloading and printing of the clinically interesting cases of a person's choice (iv) Arrangement of the cases in the computer in any preferred order for easy reference and retrieval in future according to one's own choice (v) Critical analysis of the events from different angles (vi) Learning lessons from the past occurrences given in each story (vii) Answers to standard queries to help students, researchers and clinical auditors (viii) Statistical tables based on the entire data base. (ix) The data bases should confirm to the needs and requirements of artificial intelligence programs and expert medical advice systems which would become one of the educative tools for the postgraduate medical students, junior clinicians and non-specialist practitioners (x) The data bases and the contents are required to act as a rich source of archived references for use by people in medical research work (xi) It is recommended that the software is independent of platform and feasible to work in Windows and Unix (xii) The software should have a tool to select a future case record to automatically add to the old data base of a similar nature and delete a case from the Knowledge Base if the subscriber chooses and decides so.

The critical analysis software aims to find if any essential information is missing in the present, past and family histories, investigations not done, if any atypical features of the disease were ignored, if any secondary or primary illness was missed, any complications were not recognized, whether any drug or procedure in the past had caused one or more of the present problems, etc, to find if too much unwanted information is present and unnecessary investigations done, etc, to find whether the investigations, medical and surgical procedures and treatment were correct, proper, timely and done by authentic persons or not , whether the postmortem findings tally with the clinical diagnosis, if any of the laboratory test results are contradictory to one another or not in unison with the history, findings, diagnosis etc, if any of the clinical signs are not corroborating with the symptoms, etc. In addition to the clinical care, this software checks the timeliness and righteousness of the administrative and supportive services of the hospital rendered to the patient.

Though, initially, the participation of Indian medical institutes and hospitals only is called for, there should not be any restriction for active participation and contribution by foreign medical institutes. The more the contributory members, the better purpose the PEKB serves. The authenticity aspect in PEKB should be taken care of from all angles for its total acceptance and future growth which may multiply many times in a short time to come as more and more institutional members join in and contributions pour in. Acceptance of cases that were actually treated and discussed in a properly conducted clinical meeting or conference and certified by the hospital authorities will rule out any problem in authenticity. Necessary steps are to be taken to make all the concerned to treat the PEKB equivalent if not superior to an internationally published professional journal of the corresponding subject. The strength of this electronic journal lies in its embedded software and such a superb facility like its critical analysis software can not be found in any other type of electronic journal, book or hospital information system.

The various issues concerned with case selection for knowledge base, data collection, data entry, development of software for retrieval of data and data analysis, etc are listed below :-

  1. Restrict the number of specialties to those which you can easily handle to begin withand gradually increase them. This will also be subject to the infra structure available to you to carry out all the functions until the creation of the concerned Knowledge Base to the entire satisfaction of the experts involved.

  2. Select a team of experts one or two from each specialty to be the core team to draw the criteria for selection of cases for the Knowledge Bases and to design a data collection form for each specialty in collaboration with a software development team.

  3. Circulate and spread this idea on shared knowledge bases through out the country and obtain willingness for active participation and for sharing the Past Experience Knowledge Bases with other hospitals.

  4. Select the qualifying cases out of the lot submitted by various hospitals in the fixed format and get them reviewed by peer experts in each specialty.

  5. The edited clinical notes of the selected patients are entered in the designed data base files (separate for each specialty) and stored in the respective folders.

  6. The clinical records that are stored in the data base files have to be checked for any errors and only the fully validated records should be allowed to remain. Many typographical errors in the text and numeric data and some of the important logical errors in the clinical data can be detected and removed better and quickly by a good validation software and the rest of the errors if any, may be checked and corrected finally by human experts (reviewers and editors). The validation software is required to be specially developed for this purpose by the software team.

  7. A team of software experts are to be selected and included in the development team for the Shared Knowledge Bases. The tasks to be given to the software team are (i) Design of the Knowledge Bases for each specialty in collaboration with the concerned specialist clinicians. (ii) Development of a good validation software for detection, output and correction of most commonly occurring errors. (iii) Development of a highly efficient software for analysis of the Knowledge Bases to suit different users. The same software will be used for collection of clinical records and data entry in the data base fields. (iv) It should assist the clinicians in developing clinical text book and journal data bases, drug information data base, investigation data base, logical clinical error data base.

  8. A list of feasible media which the knowledge bases can be recorded and distributed through CDs and USB Flash Drives among the members along with the software embedded in them may be selected.

  9. The PEKB making organization / institute / company should get it registered under professional journal category and get listed in the international index.

  10. The utilities of the software will include feedback form from the members who utilized these Knowledge Bases for various purposes to facilitate further improvements in the Knowledge Bases and the analysis software.

  11. To have a doubt on any aspect to be cleared, there has to be a huge Reference Knowledge Base (RKB) and a multi Text Book Knowledge Base (TBKB) readily available in the system as well as in the Past Experience Knowledge Base supplied to the subscribed members.

  12. The best medium which the PEKB may be copied and distributed is the Flash Drive for use though the USB port of the computer. The next best medium is a CD.

  13. The best feasible periodicity of publishing the PEKB is quarterly for all the selected specialties. A government institute or a private organization may take up this task on institutional and individual membership basis.

  14. The software should cater for all the requirements of all the users and is required to undergo a thorough quality check periodically, so that no bugs remain in the software and its performing efficiency remains high.

  15. The software should be subjected to review periodically to cater for more needs in future.

  16. The PEKB should have links to all the other knowledge bases including internet, MEDLARS, medical dictionary, etc for immediate reference, consultation and second opinion.

Discussion:

Many appreciations and encouragement for this kind of an unique source of knowledge for Continuing Medical Education in the form of Shared Knowledge Bases based purely on past experiences of different clinicians from different countries have been received from not only across our country, but also many foreign countries wherever the author traveled and lectured on the necessity of a Shared Knowledge Base. The author received some critical comments too on certain very valid issues. An elaborate discussion on all the advantages versus the likely problems along with the solutions are given below:-

  1. The success of the Shared Knowledge Bases is totally dependent upon the subscribed member institutes, who in their own interest for enhancing the quality of their healthcare services and in the larger interest of the medical profession which deals with life and death question of people, should opt to contribute more and more cases and episodes of clinical interest to the PEKB.

  2. The data required to be submitted for the PEKB is concise and precise without unwanted and unnecessary information, yet, it is comprehensive enough to give full picture of the patient concerned. Time consuming and difficult information like ICD is intentionally avoided. Hence, it takes very little time for filling data, entering it in to a data base file and converting in to an organized Knowledge Base. It also helps the software in making its critical analysis quickly and accurately.

  3. Due to the fact, the episodes discussed in PEKB are true stories, it is bound to enhance the professional skills and efficiency of the clinicians, both junior and senior and this will automatically result in better patient care standard of the hospital. They learn from others' experiences too.

  4. The overall quality of hospital services will improve as all the three hospital services (clinical, administrative and supportive) are covered, reviewed and analyzed in the PEKB. They are bound to become wiser through learning from their own past mistakes as well as others' in the same field.

  5. Over a period of time, a good collection of PEKB issues would help the clinicians in building artificial intelligent programs and expert medical advice systems. The software developed by for the PEKB will be constantly revised and improved to cater for the increased demands in the future.

  6. The archived references formed by the PEKB would be an eternal reference source for bibliography. The status of the PEKB will be that of a professional journal of international standard.

  7. The PEKB forms a strong base for improved medical research. The hospital and health statistics are likely to reflect correct picture.

  8. The software used for creating and analyzing the PEKB ought to be fault free and it should meet all the users' requirements. The creation of text book & journal knowledge base, pharmacology knowledge base, investigations knowledge base, logical error knowledge base etc that are essential for second opinion on diagnosis, treatment, investigations etc and for detection of all usually occurring and rare errors in the clinical data are not easy tasks, these are rather, the most difficult and time consuming tasks. This can be achieved only through dedicated and concerted effort of multitude of professionals sitting around a round table.

  9. About the question of retaining the confidentiality of the individuals concerned, the patient's identity (Name, identity numbers, photographs, nationality, address, place, insurance provider, etc) and the identity of the clinicians and hospitals who treated the patient are not revealed in the PEKB.

  10. The most essential personal information used for making crucial decisions on diagnosis and treatment like sex, age, race, occupation, country where treated, month and year of episode and the concerned medical specialty are included in the case notes. This information is also needed for the software for critical analysis of the episode.

  11. About the legal aspects that are likely to be raised by new members, it is ensured that there is no possibility of the unknown patient or his/her kith and kin seeking justice from a court of law for any deficiency in service through an act of commission or omission by an unidentified clinician or hospital as all the identities are deliberately deleted from the data bases. The reviewers and editorial board too are not provided with this information. Only the name of the member institute / hospital will be known to the Editor-in-Chief who receives the case notes contributed by them.

  12. As the identities of the healthcare provider and the patients are deleted from the data bases, development of inferiority / superiority complexes, jealous attitudes and unethical comparisons on performance and quality standards will not be seen among clinicians and hospitals.

The author however, advocates due care in collection of material for the PEKB, getting it reviewed by highly qualified and very experienced clinicians at the level of consultant and professor in the subject of his/her specialization, keeping editing of the actual episode to the minimum, giving data security and confidentiality aspects the utmost importance, removal of all typographical errors from the contents through best validation software, giving pharmacological generic names of medicines instead of brand names, mention of measuring units for all the tests, constantly reviewing the efficiency of the software, publishing the PEKB in the required international standards and with due promptness for each issue, etc. Making a medical journal out of past experiences alone is not an easy and smooth process. You have to cross lot of hurdles.

He prefers combining the validation software with the main PEKB software with links to all the other knowledge bases and AI software used for getting expert medical advice. The software should undergo quality check periodically. It would be better if it is made platform free and operating system free and internet enabled, so that a cross reference can easily be obtained through internet and immediate communication may be sent to the concerned individual or institute. The design of the data base for the PEKB should be separate for each specialty and the transfer of data during retrieval and analysis are required to be very fast.

Summary:

The knowledge potential of the past experiences of a hospital, regarding clinical care of its patients, in respect of difficult-to-diagnose & treat category irrespective of the fact whether they were treated successfully or not, is so great, that it is more valuable than many text books of medicine and journals of clinical care. It is a pity, that such a vast knowledge is not put to full utility and is allowed to go waste. It is mainly because of lack of adequate knowledge among the clinicians on its richness. The other important cause is lack of proper resources and drive to put it in to use for utilization of all clinicians. The hospital administrators also have never spent any time to think on this issue as a tool for quality enhancement of the healthcare delivery standards of their hospitals. This paper has described in detail, the importance of such past experience database, the methodology to convert it to a proper knowledge base and the ways of making it available to all the clinicians in the world including its own. The author suggests that every hospital, big or small can contribute its might to this novel knowledge base that can be shared by the whole medical world and it is never too late to start building it and put it in to practice.

Next          Back