|
QUALITY: HOW EXPERT TECHNOLOGY CAN HELP Dennis Winsten & Lonnie M. Vanderslice The "Q" word, Quality, is being discussed in virtually every laboratory in America. "Total Quality Management" (TQM) and "Continuous Quality Improvement" (CQI) are the "phrases" laboratorians are using and trying to make into reality for their labs. According to a survey done by Coopers and Lybrand, hospitals credited TQM programs with decreasing average laboratory test costs by 23 percent and decreasing laboratory test turnaround time by 50 percent.(1) The TQM initiative began with the JCAHO's Agenda for Change in 1987. The intent was, in part, "to create a health-care environment that focuses on the quality of patient care and ensures that every health-care organization's governing, administrative, managerial, and clinical leadership is devoted to CQI."(2) The JCAHO has chosen two strategies to achieve the mission of enhancing the quality of health care provided to the public. The short range strategy was the formalization of an ongoing quality surveillance process (QA) that instituted the 10-step Quality Assurance process to ensure uniform monitoring and evaluation of health-care quality. The QA program served to introduce the principles of Total Quality Management (TQM). Longer range, the JCAHO introduced the Agenda for Change, a far reaching plan that marked the transition from a formalized QA to a more modern approach of quality surveillance, quality assessment, and improvement (QI) which is serving as an industry-wide introduction to the principles of Continuous Quality Improvement (CQI). The Clinical Laboratory Management Association (CLMA) has made TQM one of the major cornerstones of its programs for the Nineties. CLMA recognized early in the process that laboratorians were in a leadership position to effectively implement TQM within their organizations. Laboratories, by and large, have done a good job of providing quality services, but today, just "good" is not good enough. Laboratories are being expected to improve the quality of their services with regard to accuracy, comprehensiveness, timeliness, and availability. More accurate, complete, understandable test results are expected to be provided faster to all consumers of the laboratory's services. Further, those results must be in a form most appropriate to the particular consumer be it physician, nurse, records administrator, business office, or administrative management. At the same time, pressure to reduce or maintain costs has increased. Labs have never had to work smarter and more efficiently than today. The Clinical Laboratory Improvement Act (CLIA '88) mandates that laboratories "establish and follow policies and procedures for an ongoing quality assurance program designed to monitor and evaluate quality; identify and correct problems; assure the accurate, reliable, and prompt reporting of test results; assure adequacy and competence of the staff."(3) Laboratories are both "Process" (event driven) and Information (data driven) entities. Simply to have a data base is insufficient. TQM and CQI cannot be implemented without accurate and relevant data. The use, distribution, and timeliness of data is critically important, in a process environment, to create useful information from random or dis-ordered data. Data is often very volatile and can be as useless as yesterday's newspaper unless delivered when it is needed in a usable, readily understandable form for evaluation and action by the particular user/consumer of laboratory services. The thrust of TQM is not to wait until errors or problems occur and then fix them, but rather to prevent such error/problems from occurring in the first place. Just to correct the error, itself, is necessary but clearly insufficient for TQM. TQM embodies a prospective approach to try to identify errors (of commission or omission), very early, before they become problems; and having identified the error, to correct the source of the error to prevent its (or its brethren's) re-occurrence. Some new technologies available in laboratory information systems (LIS) support such an approach. At the risk of oversimplification, we believe that in order to achieve TQM you need two generic things:
- The conscious adoption of a TQM "mind set" as a way of life and a way of doing things.
- The methodologies and tools necessary to implement TQM.
This article will leave you to your own devices to accomplish #1 (there are many excellent articles on this - see references) and will concentrate on #2 with an acknowledged narrow focus on LIS. We will describe what LIS can (or should be able to) do and provide some real-world, real-life examples indicating how laboratorians, not unlike you, are using LIS to assist in TQM at their institutions. Before providing some specific examples, we will describe the methodologies employed and the types of TQM supportive services being provided. The tools and methodologies we are referring to can be classified under the generic title "Artificial Intelligence." Vendors are referring to their LIS products as "Expert Systems" or "Rule-Based Systems. We will use the more generic term "Expert Technology" or "E.T." for short. Simply speaking, these capabilities are adjuncts to the LIS and act upon data entered into the LIS or data residing in the LIS data base. An important characteristic of these capabilities is that they operate in "real-time"; that is to say, now, right now. As soon as a particular event or condition is detected by the LIS, it automatically and immediately takes action(s) defined by the user as appropriate to that event or condition. The specific nature of the events or conditions and the action(s) to be taken are defined by laboratory staff in concert with institutional policies and procedures. The "expertise" in "expert technology" is yours. The kinds of things these "expert technology" applications do are: (4,5)
- Compare results from different sources for consistency; e.g., cytotech's vs. pathologist's pap smear interpretation if they are different to immediately generate a message.
- Initiate reflexive testing based on test results, patient demographics, and physician's preference.
- Monitor ordering practices to identify cases where lesser cost generic alternatives may be used in antibiotic therapy.
- Identify adverse or contraindicated conditions which may cross departmental boundaries; e.g., new Radiology IVP order for patient with high BUN and high creatinine in last 48 hours.
- Monitoring CPT codes used to assure maximum appropriate reimbursement.
- Monitoring the occurrence of critical or other "call physician" results and providing immediate notification, audit, and follow-up.
- Monitoring the occurrence of a unique set of parameters, case types, or results and triggering notices to cognizant staff and ordering follow-ups as indicated.
- Determination if certain tests (e.g., manual differentials) should or should not be performed based on test results and patient demographics.
- Automatic extraction of selected data for statistical analysis and for the detection of trends or anomalies requiring management attention.
In summary, "expert systems" can be used to:
- Monitor: events, trends, and data related to quality
- Identify: the occurrence of events, trends, or data requiring action to insure quality
- Initiate: actions including messages, audits, and specific transactions; e.g., reflex orders, etc., in response to an identified condition
Usage of "expert systems" technology on a routine basis is a relatively new phenomenon with still just a few laboratories having it operational. The increasing emphasis on TQM will serve to stimulate further development by vendors and more sophisticated and valuable applications by laboratories. The following are some examples of how some laboratories are, today, using their LIS as a proactive tool to assist in the TQM movement.(4,5)
Lipid Profile: Provides automatic interpretation based on specific patient demographics and results obtained. Provides "tailored" easy-to-understand report without extraneous data. Outreach Clients: Provides automatic screening of test requests on a per client basis with immediate notification of physician client specific actions required and/or automatic ordering of follow-up test(s) based on particular physician procedures. "Exception" Critical Checks: Based on patient characteristics and doctor instructions to automatically generate alerts, audit logs, "call physicians" notices, special reports when certain critical results are detected. Such actions being exceptions to standard policies are automatically handled as "special cases." Isoenzyme Test Ordering: Where follow-up isoenzyme testing is indicated, the appropriate timed tests are automatically ordered according to schedule based on time of day to optimize specimen collections. Inappropriate Ordering: Monitoring of test orders, over a time period, for appropriateness relative to policies, patient characteristics, and as a source of additional diagnostic information. Maintaining audit logs, by physician, of any trends in ordering tests of no clinical significance. Data Extraction for Studies: Based on user-specified selection criteria to automatically identify cases of interest and extract pertinent data for further analysis. Monitoring CPT Codes: Monitoring anatomic pathology case billing to assure that major case category CPT code has been recorded at time of case sign-out. If none found, message sent to departmental secretary to follow up.
Laboratories who have been using "expert technology" with their LIS have made the following observations:
- Decreased lab errors and increased quality of results
- Training of new personnel is easier (since policies and procedures have been embodied in the LIS)
- Consistency of compliance with policies/procedures across all shifts
The use of Expert Technology, "E.T.," is still in its infancy. As more users become involved, their feedback to vendors will serve to stimulate further extensions and improvements to their E.T. capabilities. Likewise, increased usage in more and more laboratories will stimulate new ideas, techniques, and approaches for using E.T. With time and with gained experience by user and vendor, we can expect increased sophistication in the type and extent of quality improvements obtained from the use of this technology within an LIS and substantial support for laboratory efforts toward TQM. NOTES: 1 HOSPITALS, "Quality Watch," December 20, 1991, page 16. 2 Clark, G., "Quality Assurance, An Administrative Means to a Managerial End: Part III,"
CLMA, November/December 1991, Volume 5, Number 6, page 464. 3 Foster, Hope S., "HCFA Issues Proposed CLIA '88 Regulations," CLMA Management
Briefs, June 1990. 4 3M Internal Correspondence: Whitaker, C. and Stewart M. Expert System
Technology/User Comments, October 5, 1990. 5 Cerner Corporation, Introduction to the Cerner Rule System, October 5, 1990. GENERAL REFERENCES: Elevitch, F., Lu, S., "Computerization: Key to a Successful QA Program." MLO, October
1990, p. 22-26. Fitzgibbon, R., "Is Your Hospital Ready for CQI?" MLO, December 1991, p. 9. Gardner, E., "JCAHO's 'Agenda for Change' to Place New Demands on Systems,"
Modern Healthcare, January 6, 1992, p. 72. Hunter, L. and Vanderslice L., "The LIS: The Overlooked Tool for TQM," Computers in
Healthcare, September 1991, p. 48-49. Joseph, J. and Kampa, I., "Building Quality Labs," Clinical Chemistry News, February
1991, Volume 17, No. 2, p. 14. Palmer, J., Bissell, M., and Cosman, T., "Laboratory Quality and Economic Necessity:
Values in Collision," MLO, September 1991, p. 49-54. Scheuing, E., "Creating a Quality Culture in Your Organization," CLMR, March/April 1991,
Volume 5, No. 2, p. 89. |
|
