Biomedical subjects

D P Connelly

Publications and source records attributed to D P Connelly.

At least 19 recordsLinked to original sources

PubMed1992

Computer-assisted test interpretation: considerations in patient care.

Computer-assisted test interpretation (CATI) is a set of developing technologies designed to support medical decision-making. This paper develops a taxonomy of computer-assisted test interpretation, giving specific consideration to the characteristics of the data that are to be interpreted, the nature of the interpretive task, the expected involvement of the health professional in the generation of the interpretation, the inference mechanism used for the interpretation, and the broader context of the interpretation. We go on to examine potential benefits and disadvantages of CATI systems in terms of accuracy, information management, interpretation time, patient management, medical communication, and expense. Finally, we examine electrocardiogram interpretation systems from the perspective of this taxonomy, and offer suggestions regarding areas of further inquiry into the effects of CATI on medical care.

Classification↗

PubMed1992

Nonparametric probability density estimation: improvements to the histogram for laboratory data.

The histogram has long been used in the clinical laboratory for the depiction and manipulation of frequency data. We present recent results of refinements to the usual histogram procedures along with modern alternative methods of estimating frequency distributions, including the kernel and discrete maximum penalized likelihood estimation (DMPLE) approaches. We compared these nonparametric methods on 15 different types of simulated distributions, and on several sets (greater than 1000 subjects/set) of real data, including alanine aminotransferase, aspartate aminotransferase, and lactate dehydrogenase levels. Each frequency curve estimation technique was evaluated by measuring the integrated mean square error between each technique's prediction and the true underlying distribution, using Monte Carlo techniques on sample sets with size 49 and 119. The kernel method was the clear method of choice, both in performance (best in 22/36 cases) and in practical usage.

Alanine Transaminase↗

PubMed1992

Continuous quality improvement and medical informatics: the convergent synergy.

Continuous quality improvement (CQI) and medical informatics specialists need to converge their efforts to create synergy for improving health care. Health care CQI needs medical informatics' expertise and technology to build the information systems needed to manage health care organizations according to quality improvement principles. Medical informatics needs CQI's philosophy and methods to build health care information systems that can evolve to meet the changing needs of clinicians and other stakeholders. This paper explores the philosophical basis for convergence of CQI and medical informatics efforts, and then examines a clinical computer workstation development project that is applying a combined approach.

Computer Security↗

PubMed1992

Physician use of an NICU laboratory reporting system.

Clinical workstation developers may gain useful insights from studies of physician acceptance and use of computer systems that have been incorporated into daily practice. We used a physician survey, intrinsic monitors built into the system and an observational study to assess physician acceptance and use of a touchscreen workstation that was put in place in an NICU in 1985. Each of the 87 physicians assigned to the 30 bed NICU during the two academic years beginning July, 1987 was sent a questionnaire that assessed experience and attitudes about the system. The 70 responding physicians (80 percent) were unanimous in agreeing or strongly agreeing that the system was "easy to learn" (57 reported it taking less than five minutes), "easy to use", and "integrates smoothly into patient care activities". Over 94 percent of the physicians agreed or strongly agreed that the system was "fast", "saves time", and was "reliable and dependable." Sixty-three of the responding physicians (90%) reported using the system two or more times a day with 53 using it more than five times daily. The most frequently requested new feature was that of time-trend graphs (51 occurrences). The intrinsic monitors were useful in validating design decisions and survey results but also provided new insights relevant to security issues. Similarly, the observation study reinforced some of the survey results but also highlighted an additional issue not brought out by the other two assessment methods. The overall assessment indicated that the system has been both well-accepted and well-used by its intended clinical clientele.

Attitude to Computers↗

PubMed1992

Certification of cholesterol measurements by the National Reference Method Laboratory Network with routine clinical specimens: effects of network laboratory bias and imprecision.

The National Reference Method Laboratory Network has initiated a program to certify clinical laboratory cholesterol measurement performance by using routine clinical specimens. Clinical laboratory and reference laboratory measurements of split samples are used to assess whether the clinical laboratory is meeting the Laboratory Standardization Panel's goals for accuracy and precision. We used a computer-based Monte Carlo simulation model of split-sample proficiency testing to evaluate the certification program and, in particular, to analyze the effects of reference laboratory bias and imprecision. Results of our simulations indicate that the accuracy of the certification program is strongly influenced by reference laboratory bias and less influenced by reference laboratory imprecision. The certification program is potentially highly accurate, but unless reference laboratory bias is tightly controlled, the number of classification errors may limit its utility. Moreover, the decision limit of the certification program needs to be higher than the Laboratory Standardization Panel's goal (3.5% instead of 3.0%) to ensure that an acceptably high proportion of well-performing clinical laboratories can become certified.

Certification↗

PubMed1991

Design and preliminary evaluation of an expert system for platelet request evaluation.

In spite of growing awareness of the potential risks associated with transfusion, the number of platelet units transfused in the United States continues to increase each year. There is a growing interest in ensuring that all transfusions are administered for appropriate reasons. Prospective review of requests for transfusions has been used to accomplish this goal. Although successful in reducing the number of inappropriate transfusions, this review method requires great time commitments by blood bank personnel and physicians. A knowledge-based system (ESPRE) that aids hospital blood bank personnel in the review of requests for platelet transfusions has been developed. The system automatically obtains most of the required patient data via a direct link to the hospital's main laboratory computers. The system generates a printed report that includes a list of patient-specific data, a list of the conditions for which a transfusion would be appropriate for the particular patient (given the clinical condition), and the conclusions drawn by the system. During a preliminary clinical evaluation of ESPRE, 73 randomly selected platelet transfusion requests were evaluated for approval by laboratory personnel and ESPRE. Overall, ESPRE would have approved 71 of the requests and laboratory staff would have approved 72. Forty-four percent of the requests would have been approved for the same reasons given by the staff. There were only three disagreements on final approval between ESPRE and blood bank personnel. This computerized expert system is a promising approach to the prospective review of all platelet transfusions.

Blood Banks↗

PubMed1991

Assessment of split-sample proficiency testing for cholesterol by use of a computer simulation model.

We developed a computer model to study the use of patients' specimens to assess compliance of cholesterol measurement performance with the 1992 goals of the Laboratory Standardization Panel of the National Cholesterol Education Program. The model uses Monte Carlo techniques to simulate cholesterol measurements that are subject to both systematic and random error. Split-sample measurements by a clinical laboratory and by a reference laboratory are compared by using linear regression to estimate clinical laboratory bias and imprecision; subsequently, according to specified decision limits, the performance of the clinical laboratory is classified as acceptable or deficient. We have quantified the influence of the bias and imprecision of the clinical laboratory, the imprecision of the reference laboratory, the number of split samples compared, and the decision limits on the accuracy of the classification of clinical laboratory performance. Unless the decision limits are carefully selected and a sufficient number of split samples are used, clinical laboratory performance will be frequently misclassified.

Cholesterol↗

PubMed1991

Expert systems and the clinical laboratory information system.

Many of the tasks of the clinical laboratory involve the application of knowledge and experience to patient-specific problems. Expert systems that have ready access to patient data and expert knowledge have the potential to support many laboratory processes, especially those that involve the repetitive and reasonably predictable application of knowledge. Quality assurance and utilization improvement efforts requiring constant vigilance and surveillance of an array of laboratory processes especially may benefit. Expert system technologies, trends of their use in the clinical laboratory, their role in the total testing process, and their potential for influencing physician use of the laboratory through timely feedback are discussed.

Clinical Laboratory Information Systems↗

PubMed1990

ESPRE--expert system for platelet request evaluation.

The expert system for platelet request evaluation (ESPRE) is being developed to support independent learning and decision making regarding the use of platelet transfusions while physicians are actively engaged in clinical practice. The knowledge of transfusion medicine incorporated in ESPRE has been largely gathered from blood bank physicians responsible for determining the appropriateness of blood product administration. Knowledge acquisition methods have included structured and unstructured interviews with protocol analysis using real and fabricated transfusion cases and critiques of an expert system prototype's conclusions. For knowledge representation, ESPRE uses frames with embedded production rules to provide processing efficiency and facilitate knowledge base development. ESPRE automatically acquires key laboratory findings from a laboratory information system. Systems such as ESPRE require careful evaluation and, if proven effective and accurate, present to laboratorians a new tool for delivering high-quality and individually adapted health care.

Decision Making, Computer-Assisted↗

PubMed1990

Embedding expert systems in laboratory information systems.

Expert systems (ESs) may enhance decision making throughout most steps in the clinical laboratory testing process. By embedding ES capabilities in a laboratory information system (LIS), the LIS gains the capability of symbolic reasoning whereas the ES gains use of the LIS database and communications resources. Design considerations include modular integration, performing timely inferences, sparing LIS resources, and providing a syntax that facilitates knowledge base development and verification by laboratorians. The LIS notifies the ES of specimen status changes through an event log. An event scanner finds events that are relevant to prestored knowledge frames and passes this information to an inference processor through instance records. If conditions specified in the knowledge frames are satisfied, an alert processor sends a message to a CRT, printer, file, or report. Examples of applications during order entry, specimen distribution, processing, and results reporting are presented.

Clinical Laboratory Information Systems↗

PubMed1990

Knowledge resource preferences of family physicians.

Because of the pivotal role of medical knowledge in clinical problem solving, it is important to understand how clinicians decide to seek additional knowledge for patient care decisions and how they choose among the resources available to them. Using a self-administered questionnaire, 126 family physicians reported their use of 11 types of knowledge resources for answering patient-specific questions arising in clinical practice. They reported almost daily use of the Physicians' Desk Reference and more often than weekly use of colleagues. There was little use reported of Index Medicus or computer-based bibliographic retrieval systems. The research literature of medicine was used infrequently and rated among the lowest of resources in terms of credibility, availability, searchability, understandability, and applicability. In deciding among a subset of knowledge resources for answering a clinical practice question, resource cost variables related to clinical availability and applicability of the information to the problem at hand appeared to be more influential in the minds of physicians than factors related to quality of the resource. These findings have important implications for the development and deployment of knowledge resources intended to be useful and used in clinical practice.

Diagnosis↗

PubMed1989

ESPRE: a knowledge-based system to support platelet transfusion decisions.

ESPRE is a knowledge-based system which aids in the review of requests for platelet transfusions in the hospital blood bank. It is a microcomputer-based decision support system written in LISP and utilizes a hybrid frame and rule architecture. By automatically obtaining most of the required patient data directly from the hospital's main laboratory computers via a direct link, very little keyboard entry is required. Assessment of time trends computed from the data constitutes an important aspect of this system. To aid the blood bank personnel in deciding on the appropriateness of the requested transfusion, the system provides an explanatory report which includes a list of patient-specific data, a list of the conditions for which a transfusion would be appropriate for the particular patient (given the clinical condition), and the conclusions drawn by the system. In an early clinical evaluation of ESPRE, out of a random sample of 75 platelet transfusion requests, there were only three disagreements between ESPRE and blood bank personnel.

Decision Support Techniques↗

PubMed1989

Monte Carlo simulation and the clinical laboratory.

Complex decisions regarding the scope, efficiency, and effectiveness of service coupled with advances in microcomputer workstations and modeling software have created new incentives and opportunities for the application of powerful simulation methodologies in the clinical laboratory. Monte Carlo techniques that involve a large number of repetitive simulations of a system that has probabilistic characteristics may be especially useful. These techniques can be applied to problems where complex interactions, the nondeterministic nature of medical problems, and the incompleteness of medical knowledge render traditional analytic techniques impotent. Examples include complex clinical laboratory data analysis, clinical strategy development using advanced decision analysis methodologies, and prospective evaluation of the effects of proposed changes in laboratory operations. Although Monte Carlo simulation techniques appear to be promising for clinical laboratory use, limitations must be noted. Care and effort are required in model specification if meaningful results are to be obtained and their significance convincingly conveyed.

Computer Simulation↗

PubMed1989

Does compliance mismatch alone cause neointimal hyperplasia?

To define the relationship between compliance mismatch and the development of neointimal hyperplasia, one 3 cm segment of common iliac artery was externally banded in seven dogs, thereby fixing the arterial diameter at end diastole. To quantify compliance, end-diastole diameter and its change with pulse pressure were measured by induction angiometry. This technique uses intravascular soft trifilar wire probes introduced through distally placed polytetrafluoroethylene sidearms. Compliance was checked in the banded and contralateral undissected unbanded control iliac arteries at 3 and 6 months, at which times the vessels were fixed by perfusion, excised, and examined histologically. Sustained (6-month) compliance mismatch was successfully induced within the banded segments (p less than 0.0001), and no compliance mismatch was seen in the control segments (p = 0.357). The intima of all banded vessels was virtually indistinguishable from that in controls grossly and histologically. Mild focal intimal thickening, less than 3 cell layers thick involving less than 5% of the vessel circumference, was typically seen in both banded and control vessels (range 6.57 +/- 6.80 micron to 38.86 +/- 57.16 micron). In marked contrast, at the sites of the polytetrafluoroethylene-to-femoral artery anastomosis, near-occlusive neointimal hyperplasia (1714 +/- 415.47 micron) was seen in all animals. Residual lumen area in the banded and control vessels was only minimally abnormal (range 98.65% +/- 2.18% to 99.96% +/- 0.08%). These data indicate that compliance mismatch alone is an insufficient stimulus for the development of neointimal hyperplasia in the canine model.

Animals↗

PubMed1988

Rate and delta checks compared for selected chemistry tests.

We compared delta and "rate" check methods for 12 selected chemistry tests. Rate checks were determined by dividing delta checks by inter-specimen interval time. The delta and rate check methods were based on differences and percent change of untransformed and absolute value-transformed values. The distribution of delta differences was not symmetrical for calcium, alkaline phosphatase, aspartate aminotransferase, or phosphorus, which led to different check limits between untransformed and absolute value-transformed methods. The dispersion of rate checks was large. The interval time between two consecutive tests was multimodal, which probably reflected adherence to fixed testing protocols.

Alkaline Phosphatase↗

PubMed1987

Knowledge-based systems in laboratory medicine and pathology. A review and survey of the field.

Knowledge-based systems are computer systems designed to handle knowledge-intensive tasks, usually involving reasoning and inference. They are increasingly being applied to problems in laboratory medicine and pathology. In this article we provide a brief introduction to the basic concepts of knowledge-based systems, review some of their published applications, and report on an informal survey of specialists in laboratory medicine and pathology. The survey, sent to 102 individuals, indicated that 24% were involved in developing knowledge-based systems, with most systems at an early stage of development. Recent advances in knowledge-based systems research as well as survey responses suggest that this technology will have increasing value in laboratory medicine and pathology.

Artificial Intelligence↗

PubMed1986

Probabilistic sensitivity analysis methods for general decision models.

Probabilistic sensitivity analysis has previously been described for the special case of dichotomous decision trees. We now generalize these techniques for a wider range of decision problems. These methods of sensitivity analysis allow the analyst to evaluate the impact of the multivariate uncertainty in the data used in the decision model and to gain insight into the probabilistic contribution of each of the variables to the decision outcome. The techniques are illustrated using Monte Carlo simulation on a trichotomous decision tree. Application of these powerful tools permits the decision analyst to investigate the structure and limitations of more complex decision problems with inherent uncertainties in the data upon which the decisions are based. Probabilistic sensitivity measures can provide guidance into the allocation of resources to resolve uncertainty about critical components of medical decisions.

Decision Theory↗

PubMed1985

Evidence for an informal clinical policy resulting in high use of a very-low-yield test.

The actual and self-reported practice regarding the use of cerebrospinal fluid cultures for Mycobacterium tuberculosis was examined. All neurology house staff members surveyed, 62 percent of internal medicine house staff members, and none of pediatric house staff members reported that they would order mycobacterial cultures of cerebrospinal fluid routinely. The actual practice was comparable, with 71 percent of cerebrospinal fluid specimens being subjected to culture for mycobacteria on the neurology service, 65 percent on the internal medicine service, and 6 percent on the pediatric service. In this practice, medicine and neurology house staff differ significantly from their pediatric colleagues (p less than 0.001) and from the stated practice of their respective faculties (p less than 0.01). For at least six years, most medicine and neurology house staff have commonly applied an informal clinical policy of routinely culturing cerebrospinal fluid specimens for mycobacteria, despite a low suspicion of disease, lack of faculty support for the practice, and a zero yield for the test. Informal clinical policies such as this may be an important contributor to the problem of technology overuse.

Bacteriological Techniques↗