Computerized controlled-substance inventory management in a forensic toxicology laboratory: practical application of a state-change model.

Efficient, accurate, secure, comprehensively audited, and legally defensible maintenance of an electronic chain-of-custody applied to a controlled-substance inventory system in a busy forensic toxicology lab (FTL) has been achieved with custom-written software which is based on a general state-change model. Prior to the use of this program we were unfortunately accustomed to being surprised by incomplete drug use documentation leading to pseudo-shortages, expired or uncertified lots, the lack of an adequate chain of custody, and the risk of legal challenges. This resulted in confusion, delays, and emergency orders, with the inevitable waste of time and money. The control strategy implemented in the INVEN subsystem of our main system, TOXLAB, is based on four principles. First, access to the drug inventory is strictly limited. Second, controlled drugs are dispersed to technical staff only when a computer-authorized request form is presented. Third, every request for addition/subtraction must be pre-authorized by the supervisor. Fourth, the program enforces a strict progression of requests and actions from one state to another and produces request, authorization and certification, and/or approval logs ready for signature at every step. This approach has given us a comprehensive, chronological record of all events involving drug inventory transactions and promises to improve resource utilization. Our system appears to fulfill the basic requirements for the legal adequacy of computer-resident data.

Nuclear DNA content of altered hepatic foci in a rat liver carcinogenesis model.

In individual altered hepatic foci (AHF), aneuploidy occurs before malignant changes can be diagnosed histologically (O. Sudilovsky and T. K. Hei. Fed. Proc., 42:2225, 1983). In the current experiments Sprague-Dawley rats of both sexes were given i.p. injections of diethylnitrosamine (50 mg/kg body weight) 18 h after partial hepatectomy and were given a choline-sufficient diet (CS) for 1 wk. Four treatment groups were then formed and fed CS, CS containing 0.05% phenobarbital (PHB), choline-deficient diet (CD), and CD with 0.05% PHB. An extra female group received infusions of saline after the hepatectomy and fared CD. Control animals were partially hepatectomized, inoculated i.p. with saline, and placed on CS. The rats were sacrificed 16 wk later, liver sections were stained with a combined Feulgen-gamma-glutamyl transpeptidase stain, and the DNA content of gamma-glutamyl transpeptidase-positive foci was measured cytospectrophotometrically. There were no AHF in the control animals. Hepatocytes from control livers and cells adjacent to foci in treated livers had peaks corresponding to the 2C, 4C, and 8C range. In AHF the ploidy, however, was predominantly diploid, tetraploid, or heterogeneous. The ratio of diploid to tetraploid cells in foci of rats provided with CS + PHB was 5.5 and in those supplied with CD + PHB was 0.09. This suggested that dietary manipulations change the nuclear DNA distribution of AHF. Aneuploidy was also present, as expected, in 4 of 33 AHF in the animals placed on CD + PHB. It was observed as well in 2 of 26 AHF of rats given CD but in none of the 20 AHF fed CS + PHB. These data indicate that CD (which acts as both initiator and promoter) may be responsible for the appearance of aneuploidy. A general model, based on these results and the clonality of each individual focus, is proposed for the development of cells through the preneoplastic stage.

Automated rule-based decision systems in forensic toxicology using expert knowledge: basic principles and practical applications.

This paper presents the basic principles and practical benefits of the application of expert systems (ES) and artificial intelligence (AI) to problem solving in forensic toxicology. We acknowledge the complexity and elegance of the theoretical substance and program algorithms of existing work in these disciplines, while simultaneously observing that many presentations of this material cloak the essential facts and concepts in unnecessary jargon and hyperbole. We attempt to remove the cloak without misrepresenting or oversimplifying the underlying structures. We first present a summary of the history, basic functions, technical fundamentals, and typical applications in three major categories of established ES/AI systems. We then assess the status of ES/AI in the forensic toxicology laboratory (FTL) with emphasis on potential applications. We conclude with an analysis of experiences with ESs in our laboratory where we have used an integrated expert system to reduce laboratory errors, detect internal inconsistencies in data, discover new substance abuse subpopulations, and reduce the frequency of sample reprocessing. We have minimized specimen processing time and instrument wear while maximizing technician efficiency and thus performing more tests for the same or reduced costs.

Prevalence and significance of methicillin-resistant Staphylococcus aureus in patients with cystic fibrosis.

Recent reports indicate that methicillin-resistant Staphylococcus aureus (MRSA) may be emerging as a significant pediatric nosocomial pathogen. Children with cystic fibrosis (CF) pulmonary disease are subject to many of the risk factors for MRSA colonization and/or infection. We retrospectively investigated the prevalence and significance of MRSA from sputum and throat cultures in 452 patients with CF followed during 1986. No MRSA had been isolated during 1984 or 1985. Although S. aureus was isolated from 212 patients (47%) in 1986, only 14 (3%) showed MRSA. The MRSA strains had 11 different antimicrobial susceptibility patterns. Neither age, clinical condition, nor recent prior hospitalization correlated with MRSA acquisition. Acquisition did not appear to directly affect the course of the pulmonary disease in these patients even though no patient received any treatment for their MRSA. The prevalence of MRSA is low, although patients with CF are subject to many risk factors. MRSA appears to be mainly community-acquired and to represent colonization rather than infection. However, the potential for nosocomial MRSA infection is present, and vigilance is required in monitoring any changes in frequency of isolation or infection with these organisms.

A multifunction network of computers in a large pathology department. Integration of word processing, data base management, and general purpose computing using network principles.

We report the strategies for design and implementation of a system of compatible mini- and microcomputers intended to improve efficiency of data handling and research in several clinical and academic divisions of a large university hospital department of Pathology. A dedicated, preprogrammed, multiuser, word processing computer and small standalone word processing stations connected to a general purpose minicomputer, using standard network protocols for communication, permit virtually error-free movement of data between computers doing functionally distinct tasks of word processing, data base management, and general scientific computing. User programming is minimized and is done in a simple, well-known, high-level language. We show that cost, thruput, speed, document volume, document size, revision cycle frequency, data base size, and frequency of use are important design criteria and that existing staff can be trained easily to operate the systems.

Proposal for a national autopsy data bank: a potential major contribution of pathologists to the health care of the nation.

This report highlights the salient deliberations of the Anatomic Pathology Committee of the College of American Pathologists (CAP) and of its consultants and advisors in formulating a proposal for the establishment of a National Autopsy Data Bank (NADB). It has constituted the major thrust of this Committee for the past five years and is presented in considerable detail in order to familiarize the pathology community and other interested parties with the plan, which, if implemented, would enable pathologists to contribute even further to the health care of the nation while concomitantly realizing benefits both individually and collectively. In order for the proposal to culminate in an optimally functioning NADB, the support and participation of a majority of, if not all, pathologists with autopsy responsibilities would be needed. Goals, objectives, justification, and merits, background, legal considerations, pilot studies, and management plans are discussed and the singular advantages that could accrue with implementation are delineated. If achieved, it would be the first time that any medical discipline or organization would have amassed such kinds of accurate biomedical information potentially so useful to so many for so little.

Misdiagnosis of bronchogenic carcinoma: the role of cigarette smoking, surveillance bias, and other factors.

The effects of cigarette smoking and other factors on the accuracy of clinical diagnosis of bronchogenic carcinoma were studied retrospectively in 14,074 autopsies performed over 26 years (1948-1973) at University Hospitals of Cleveland. Within a selected study group of 415 cases diagnosed as bronchogenic carcinoma either clinically, at atuopsy, or both, the disease was diagnosed accurately in 260 cases (63%), overdiagnosed in 38 cases (9%), and underdiagnosed in 117 cases (28%). Misdiagnoses occurred in female patients nearly twice as frequently as in male patients. Elderly men were over- and underdiagnosed more frequently than were young men. An accurate diagnosis of this neoplasm was strongly associated with a histroy of smoking and was also related to the number of hospital admissions, the diagnostic procedures used, and surveillance bias associated with a history of smoking or coughing. In 88% of the misdiagnosed cases, the tumor was either simulated or masked by other diseases. It is estimated from these data that the rate of unavoidable clinical misdiagnosis of the disease is 32% and the true error (overt misjudgment) in clinical diagnosis, 5%. Previous estimates of causal association between smoking and lung cancer would not be affected by the findings of this study.

Storage and retrieval of SNOP-coded pathologic diagnoses using offsite computing and optical character recognizing systems.

A computerized cross-reference system for retrieving autopsy and surgical pathology cases on the basis of case number or diagnosis has been implemented. The system achieves economy and flexibility by using offsite computer service bureaus for job production, eliminating the need for expensive onsite equipment. Coded diagnoses may be typed using the OCR (Optical Character Recognition) font simultaneously with or separately from the clinical documentation. The flexibility of new OCR equipment permits production of machine-readable code sheets with an ordinary pencil and completely eliminates the need for typing. The system produces year-to-date books that list all diagnoses, on an accumulating basis, in alphabetic order by SNOP* topology, morphology, etiology and function, and will be compatible with SNOMed. Because all data are stored on magnetic tape, they may be manipulated and retrieved as desired through user programming. The initial setup cost was dollar 1,000 for programming and testing, and production runs and all report printing cost about dollar 1,000 per year (autopsies and surgical pathology cases), which is about 1.1 cents per diagnosis.