Cytochemical staining and flow cytometry methods applied to the diagnosis of acute leukemia in the pediatric population: an assessment of relative usefulness.

BACKGROUND: Cytochemical staining has been used in the diagnosis of acute leukemia for more than 20 years. The general availability of flow cytometers and an extensive panel of antibody reagents useful for characterizing blood cell lineage question the usefulness of continuing routine use of the cytochemical staining for the diagnosis of acute leukemia. PATIENTS AND METHODS: Test results were evaluated in 122 (n = 122; 112 with acute lymphocytic leukemia and 10 with acute myeloid leukemia) patients selected from among 320 patients with acute leukemia at Texas Children's Hospital in 1997 and 1998. Results were selected for review if the clinical encounter represented the initial diagnostic work-up and if data were available from cytochemical staining and flow cytometry studies. RESULTS: Cell lineage classification derived from flow cytometry and cytochemical stains were in agreement in all cases. Definitive diagnoses were feasible using flow cytometry results alone in 120 of 122 patients (98.4%) as compared with only 99 of 122 patients (81.2%) when only cytochemical staining results were considered. In two patients with inconclusive flow cytometry results, cytochemical staining alone provided information sufficient for diagnosis. CONCLUSIONS: Results from this study indicate that with few exceptions, flow cytometry studies alone provide sufficient information for diagnosis and management of acute leukemia in children. Nevertheless, cytochemical staining should be available for those cases in which flow cytometry results fail to allow a definitive diagnosis. A modified testing protocol is recommended.

The metrics process: Baylor's development of a "report card" for faculty and departments.

In 1996, Baylor College of Medicine began the first year of its "metrics process," collecting, analyzing, and reporting data on the performance of each individual faculty member and each department in achieving the school's missions of education, patient care, research, service, and finance. This article is a report of the first two years of the process, with updates about the 1999 process, future plans, and lessons learned. The primary goal of the metrics process is to provide meaningful data to assess and improve the performance of faculty and departments across all missions. The authors (1) indicate the categories chosen, within each mission of the school, for measuring faculty time and effort (e.g., patient care, with or without learners) and state the measures chosen (e.g., percentage of time); (2) describe the development of questionnaires in 1996 and 1997 to acquire data from faculty, in the chosen categories and measures, about the time and effort they spent; and (3) report highlights of the resulting departmental data that were gathered in 1997. Among the key categories and units of measure chosen for measuring faculty (and departmental) time and effort are research grant dollars (total and per research full-time equivalent, or FTE); basic research grant dollars per square foot of laboratory space; percentage of faculty who spend at least 50% of their time in research who are National Institutes of Health principal investigators; numbers of inpatient and outpatient visits per evaluation and management FTE; total relative value units (RVUs) per patient-care FTE; patient-care income/RVU and expense/RVU for total faculty and support staff; percentage of faculty with at least one leadership position in a state or national organization; and income in excess of expense, by mission (e.g., patient care). Results of comparing data from the first two years of the metrics process demonstrate marked improvements in performance for most research measures (i.e., items of measurement agreed upon for the metrics process). The process is continually being redeveloped; the ultimate challenge is to place the objective measurements in a context where less objective qualities (e.g., innovation) also figure importantly in the evaluation and fostering of excellence. The metrics process is providing important management data, encouraging significant discussions among faculty and chairs about performance and accountability, and aiding greatly in departmental goal-setting and ultimately in determining the overall performance of the school.

A conceptual framework for managing clinical processes.

Reengineering of the health care delivery system is underway, as is the transformation of the processes and methods used for recording information describing patient care (i.e., the development of a computer-based record). This report describes the use of object-oriented analysis and design to develop and implement clinical process reengineering as well as the organization of clinical data. In addition, the facility of the proposed framework for implementing workflow computing is discussed.

Blood lead levels in a continuity clinic population.

INTRODUCTION: Lead toxicity is well recognized as a significant cause of morbidity in children, especially those under the age of six years. While lead toxicity has not been recognized as a public health problem in Houston, it is possible that children in the area suffer from low-level lead effects on the central nervous system. OBJECTIVES: To detect asymptomatic cases of lead toxicity in one population of Houston children and to assess the effectiveness of the lead risk questionnaire. DESIGN: Venous blood samples for quantitative lead were analyzed utilizing the Anodic Stripping Voltameter. The Centers for Disease Control's lead risk assessment questionnaire was administered to each patient. SETTING: Baylor College of Medicine Continuity Clinic at Texas Children's Hospital. SUBJECTS: All patients, ages 9-72 months, seen for routine care between December 1992 and June 1994 were screened once. RESULTS: Blood lead levels were obtained on 801 children; all but 47 completed lead risk questionnaires. The mean age of the study group was 2.37 years (SD 1.84) and they were 54% male. They were 39% Hispanic, 39% Black, and 18% White. Eighty-eight percent reported an annual income of < $20,000. They lived in 127 separate zip codes. Twenty-five (3.1%) patients had elevated blood lead, 21 between 10-14 micrograms/dL and 4 between 15-19 micrograms/dL. No patients had blood lead levels of > or = 20 micrograms/dL. No statistically significant differences were found between patients with blood lead < 10 micrograms/dL and those with > or = 10 micrograms/dL when comparing for age, sex, ethnicity, income, and zip code. Only those children living in or regularly visiting a pre-1960 home with peeling or chipping paint were significantly more likely to have elevated blood lead (p = .045). CONCLUSION: Although the majority of children in our setting were poor and urban, the prevalence of blood lead > or = 10 micrograms/dL was 3.1%, well below the estimated 17% quoted by the Centers for Disease Control in recommending stringent screening guidelines. The lead risk assessment questionnaire failed to identify 32% of children with elevated blood lead levels. Since this questionnaire is critical to screening populations at low risk for lead toxicity, it is important to determine whether a revised questionnaire or a more careful elicitation of parental responses will improve identification of those children at risk.

Workflow computing. Improving management and efficiency of pathology diagnostic services.

Traditionally, information technology in health care has helped practitioners to collect, store, and present information and also to add a degree of automation to simple tasks (instrument interfaces supporting result entry, for example). Thus commercially available information systems do little to support the need to model, execute, monitor, coordinate, and revise the various complex clinical processes required to support health-care delivery. Workflow computing, which is already implemented and improving the efficiency of operations in several nonmedical industries, can address the need to manage complex clinical processes. Workflow computing not only provides a means to define and manage the events, roles, and information integral to health-care delivery but also supports the explicit implementation of policy or rules appropriate to the process. This article explains how workflow computing may be applied to health-care and the inherent advantages of the technology, and it defines workflow system requirements for use in health-care delivery with special reference to diagnostic pathology.

Laboratory computing--process and information management supporting high-quality, cost-effective healthcare.

One currently observes many healthcare institutions rushing to reengineer and install information systems with the expectation of achieving enhanced efficiency, competitiveness, and, it is hoped, higher patient satisfaction resulting from timely, high-quality care. Unfortunately, information system concepts, design, and implementation have not yet addressed the complexity of representing and managing clinical processes. As a result, much of the synergy one might expect to derive from understanding and designing clinical processes to gain efficiency and quality while maintaining humanness is not readily achievable by implementing traditional information systems. In this presentation, with laboratory services as an example, we describe a conceptually different information systems model, which we believe would aid care-givers in their efforts to deliver compassionate, quality care while addressing the highly competitive nature of market-driven healthcare.

A server architecture for ambulatory patient record systems.

Baylor College of Medicine is developing the Collaborative Social and Medical Services System (CSMSS) to provide a computerized patient record (CPR) system for the Teen Health Clinics (THC) [1]. The THCs consist of five geographically distributed clinics providing health care and social services to teenagers in the Harris County Hospital District. The CSMSS is the first application built upon the Ambulatory Systems Architecture (ASA) [2]. The ASA is aimed at providing an architecture and application framework for the development and deployment of CPR systems in ambulatory care settings. This paper describes the ASA server architecture.

A proposed architecture for ambulatory systems development.

We have developed an architecture and application framework known as the Ambulatory Services Architecture (ASA) for computerized management of patient records for ambulatory care. Our primary design goals included the development of an architecture that will be readily adaptable to advances in technology needed to enhance computerized patient record systems (e.g., multimedia, human-computer interfaces such as voice-to-text, and a fully distributed objects, etc.) and a data model and database implementation capable of providing the flexibility and extensibility needed to support a broad spectrum of specialty medical practices. This report describes the data model, access control, and the client/server components of the Ambulatory Services Architecture.

Collaborative social and medical service application.

Baylor College of Medicine has five Teen Health Clinics (THC) dispersed throughout Harris county. The population served by the clinics includes inner-city adolescent boys and girls 19 years of age and under. Patients receive services such as family planning, sexually transmitted disease screening and treatment, perinatal care, counseling, and support services. Adolescents may receive services at any one of the clinics at no cost to the adolescent or their dependents. Given the geographical distribution of the clinics and the reliance on paper-based records, client services cannot be provided efficiently or expeditiously. According to the statistics developed by Clinic staff, ineffective coordination of service needs and client schedules undermine the follow-up needed for effective care. For example, a counselor will often need to balance a school schedule, clinic visits, well baby follow-up, and the Best Friends Program for a new mother. In addition, the lack of ready access to patient information impairs the ability of clinical and social service staff to provide continuity of care. In fact, some cases of client dropout are attributable to these difficulties. We have developed the Collaborative Social and Medical Service Application (CSMSA) to facilitate the provision of social and medical services to this population. The CSMSA is a domain-specific application based on a robust infrastructure known as the Ambulatory Services Architecture (ASA). This system is designed to support integrated social and ambulatory care. The ASA is a Baylor developed application framework and architecture for the computerization of the patient medical record in the ambulatory care setting. The working environment for the CSMSA user is an integrated desktop which provides an operating environment for both third-party applications and the CSMSA, as well as a fundamental set of services. The integrated desktop services include a mechanism for object organization or grouping, a facility for the management of desktop objects including disposal and storage, and an embedded search utility to assist in the location of desktop objects as well as other application objects. The access control mechanism will provide the security for the desktop environment by requiring the user to log into and out of the environment. This security mechanism will also enable/disable CSMSA tools based on the user's role at the clinic. The CSMSA is designed to provide work process, functionality and data access appropriate to the responsibilities of the user, e.g., a THC clerk may have permission to view a patient's orders but does not have the facility to create an order. Patient context is defined through a patient browser containing the master patient index or a user defined patient list. Patient encounters are managed using forms based on the process for each encounter type (e.g., initial visit). Forms are used for data entry and for queries. Data entry forms are only committed to the patient database (i.e., making the patient data available to other users who have permission to access the data) when the user approves the data. The CSMSA provides a means to suspend an entry into a chart by saving incomplete or non-committed data entry forms as part of the desktop state which is restored when the user logs into the system. CSMSA was designed to provide a robust, expandable application capable of accommodating the changes in health care and social services delivery while encompassing evolving software standards and new technology. This goal was achieved using object-oriented methodologies and technologies, combined with an object-oriented database management system as the foundation of our server to facilitate the evolution of our data model.

Cost-effectiveness analysis for evaluation of screening programs: hereditary hemochromatosis.

A significant body of research over the last 10-20 years supports the hypothesis that screening for hereditary hemochromatosis (HH) may be cost-effective, given the low-cost, low-risk therapeutic options available for most homozygous individuals. The factors that confound a straightforward test of this hypothesis include the fact that the disease is not fully penetrant and that, to achieve the anticipated life-year gains, therapy must be instituted before disease complications become irreversible. Recent articles and editorials, as well as practice guidelines prepared by the College of American Pathologists, recommend screening for HH with transferrin saturation and ferritin testing, and with percutaneous liver biopsy for those with positive laboratory test results. Patients at risk would be treated with phlebotomy for life and monitored with ferritin testing. We present a cost-effectiveness analysis that evaluates the efficacy of using a screening strategy to accomplish the desired healthcare goals.

Polymerase chain reaction to detect cytomegalovirus DNA in the cerebrospinal fluid of neonates with congenital infection.

To determine if cytomegalovirus (CMV) DNA could be detected in the cerebrospinal fluid (CSF) of infants with congenital infection with CMV, polymerase chain reaction (PCR) was done on CSF samples from 13 infants (10 with confirmed, 1 with possible, and 2 with asymptomatic congenital CMV infection) and on CSF samples from 100 control patients with alternative diagnoses. By use of two sets of primers that targeted different areas of the CMV genome, CMV DNA was amplified in the CSF of 6 of 10 symptomatic infants, 0 of 3 infants with asymptomatic CMV infection or possible congenital CMV disease, and 2 control patients (one primer set only). A positive CSF PCR result at birth correlated with a poor neurodevelopmental outcome (P = .048; two-tailed Fisher's exact test).

Facilitation of health care delivery through enhanced communications.

To realize the advantages of advanced computing technology in medicine we will have to blend computing and communications facilities into a seamless entity that can support ubiquitous computing for medical use. Work-group computing applications, high speed networks linking a multiplicity of servers, and mobile computing platforms can provide the next generation of devices and services for the physician. Ultimately, the biggest challenge for communications and computing in medicine will be found in the requirement for standards in semantics and syntax in the highest levels of the OSI model. We present here our view of the functional requirements for physician communications and computing to meet standard medical practice in Western societies.

Analysis, requirements and development of a collaborative social and medical services data model.

In any medical and social service setting, patient data must be readily shared among multiple providers for delivery of expeditious, quality care. This paper describes the development and implementation of a generalized social and medical services data model for an ambulatory population. The model, part of the Collaborative Social and Medical Services System Project, is based on the data needs of the Baylor College of Medicine Teen Health Clinics and follows the guidelines of the ANSI HISPP/MSDS JWG for a Common Data Model. Design details were determined by informal staff interviews, operational observations, and examination of clinic guidelines and forms. The social and medical services data model is implemented using object-oriented data modeling techniques and will be implemented in C++ using an Object-Oriented Database Management System.

The development of a data security model for the Collaborative Social and Medical Services System.

This paper presents the development of the Collaborative Social and Medical Services System's (CSMSS) data security mechanism. This mechanism was synthesized from an analysis of the CSMSS problem domain, and from a study of the methods used by modern operating systems and database management systems. The resulting mechanism is more flexible and expressive than traditional access control methods and is generally applicable to the management of privacy and multi-provider access.

The development of a client application for the collaborative social and medical services system.

This paper describes the design and implementation of a client application for the Baylor College of Medicine Teen Health Clinics. The application is the front end to the Collaborative Social and Medical Services System (CSMSS) under development by Baylor's Medical Informatics and Computing Research Program [8]. The application provides distributed access to an underlying object oriented database system. A process driven and patient centered design will provide staff members with a complete set of services, including forms for data entry and viewing, query, and access management to facilitate efficient and effective delivery of services. Role-specific interfaces will be supplied for clerks, nurses, nurse practitioners, physicians, and social workers. The client application is being designed using object oriented methodologies and technologies with the C++ programming language, and will operate within a Microsoft Windows operating environment utilizing Object Linking and Embedding for application interoperability.

Rapid diagnosis of herpes simplex virus encephalitis by using the polymerase chain reaction.

To determine the diagnostic value of the polymerase chain reaction (PCR) in establishing the rapid diagnosis of herpes simplex virus encephalitis (HSE) in the pediatric age group, we performed PCR to detect herpes simplex virus (HSV) in the cerebrospinal fluid of 8 neonates with HSV infection (4 with central nervous system involvement), 11 infants and children with suspected HSE (4 proved, 1 presumed, 6 excluded), and 105 control patients who had cerebrospinal fluid obtained as part of the evaluation for other diagnoses. The HSV DNA was amplified and typed by using primers specific for the DNA polymerase gene of HSV types 1 and 2. Herpes simplex virus DNA was detected in the cerebrospinal fluid of 3 of 4 neonates with CNS involvement (all with HSV type 2) and 3 of the 4 patients with proved HSE (all with HSV type 1). No HSV DNA was detected in the 4 neonates without CNS disease, the 1 patient with presumed HSE, the 6 patients who had HSE excluded from the diagnosis, and the 105 control patients. Overall, HSV PCR had a sensitivity of 75%, a specificity of 100%, a positive predictive value of 100%, and a negative predictive value of 98%. These results indicate that PCR is a useful noninvasive test in establishing the diagnosis of acute HSE, but a negative result did not exclude the diagnosis.

The diagnosis of CMV pneumonitis in lung and heart/lung transplant patients by PCR compared with traditional laboratory criteria.

Polymerase chain reaction (PCR) amplification of CMV DNA recovered from bronchial alveolar lavage (BAL) and peripheral blood samples was compared with tissue culture, cytology, and/or histology for the earlier detection of CMV pneumonitis in 12 recipients of single-lung or heart/lung transplants. In patients with confirmed CMV pneumonitis, cytological evidence of CMV disease in BAL samples was detected 38 +/- 14 days posttransplantation, while tissue culture and PCR-positive results were noted as early as 30 +/- 4.0 days and 18 +/- 4.6 days, respectively. While PCR was positive earlier than culture in a number of cases, culture-positive results were subsequently obtained in each case, consistent with earlier detection of viral replication by PCR as opposed to detection of latent virus. CMV was detected by PCR in 6 of 24 blood samples from patients with confirmed or suspected CMV pneumonitis, while results of all 24 blood samples were negative when assayed by tissue culture. PCR-based testing was more sensitive than traditional tests, allowing detection of viral replication earlier than tissue culture in the posttransplant period. PCR could provide a powerful means of monitoring the immunocompromised patients in whom preemptive therapeutic intervention for CMV disease is desirable.

Informatics. A subspecialty in pathology.

The emerging discipline of pathology informatics is reviewed, and its placement as a subspecialty within the broader field of pathology and laboratory medicine is proposed. Informatics concepts should guide the development of the next generation of laboratory information systems. Advanced laboratory systems will incorporate decision support leading to improvements in quality and in interpretive reporting providing support for clinical diagnosis and decision making. Training programs in pathology should take into account the need for expertise in informatics and develop fellowships in this area to adequately prepare junior faculty members for their future professional role.