Automation of key functions promises to dramatically improve the way that laboratories operate in the future. Internal operations, epidemiologic surveillance, communication with private laboratories, and use of test results could all be enhanced through the implementation of consistent laboratory information systems (LISs). In the private sector, many laboratories are investing in the infrastructure necessary to communicate with MCOs and other important constituencies. A critical question from the perspective of this study is the degree to which PHLs are involved in automation of information systems.
The vision for the future of automated LIS in the context of PHLs is that of one element in a seamless web of public health data. Communication of test results and other information between the PHL and both public and private sector organizations will be accomplished through automated electronic transactions. An automated system would also be linked into sophisticated epidemiology information that would assist in tracking the geographic patterns associated with outbreaks. This vision was articulated by Washington State in their Draft Information Services Plan, as reproduced in Figure 14.
Figure 14: Future Electronic Laboratory Data Network, Five Year Vision
The CDC has recognized the public health need to improve information infrastructure, and has articulated the need for involvement from laboratories. The CDC's vision for integrated disease information reporting is discussed in MMWR (February 16, 1996), and grants to improve information infrastructure have been established through the National Center for Infectious Disease (NCID) and Information Network for Public Health Officials (INPHO) program.
Some leadership in this area has also emerged from a coalition between CDC, CSTE, and ASTPHLD focusing on electronic reporting of clinical laboratory data. In March 1997, this group convened to develop draft recommendations for standardization of electronic reporting of laboratory results, which articulated problems and solutions in the areas of data flow, message format, and message content. These draft recommendations have been formally approved by CSTE and ASTPHLD, and have also been reviewed and accepted by the Health Information Systems and Surveillance Board (HISSB) within the CDC.
Our interviews suggest that there may be a growing gap between the development of needed infrastructure in the public and private sectors. Despite the CDC's current programs, development activity in PHLs is not robust. While some PHLs operate LISs to automate information transfer within the laboratory, few reported the ability to efficiently report test results to providers, other public health agencies and the CDC. Many PHLs indicated they still used paper for reporting laboratory results, and some PHLs were only recently beginning to computerize the process of billing customers for testing services associated with a fee.
The following section reviews the status of PHL information systems development relative to that of the private sector. We begin by describing the status of a common laboratory reporting nomenclature B a prerequisite to automation. We then describe current activity along three dimensions: (1) internal laboratory information systems; (2) inter-laboratory communication; and (3) creation of integrated knowledge bases.
Universal Coding Systems for Clinical Laboratories
The creation of a standardized set of codes to describe laboratory tests and their results is a necessary prerequisite to efficient information exchange, and there have been a number of efforts to establish such codes. While laboratory services have been included in the establishment of general medical standards (e.g., ASTM, SNOMED, HL-7), the logical observation identifier names and codes (LOINC) database merits special mention.
LOINC is a public use set of codes and names for electronic reporting of clinical laboratory results developed by a committee consisting mostly of pathologists and laboratorians representing medical schools and commercial reference laboratories.19 The ultimate aim of LOINC was to create new efficiencies in the reporting of test results across institutions in an era of integration and managed care. The original goal of the committee was to develop a system of codes that would encompass at least 98 percent of the average laboratory's results, and would include an appropriate level of descriptive specificity. Tests that occupied their own columns on a clinical report or had clinical significance distinct from other closely related tests were assigned separate names and codes under the system.
Each test coded under the LOINC system is designated by a long test descriptive code and an associated numerical code. In addition to these descriptive and numerical codes, the LOINC name given to each test is associated with a six part description of the test, which includes the following name fields:
- Analyte/component name (i.e., what is the test looking for -- potassium, hemoglobin, Hepatitis C antigen)
- Challenge information (e.g., amount, route and timing of challenges)
- Special adjustments (e.g., pH)
- Special distinctions about the relationship of the measured entity to the patient
- Kind of property measured or observed (e.g., blood product unity, mass, mass concentration, enzyme concentration)
- Time aspect of measurement or observation (e.g., point observation v. observation integrated over time)
- System/sample type (e.g., urine, blood, serum)
- Type of scale of the measurement or observation (e.g., quantitative, semi-quantitative, or qualitative)
- Type of method used to obtain measurement or observation
To date, LOINC has been used predominantly as a means for communicating laboratory result information among hospitals, commercial laboratories, and providers who use the HL7 system as a convention for transmitting data to and from different computer systems; it has also been widely used throughout the LIS community and has been accepted by the American Clinical Laboratory Association (ACLA) as the recommended standard universal code for submitting laboratory results. Included in the list of institutions currently using LOINC to transmit laboratory data are:
- MetPath Laboratories
- Mayo Medical Laboratory
- ARUP Labs
- Quest Diagnostics Incorporated
- Laboratory Corporation of America
- LifeChem Laboratory Services
- University of Colorado
- University of Missouri
- Indiana University Laboratories
- Intermountain Healthcare
- LDS Hospital
The advantage committee members claim LOINC has over other systems for coding laboratory results is its multi-axis approach for identifying laboratory tests. Other laboratory test coding systems (e.g., SNOMED) lack, in some cases, the level of detail necessary in making clinically significant distinctions between the results of similar tests. In addition to the six part name description, the LOINC database contains 23 additional fields, including, among other things, synonymous terms allowing linkage to other systems for coding laboratory results.
While standardization of laboratory test reporting nomenclature through LOINC has been touted for its potential role in helping public health officials work with laboratory data for epidemiological purposes, it has not had an observable effect on computing within PHLs. Of the 46 members of the LOINC Committee, only one represented a state health agency, and an interviewee from the LOINC Committee was unaware of any specific activity linking LOINC to the recovery of utilization data for public health purposes.
Internal Laboratory Information Systems
Virtually all interviewees believe that automation can increase the efficiency of a laboratory. Automation of the commercial reference laboratory market is a well established industry whose vendors include Cerner Corporation, Citation, MediQuest, SMS, and SunQuest Information Systems, Inc. Commercial LIS systems were originally designed to assist automation and efficiency in reporting results back to providers. Most recently, LIS operators have focused on helping reference laboratories adapt to the demands of their managed care customers. One LIS focus involves helping laboratory service vendors use robotics to automate testing and reporting functions centralized on a regional or national level. Another, somewhat competing focus, involves satisfying the increasing demand for point of care testing facilities, which are monitored for quality through a central LIS system and are able to transfer data to and from a central database.
A few products are specifically targeted to PHLs. Epic Systems Corporation is one company offering such services to PHLs (current clients include the state PHLs in Michigan, South Dakota, Wisconsin and Texas). The standard LIS product is able to track specimens from collection to result reporting, and can produce detailed reports using statistics generated from the data. Additionally, the Epic LIS system's open architecture allows for the creation of additional fields and screens for the integration of data from non-laboratory sources (e.g., medical or pharmacy data). A representative from Epic reported that most of its PHL clients are able to use the open architecture to support broader data integration efforts within their state health agencies.
Some states have not relied on comprehensive products, but rather have developed the LIS environment internally (e.g., Washington). There are also a few more products for specialized laboratory areas. For example, Washington State recently purchased NeoMetrics, an automated system for use by PHLs in tracking neonatal screening samples (16 other states have this software currently). Finally, the CDC has also created a Laboratory Information Tracking System (LITS) for use by PHLs.
The State of Automation in Tennessee
Most environmental testing is automated. However, with the exception of newborn screening most microbiology testing is not automated. In part, this is because the more comprehensive testing offered by the state PHL does not readily lend itself to automation. In addition, the PHL's ability to obtain new equipment has been hampered by budget constraints (last major capital acquisition was in 1995).
The PHL does not use an LIS for its microbiology services. The laboratory is developing a contract with the company that created the Patient Tracking Billing Management Information System (PTBMIS) to obtain a system that can electronically transmit billing, ordering, and lab test information. The system will give the state PHL connectivity with its clients, including the county health departments and the CDC (for disease reporting). The acquisition of this system will depend on decision-making in the state's IRM budget planning.
The environmental side used its equipment budget to develop a LMIS, which has slowly replaced paper based records within the laboratory (including branch laboratories). Currently, the system is used to produce paper reports, which are disseminated to lab clients.
Despite the current lack of automation, the PHL reports no complaints about its turnaround time. Several years ago, the environmental side was generating complaints over slow turnaround times, but it has since corrected this problem.
An important element of the public and private laboratory function is to communicate with other laboratories, health care practitioners, and public health officials. Electronic reporting of laboratory test results to the ordering physician has become routine in most large hospitals. But this function is also critical in PHLs, since much of the testing is referred from other laboratories, and a primary role is to support epidemiological surveillance.
The CDC has made efforts to help state and local public health laboratories comply with its own disease surveillance projects through the development of the Public Health Laboratory Information System (PHLIS).20 PHLIS is information transfer software that allows officials at SHAs, local health agencies, and PHLs to electronically report information to the CDC in support of CDC surveillance programs. Additionally, PHLIS allows the CDC to issue tables and summary reports to all sites. The modest requirements for operating PHLIS (a PC with 512 kilobytes of memory and a modem) increase its potential effectiveness as a tool for collecting and disseminating information on disease incidence and potential outbreaks for public health laboratories with dissimilar or non-existent laboratory information systems.
Currently, PHLIS is used primarily by local and state PHLs and the CDC to communicate test results electronically. Within SHAs, the CDC has developed the NETTS system, which state epidemiologists use to transmit information about reportable diseases originating from PHLs, private sector laboratories, or other sources.
PHLIS operates using a modular approach, with a separate module dedicated to the subject of each CDC disease surveillance program. Modules are designed uniquely, and users interfacing with a particular module receive prompts detailing what information to enter and where within the module to enter it. Decisions regarding what types of information would be required for a given program were made by the CDC in conjunction with ASTPHLD and other medical epidemiologists. After entering the information required in a given module, the user is able to select "SEND TO CDC," which prompts the PHLIS to call a central communications computer, log on by responding to identification prompts, and transfer data. PHLIS currently includes modules for CDC programs designed to track incidence of Salmonella, Shigella, Campylobacter, and Mycobacteria, and all state and territorial public health laboratories have agreed to participate in the program.
The Washington State PHL is also involved in the creation of an innovative system for electronic disease reporting between private laboratories and the state health agency. The goal of the system is to automate reporting in hopes of lessening the burden of reporting, improving accuracy, and expediting reporting. The Office of Epidemiology has received a grant from the National Center for Infectious Disease (NCID) at the CDC to pilot the system. The Office of Epidemiology is working closely with the PHL, Group Health Cooperative of Puget Sound, and Laboratory Pathology Associates on clinical and technical issues involved in automated reporting. Current issues being addressed by this project include establishing a mechanism for data transfer between private sector and public sector organizations, protocols for data recording, and encryption / confidentiality of records.
Creation of Integrated Knowledge Bases
Many laboratorians interviewed suggested that the future of the commercial reference laboratory industry centered on the development of sophisticated informatics products that will be able to integrate laboratory results data with outcomes data from other health system players. Some large clinical laboratory companies have reported making large investments in information technology. For example, Quest Informatics has reported investing $20 million in relational databases in recent years. A fundamental issue for these systems is the lack of coordination with other sources of patient data (e.g., hospital, physician, and prescription pharmaceutical observations).
Despite some limitations, major clinical reference laboratory providers are already actively marketing their ability to provide software that allows MCOs to aggregate and index laboratory data, which is useful for assessing overall health status among members and developing disease management programs for specific therapeutic areas. In addition to the software products, laboratory services vendors may offer consulting services designed to implement Aquality improvement opportunities identified by their software. While it is likely that clinical reference laboratories will continue to market informatics products, the effectiveness of their tools have yet to be widely tested in health markets.
The longer-term vision for PHLs in the context of the broader public health safety net is to develop a comprehensive and integrated LIS that would facilitate epidemiologic surveillance as well as public health case support.21 The vision includes development of common community health indicators that could be monitored on a real-time basis to improve public health. Such a system would clearly require cooperation from other offices within a Department of Health (e.g., Epidemiology), as well as other state organizations involved in health (e.g., Medicaid) that track patient information. Some states, such as Washington, Missouri, and South Carolina, have created working groups to support realization of this vision.
An example of current PHL involvement in information transfer technology is the internet effort of Iowa's University Hygienic Laboratory, which serves as the state's PHL. While Iowa's site is still in development, it currently includes a description of the laboratory's activities, as well as a state-wide disease reporting system for respiratory tract pathogens. The system allows physicians using the site to click on the county in which they practice in order to view an updated list of the number of cases of respiratory infections reported in the county, along with the pathogens that caused the infections.
Other state PHLs are developing a presence on the web through SHAs, including internet accessible systems to provide health status indicator information to medical providers and local health agencies. These efforts have been monitored by the NCVHS. An example of this type of activity can be found at the Massachusetts State Laboratory Institute website.22
Environmental Scan Conclusions
The growth of managed care is forcing clinical laboratories to produce rapid, efficient testing services and to focus on cost reduction. The clinical testing services market is highly competitive, with large independent reference laboratories increasing the standard in terms of technology, speed of testing, and price.
Public health laboratories need to be agile in adapting to change. They are also challenged by governmental change, such as fluctuations in state funding streams for laboratory services and change in expectations vis-à-vis personal testing in the Medicaid population. Other environmental factors such as the computerization of clinical laboratories and the federal government role in PHL-related issues will continue to challenge the ability of PHLs to accomplish their objectives. These factors contribute to the climate of uncertainty reported by many PHL directors, and underscore the importance of our key study questions