Mark LoBue

The PDMS as a focal point for distributed patient data

SummaryDistributed data links are essential requirements for a successful patient data management, system (PDMS). These links funnel clinically needed data into bedside ICU workstations. We have constructed four data links which acquire most of the objective data required for direct patient care. Data from bedside monitors is acquired via a standard HP Signal Distribution Network. Urine volumes and core bladder temperatures are acquired over a link to 20 electronic urimeters. Clinical laboratory data is obtained over an HP General Purpose Data Link (GPDL) to a VAX 11/785 laboratory system. Blood gas data is obtained over a second GPDL link to a DEC 11/23 computer. The ICU staff is notified of incoming lab results with bedside video messages. Combined with automated calculations, these data links eliminate thousands of data entry keystrokes daily and allow the PDMS to serve as the focal point for real-time patient care.

Decision support alerts for clinical laboratory and blood gas data.

SummaryWe have designed and implemented a computerized Intensive Care Unit (ICU) decision support alerting system which analyzes all incoming laboratory and blood gas data for critically abnormal values and trends. A computerized patient data management system (HP 78709A PDMS, Hewlett-Packard Co., Waltham, MA) serving 20 Surgical ICU beds is networked to a Clinical Laboratory Information System and a blood gas computer system. The ALERTS subsystem operates on the PDMS as an automatic program triggered by the receipt of fresh laboratory data. Three types of ALERTS are detected: (1) high and low critical values, (2) calculation-adjusted critical values, and (3) critical trends. Once detected, a specific ALERT message is displayed at the bottom of the patient’s bedside PDMS terminal and at the central station.Over an eight month period a total of 1,515 ALERTS were detected from amongst approximately 115,000 laboratory data results transmitted to the Surgical Intensive Care Unit (SICU). Slightly over half of all ALERTS were caused by critical blood gas values. ALERTS were found to be a sensitive indicator of severity of illness: patients with one or more ALERTS suffered an ICU mortality of 9.52%, compared to 0% mortality in patients with no ALERTS. We conclude that automated laboratory data ALERTS represent a valuable decision support tool for the management of high risk ICU patients.

Automatic extraction of intensity-intervention scores from a computerized surgical intensive care unit flowsheet

Systems that objectively score severity of illness and intensity of patient care interventions have been used to guide the appropriate use of intensive care facilities, provide information on nurse staffing ratios, validate subjective classifications of patient illness, and normalize scientific and financial studies for severity of illness. Existing scoring systems require a well-trained observer to perform a thorough chart review to complete manual scoring forms. We have designed a new system in which computerized intensity-intervention scores are automatically extracted from electronic intensive care unit flowsheets, eliminating both manual labor and potential observer variation. In prospective studies, these computerized scores correlated well with manual TISS scores, intensive care unit mortality, intensive care unit length of stay, hospital length of stay, and a subjective classification of patients to graded levels of hospital care. Such automated scores may be used for real-time allocation of health care resources and normalization of prospective studies for severity of illness.

Wireless clinical alerts for critical medication, laboratory and physiologic data

Clinical information systems (CIS) are increasingly employed to manage the information associated with hospital and Intensive Care Unit (ICU) patients. CIS are typically interfaced to a variety of other systems which provide bedside physiologic data, laboratory results and medication information for video displays and reports. However, having all this information together in electronic format provides an opportunity to detect critically adverse patient conditions, which may be complex. The authors have devised a software system which extracts all pertinent information from the CIS on a continuous basis and sends the data through a series of event detection algorithms. These algorithms are configured to detect critically abnormal physiologic and laboratory values, critical trends and critical indicators of drug reactions and side effects. Once an alert is detected, the software system codes it into a readable alphanumeric alert message and automatically sends it to a commercial paging system. Alerts are received on pagers carried by designated physicians and pharmacists who can take immediate actions to reverse the alert condition.

Inferencing Strategies for Automated Alerts on Critically Abnormal Laboratory and Blood Gas Data

A relatively insignificant amount of human thought is required to recognize critically abnormal events. After a few weeks of training on the ward, most medical students can recognize seriously abnormal results of common laboratory tests and take some definitive action, such as calling a supervising physician. The “gestalt” by which laboratory results are appreciated as clinically dangerous is complex and challenging to duplicate in a modern digital computer.

Integrated computerized databases for medical data management beyond the bedside.

SummaryComputers are beginning to be utilized extensively for direct patient care, assisting nursing and medical staff with data collection and review at the bedside. However, most clinical data management systems are optimized for bedside patient care and offer limited resources for multi-patient data analysis. At Cedars-Sinai Medical Center, a network of computer systems has been developed to provide linkages between clinical, administrative and outcome data for Surgical Intensive Care Unit (SICU) patients. Increasingly, such data is needed to evaluate the relationship between severity of illness and patient outcome and the utilization of expensive critical resources.Over the parts 3 years, comprehensive data on 6,755 consecutive SICU patients receiving 18,394 days of care have been accumulated by our PDMS. Using linkages constructed to other hospital systems and databases, trends for severity of illness, severity adjusted survival, census, bed utilization, nursing utilization and many other parameters have been constructed. These linkages are valuable in documenting cost-effective and medically-effective patient care practices.

An automatic PDMS interface for the Urotrack Plus 220 urimeter

SummaryPeriodic measurements of urine volume and temperature in critically ill patients are time consuming, unclean procedures. These measurements may be automated with an electronic urimeter (Urotrack Plus 220, C.R. Bard Company, Murray Hill, NJ). This device contains an RS-232 output port which transmits a complete device status report once per second. We interfaced 20 Urotrack Plus urimeters to a single I/O port of a computerized patient data management system (Hewlett-Packard 78709A PDMS, Hewlett-Packard Company, Waltham, MA). This interface required daisy chained controllers for port switching and a communications adapter for flow control. The urimeters have proven to be cost-effective, labor-saving devices. The PDMS interface provides a continuous display of measured variables and completely automates data entry for flowsheet documentation. Automatic urimetry data acquisition is estimated to save approximately 27 nursing minutes per patient per day.

The effect of surgical ICU triage patterns on differing severity adjusted outcomes in France and the United States

IntroductionSurgical patients treated in French intensive care units (ICU’s) appear to have higher mortality rates than patients in the United States. We hypothesized that this may be due to the French practice of not transferring dying patients from the ICU. We wished to determine if the different mortality rates could be explained by transfer practices for dying patients or other factors such as severity of illness.MethodsFlowsheet data for 6,787 consecutive surgicial ICU (SICU) patients from our institution over a 31 month period was entered into an ICU Clinical Information System which calculated the Day 1 Simplified Acute Physiology Score (SAPS) for each patient upon admission to the SICU. SICU and overall hospital mortality data were matched with severity data and the complete data set was analyzed against results for 2,604 surgical patients in French ICU’s. Since terminally ill patients in France are not transferred to floor care, we also compared the French ICU mortality rate with both our SICU mortality rate and combined SICU and surgical floor mortality rates.ResultsOur overall SICU mortality was 1.7% and our combined SICU and hospital mortality was 4.2%, while the French ICU mortality was 14.1%. The French ICU’s had more patients with higher severity of illness as measured by SAPS. When the effects of ICU transfer practices and severity of illness were considered, there were no mortality differences seen among patients admitted to the different units after elective surgery. Significant differences in mortality were seen when patients admitted emergently were studied.ConclusionsThe differences in severity adjusted ICU mortality between French ICU’s and our SICU are explained by different triage practices for terminally ill patients following elective ICU admission. These triage differences do not fully explain the mortality differences seen among patients emergently admitted to the ICU. Other factors such as the presence of trauma, ICU staffing practices, patient mix or other unidentified factors may be responsible for the severity adjusted differences in mortality among emergency surgical ICU patients.

Quality Assurance and Utilization Assessment: The Major By-Products of an Intensive Care Unit Clinical Information System

The power of computerized Clinical Information Systems (CIS) has yet to be tapped by most hospital Quality Assurance (QA) and Utilization Review (UR) departments. The CIS provides an economical and reliable means by which key clinical data can be extracted from the electronic chart and utilized for quality and utilization analyses. In comparison with current manual methods of extracting data by chart audits, the electronic method is not only faster, it also allows for every chart to be audited against standards for efficiency and quality of care. The science of industrial quality management is well known and appreciated in most other industries—many agree that the time is at hand for using these techniques in health care institutions [1]. The Joint Commission for Accredation of Health Care Organizations (JCAHO), the Health Care Financing Authority (HCFA), and other regulatory agencies now require detailed information and trends about outcomes that can not be easily obtained by traditional, tedious methods of manual chart review. However, this volume of data can objectively be extracted from the electronic record provided by a comprehensive CIS. A reduction in the number of hours spent by QA and UR nurses culling data from charts could be channeled into more meaningful activities of data interpretation and reporting. In this paper we describe the use of CIS-derived data for secondary QA and UR activities. ICUs that have a CIS are ready to enjoy the benefits such a system can provide for daily monitoring of patient care and resource activities.