Reinventing patient-centered computing for the twenty-first century.

Despite evidence over the past decade that patients like and will use patient-centered computing systems in managing their health, patients have remained forgotten stakeholders in advances in clinical computing systems. We present a framework for patient empowerment and the technical realization of that framework in an architecture called CareLink. In an evaluation of the initial deployment of CareLink in the support of neonatal intensive care, we have demonstrated a reduction in the length of stay for very-low birthweight infants, and an improvement in family satisfaction with care delivery. With the ubiquitous adoption of the Internet into the general culture, patient-centered computing provides the opportunity to mend broken health care relationships and reconnect patients to the care delivery process. CareLink itself provides functionality to support both clinical care and research, and provides a living laboratory for the further study of patient-centered computing.

Cytomegalovirus immune globulin after liver transplantation: a cost-effectiveness analysis.

OBJECTIVE: Cytomegalovirus (CMV) immune globulin (CMVIG) has been shown to significantly reduce severe CMV-associated disease complicating orthotopic liver transplant (OLT). We evaluated the economic impact of severe CMV-associated disease and calculated the marginal cost-effectiveness (C/E) of routine prophylaxis with CMVIG after OLT. DESIGN: C/E analysis. SETTING: Four teaching hospitals in Boston. PATIENTS: Patients who underwent OLT from January 1988 through June 1990. MEASUREMENTS: We gathered actual cost data of hospital care for patients enrolled in a clinical trial of CMVIG prophylaxis in OLT. We calculated average outpatient expenses from a separate group of patients undergoing OLT and developed a regression model to estimate costs during the first year post-transplant (R2 = 0.77). Based on this model, we calculated variable costs (in 1999 US dollars) for all patients in the randomized trial. From the published literature we obtained the probability of CMV outcomes and of long-term survival after OLT. We then developed a decision analytical model to determine an incremental C/E ratio, using a Markov simulation to estimate long-term survival and long-term costs. We discounted costs and life-years at 3% and 5% per yr. RESULTS: Based on the efficacy rate of 54% in the controlled trial, we estimate that CMVIG will increase life expectancy by 0.65 discounted years at an additional cost of $11,600, providing a marginal C/E ratio of $17,900/yr life saved. Examining the confidence limits of efficacy, we estimate that CMVIG will have a marginal C/E ratio of $66,200 gained/yr at an efficacy of 11% and $14,000 gained/yr at an efficacy of 83%. CONCLUSION: After OLT, prophylactic CMVIG has an incremental C/E ratio comparable to that of other well-accepted medical therapies and should be used routinely in these patients.

Knowledge-mediated retrieval of laboratory observations.

Intelligent medical applications including agents, clinical decision support systems, and expert systems can benefit from components that expose the meanings of medical concepts. We have endeavored to create an ontology for laboratory observations and to make the ontology accessible in a distributed environment through a knowledge mediator offering several services. To date we have created two such services, one service to mediate the retrieval of laboratory observations and an auxiliary service to facilitate the mapping of units of measure to LOINC property-types. We report progress and insights on the development of our ontology and related knowledge mediator.

A component-based problem list subsystem for the HOLON testbed. Health Object Library Online.

One of the deliverables of the HOLON (Health Object Library Online) project is the specification of a reference architecture for clinical information systems that facilitates the development of a variety of discrete, reusable software components. One of the challenges facing the HOLON consortium is determining what kinds of components can be made available in a library for developers of clinical information systems. To further explore the use of component architectures in the development of reusable clinical subsystems, we have incorporated ongoing work in the development of enterprise terminology services into a Problem List subsystem for the HOLON testbed. We have successfully implemented a set of components using CORBA (Common Object Request Broker Architecture) and Java distributed object technologies that provide a functional problem list application and UMLS-based "Problem Picker." Through this development, we have overcome a variety of obstacles characteristic of rapidly emerging technologies, and have identified architectural issues necessary to scale these components for use and reuse within an enterprise clinical information system.

Validation of clinical problems using a UMLS-based semantic parser.

The capture and symbolization of data from the clinical problem list facilitates the creation of high-fidelity patient resumes for use in aggregate analysis and decision support. We report on the development of a UMLS-based semantic parser and present a preliminary evaluation of the parser in the recognition and validation of disease-related clinical problems. We randomly sampled 20% of the 26,858 unique non-dictionary clinical problems entered into OMR (Online Medical Record) between 1989 and August, 1997, and eliminated a series of qualified problem labels, e.g., history-of, to obtain a dataset of 4122 problem labels. Within this dataset, the authors identified 2810 labels (68.2%) as referring to a broad range of disease-related processes. The parser correctly recognized and validated 1398 of the 2810 disease-related labels (49.8 +/- 1.9%) and correctly excluded 1220 of 1312 non-disease-related labels (93.0 +/- 1.4%). 812 of the 1181 match failures (68.8%) were caused by terms either absent from UMLS or modifiers not accepted by the parser; 369 match failures (31.2%) were caused by labels having patterns not recognized by the parser. By enriching the UMLS lexicon with terms commonly found in provider-entered labels, it appears that performance of the parser can be significantly enhanced over a few subsequent iterations. This initial evaluation provides a foundation from which to make principled additions to the UMLS lexicon locally for use in symbolizing clinical data; further research is necessary to determine applicability to other health care settings.

DecisionNET: database/network support for clinical decision analysts.

DecisionNET is a proposed remote database/wide-area computer network that should cultivate quantitative approaches to medicine by providing rapid access to medical facts, literature citations, decision models, and commentary of specific use to those involved with quantitative clinical decision making. Increasing accessibility to data and providing a forum for exchange of ideas should prove to be a significant asset to clinicians, either as modelers or as clients. Further work will address specific database implementation, development of a query engine, and coordination of the communications infrastructure. DecisionNET provides a unique model for the growth and maintenance of other medical knowledge bases.

Prostaglandin E1 as a screening vasodilator in primary pulmonary hypertension.

Pulmonary vasodilators are variably efficacious in primary pulmonary hypertension (PPH). None has consistently improved hemodynamics enough to obviate the need for complex and potentially hazardous testing of several vasodilators. Prostaglandin E1 (PGE1), a potent, short-acting pulmonary vasodilator, was administered to seven patients with PPH in order to determine whether PGE1 could accurately predict the hemodynamic and gas exchange effects of other commonly used vasodilators. Prostaglandin E1, nifedipine and hydralazine were administered to the patients while measuring pulmonary and systemic hemodynamics and arterial blood gases. Prostaglandin E1 was easily titrated but was inconsistent as a predictor of the effects of the other vasodilators with respect to pulmonary artery pressure, cardiac output and adverse effects on arterial oxygenation. This study suggests that patients with PPH must still receive carefully monitored trials of several vasodilators to determine whether there is a beneficial response and to select the appropriate treatment.

Zone 2 and zone 3 pulmonary blood flow.

In the West model of zonal distribution of pulmonary blood flow, increases in flow down zone 2 are attributed to an increase in driving pressure and a decrease in resistance resulting from recruitment and distension. The increase in flow down zone 3 is attributed to a decrease in resistance only. Recent studies indicate that, besides the pressure required to maintain flow through a vessel, there is an added pressure cost that must be overcome in order to initiate flow. These additional pressure costs are designated critical pressures (Pcrit). Because Pcrit exceed alveolar pressure, the distinction between zones in the West model becomes less secure, and the explanation for the increase in flow even in West zone 3 requires reexamination. We used two methods to test the hypothesis that the Pcrit is the pertinent backpressure to flow even in zone 3, when the pulmonary venous pressure (Ppv) exceeds alveolar pressure (PA) but is less than Pcrit in the isolated canine left caudal lobe. First, PA was maintained at 5 cmH2O, and pressure flow (P-Q) characteristics were obtained in zone 2 and zone 3. Next, with PA still at 5 cmH2O, we maintained a constant flow and measured the change in pulmonary arterial pressure as Ppv was varied. Both techniques indicated that the pertinent backpressure to flow was the greater of either Pcrit or Ppv and that PA was never the pertinent backpressure to flow. Also, our results indicate no significant change in the geometry of the flow channels between zone 2 and zone 3. These findings refine the zonal model of the pulmonary circulation.

Cost-benefit comparison of aerosol bronchodilator delivery methods in hospitalized patients.

We compared two modes of aerosol bronchodilator delivery in 34 patients hospitalized with obstructive airways diseases. The standard mode, therapist-administered up-draft nebulization (UDN), is labor-intensive and therefore relatively costly. The alternative mode, self-administration by a metered dose inhaler (MDI), is less costly, but its efficacy over an entire hospitalization has heretofore not been established. Patients were enrolled after transfer to the pulmonary ward from the emergency room or intensive care units (ICU). We then randomized them to receive metaproterenol q4h either via MDI or UDN. Daily spirometry revealed that MDI and UDN were associated with equivalent bronchodilation initially and equivalent improvement at discharge. The duration of hospitalization for the two groups was also the same. Thus, the two delivery methods were equally effective. We could not attribute this equivalence to pretreatment intergroup differences or to differences in concomitant therapy with steroids, theophylline, other bronchodilators, or antibiotics. Routine use of MDI rather than UDN in all non-ICU adult patients would save $253,487 per year at our institution alone.

The effect of body position on the respiratory rate of infants with tachypnea.

In neonatal disease states where lung compliance is reduced (e.g., inadequate resorption of fetal lung fluid or, surfactant deficiency) an infant's normally low functional residual capacity (FRC) decreases even further. Tachypnea is an efficient compensatory maneuver for the newborn. We evaluated the effect of different bed and body positions on the increased respiratory rate observed in infants with transient tachypnea of the newborn (TTN), infant respiratory distress syndrome (RDS), and bronchopulmonary dysplasia (BPD). Seventeen infants were studied (TTN, n = 6; RDS, n = 6; BPD, n = 5) in four different positions: supine flat, supine elevated, prone flat, and prone elevated. Respiratory rate and heart rate were evaluated in each position. Analysis of variance for the three patient groups showed a lower respiratory rate when the bed was elevated 20-30 degrees compared to flat (P = 0.0001), in the prone posture compared to supine (P = 0.031), and no significant difference in heart rate. The lowest mean respiratory rate occurred when patients were in the prone elevated position. The significant improvement in tachypnea seen in the prone and elevated positions was likely related to improved FRC resulting from reduced cephalad stress on the diaphragm from the abdomen. Positioning neonatal patients with respiratory insufficiency was a simple and safe therapeutic maneuver with prompt and demonstrable benefit.

Relationship of arterial wedge pressure to closing pressure in the pulmonary circulation.

We studied the relationship between the pulmonary artery wedge pressure (Pw) and pulmonary venous pressure (Ppv) at 2 alveolar pressures (PA) in 7 isolated perfused dog lobes. If PA were the critical pressure in the pulmonary circulation, one would expect Pw to equal Ppv for all Ppv greater than PA. Relative to the hilum, the average critical pressure in these lobes was 15.07 +/- 0.40 cmH2O at PA = 5 cmH2O and increased significantly to 17.23 +/- 0.82 cmH2O at PA = 7 cmH2O. Because the critical pressure in fact exceeded PA, Pw was found to be relatively constant and independent of Ppv even when Ppv exceeded PA by 5 cmH2O or more. For example, at PA = 5.13 +/- 0.04 cmH2O and Ppv = 9.64 +/- 0.28, the mean value for Pw was 13.30 +/- 0.59 cmH2O. Pw is equal neither to the average critical pressure nor to PA, but instead lies between these two values. It is determined by the spectrum of closing pressures in the pulmonary circulation, and the time-constants for drainage of beds downstream from the occluded pulmonary arterial branch.

Effect of topical indomethacin and timolol maleate on intraocular pressure in normal subjects.

In a prospective randomized double-masked crossover study in ten healthy volunteers, conducted to determine if the intraocular hypotensive effect of timolol eyedrops was significantly affected by the concomitant use of the topically administered prostaglandin inhibitor, indomethacin, we found that significant ocular hypotension was achieved with timolol alone. This was not changed by the concurrent administration of indomethacin. The results provide further evidence that prostaglandins are not involved in the control of normal intraocular pressure.

Mechanical vibration and conventional chest physiotherapy in outpatients with stable chronic obstructive lung disease.

To assess the effect of conventional chest physiotherapy and mechanical chest vibration on arterial blood gas levels, spirometry, and sputum production, we studied 20 stable outpatients with chronic obstructive lung disease. All patients had severe obstructive ventilatory defects with a mean FEV1/FVC ratio of 30 percent and all produced moderate amounts of sputum. Patients were randomized and received chest physiotherapy or mechanical vibration for 20 minutes. Arterial blood gas levels and spirometry obtained 20 minutes and 40 minutes after completion of the treatment did not show any significant change compared to the baseline. We conclude that neither chest physiotherapy nor chest mechanical vibration improved gas exchange, flow rates, or clearance of secretions in stable outpatients with severe chronic obstructive lung disease.

Pulmonary interstitial compartments and tissue resistance to fluid flux.

We have produced interstitial fluid exchange in six isolated plasma-perfused canine lobes by introducing small increases in microvascular hydrostatic pressure. We measured early fast fluid exchange with a colorimetric technique and used weight changes to follow slow exchange. The observed biphasic time course suggested fluid flux across the microvascular membrane into two interstitial compartments in series (perimicrovascular and central). We related the initial rate of fluid flux into each compartment to the applied hydrostatic pressure change to obtain membrane (Kf1) and tissue conductances (Kf2) and to the exchanged volume to determine perimicrovascular (C1) and central (C2) interstitial compliances. C2 (0.25 +/- 0.193) was twice C1 (0.10 +/- 0.031 ml X cmH2O-1 X g DW-1, where DW represents dry weight. C2 increased significantly with hydration (C2 = 0.06 X WW/DW - 0.15) ml X cmH2O-1 X g DW-1 (WW/DW, wet-to-dry weight ratio), whereas C1 did not. Kf1 (0.26 +/- 0.17) was one order of magnitude larger than Kf2 (0.027 +/- 0.014 ml X min-1 X cmH2O-1 X g DW-1). Kf2 increased with hydration (Kf2 = 0.005 X WW/DW - 0.007) ml X min-1 X cmH2O-1 X g DW-1, whereas Kf1 did not. Our data point to the tissues and not the microvascular membranes as the major rate-limiting structure. Our data suggest an interstitium composed of a smaller rigid perimicrovascular space which communicates to a larger looser downstream space by a high-resistance pathway. As hydration increases, fluid accumulation becomes easier because tissue resistance to fluid flux drops and the compliance of the downstream compartment doubles.

Change in extra-alveolar perimicrovascular pressure with lung inflation.

In eight isolated dog lobes, we examined the change in extra-alveolar perimicrovascular hydrostatic pressure (Pis) due to lung inflation. The vasculature was filled with autologous plasma. Pulmonary arterial and venous lines were connected to a common plasma reservoir. Perimicrovascular volume change (delta Vis), compliance (Cis), and the microvascular filtration coefficient (Kf) were derived from the change in lobe mass over time following a step increase in vascular pressure (Piv). Initially, transpulmonary pressure (PL) was 5 cmH2O and Piv = 0 cmH2O. At constant Piv, two sequential 5-cmH2O increases in PL increased Vis; division of delta Vis by Cis yielded the change in Pis attributable to lung inflation. Cis was 0.035 +/- 0.018 g X cmH2O-1 X g dry mass-1 (mean +/- SD) at PL = 15 cmH2O. Kf was 0.019 +/- 0.023 g X min-1 X cmH2O-1 X g dry mass-1. With inflation from PL = 5 to PL = 10 cmH2O, Pis = -2.15 +/- 1.76 cmH2O; from PL = 10 to PL = 15 cmH2O, Pis = -2.25 +/- 1.50 cmH2O. This perimicrovascular pressure change is very close to the perihilar interstitial pressure change reported by others. Such near equality suggests that the stress of lung inflation is very uniformly applied to the interstitial continuum.

Recent developments in pulmonary edema.

Research on lung fluid balance and pulmonary edema has increased during the last decade. New approaches have led to insights into the role of each component of the alveolar-capillary barrier. The capillary endothelium is the first line of defense against lung fluid accumulation. The interstitium may play a more important role in lung fluid balance than previously appreciated. Active and passive transport properties of alveolar epithelium may be important in the pathogenesis and resolution of alveolar edema. New methods for the determination of epithelial permeability and lung water are being evaluated. The developments reviewed here may have an impact on the institution of new diagnostic and therapeutic approaches to pulmonary edema during the next decade.

The effect of lung edema on pulmonary vasoactivity of furosemide.

Previous data suggest that furosemide improves gas exchange in pulmonary edema by preferential perfusion of nonedematous lung units. To test whether this is a direct effect of furosemide on the pulmonary vasculature as opposed to a secondary phenomenon resulting from the known peripheral effects of this drug, the effect of furosemide on the pressure-flow characteristics of the pulmonary vasculature was studied in six isolated perfused canine lungs with different degrees of gravimetrically determined edema. Furosemide shifted the pressure-flow curve by decreasing the mean intercept or average closing pressure of the pulmonary vascular bed from 13.8 +/- 5.3 to 9.5 +/- 5.4 cm H2O and the zero-flow critical closing pressure from 9.3 +/- 4.3 to 4.7 +/- 3.5 cm H2O (P less than 0.05). The slopes of these curves did not change between control and furosemide treatment. The decrease in intercept and the decrease in zero-flow critical closing pressures were closely correlated with the increase in edema (r = 0.895 for average closing pressure and r = -0.928 for critical closing pressure) (P less than 0.05). Furosemide doubled the pulmonary blood flow in the isolated lobe for the same driving pressure and the greater the amount of lobar edema the less pronounced was this furosemide-associated increase in blood flow. This direct effect of furosemide on the pulmonary vasculature could explain the improved gas exchange seen before a decrease in pulmonary edema, since this pulmonary vasoactivity would result in preferential perfusion of nonflooded alveolar units.

Dopamine, dobutamine, and phentolamine effects on pulmonary vascular mechanics.

The pressure-flow (P-Q) relationship of the pulmonary vasculature, in an isolated canine lobe perfused under classical zone II conditions, can be characterized by a rectilinear segment at high flow, a curvilinear segment at low flow, and a pulmonary arterial pressure (Ppa) that exceeds alveolar pressure at zero flow. This demonstrates the presence of critical closure in the pulmonary vascular bed. Effects of drugs on pulmonary vascular resistance (PVR) must take the normal P-Q relationship into account. We examined the effect of dopamine (D) and dobutamine (DB), alone and in combination with phentolamine (P), on the slope of the rectilinear segment of the P-Q curve (equivalent to vascular conductance), the extrapolated Ppa intercept (Ppai), and the Ppa at zero flow (Ppaz). Low-dose D (0.4-0.8 mg) and DB (1.0-5.0 mg) did not significantly alter any parameter from control. Higher-dose D (1.2-6.8 mg) and DB (13-38 mg) decreased vascular conductance 32.3 +/- 12.1 (SE) to 50.45 +/- 5.6% (P less than 0.05), and P alone increased conductance 12.0 +/- 2.6% (P less than 0.01) from control with no significant effect on Ppai or Ppaz. The change in conductance with D and DB alone was abolished when either drug was given in combination with P. Ppaz and Ppai decreased significantly from control with DB in combination with P when no significant effect on vascular conductance was noted. The results suggest that lung vessels determining changes in flow resistance are pharmacologically distinct from those subserving critical closure in the pulmonary vascular bed.