Marginal Cost of Emergency Department Outpatient Visits

Neighborhood Child Opportunity and Emergency Department Utilization

The Child Opportunity Index measures the structural neighborhood context that may influence a child's healthy development. We examined relationships between the Child Opportunity Index and emergency department utilization.

BACKGROUND AND OBJECTIVES

The Child Opportunity Index (COI) is a multidimensional measure of structural neighborhood context that may influence a child's healthy development. Our objective was to determine if COI is associated with children's emergency department (ED) utilization using a national sample.

METHODS

This was a retrospective cohort study of the Pediatric Health Information Systems, a database from 49 United States children's hospitals. We analyzed children aged 0 to 17 years with ED visits from January 1, 2018, to December 31, 2019. We modeled associations between COI and outcomes using generalized regression models that adjusted for patient characteristics (eg, age, clinical severity). Outcomes included: (1) low-resource intensity (LRI) ED visits (visits with no laboratories, imaging, procedures, or admission), (2) ≥2 or ≥3 ED visits, and (3) admission.

RESULTS

We analyzed 6 810 864 ED visits by 3 999 880 children. LRI visits were more likely among children from very low compared with very high COI (1 LRI visit: odds ratio [OR] 1.35 [1.17-1.56]; ≥2 LRI visits: OR 1.97 [1.66-2.33]; ≥3 LRI visits: OR 2.4 [1.71-3.39]). ED utilization was more likely among children from very low compared with very high COI (≥2 ED visits: OR 1.73 [1.51-1.99]; ≥3 ED visits: OR 2.22 [1.69-2.91]). Risk of hospital admission from the ED was lower for children from very low compared with very high COI (OR 0.77 [0.65-0.99]).

CONCLUSIONS

Children from neighborhoods with low COI had higher ED utilization overall and more LRI visits, as well as visits more cost-effectively managed in primary care settings. Identifying neighborhood opportunity-related drivers can help us design interventions to optimize child health and decrease unnecessary ED utilization and costs.

Data Resources for Evaluating the Economic and Financial Consequences of Surgical Care in the United States

LEVEL OF EVIDENCE

V.

Urgent Care Centers Deter Some Emergency Department Visits But, On Net, Increase Spending

There is substantial interest in using urgent care centers to decrease lower-acuity emergency department (ED) visits. Using 2008-19 insurance claims and enrollment data from a national managed care plan, we examined the association within ZIP codes between changes in rates of urgent care center visits and rates of lower-acuity ED visits. We found that although the entry of urgent care deterred lower-acuity ED visits, the impact was small. We estimate that thirty-seven additional urgent care center visits were associated with a reduction of a single lower-acuity ED visit. In addition, each $1,646 lower-acuity ED visit prevented was offset by a $6,327 increase in urgent care center costs. Therefore, despite a tenfold higher price per visit for EDs compared with urgent care centers, use of the centers increased net overall spending on lower-acuity care at EDs and urgent care centers.

Randomized Controlled Trial of Acute Illness Educational Intervention in the Pediatric Emergency Department: Written Versus Application-Based Education

OBJECTIVES

The aim of this study was to determine the (1) feasibility, (2) demand, (3) acceptability, and (4) usefulness of a mobile health (mHealth) application (app) compared with a written intervention distributed in a pediatric emergency department (ED).

METHODS

This was a randomized controlled trial with parents of children 12 years or younger presenting to the ED for nonurgent complaints. Parents were randomized to receive a (1) low literacy pediatric health book with video, (2) pediatric mHealth app, (3) both 1 and 2, or (4) car-seat safety video and handout (control). Demand, acceptability, and usefulness were assessed at 1-, 3-, and 6-month follow-ups. Modified intention-to-treat analysis was completed for proportional data.

RESULTS

Ninety-eight parents completed randomization (83% approached). One or more follow-up was completed for 80.6% of parents. Only 57.1% downloaded the app. Parents used the app less than the book (35.1% vs 73.0%, P < 0.01), found the app to be harder to understand (26.0% vs 94.6%, P < 0.001) and less useful (37.8% vs 70.3%, P < 0.01), and were less likely to recommend the app to others (48.7% vs 100%, P < 0.01). No parent who received both book and app would prefer to have only the app; 88.9% of parents wanted either the book or both.

CONCLUSIONS

There was low demand for an mHealth app with parents who prefer, accept, and use the book more. Giving written health information to vulnerable populations in a pediatric ED has the capacity to empower parents with knowledge to care for a child and potentially decrease future nonurgent ED use with translation into a larger study.

The next step to reducing emergency department (ED) crowding: Engaging specialist physicians

BACKGROUND

Much work on reducing ED utilization has focused on primary care practices, but few studies have examined ED visits from patients followed by specialists, especially when the ED visit is related to the specialist's clinical practice.

OBJECTIVE

To determine the proportion and characteristics of patients that utilized the ED for specialty-related diagnosis.

METHODS

Retrospective, population-based, cohort study was conducted using information from electronic health records and billing database between January 2016 and December 2016. Patients who had seen a specialist during the last five years from the index ED visit date were included. The identification of ED visits attributable to specialists was based on the primary diagnosis of ED visits and the frequency of visit with specialists within a given timeframe.

RESULTS

Approximately 28% of ED visits analyzed were attributable to specialists. ED visits attributed specialists were represented by older patients and occurred more during working hours and early days of week. The most common diagnoses related to ED visits attributed to specialists were Circulatory, Musculoskeletal, Skin, Breast and Mental. Multiple departments, subdivisions and specialists were involved with each ED visit. The number of specialists following the patients who visited the ED ranged from one to six and the number of departments/subdivisions ranged from one to four. Patients that used the ED often were more likely to belong to departments (OR = 1.53) and specialists (OR = 1.18) associated with high ED utilization patterns.

CONCLUSION

Patients coming to the ED with specialty-related complaints are unique and require full engagement of the specialist and the specialty group. This study offers a new view of connections patients have with their specialists and engaging specialists both at department level and individual specialist level may be an important factor to reduce ED overcrowding.

Barriers and enablers to a physician-delivered educational initiative to reduce low-acuity visits to the pediatric emergency department

BACKGROUND

Use of the pediatric emergency department (PED) for low-acuity health issues is a growing problem, contributing to overcrowding, longer waits and higher health system costs. This study examines an educational initiative aimed at reducing low-acuity PED visits. The initiative, implemented at an academic pediatric hospital, saw PED physicians share a pamphlet with caregivers to educate them about appropriate PED use and alternatives. Despite early impacts, the initiative was not sustained. This study analyzes the barriers and enablers to physician participation in the initiative, and offers strategies to improve implementation and sustainability of similar future initiatives.

METHODS

Forty-two PED physicians were invited to participate in a semi-structured individual interview assessing their views about low-acuity visits, their pamphlet use, barriers and enablers to pamphlet use, and the initiative's potential for reducing low-acuity visits. Suggestions were solicited for improving the initiative and reducing low-acuity visits. Constant comparative method was used during analysis. Codes were developed inductively and iteratively, then grouped according to the Theoretical Domains Framework (TDF). Efforts to ensure study credibility included seeking participant feedback on the findings.

RESULTS

Twenty-three PED physicians were interviewed (55%). Barriers and enablers for pamphlet use were identified and grouped according to five of the 14 TDF domains: social/professional role and identity; beliefs about consequences; environmental context and resources; social influences; and emotions.

CONCLUSIONS

The TDF provided an effective approach to identify the key elements influencing physician participation in the educational initiative. This information will help inform behavior change interventions to improve the implementation of similar future initiatives that involve physicians as the primary educators of caregivers.

Health Information Preferences of Parents in a Pediatric Emergency Department

Parents of children seeking nonurgent care in the emergency department completed surveys concerning media use and preferences for health education material. Results were compiled using descriptive statistics, compared by health literacy level with logistic regression, adjusting for race/ethnicity and income. Semistructured qualitative interviews to elicit reasons for preferences, content preference, and impact of health information were conducted and analyzed using content analysis. Surveys (n = 71) showed that despite equal access to online health information, parents with low health literacy were more likely to use the internet less frequently than daily ( P < .01). Surveys and interviews (n = 30) revealed that health information will be most effective when distributed by a health care professional and must be made available in multiple modalities. Parents requested general information about childhood illness, including diagnosis, treatment, and signs and symptoms. Many parents believed that appropriate health information would change their decision-making regarding seeking care during their child's next illness.

Concentrating Emergency Rooms: Penny-Wise and Pound-Foolish? An Empirical Research on Scale Economies and Chain Economies in Emergency Rooms in Dutch Hospitals

In this paper, we address the issue of whether it is economically advantageous to concentrate emergency rooms (ERs) in large hospitals. Besides identifying economies of scale of ERs, we also focus on chain economies. The latter term refers to the effects on a hospital's costs of ER patients who also need follow-up inpatient or outpatient hospital care. We show that, for each service examined, product-specific economies of scale prevail indicating that it would be beneficial for hospitals to increase ER services. However, this seems to be inconsistent with the overall diseconomies of scale for the hospital as a whole. This intuitively contradictory result is indicated as the economies of scale paradox. This scale paradox also explains why, in general, hospitals are too large. There are internal (departmental) pressures to expand certain services, such as ER, in order to benefit from the product-specific economies of scale. However, the financial burden of this expansion is borne by the hospital as a whole. The policy implications of the results are that concentrating ERs seems to be advantageous from a product-specific perspective, but is far less advantageous from the hospital perspective. © 2016 The Authors. Health Economics Published by John Wiley & Sons, Ltd.

Can improved specialty access moderate emergency department overuse?: Effect of neurology appointment delays on ED visits

Background

Delayed access to specialty care may increase inappropriate emergency department (ED) visits. However, the details of this relationship after referral to a specialist are unknown.

Methods

The correlations in an academic medical center between time to new neurology patient appointments and nonurgent ED use are explored in this study. Access was measured as the number of days between the scheduling and outpatient appointment dates. A series of statistical analyses including correlation analysis, regressions, and hypothesis tests were conducted to investigate possible associations between delayed access to specialty care and ED visits, as well as the effect of ED visits on specialty care cancellation and no-show rates.

Results

Of 19,505 new neurology patients, 310 visited an ED prior to their appointment, 95.2% (295) of whom had poor access (defined here as exceeding 21 days). Patients with access >21 days for new visits were 6.6 times more likely to visit the ED before their appointment date, 19% within the first week after scheduling. Patients who visited the ED between their booking and appointment dates were 2.3 times more likely to cancel or fail to attend their appointment.

Conclusion

These results suggest that long access delays in specialty referrals can significantly increase ED costs and congestion. Further studies in other specialties to explore this relationship are warranted.

Ambulatory Care Sensitive Hospitalizations Through the Emergency Department by Payer: Comparing 2003 and 2009

BACKGROUND

Ambulatory care sensitive hospitalizations (ACSHs) are hospitalizations that may have been preventable with timely and effective outpatient care. Approximately 75% of all ACSHs occur through the emergency department (ED). ACSHs through the ED (ED ACSHs) have significant implications for costs and ED crowding.

OBJECTIVE

This study compares rates of ED ACSHs for 2003 and 2009 among patients 18 to 64 years of age with private insurance, Medicaid, or no insurance.

METHODS

Nationally representative estimates of ED ACSHs, defined by the Agency for Healthcare Research and Quality (AHRQ) prevention quality indicators (PQIs), were generated from the 2003 and 2009 Nationwide Inpatient Samples. Census data were used to calculate direct age- and sex-standardized ACSH rates by non-Medicare payers for both years.

RESULTS

Between 2003 and 2009, the overall rate of ED ACSHs decreased from 7.6 (95% confidence interval [CI] 7.57-7.75) to 7.3 (95% CI 7.2-7.4) per 1000 18- to 64-year-old non-Medicare patients. ED ASCH rates declined significantly from 42.4 (95% CI 42.0-42.8) to 25.3 (95% CI 25.0-25.6) per 1000 patients with Medicaid, and declined modestly from 3.8 (95% CI 3.8-3.9) to 3.3 (95% CI 3.2-3.4) per 1000 patients with private insurance. However, the ED ACSH rate increased for the uninsured population from 5.4 (95% CI 5.2-5.7) to 6.2 (95% CI 5.9-6.4) per 1000 patients.

CONCLUSION

Expansion of Medicaid over the study period was not associated with an increase in ED ACSHs for Medicaid patients. However, an increase in the uninsured population was associated with an increase in the rate of ED ACSH for uninsured patients.

The human health effects of Florida red tide (FRT) blooms: an expanded analysis

Human respiratory and digestive illnesses can be caused by exposures to brevetoxins from blooms of the marine alga Karenia brevis, also known as Florida red tide (FRT). K. brevis requires macro-nutrients to grow; although the sources of these nutrients have not been resolved completely, they are thought to originate both naturally and anthropogenically. The latter sources comprise atmospheric depositions, industrial effluents, land runoffs, or submerged groundwater discharges. To date, there has been only limited research on the extent of human health risks and economic impacts due to FRT. We hypothesized that FRT blooms were associated with increases in the numbers of emergency room visits and hospital inpatient admissions for both respiratory and digestive illnesses. We sought to estimate these relationships and to calculate the costs of associated adverse health impacts. We developed environmental exposure-response models to test the effects of FRT blooms on human health, using data from diverse sources. We estimated the FRT bloom-associated illness costs, using extant data and parameters from the literature. When controlling for resident population, a proxy for tourism, and seasonal and annual effects, we found that increases in respiratory and digestive illnesses can be explained by FRT blooms. Specifically, FRT blooms were associated with human health and economic effects in older cohorts (≥55 years of age) in six southwest Florida counties. Annual costs of illness ranged from $60,000 to $700,000 annually, but these costs could exceed $1.0 million per year for severe, long-lasting FRT blooms, such as the one that occurred during 2005. Assuming that the average annual illness costs of FRT blooms persist into the future, using a discount rate of 3%, the capitalized costs of future illnesses would range between $2 and 24 million.

Understanding the value of emergency care: a framework incorporating stakeholder perspectives

BACKGROUND

In the face of escalating spending, measuring and maximizing the value of health services has become an important focus of health reform. Recent initiatives aim to incentivize high-value care through provider and hospital payment reform, but the role of the emergency department (ED) remains poorly defined.

OBJECTIVES

To achieve an improved understanding of the value of emergency care, we have developed a framework that incorporates the perspectives of stakeholders in the delivery of health services.

METHODS

A pragmatic review of the literature informed the design of this framework to standardize the definition of value in emergency care and discuss outcomes and costs from different stakeholder perspectives. The viewpoint of patient, provider, payer, health system, and society is each used to assess value for emergency medical conditions.

RESULTS

We found that the value attributed to emergency care differs substantially by stakeholder perspective. Potential targets to improve ED value may be aimed at improving outcomes or controlling costs, depending on the acuity of the clinical condition.

CONCLUSION

The value of emergency care varies by perspective, and a better understanding is achieved when specific outcomes and costs can be identified, quantified, and measured. Using this framework can help stakeholders find common ground to prioritize which costs and outcomes to target for research, quality improvement efforts, and future health policy impacting emergency care.

Resource use, costs, and utility estimates for patients with cystic fibrosis with mild impairment in lung function: analysis of data collected alongside a 48-week multicenter clinical trial

OBJECTIVES

Transport of ions to generate epithelial rehydration (TIGER)-1 was a randomized trial conducted to evaluate the safety and efficacy of denufosol versus placebo in patients with cystic fibrosis with mild impairment in lung function. The trial met its primary end point at 24 weeks, but a subsequent trial did not show a sustained effect of denufosol at 48 weeks. By using the 48-week data, we characterized resource use, direct medical costs, indirect costs, and utility estimates.

METHODS

Data on medications, outpatient and emergency visits, hospital admissions, tests, procedures, and home nursing were captured on study case report forms. Sources for unit costs included the Medicare Physician Fee Schedule, the Nationwide Inpatient Sample, and the Red Book. Health utilities were derived from the Health Utilities Index Mark 2/3. We used multivariable regression to evaluate the impact of baseline covariates on costs.

RESULTS

Characteristics of the 352 participants at enrollment included mean age of 14.6 years, history of Pseudomonas aeruginosa colonization in 45.2%, use of dornase alfa in 77.0%, and long-term use of inhaled antibiotics in 37.2%. Over 48 weeks, 22.4% of participants were hospitalized and, on average, participants missed 7.4 days of school or work. Mean total costs (excluding denufosol) were $39,673 (SD $26,842), of which 85% were attributable to medications. Female sex and P. aeruginosa colonization were independently associated with higher costs.

CONCLUSIONS

Prospective economic data collection alongside a clinical trial allows for robust estimates of cost of illness. The mean annual cost of care for patients with cystic fibrosis with mild impairment in lung function exceeds $43,000 and is driven by medication costs.

Emergency severity index triage system correlation with emergency department evaluation and management billing codes and total professional charges

OBJECTIVES

All services provided by physicians to patients during an emergency department (ED) visit, including procedures and "cognitive work," are described by common procedural terminology (CPT) codes that are translated by coders into total professional (physician) charges for the visit. These charges do not include the technical (facility) charges. The objectives of this study were to characterize associations between Emergency Severity Index (ESI) acuity level, ED Evaluation and Management (E&M) billing codes 99281-99285 and 99291, and total ED provider charges (sum of total procedure and E&M professional charges). Secondary objectives were to identify factors that might affect these associations and to evaluate the performance of ESI and identified variables to predict E&M code and average total professional charges.

METHODS

The authors reviewed 276,824 patient records for calendar year 2007, of which 193,952 adult ED visits from three different ED types (community, university-based academic, and non-university-based academic) met inclusion criteria. Correlations between 1) ESI level and E&M billing code per visit by institution and 2) ESI and total professional charges were analyzed using Spearman rank correlation. Linear regression analysis was performed to identify variables that significantly affected these correlations.

RESULTS

ESI level and E&M codes were moderately correlated (Spearman r = 0.51). ESI levels corresponded proportionately to higher E&M codes. ESI 1, 2, and 3 most frequently corresponded with E&M level 5 (50, 62, and 45%, respectively), and ESI 4 and 5 most frequently corresponded with E&M level 3 (56 and 67%, respectively). Only age by decade significantly affected the association between ESI level and E&M billing code. The mean total professional charge for all patient encounters was $421 (SD ± $204) with increasing mean charges per patient by increasing ESI acuity. Race and E&M code significantly affected the relationship between ESI level and total ED professional charges per patient (adjusted r(2) = 0.66).

CONCLUSIONS

A moderate, nonlinear correlation exists between ESI acuity levels and ED E&M billing codes. Increasing age affects this correlation. Race and E&M code affect the correlation between ESI level and total professional charges. As such, basic triage data can be used to estimate E&M code and total professional charges. Future studies are needed to validate these findings across other institutional settings.

Emergency department visits: the cost of trauma centers

Crowded emergency departments (EDs) have become a serious problem in the current U.S. healthcare system. Patient wait times and periods of ED diversion have increased, raising concerns about the timeliness, efficiency, and quality of ED treatment. This study addresses the question of whether there are economies of scale (EOS) in ED care, and the extent to which such economies vary across different types of EDs. A hospital cost function approach is taken to evaluate average and marginal costs of EDs designated as trauma centers. Data comes from acute care hospitals in Texas for the period 1998-2004. Cost functions corresponding to four different levels of ED trauma care are estimated using a translog panel data model with hospital fixed effects. The marginal costs (in 2004 dollars) of each trauma center level are: $53 (Level I), $177 (Level II), $119 (Level III), and $258 (Level IV). Average cost per ED visit for trauma centers exceeds marginal cost at all Levels, indicating the presence of EOS. The results support a possible expansion of ED size policy in order to improve the cost efficiency of ED services.

Resource use and costs of treatment with anticoagulation and antiplatelet agents: results of the WATCH trial economic evaluation

BACKGROUND

The Warfarin and Antiplatelet Therapy in Chronic Heart Failure (WATCH) trial revealed no significant differences among 1587 symptomatic heart failure patients randomized to warfarin, clopidogrel, or aspirin in time to all-cause death, nonfatal myocardial infarction, or nonfatal stroke. We compared within-trial medical resource use and costs between treatments.

METHODS AND RESULTS

We assigned country-specific costs to medical resources incurred during follow-up. Annualized rates of hospitalizations, inpatient and outpatient procedures, and emergency department visits did not differ significantly between groups. Annualized total costs averaged $5901 (95% confidence interval [CI], $4776-$7520) for the aspirin group, $5646 (95% CI, $4903-$6584) for the clopidogrel group, and $5830 (95% CI, $4838-$7400) for the warfarin group.

CONCLUSIONS

Consistent with clinical findings, our analyses did not identify significant cost differences between treatments.

Impact of a health literacy intervention on pediatric emergency department use

OBJECTIVE

The aim of this study was to measure the impact of a simple parent health literacy intervention on emergency department and primary care clinic usage patterns.

METHODS

Study participants consisted of parents who brought their children to the Harbor-UCLA Medical Center pediatric emergency department for nonurgent complaints. Study participants filled out questionnaires regarding their management of children's mild health complaints and where respondents first seek help when their children become sick. After completing the questionnaires, participants were educated about how to use the health aid book What to Do When Your Child Gets Sick and provided a free copy. After 6 months, telephone follow-up interviews were conducted to assess whether the health literacy intervention had influenced the participants' management of their children's mild health complaints and their health care resource usage patterns.

RESULTS

One hundred thirteen parents were enrolled in the preintervention phase, and 61 were successfully interviewed at 6 months by telephone. Before and after comparisons demonstrated a 13% reduction in the percentage of respondents who stated they would go to the emergency department first if their child became sick. In addition, 30% fewer respondents reported actual visits to the emergency department in the previous 6 months. Regarding specific low-acuity scenarios, significantly fewer participants would take their child to the emergency department for a low-grade fever with a temperature of 99.5 degrees F and for vomiting for 1 day. There was no significant change in the proportion of parents who would take their child to the emergency department for earache or cough.

CONCLUSIONS

Health literacy interventions may reduce nonurgent emergency department visits and help mitigate emergency department overcrowding and the rising costs of health care.

Cost Effectiveness of Ixabepilone Plus Capecitabine for Metastatic Breast Cancer Progressing After Anthracycline and Taxane Treatment

Purpose

Using data from a recent randomized trial, we evaluated the cost effectiveness of ixabepilone plus capecitabine versus capecitabine alone in patients with predominantly metastatic breast cancer considered to be taxane-resistant and previously treated with or resistant to an anthracycline.

Methods

We developed a stochastic decision-analytic model to represent data collected in the trial on medical resource use, health-related quality of life, and clinical outcomes. Estimates of overall survival were conditional on level of tumor response. We assigned monthly costs and utility weights according to periods defined by the duration of study treatment, time from discontinuation of the study drug until disease progression, and from progression until death and were specific to the level of response and receipt of subsequent therapy. Medical resources were valued in 2008 US dollars. We performed Monte Carlo simulations and sensitivity analyses to evaluate model uncertainty.

Results

Overall survival was significantly associated with level of tumor response (P < .001). Total costs were estimated at $60,900 for patients receiving ixabepilone plus capecitabine and $30,000 for patients receiving capecitabine alone. The estimated gain in life expectancy with ixabepilone was 1.96 months (95% CI, 1.36 to 2.64 months); the estimated gain in quality-adjusted survival was 1.06 months (95% CI, 0.09 to 2.03 months). The resulting incremental cost-effectiveness ratio was $359,000 per quality-adjusted life-year (95% CI, $183,000 to $4,030,000). In sensitivity analyses, the results were robust to changes in numerous inputs and assumptions.

Conclusion

Addition of ixabepilone to capecitabine adds approximately $31,000 to overall medical costs and affords approximately 1 additional month of quality-adjusted survival.

Cost savings attributable to reductions in intensive care unit length of stay for mechanically ventilated patients

OBJECTIVES

To estimate the actual cost savings that could be achieved through reductions in intensive care unit (ICU) length of stay and duration of mechanical ventilation by determining the short-run marginal variable cost of an ICU and ventilator day.

RESEARCH DESIGN

Retrospective cohort study in a university-affiliated teaching hospital.

SUBJECTS

All patients receiving mechanical ventilation in the ICU for more than 48 hours (n = 1778) from July 1, 2005 to June 30, 2006.

MEASURES

The hospital's administrative and cost databases were used to determine total costs, variable costs, and direct-variable costs for each patient on each individual ICU and hospital day.

RESULTS

Direct-variable costs comprised 19.3% of total ICU costs and 18.4% of total hospital costs. Marginal direct-variable costs (the cost of each additional ICU day) were small compared with the average daily total cost ($649 to $839 vs. $1751, in US dollars). In survivors with ICU lengths of stay more than 3 days, the mean direct-variable cost of the last ICU day was $397, while the mean direct-variable cost of the first ward day was $279, for a mean cost difference of $118 (95% CI, $21-$190). Reducing ICU and hospital length of stay by 1 day in all survivors with ICU lengths of stay more than 3 days would result in an immediate cost savings of only 0.2% of all hospital expenditures for these patients.

CONCLUSIONS

Marginal variable ICU costs are relatively small compared with average total costs and are only slightly greater than the cost of a ward day.