A Multicenter Description of Intermediate-Care Patients - CiteSeerX

A Multicenter Description of Intermediate-Care Patients* Comparison With ICU Low-Risk Monitor Patients Christopher Junker, MD; Jack E. Zimmerman, MD; Carlos Alzola, MS; Elizabeth A. Draper, MS; and Douglas P. Wagner, PhD

Study objectives: To describe the characteristics and outcomes of patients admitted to intermediate-care areas (ICAs) and to compare them with those of ICU patients who receive monitoring only on day 1 and are at a low risk (ie, < 10%) for receiving subsequent active life-supporting therapy (ie, low-risk monitor patients). Design: Nonrandomized, retrospective, cohort study. Setting: Thirteen US teaching hospitals and 19 nonteaching hospitals. Patients: A consecutive sample of 8,971 patients at 37 ICAs and 5,116 low-risk (ie, < 10%) monitor patients at 59 ICUs in 32 US hospitals. Interventions: None. Measurements and results: We recorded demographic and clinical characteristics, resource use, and outcomes for the ICA and ICU low-risk monitor patients. Patient data and outcomes for this study were collected concurrently or retrospectively. ICA and ICU low-risk monitor patients were similar in regard to gender, race, and frequency of comorbitities, but ICA patients were significantly (p < 0.001) older, had fewer physiologic abnormalities (mean acute physiology score, 16.7 vs 19.8, respectively), and were more frequently admitted due to nonoperative diagnoses. The mean length of stay for ICA patients was significantly longer (3.9 days) than for ICU low-risk monitor patients (2.6 days; p < 0.001). The hospital mortality rate was significantly higher for ICA patients (3.1%) compared to ICU low-risk monitor patients (2.3%; p ⴝ 0.002). Conclusions: The clinical features of ICA patients are similar, but not identical to, those of less severely ill ICU monitor patients. Comparisons of hospital death rates and lengths of stay for these patients should be adjusted for characteristics that previously have been shown to influence these outcomes. (CHEST 2002; 121:1253–1261) Key words: critical care; facility design and construction; high dependency units; ICU; intermediate care; length of stay; resource allocation; triage Abbreviations: APACHE ⫽ Acute Physiology and Chronic Health Evaluation; COTH ⫽ Council of Teaching Hospital; ICA ⫽ intermediate-care area

atients who are admitted to intermediate-care P areas (ICAs) of a hospital do not require full intensive care but need more services than those *From the Department of Anesthesiology and Critical Care Medicine (Dr. Junker), George Washington University Medical Center, Washington, DC; the Department of Health Evaluation Sciences (Dr. Wagner), University of Virginia, Charlottesville, VA; the Dyne Corp (Ms. Draper) Reston, VA; and APACHE Medical Systems, Inc, (Dr. Zimmerman and Mr. Alzola), McLean, VA. This research was supported by the Department of Anesthesiology, George Washington University Medical Center, and by APACHE Medical Systems (AMS), Inc. AMS markets a clinical information system and holds the commercial copyright on the risk of active treatment equation and the APACHE database. Drs. Zimmerman and Wagner and Ms. Draper are founders and former shareholders of AMS. Drs. Zimmerman and Wagner have received financial support from AMS in the form of a research grant. Mr. Alzola is an employee of AMS. www.chestjournal.org

provided on a hospital ward.1–3 The type and amount of services provided by ICAs, however, varies depending on resource availability,1,4,5 and the philosophy, organization, and funding of different healthcare systems.2,6 As a result, there are marked variations in the roles and capabilities of ICAs internationally and even within individual hospitals. In the United Kingdom, where ICU beds are less available than in the United States,4,7 intermediate or high-dependency units provide care for patients who have experienced acute myocardial infarctions Manuscript received April 5, 2001; revision accepted October 19, 2001. Correspondence to: Christopher Junker, MD, Department of Anesthesiology and Critical Care Medicine, The George Washington University Medical Center, 901 23rd St NW, Washington, DC 20037; e-mail:[email protected] CHEST / 121 / 4 / APRIL, 2002

1253

or those who are at risk for or have already developed single-organ failure.8,9 The role and capabilities of ICAs in US hospitals also differ. Some are multipurpose areas that act as a step-up or step-down from care delivered on wards and in ICUs, and others provide specialty care for medical, cardiac, surgical, neurologic, respiratory, or chronic ventilator patients.3 There are several reasons why hospitals have developed ICAs. First, they facilitate earlier hospital discharges of stable ICU patients who are thought to need more care than can be provided on wards.10 –13 Second, they may decrease the need for ICU readmission by providing more monitoring and nursing care than is available on hospital wards.14 –16 Third, they appear to decrease patient mortality rates in hospitals that are experiencing marked pressure on the availability of beds in ICUs.17–19 Fourth, they may reduce the cost of treating patients who do not need the unique services of an ICU6,20 –24 by providing care in areas with a lower nurse/patient ratio and less complex technology.6,25–27 Although many US hospitals have developed ICAs, there have been no multicenter descriptions of ICA patients or their outcomes. In addition, patients admitted to the hospital for monitoring who are at a low risk for receiving life support (ie, low-risk monitor patients)28 –31 often receive care in ICAs rather than in ICUs.32–35 There have not been any comparisons of outcomes for these low-risk patients in ICAs vs ICUs. The purpose of this study was to describe the demographic and clinical characteristics of a multihospital population of US ICA patients. We also compared the characteristics, resource use, and outcomes of ICA patients with those of ICU low-risk monitor patients who were being treated at the same hospitals.

Materials and Methods The patient data and outcomes for this study were collected, concurrently or retrospectively, between May 1993 and September 1998. The hospitals were selected because each had collected data in an ICA. Data for ICU patients were collected in the same hospitals. Twenty-five of the hospitals collected patient data to assess and compare risk-adjusted mortality rates and resource use to national benchmarks, and 10 hospitals were clients of a private company (APACHE Medical Systems; McLean, VA). For each client hospital, the analysis was limited to the data collected during the first 4 to 12 months to ensure that reliability testing had been completed and to minimize any potential bias introduced by the information collected. The average duration of data collection for the other 25 hospitals that conducted benchmark comparisons was 3.1 months. The methods, training, and support provided for data collection have been reported previously.36 Each hospital provided information about bed size and teaching status as defined by membership in the Council of Teaching Hospitals (COTH), the presence of one or more accredited residency programs (ie, non-COTH teaching hospitals), or nei1254

ther (nonteaching hospitals). The type of ICA and ICU that was used was reported by each hospital. To eliminate uncertainties in definition, we excluded data from any ICA that cited a nurse/ patient ratio of ⬎ 1:231,37 or provided active life-supporting therapy for ⬎ 20% of hospital admissions. Table 1 lists the active life-supporting therapies used for this study and in previous work by Zimmerman and colleagues.31 We also excluded ICAs that failed to meet previously published data reliability criteria.38 Informed consent was not obtained because of the existence of prior institutional review board waivers for identical data collection efforts during previous studies.36,38 Patient Information The data that were generated as a result of patient care and were recorded in the medical record were collected concurrently or retrospectively for consecutive hospital admissions at each ICA and ICU. For each patient, we recorded age, gender, race, and the comorbidity and physiologic data required to calculate an acute physiology and chronic health evaluation (APACHE) III score. We also recorded whether active life-supporting therapy was received on the unit on day 1, patient location and length of stay in the hospital before admission to the unit, operative status, and whether there was emergency surgery or unit readmission. Definitions of the above variables have been reported else-

Table 1—Active Life-Supporting Measures* Treatments Respiratory

GI

Neurologic

Cardiovascular

Renal Miscellaneous

Description Controlled ventilation IMV or assisted ventilation Spontaneous PEEP or CPAP Nasal or oral intubation Fresh tracheostomy Emergency bronchoscopy IV pitressin infusion Continuous arterial drug infusion Balloon tamponade for esophageal varices Continuous nasogastric lavage Emergency endoscopy Mannitol infusion Ventriculostomy Treatment of seizure Induced hypothermia Barbiturate anesthesia Atrial or ventricular pacing Intraaortic balloon Vasoactive drugs IV fluids (⬎ 6 L/d) Rapid blood transfusion Post arrest (⬎ 24 h) Trauma suit Cardioversion Pericardiocentesis Hemodialysis Treatment of metabolic acidosis or alkalosis Emergency operation IV concentrated K⫹ Complex metabolic balance (ie, frequent I and O) Active diuresis for fluid overload

*IMV ⫽ intermittent mandatory ventilation; PEEP ⫽ positive endexpiratory pressure; CPAP ⫽ continuous positive airway pressure; I and O ⫽ ins and outs, recording of fluids administered and put out by the patient in the form of urine, nasogastric suction, etc. Clinical Investigations in Critical Care

where.31,38 Each patient’s primary reason for unit admission (ie, the admission diagnosis) was recorded by selecting 1 of 430 diseases, injuries, surgical procedures, or events that was most immediately threatening to the patient and that required ICA or ICU services. The methods used to record and track unit admission diagnoses have been previously reported.36 For this study, we excluded patients who had been admitted to the hospital for ⬍ 4 h, patients with burns, patients who were ⬍ 16 years of age, and unit readmissions. We also excluded ICU patients who during day 1 received active life-supporting therapy and ICU patients who received only monitoring services during day 1 but were at a high risk (ie, ⱖ 10%) for subsequently receiving active therapy. To do this, we used ICU day 1 active therapy data and a previously validated multivariate predictive equation to identify ICU monitor patients who were at a low risk (ie, ⬍ 10%) for receiving subsequent active therapy.31 We used the same predictive equation to assess the risk of active therapy for ICA patients who did not receive active therapy on day 1. We emphasize that the risk predicted was for receiving subsequent active therapy, not the risk of mortality. The analysis of risk for active therapy excluded patients who had been admitted to the hospital for chest pain to rule out myocardial infarction and after coronary artery bypass surgery because the predictive equation has not been validated for patients with these diagnoses. Outcomes recorded for each patient included the duration of unit and hospital stay, unit readmission, transfer to another ICA or ICU, and survival at unit and hospital discharge. To eliminate patients for whom the unit length of stay was uncertain, we excluded patients who were discharged to another ICA or ICU from all analyses of length of stay. We also recorded whether active therapy was received after unit day 1 for 3,374 of the ICU low-risk monitor patients (66%) and 4,975 of the ICA patients (55%). The type of active therapy was not recorded, and these data were collected only at the seven hospitals that were clients of a private company (APACHE Medical Systems). Data Analysis and Statistics Univariate statistical comparisons of institutional and patient characteristics were performed using a t test with a significance criterion of p ⬍ 0.05 for hospitals and units and p ⬍ 0.001 for patients to avoid the overreporting of small differences as significant. Data on length of stay were analyzed using mean values to describe total resource use, median values to describe typical patterns, and visual examination of frequency distributions to describe typical patterns and information on outliers, as reflected by rightward skewing.39

Results Hospital and Unit Characteristics We collected data in 40 ICAs and 59 ICUs at 35 hospitals, but we excluded data from 2 ICAs because they did not meet the definition of intermediate care used in our study. We also excluded data from one ICA because of the failure to meet data reliability criteria. This left 37 ICAs and 59 ICUs at 32 hospitals for analysis. Geographically, 12 of the 32 hospitals (38%) were located in the West, 11 (34%) were located in the Midwest, 7 (22%) were located in the Southeast, and 2 (6%) were located in the Northeast. The mean number of hospital beds was 413 (range, 145 to 900 beds). Thirteen hospitals www.chestjournal.org

(41%) met our definition of a teaching hospital (COTH, 3; non-COTH, 10), and 19 hospitals (59%) were nonteaching hospitals. Table 2 summarizes the characteristics of the 37 ICAs and 59 ICUs. Of the ICAs, a majority (81%) were mixed medical-surgical, whereas 51% of the ICUs were specialized. Patients Data were collected for 8,971 patients who had been admitted consecutively to the 37 ICAs and for 14,697 patients who had been admitted consecutively to the 59 ICUs. We excluded 7,690 ICU patients who received active life-supporting therapy, 1,641 ICU monitor patients who were at high risk (ie, ⱖ 10%) of receiving active therapy on ICU day 1, and 250 patients with missing active treatment data. The exclusions described above resulted in a nonrandomized observational sample of 8,971 ICA patients and 5,116 ICU low-risk monitor patients. Among the 8,971 ICA admissions, data were collected for 3,996 patients to compare outcomes to national benchmarks and for 4,975 patients by clients of a private company (APACHE Medical Systems). Among the 5,116 ICU low-risk monitor admissions, data were collected for 1,742 patients to compare outcomes to national benchmarks and for 3,374 patients by clients of a private company (APACHE Medical Systems). The demographic and clinical characteristics of the ICA and ICU low-risk monitor patients are shown in Table 3. ICA and ICU low-risk monitor patients were similar in regard to gender, race, and frequency of comorbidities. In contrast, ICA patients were significantly (p ⬍ 0.001) older and had been admitted to the ICA more frequently for nonoperative diagnoses compared to ICU low-risk monitor patients. The severity of physiologic abnormality was also significantly lower for ICA patients than for ICU low-risk monitor patients (mean acute physiology scores, 16.7 vs 19.8, respectively; p ⬍ 0.001). Figure 1 shows the distribution of APACHE III scores for ICA and ICU low-risk monitor patients. Table 2—Description of ICAs and ICUs Studied at 33 US Hospitals* Type of Unit

ICAs (n ⫽ 37)

ICUs (n ⫽ 59)

Medical-surgical Surgical Medical Cardiac Telemetry Cardiothoracic Neurologic

30 (81.1) 0 0 1 (2.7) 6 (16.2) 0 0

30 (50.8) 7 (11.9) 6 (10.2) 10 (16.9) 0 4 (6.8) 2 (3.4)

*Values given as No. (%). CHEST / 121 / 4 / APRIL, 2002

1255

Table 3—Demographic and Clinical Characteristics of ICA Patients and ICU Low-Risk Monitor Patients* Characteristics Age, yr Mean ⫾ SD ⬍ 45 ⬎ 65 Female gender Race White Black Hispanic APACHE III comorbidities† Operative status‡ Nonoperative Postoperative Surgery§ Emergency Elective Location before admission Emergency department Recovery room Operating room Hospital ward ICU Other hospital ICA Direct admission Severity of illness (day 1) APACHE III Acute physiology score

ICA Patients (n ⫽ 8,971)

ICU Low-Risk Monitor Patients (n ⫽ 5,116)

64.0 ⫾ 16.6 1,262 (14.1) 5,021 (56.0) 4,089 (45.6)

59.4 ⫾ 18.4 1,158 (22.6) 2,405 (47.0) 2,320 (45.3)

6,793 (75.7) 602 (6.7) 337 (3.8) 693 (7.7)

3,941 (77.0) 407 (7.9) 228 (4.5) 405 (7.9)

7,465 (83.2) 1,506 (16.8)

2,991 (58.5) 2,125 (41.5)

305 (22.5) 1,048 (77.5)

320 (15.1) 1,805 (84.9)

3,369 (37.6) 1,181 (13.3) 174 (1.9) 559 (6.2) 1,492 (16.6) 426 (4.7) 31 (0.3) 1,176 (13.1)

1,940 (37.9) 1,679 (32.8) 426 (8.3) 366 (7.2) 160 (3.1) 96 (1.9) 70 (1.4) 359 (7.0)

28.9 ⫾ 15.5 16.7 ⫾ 12.3

30.2 ⫾ 13.9 19.8 ⫾ 10.9

p Value ⬍ 0.001

0.79 0.02

0.68 0.001

0.001

0.001

⬍ 0.001 ⬍ 0.001

*Values given as mean ⫾ SD or No. (%), unless otherwise indicated. †Signifies patients with one or more comorbidities. ‡No data exist for 125 ICA patients and for the location before unit admission for 563 ICA patients and 20 ICU low-risk monitor patients. §There was no information available on whether ICA surgical patients had undergone elective or emergency surgery. Those patients were included in all other reporting.

Table 4 shows the most frequent nonoperative admission diagnoses for ICA and ICU low-risk monitor patients. The 12 nonoperative admission diagnoses in Table 4 accounted for 55.8% of the nonoperative ICA patients and for 68.8% of the nonoperative ICU low-risk monitor patients. Cardiovascular diagnoses (ie, unstable angina, cardiac dysrhythmia, congestive heart failure, and acute myocardial infarction) accounted for 39.1% of those in nonoperative ICA patients and for 39.2% of those in nonoperative ICU low-risk monitor patients. Admissions to the unit for chest pain and to rule out myocardial infarction accounted for 569 of the 4,765 nonoperative ICA patients (7.6%), but the risk for active therapy could not be assessed among ICU or ICA patients. Table 5 shows the most frequent postoperative admission diagnoses among ICA and ICU low-risk monitor patients. The 12 postoperative admission diagnoses accounted for 60.7% of diagnoses in the postoperative ICA patients and for 59% of diagnoses in the postoperative ICU low-risk monitor patients. 1256

Vascular surgery (ie, carotid endarterectomy, femoral-popliteal artery graft, and operations for peripheral vascular ischemia and aortoiliac or aortofemoral obstructions) accounted for 24.4% of the postoperative ICA patients, and for 24.7% of the postoperative ICU low-risk monitor patients. Resource Use Resource use for the 8,971 ICA and 5,116 ICU low-risk monitor patients are summarized in Table 6. The unit length of stay was significantly longer for ICA patients than for ICU low-risk monitor patients (mean, 3.9 vs 2.6 days, respectively; p ⬍ 0.001). Figure 2 displays the distribution of unit length of stay for the two patient groups. The type of unit and the patient’s admission diagnosis had a substantial influence on the mean unit length of stay (Tables 3, 4). For example, the mean unit length of stay was 4 days for ICA patients vs 2.2 days for ICU low-risk monitor patients who had been admitted to a unit for cardiac dysrhythClinical Investigations in Critical Care

Figure 1. The distribution of severity of illness estimated by the unit day 1 APACHE III scores for 8,971 ICA and 5,116 ICU low-risk monitor patients.

mia, and 3.5 vs 2.2 days, respectively, for patients admitted to a unit following surgery for multiple trauma. Active life-supporting therapy was received by 747 of the 8,971 ICA patients (8.3%) on day 1. The most prevalent nonoperative and postoperative admission diagnoses among ICA patients who received active therapy on day 1 are shown in Table 7. Among 7,579 ICA patients who received only monitoring services on day 1, 6457 (85.3%) were at a low risk (ie, ⬍ 10%) for subsequent active treatment. Among the 4,237 ICA monitor patients who had active therapy information recorded after unit day 1, 3,646 (86%) were at a low risk on day 1 for active therapy, and 95.7% of these patients never received subsequent active treatment. Of the 3,374 ICU low-risk monitor patients who had active therapy information recorded after unit day 1, 94.3% never received subsequent active treatment. Patient Outcomes Outcomes for the 8,971 ICA and 5,116 ICU low-risk monitor patients are also shown in Table 6. Unit readmission and transfer to another ICU were infrequent for ICA and ICU low-risk monitor patients. The mean in-unit death rate was significantly higher among ICA patients (1.1%) compared to ICU low-risk monitor patients (0.6%; p ⬍ 0.004). The mean hospital mortality rate was also significantly higher for ICA patients (3.1%) compared to ICU low-risk monitor patients (2.3%; p ⬍ 0.003). Discussion This study of 8,971 patients who were admitted to 37 ICAs at 32 US hospitals is unique because www.chestjournal.org

previous descriptions22–27,32–35 have been limited to individual units or hospitals. ICA patients had a mean age of 64 years and low severity of illness, with cardiovascular diagnoses accounted for 47% of nonoperative ICA admissions and vascular surgery accounting for 24% of postoperative ICA admissions. We compared the characteristics and outcomes of these ICA patients to those of 5,116 ICU monitor patients who received only monitoring and concentrated nursing care during their first ICU day and were at a low risk (ie, ⬍ 10%) for receiving subsequent active life-supporting therapy.29,31 We performed this comparison because we29 –31,37 and others11,40 – 44 have suggested previously that physicians could identify ICU low-risk monitor patients and treat them safely and effectively in an ICA. When compared to ICA patients, the ICU low-risk monitor patients were similar in regard to gender, race, comorbidities, and the most frequent admission diagnoses. The ICA patients were, on average, 4 years older and had fewer physiologic abnormalities (mean APACHE III score, 16.7 vs 19.8, respectively) than did ICU low-risk monitor patients. A higher proportion of ICA patients were nonoperative patients, a finding that is probably related to the comparison of patients who were admitted to telemetry and medical-surgical ICAs (97.3%) vs a sample of patients in which 51% of the ICUs to which they were admitted were specialized. The vast majority of ICA patients (91.7%) received only monitoring and concentrated nursing care on day 1, and 85.3% of these patients were at a low risk (ie, ⬍ 10%) for receiving subsequent active therapy. Because 95.7% of these patients never received subsequent active therapy, our results suggest that the method used to assess the risk for active therapy among ICU monitor patients is also applicable in ICAs.31 A similar proportion of ICU low-risk monitor patients (94.3%) never received subsequent active therapy, a finding that is similar to the previously reported value (95.6%)31 and confirms the validity of this method for predicting a low risk (ie, ⬍ 10%) of active therapy in an independent sample of ICU monitor patients. Low-risk monitor patients with identical diagnoses were frequently admitted to either an ICA or an ICU at the same 32 hospitals. This suggests that physicians use different selection criteria for ICA and ICU admissions among low-risk monitor patients with similar conditions. A greater extent of physiologic abnormality appears to be an important factor in the decision to admit low-risk monitor patients to an ICU rather than an ICA, but other factors that might account for such decisions may not be captured by our data. For example, a surgeon might monitor a patient in the ICU after the patient had undergone an carotid endarterectomy if the obstruction was CHEST / 121 / 4 / APRIL, 2002

1257

Table 4 —Most Frequent Nonoperative Hospital Admission Diagnoses and Unit Lengths of Stay Among ICA Patients and ICU Low-Risk Monitor Patients ICA Patients (n ⫽ 7,465)

ICU Low-Risk Monitor Patients (n ⫽ 2,991)

Admission Diagnosis*

Patients, No. (%)

Mean Stay, d

Median Stay, d

Patients, No. (%)

Mean Stay, d

Median Stay, d

Unstable angina Cardiac dysrrhythmia Congestive heart failure Acute myocardial infarction Bacterial/viral pneumonia Cerebral vascular accident Multiple trauma Upper GI bleeding (ulcer or laceration) Head trauma (with or without multiple trauma) Chronic bronchitis or emphysema Self-inflicted drug overdose Asthma

1,297 (17.4) 635 (8.5) 631 (8.4) 356 (4.8) 206 (2.8) 211 (2.8) 193 (2.6) 162 (2.2)

3.4 4.0 4.9 4.3 5.7 4.9 3.9 3.6

3.0 3.0 4.0 4.0 4.0 4.0 3.0 3.0

641 (21.4) 149 (5.0) 194 (6.5) 188 (6.3) 64 (2.1) 83 (2.7) 124 (4.1) 48 (1.6)

2.3 2.2 3.0 3.0 3.0 3.4 2.9 2.5

2.0 2.0 2.0 3.0 3.0 3.0 2.0 2.0

137 (1.8)

4.3

3.0

106 (3.5)

3.5

2.0

132 (1.8) 124 (1.7) 77 (1.0)

4.4 2.3 4.0

4.0 2.0 3.0

59 (1.9) 348 (11.6) 63 (2.1)

2.6 1.9 2.2

2.0 2.0 2.0

*Admission diagnoses are displayed according to frequency among ICA patients. See Zimmerman et al36 for a complete list of the 39 nonoperative admission diagnoses included in the APACHE III prognostic system.

highly symptomatic or was complicated by significant coronary artery disease, whereas an identical but asymptomatic patient without coronary artery disease might be admitted to an ICA. Although the selection criteria are uncertain, our findings support previous suggestions3,31,42– 44 that physicians can identify low-risk monitor patients who can be treated in an ICA rather than an ICU. Prerequisites for advocating the use of ICAs are patient safety and the ability to improve efficiency

and to reduce cost. Our most striking findings were that ICA patients had a significantly higher hospital death rate than ICU low-risk monitor patients (3.1% vs 2.3%, respectively; p ⬍ 0.003) and a longer mean unit length of stay (3.9 vs 2.6 days, respectively; p ⬍ 0.001). This suggests that ICAs may be associated with an excessive death rate and a more prolonged length of stay. Before reaching this conclusion, however, several factors that might explain a higher mortality rate and longer unit length of stay

Table 5—Most Frequent Postoperative Admission Diagnoses and Unit Lengths of Stay Among ICA Patients and ICU Low-Risk Monitor Patients Intermediate Care Area Patients (n ⫽ 7,465)

ICU Low Risk Monitor Patients (n ⫽ 2,991)

Admission Diagnosis*

Patients, No. (%)

Mean Stay, d

Median Stay, d

Patients, No. (%)

Mean Stay, d

Median Stay, d

Carotid endarterectomy Laminectomy Surgery for peripheral vascular ischemia Surgery for GI neoplasm Surgery for lung neoplasm Surgery for multiple trauma Femoral-popliteal artery graft Surgery for cholecystitis or cholangititis Surgery for GI obstruction Surgery for renal neoplasm Surgery for GI perforation Surgery for aortoiliac or aortofemoral obstruction

170 (11.3) 162 (10.7) 108 (7.2)

2.3 3.6 3.2

2.0 3.0 2.0

201 (9.5) 128 (6.1) 139 (6.6)

2.1 2.5 2.2

2.0 2.2 2.0

97 (6.4) 76 (5.0) 76 (5.0) 62 (4.1) 42 (2.8)

3.5 2.7 3.5 2.9 3.1

3.0 2.0 2.0 2.0 2.5

118 (5.6) 300 (14.2) 72 (3.4) 128 (6.1) 30 (1.4)

2.8 2.4 2.2 2.2 2.5

2.0 2.0 2.0 2.0 2.0

41 (2.7) 29 (1.9) 28 (1.8) 27 (1.8)†

3.6 2.7

3.0 2.0

3.3 2.6

3.0 2.0

2.6

2.0

29 (1.4) 41 (1.9) 6 (0.3) 52 (2.5)‡

2.9

2.0

*Admission diagnoses are displayed according to frequency among intermediate care area patients. See Zimmerman et al36 for a complete list of the 26 postoperative admission diagnoses included in the APACHE III prognostic system. †Includes 11 aortofemoral, 2 aortoiliac, and 14 femoral-femoral grafting procedures. ‡Includes 26 aortofemoral, 4 aortoiliac, and 22 femoral-femoral grafting procedures. 1258

Clinical Investigations in Critical Care

Table 6 —Resource Use and Outcomes for ICA Patients and ICU Low-Risk Monitor Patients* Variables Unit length of stay, d Mean Median 25th percentile 75th percentile Active treatment on unit day 1 Monitoring only on unit day 1 Risk for active treatment‡ ⬍ 5% 5–10% ⬎ 10% Transfer to another ICU Unit readmission Mean hospital length of stay, d Deaths in unit Deaths in hospital

ICA Patients (n ⫽ 8,971)

ICU Low-Risk Monitor Patients (n ⫽ 5,116)

3.9 ⫾ 3.1 3.0 2.0 5.0 747 (8.3) 7,579 (84.5)

2.6 ⫾ 2.2 2.0 2.0 3.0 0† 5,116 (100)†

3,998 (52.8) 2,459 (32.5) 1,120 (14.8) 98 (1.1) 446 (5.0) 8.9 ⫾ 9.1 100 (1.1) 281 (3.1)

3,217 (62.9) 1,899 (37.1) 0† 32 (0.6) 205 (4.0) 8.7 ⫾ 8.2 32 (0.6) 116 (2.3)

p Value 0.005

0.005 0.009 0.18 0.004 0.003

*Values given as mean ⫾ SD or No. (%), unless otherwise indicated. †By definition. ‡Unit day 1. The values exclude 569 patients admitted for chest pain to rule out myocardial infarction and 78 patients with missing day 1 active treatment data.

for ICA patients must be considered. First, there are significant differences in case mix, particularly age, severity of physiologic abnormality, and disease distribution between the two patient groups. Second, because the published guidelines for unit admission are different,3,44 the selection criteria for ICAs vs ICUs may have varied. For example, some patients admitted to ICAs may have had a chronic critical illness, a condition associated with a poor prognosis and prolonged care.45,46 Third, there were differences in prior therapy and in the timing of admission for ICA and ICU low-risk monitor patients. These differences in lead-time could have an important impact on mortality rate and length of stay for ICA patients, particularly those who are transferred from an ICU or from another hospital. As previously

shown in ICU populations, an adjustment for each of the above differences is required before mortality rate and length of stay can be compared36,47,48 A preliminary analysis49 showed that predicted outcomes for ICA patients, which are adjusted for the differences in case mix, patient selection, and lead-time based on their implications for ICU patients, differ from observed values (ie, they are imperfectly calibrated). Thus, APACHE III hospital

Table 7—Most Prevalent Admission Diagnoses Among ICA Patients Who Received Active Life-Supporting Therapy on Unit Day 1 Patients Actively Treated, No. Patients, No. (%)

Admission Diagnosis Nonoperative diagnoses Seizure Cardiac dysrrhythmia Upper GI bleeding due to ulcer or laceration Pneumonia (bacterial/viral) Sepsis (including urinary tract) Postoperative diagnoses Surgery for GI perforation Carotid endarterectomy Surgery for multiple trauma Surgery for cholecystitis or cholangitis Surgery for peripheral vascular disease* Surgery for lung neoplasm

Figure 2. Distribution of the unit length of stay for 8,971 ICA and 5,116 ICU low-risk monitor patients. www.chestjournal.org

69 637 162

11 (15.9) 88 (13.8) 22 (13.6)

227 116

23 (10.1) 10 (10.2)

28 170 76 42

7 (25.0) 40 (23.5) 11 (14.5) 5 (11.9)

192

18 (9.4)

76

5 (6.6)

*Includes femoral-popliteal, aortofemoral, aortoiliac and other vascular grafts. CHEST / 121 / 4 / APRIL, 2002

1259

mortality rate and unit length of stay predictions50 do not provide accurate benchmarks for assessing these outcomes for ICA patients. Although the same factors are likely to impact outcomes for ICA patients, their implications are probably different than in an ICU population. For this reason, we plan to examine the impact of recalibrating these predictive equations within an ICA population. Our study has several limitations. First, our data cannot be considered to be representative of intermediate care (ie, high-dependency care) in other countries. Studies in the United Kingdom,9,51,52 France,53,54 Hong Kong,55 and New Zealand56 show that high-dependency units in these countries provide more life support and care for patients with a higher severity of illness and risk of mortality. Second, our data also may not be representative of intermediate care in many US hospitals because of the small number of units studied and due to the nonrandom selection of units for study. Third, we did not collect data about limits on care or about do-not-resuscitate orders, factors that might have a substantial impact on patient survival and triage in ICAs.36,57,58 Fourth, we have no information about the structure or management of ICAs. Factors such as bed availability and triage pressures may have an important impact on unit resource use and patient outcome.59 In conclusion, this observational sample of ICA patients demonstrates a substantial overlap with those of less severely ill ICU patients. The vast majority of ICA patients are admitted to the unit for monitoring, are at a low risk for receiving subsequent active therapy, and few require transfer to an ICU. Comparisons of unit length of stay and hospital death rate for ICA and ICU low-risk monitor patients, however, will require adjustment for factors that have been shown previously to influence these outcomes. The growing demand for ICU services, coupled with reductions in nurse staffing on floors is likely to increase the need to evaluate the safety and efficiency of ICAs. A description of the clinical features and outcomes for this patient sample represents a first step toward examining patient needs and in developing benchmarks for evaluating resource use and outcomes for ICA patients. References 1 Ridley SA. Intermediate care: possibilities requirements and solutions. Anaesthesia 1998; 53:654 – 664 2 Vincent JL, Burchardi H. Do we need intermediate care units? Intensive Care Med 1999; 25:1345–1349 3 American College of Critical Care. Medicine of the Society of Critical Care Medicine guidelines on admission and discharge for adult intermediate care units. Crit Care Med 1998; 26:607– 610 1260

4 Bion J. Rationing intensive care: preventing critical illness is better, and cheaper, than cure. BMJ 1995; 310:682– 683 5 Rosenthal GE, Sirio CA, Shepardson LB, et al. Use of intensive care units for patients with low severity of illness. Arch Intern Med 1998; 158:1144 –1151 6 Keenan SP, Massel D, Inman KJ, et al. A systematic review of the cost effectiveness of noncardiac transitional care units. Chest 1998; 113:172–177 7 Thijs LG. Geographical differences in outcome. In: Sibbald WJ, Bion JF, eds. Evaluating critical care: using health services research to improve quality. Berlin, Germany: Springer-Verlag, 2000; 292–308 8 Lyons RA, Wareham K, Hutchings HA, et al. Population requirement for adult critical care beds: a prospective quantitative and qualitative study. Lancet 2000; 355:595–598 9 Thompson FJ, Singer M. High dependency units in the UK: variable size, variable character, few in number. Postgrad Med J 1995; 71:217–221 10 Weissman C. Factors influencing changes in surgical intensive care unit utilization. Crit Care Med 2000; 28:1766 –1771 11 Bone RC, McElwee NE, Eubanks DH, et al. Analysis of indications for early discharge from the intensive care unit. Chest 1993; 104:1812–1817 12 Groeger JS, Guntupalli KK, Strosberg M, et al. Descriptive analysis of critical care units in the United States: patient characteristics and intensive care unit utilization. Crit Care Med 1993; 21:279 –291 13 Ryan DW, Bayly PJM, Weldon OGW, et al. A prospective two month audit of the lack of provision of a high dependency unit and its impact on intensive care. Anaesthesia 1997; 52:265– 275 14 Durbin CG, Kopel RF. A case-control study of patients readmitted to the intensive care unit. Crit Care Med 1993; 21:1547–1553 15 Chen LM, Martin CM, Keenan SP, et al. Patients readmitted to the intensive care unit during the same hospitalization: clinical features and outcomes. Crit Care Med 1998; 26: 1834 –1841 16 Rosenberg AL, Watts C. Patients readmitted to intensive care units: a systematic review of risk factors and outcomes. Chest 2000; 118:492–502 17 Franklin CM, Rackow EC, Mamdani B, et al. Decreases in mortality on a large urban medical service by facilitating access to critical care. Arch Intern Med 1988; 148:1403–1405 18 Porath A, Reuveni H, Grinberg G, et al. The intermediate care unit as a cost effective option for the treatment of medical patients in critical condition. Isr J Med Sci 1995; 31:674 – 680 19 Coggins R, de Cossart L. Improving postoperative care: the role of the surgeon in the high dependency unit. Ann R Coll Surg Engl 1996; 78:163–167 20 Tosteson ANA, Goldman L, Udvarhelyi S, et al. Cost effectiveness of a coronary care unit versus an intermediate care unit for emergency department patients with chest pain. Circulation 1996; 94:143–150 21 Farkouh ME, Smars PA, Reeder GS, et al. A clinical trial of a chest pain observation unit for patients with unstable angina. N Engl J Med 1998; 339:1882–1888 22 Krieger BP, Ershowsky P, Spivack D. One year’s experience with a noninvasively monitored intermediate care unit for pulmonary patients. JAMA 1990; 264:1143–1146 23 Gracey DR, Hardy DX, Koenig GE. The chronic ventilator dependent unit: a lower cost alternative to intensive care. Mayo Clin Proc 2000; 75:445– 449 24 Douglas S, Daly B, Rudy E, et al. The cost effectiveness of a special care unit to care for the chronically critically ill. J Nurs Adm 1995; 25:47–53 Clinical Investigations in Critical Care

25 Elpern EH, Silver MR, Rosen RL, et al. The noninvasive respiratory care unit: patterns of use and financial implications. Chest 1991; 99:205–208 26 Cullen DJ, Nemeskal AR, Zaslavsky AM. Intermediate TISS: a new therapeutic intervention scoring system for non-ICU patients. Crit Care Med 1994; 22:1406 –1411 27 Dasgupta A, Rice R, Mascha E, et al. Four year experience with a unit for long term ventilation (respiratory special care unit) at the Cleveland Clinic Foundation. Chest 1999; 116: 447– 455 28 Kalb PE, Miller DH. Utilization strategies for intensive care units. JAMA 1989; 261:2389 –2395 29 Wagner DP, Knaus WA, Draper EA. Identification of low risk monitor admissions to medical-surgical ICUs. Chest 1987; 92:423– 428 30 Zimmerman JE, Junker CD, Becker RB, et al. Neurological intensive care admissions: identifying candidates for intermediate care and the services they receive. Neurosurgery 1998; 42:91–102 31 Zimmerman JE, Wagner DP, Knaus WA, et al. The use of risk predictions to identify candidates for intermediate care units. Chest 1995; 108:490 – 499 32 Willmitch B, Egol A. Utilization of a step down unit for patients undergoing carotid endarterectomy [abstract]. Crit Care Med 1998; 26(suppl):A41 33 Lustbader D, Cooper D, Reiser P, et al. Methodology for improved ICU resource utilization and quality of care [abstract]. Chest 1998; 114(suppl):254S 34 Lustbader D, Dlugacz Y, Weissman G, et al. Regional benchmarking for critical care: methodology for quality improvement [abstract]. Chest 1998; 114(suppl):342S 35 Barrie PS, Eachempati SR, Hydo LJ. Impact of a new intermediate care unit on utilization and outcomes of a surgical intensive care unit [abstract]. Crit Care Med 1999; 27(suppl):A28 36 Zimmerman JE, Wagner DP, Draper EA, et al. Evaluation of acute physiology and chronic health evaluation III predictions of hospital mortality in an independent database. Crit Care Med 1998; 26:1317–1326 37 Zimmerman JE, Wagner DP, Sun X, et al. Planning patient services for intermediate care units: insights based on care for intensive care unit low risk monitor admissions. Crit Care Med 1996; 24:1626 –1632 38 Knaus WA, Wagner DP, Draper EA, et al. The APACHE III prognostic system: risk prediction of hospital mortality for critically ill hospitalized adults. Chest 1991; 100:1619 –1636 39 Weissman C. Analyzing intensive care unit length of stay data: problems and possible solutions. Crit Care Med 1997; 25: 1594 –1600 40 Bone RC, McElwee NE, Eubanks DH, et al. Analysis of indications for intensive care unit admission. Chest 1993; 104:1806 –1811 41 Henning RJ, McClish D, Daly B, et al. Clinical characteristics and resource utilization of ICU patients: implications for organization of intensive care. Crit Care Med 1987; 15:264 – 269 42 Byrick RJ, Power JD, Ycas JO, et al. Impact of an interme-

www.chestjournal.org

43 44

45

46

47 48 49 50 51 52 53 54 55 56 57

58 59

diate care area on ICU utilization after cardiac surgery. Crit Care Med 1986; 14:869 – 872 Swann D, Houston P, Goldberg J. Audit of intensive care unit admissions from the operating room. Can J Anaesth 1993; 40:137–141 American College of Critical Care Medicine. Guidelines for intensive care unit admission, discharge, and triage: Task force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Crit Care Med 1999; 27:633– 638 Nasraway SA, Button GJ, Rand WM, et al. Survivors of catastrophic illness: outcome after direct transfer from intensive care to extended care facilities. Crit Care Med 2000; 28:19 –25 Seneff MG, Wagner DP, Thompson D, et al. The impact of long term acute care facilities on the outcome and cost of care for patients undergoing prolonged mechanical ventilation. Crit Care Med 2000; 28:342–350 Rosenberg AL, Zimmerman JE, Alzola C, et al. Intensive care unit length of stay: recent changes and future challenges. Crit Care Med 2000; 28:3465–3473 Randolph AG, Guyatt GH, Carlet J, et al. Understanding articles comparing outcomes among intensive care units to rate quality of care. Crit Care Med 1998; 26:773–781 Zimmerman JE, Alzola C, Junker C. Effectiveness and efficiency of intermediate care units: assessment based on intensive care benchmarks [abstract]. Chest 1999; 116(suppl):239S Knaus WA, Wagner DP, Zimmerman JE, et al. Variations in mortality and length of stay in intensive care units. Ann Intern Med 1993; 118:753–761 Shakir T, Toosy N, Ridley SA. A survey of adult general high dependency units in the United Kingdom. Clin Intensive Care 2000; (special issue):15–22 Garfield M, Jeffrey R, Ridley S. An assessment of the staffing level required for a high dependency unit. Anaesthesia 2000; 55:137–143 French Multicentric Group of ICU. Description of various types of intensive and intermediate care units in France. Intensive Care Med 1989; 15:260 –265 Auriant I, Vinatier I, Thaler F, et al. Simplified acute physiology score II for measuring severity of illness in intermediate care units. Crit Care Med 1998; 26:1368 –1371 Ip SPS, Leung YF, Ip CY, et al. Outcomes of critically ill elderly patients: is high dependency care for geriatric patients worthwhile? Crit Care Med 1999; 27:2351–2357 Havill JH, Cranston D. The place of the high dependency unit in a modern New Zealand hospital. N Z Med J 1998; 111:203–205 Kollef MH, Ward S. The influence of access to a private attending physician on the withdrawal of life sustaining therapies in the intensive care unit. Crit Care Med 1999; 27:2125–2132 Jayes RL, Zimmerman JE, Wagner DP, et al. Variations in the use of do not resuscitate orders in ICUs. Chest 1996; 110:1332–1339 Cheng DCH, Byrick RJ, Knobel E. Structural models for intermediate care areas. Crit Care Med 1999; 27:2266 –2271

CHEST / 121 / 4 / APRIL, 2002

1261