Ambulatory care-sensitive conditions: clinical outcomes and impact on intensive care unit resource use. (Original Article).

Background: We identified risk factors and clinical outcomes associated with ambulatory care-sensitive conditions requiring intensive care unit (ICU) admission.

Methods: This prospective cohort study included 4,144 patients admitted to the medical ICU of an urban teaching hospital during a 3-year period.

Results: A total of 627 patients were classified as having ambulatory care-sensitive conditions (ie, potentially preventable ICU admissions). Black race, decreasing Acute Physiology and Chronic Health Evaluation II (APACHE II) score, younger age, female sex, and absence of immunodeficiency were independently associated with ambulatory care-sensitive conditions. Patients classified as having ambulatory care-sensitive conditions accounted for 2,006 ventilator days, 2,508 ICU days, and 5,392 hospital days. The hospital mortality rate was statistically lower for patients with ambulatory care-sensitive conditions than for patients without these conditions. Patients classified as having ambulatory care-sensitive conditions were also statistically more likely than other patients to lack health insurance and to sign out of the hospital against medical advice.

Conclusion: Patients with ambulatory care-sensitive conditions account for a substantial portion of all admissions to the intensive care unit. These data suggest that interventions aimed at preventing such admissions could improve ICU bed use.

**********

The aging population of the United States has contributed to a forecast that the proportion of care provided by intensive care unit (ICU) specialists and pulmonologists in the United States will decrease below current standards within the next 10 years. (1) Medicare enrollment is also expected to increase by more than 50% during the next 30 years. (2) This demographic change will have a direct impact on the use of ICUs, because the elderly consume disproportionately more health care resources, (3) and it may influence patient outcomes if ICUs are unable to accommodate the increasing number of patients. A recent study performed in the United Kingdom showed that demand for ICU beds has outpaced available resources in recent years, resulting in premature ICU discharges and increased hospital mortality. (4) Therefore, appropriate use of ICU beds is important to optimizing the benefits of this resource.

Several groups of investigators have identified groups of patients admitted to ICUs for the treatment of ambulatory care-sensitive conditions. Iatrogenic complications, including medication errors and hospital-acquired infections, are non-ambulatory-care-sensitive conditions that also can result in potentially preventable admissions to ICUs. (5-7) Ambulatory care-sensitive conditions that require hospital admission are important to target, because some of these admissions are potentially preventable. (8-10) Therefore, we performed an investigation with two main goals. First, we wanted to identify the number of patients classified as having ambulatory caresensitive conditions who were admitted to our institution's medical ICU. Our second goal was to compare the outcomes and the use of hospital resources among patients with and without ambulatory care-sensitive conditions.

Patients and Methods

This study was conducted at an urban university-affiliated teaching hospital with 1,000 beds. During a 3-year period (August 1997 to June 2000), all patients admitted to the medical ICU were eligible for inclusion in this investigation. The medical ICU is a closed unit with medical house officers and critical care fellows providing patient care under the direction of physicians who are board-certified in critical care medicine. Patients were entered into the study if they were admitted in the medical service at the time of admission to the ICU. Patients were excluded from enrollment into this study if they were transferred to the medical ICU temporarily because of a lack of available beds in one of the other hospital ICUs. The Washington University School of Medicine Human Studies Committee approved this study.

Study Design and Data Collection

A prospective cohort study design was used, with patients segregated according to whether they were classified as having an ambulatory care-sensitive condition. The primary outcome evaluated in this investigation was length of stay for medical care (mechanical ventilation, ICU, and hospital). Secondary outcome measures included hospital mortality, the occurrence of hospital-acquired infection (bacteremia, pneumonia, and acquisition of vancomycin-resistant enterococci), the development of acute renal failure, disposition at time of hospital discharge (home, nursing home, or extended care facility), and discharge against medical advice.

One of the nonphysician investigators made daily rounds in the medical ICU to identify eligible patients and record relevant data from medical records, bedside flow sheets, computerized bedside nursing stations (EMTEK Health Care Systems, Inc., Tempe, AZ), computerized radiography reports, and reports of microbiologic studies (Gram stain of sputum and cultures of sputum, blood, pleural fluid, and stool). Study patients were prospectively followed until they were discharged from the hospital or until death. The following patient characteristics were recorded at the time of ICU admission: age, sex, race, diagnosis at time of ICU admission, ratio of arterial blood oxygen tension to fractional concentration of inspired oxygen, severity of illness according to Acute Physiology and Chronic Health Evaluation (APACHE) II score, (11) hematocrit value, creatinine level (in mg/dl), health insurance status, and presence of chronic obstructive pulmonary disease, underlying malignancy, human immunodeficiency virus seroposi tivity, immunodeficiency, chronic organ insufficiency, and need for chronic hemodialysis. Specific processes of medical care examined during the patient's stay in the ICU included the use of mechanical ventilation, reintubation, tracheostomy, central venous cannulation, administration of enteral nutrition, administration of corticosteroids, histamine Type 2 receptor antagonists, sedative infusions, and acute hemodialysis.

Definitions

All definitions were selected prospectively as part of the original study design. We defined ambulatoty care-sensitive condition as either an intentional drug overdose or one of five chronic conditions for which timely and effective treatment in an outpatient setting usually prevents hospitalization (asthma, hypertension, diabetic ketoacidosis and nonketotic hyperosmolar coma, congestive heart failure, and epilepsy). (12,13) We also included within this definition acute pulmonary edema not associated with acute lung injury, acute myocardial infarction, or aspiration. This definition for acute pulmonary edema captures patients with chronic renal disease and/or congestive heart failure who were admitted to the ICU with cardiogenic or hydrostatic pulmonary edema.

Immunosuppression was considered present in patients who were administered corticosteroids, had a positive serum immunodeficiency virus antibody, had been treated with chemotherapy within the previous 45 days, had neutropenia (absolute neutrophil count, <0.5 >( 10 (9)/L) as a result of the administration of chemotherapy, or had received an organ transplant that required the administration of immunosuppressive agents. The definition of chronic organ insufficiency was adapted from APACHE II as cirrhosis with portal hypertension or encephalopathy; Class IV angina; pulmonary disease associated with chronic hypoxemia, hypercapnia, polycythemia, pulmonary artery hypertension, or home ventilator use; chronic peritoneal dialysis or hemodialysis; and immunosuppression. (11) Health insurance status was classified as none for individuals without any type of health insurance (uninsured, medically indigent, or patient-payer), public insurance for patients with Medicaid or Medicare alone, and private health insurance. Priv ate health insurance included fee-for-service health insurance and capitated health insurance obtained through either a health maintenance organization or a preferred provider organization.

Statistical Analysis

Univariate analysis was used to compare the study variables for the groups of interest (patients with or without ambulatory care-sensitive conditions). All comparisons were unpaired, and all tests of significance were two-tailed. Continuous variables were compared with the use of Student's t test for normally distributed variables and the Wilcoxon rank-sum test for nonparametric variables. The statistic was used to compare categorical variables. All values are expressed as the mean [+ or -] standard deviation (ie, continuous variables) or as a percentage of the group from which they were derived (ie, categorical variables).

Multiple logistic regression analysis was performed with a commercial statistical package (SAS/STAT; SAS Institute Inc., Gary, NC) to identify predictor variables that were significantly related to a patient's having an ambulatory care-sensitive condition. Multiple logistic regression analysis was performed with the use of models that were judged a priori to be clinically sound. This step was prospectively determined to be necessary to avoid producing spuriously significant results with multiple comparisons. A stepwise approach was used to enter new terms into the logistic regression models, where 0.05 was set as the limit for the acceptance or removal of new terms. The results of the logistic regression analysis are reported as adjusted odds ratios (AORs) with 95% confidence intervals (CIs). All significance values were two-tailed, and P < 0.05 was considered statistically significant.

Results

Patients

A total of 4,114 consecutive patients who required admission to the medical ICU were entered into the study. The mean age of the patients was 58.5 [+ or -] 18.4 years (age range, 14-106 yr), and the mean APACHE II score was 20.4 [+ or -] 8.9 (range, 0-53). There were 2,140 females (52%) and 1,974 males (48%). The diagnoses leading to ICU admission are shown in Table 1. A total of 627 patients (15%) were classified as having ambulatory care-sensitive conditions. Statistically, these patients were younger, more likely to be female and black, and had a lower incidence of chronic obstructive pulmonary disease, underlying malignancy, chronic organ insufficiency, immunodeficiency, and private health insurance than patients who were classified as having nonpreventable ICU admissions (Table 2). Patients classified as having ambulatory care-sensitive conditions also had statistically lower ratios of arterial blood oxygen tension to the fractional concentration of inspired oxygen, lower APACHE II scores, lower predicte d hospital morality, and statistically greater hematocrit values. Evaluation of the medical care provided to patients showed that the use of mechanical ventilation, reintubation, tracheostomy, central venous catheterization, enteral nutrition, corticosteroids, and histamine Type 2 receptor antagonists was statistically less common among patients classified as having ambulatory care-sensitive conditions than among patients who were classified as having nonpreventable ICU admissions (Table 3). Multiple logistic regression analysis identified decreasing (in 1-point increments) APACHE II scores (AOR, 1.09; 95% CI, 1.08-1.10; P < 0.001), younger (in 1-yr increments) age (AOR, 1.02; 95% CI, 1.01-1.02; P < 0.001), black race (AOR, 1.70; 95% CI, 1.55-1.85; P < 0.001), female sex (AOR, 1.25; 95% CI, 1.14-1.37; P = 0.017), and the absence of immunodeficiency (AOR, 1.71; 95% CI, 1.51-1.93; P < 0.001) as being independently associated with the presence of an ambulatory care-sensitive condition.

Outcomes and Disposition

Patients who were classified as having an ambulatory care-sensitive condition had statistically shorter durations of mechanical ventilation and stays in the hospital and the ICU (Table 4). These patients accounted for 2,006 (10.9%) of the total ventilator days, 2,508 (10.3%) of the total ICU days, and 5,392 (8.9%) of the total hospital days for all admissions to the medical ICU during the period of investigation. Patients with ambulatory care-sensitive conditions had statistically lower rates of hospital mortality, bacteremia, hospital-acquired pneumonia, acquisition of vancomycin-resistant enterococci, and the development of acute renal failure than patients without ambulatory care-sensitive conditions (Table 4). Patients who were classified as having ambulatory care-sensitive conditions were statistically more likely to leave the hospital against medical advice and were statistically less likely to require nursing home admission after hospital discharge. Multiple logistic regression analysis identified incr easing (in 1-point increments) APACHE II score (AOR, 1.15; 95% CI, 1.14-1.15; P < 0.001), bacteremia (AOR, 2.22; 95% CI, 1.85-2.66; P < 0.001), mechanical ventilation (AOR, 1.89; 95% CI, 1.72-2.08; P < 0.001), immunodeficiency (AOR, 1.45; 95% CI, 1.32-1.59; P < 0.001), and private health insurance (AOR, 1.23; 95% CI, 1.11-1.36; P = 0.046) as being independently associated with hospital mortality.

Discussion

This study shows that ambulatory care-sensitive conditions that require admission to a medical ICU are relatively common and account for more than 15% of total ICU admissions. Our multivariate analysis shows that patients who were classified as having ambulatory care-sensitive conditions were statistically younger and less likely to have immunodeficiency, had lower APACHE II scores, and were more likely to be female and black. In addition, patients with ambulatory care-sensitive conditions were statistically more likely to have no health insurance and statistically less likely to have private health insurance. The hospital mortality rate among patients with ambulatory care-sensitive conditions was also statistically lower than that of patients without ambulatory care-sensitive conditions.

Our results are in agreement with and extend the findings of previous investigations. Bindman et a1 (13) used statewide hospital discharge data and census information to identify patients with potentially preventable hospital admissions. They found that potentially preventable hospitalizations were more common in ZIP code areas with higher proportions of uninsured and Medicaid patients and among blacks. They also found that communities in which people perceive that they have poor access to medical care have higher rates of hospitalization for chronic diseases. Similarly, Gage et al (14) showed that antithrombotic therapy in Medicare beneficiaries is underused and is associated with measurable adverse outcomes, including hospitalizations that may be preventable. In the same study, patients older than 75 years of age, female patients, and individuals from rural rather than metropolitan areas were less likely to receive antithrombotic therapy despite the likely benefit of such therapy. Both of these studies supp ort the premise that socioeconomic factors may play an important role in determining access to outpatient medical care among individuals with chronic medical conditions. It is this lack of access to reasonable outpatient medical care that seems to predispose individuals to subsequent hospitalization because of deterioration of these chronic medical conditions.

Parchman and Culler (15) found that individuals in fair or poor health who live in areas designated as a primary care shortage areas are more likely to require hospitalization than are patients who live in adequately served communities. Even within the same metropolitan area, however, patients who fit specific socioeconomic profiles are less likely to have access to outpatient medical care, which can result in greater frequency of hospitalization for ambulatory care-sensitive conditions. (16-18) Black race, low income level, female sex, lack of secondary education, absence of health insurance, and specific area of residence are demographic markers of socioeconomic status that have been associated with reduced access to health care. (19-21) The findings of our study are consistent with these earlier observations in demonstrating that blacks and women are more likely to have ambulatory care-sensitive conditions that lead to admission to the ICU. We also have found that black patients were statistically less lik ely than white patients to have private health insurance (16.1 versus 31.4%; P < 0.001), whereas women and men had similar access to private health insurance (24.9 versus 24.0%; P = 0.527). These findings suggest that factors other than access to private health insurance may predispose individuals to seek ICU admission for ambulatory care-sensitive conditions.

The identified differences in access to health care among patients with specific demographic profiles have resulted in a number of interventions aimed at improving patient access to both inpatient and outpatient medical services. Unfortunately, several of these interventions have been unsuccessful, because they did not address the social issues responsible for patients' inability to access health care. (22,23) For example, making mammography available to lower socioeconomic status communities does not ensure the successful use of mammography if women do not have access to affordable, convenient transportation to the mammography center, the ability to provide for child care, or the necessary flexibility in taking time off from work. A number of authors have pointed out that making health insurance coverage available to patients from underserved areas does not ensure high-quality medical care in all sectors of society. (24) This point has been demonstrated in Canada, where socioeconomic differences in relation to the use of health care services can be demonstrated despite the provision of universal health insurance coverage. (25) When health insurance status is linked to other markers of socioeconomic status, however, it has been shown to be associated with the use of preventive and early diagnostic health care strategies. (26,27)

Our study has several limitations. First, we examined patients admitted to a single ICU. Therefore, our study results may not be applicable to other ICUs with different organizational structures and patient referral patterns. Other studies have identified patients with ambulatory care-sensitive conditions that required hospital admission, however, which makes our observations more universal. (8,9) Second, this analysis used a previously determined definition of ambulatory care-sensitive conditions resulting in potentially preventable admission to the ICU. (13) It is possible that the use of this definition may have either underestimated or overestimated the magnitude of this problem. The former possibility seems more likely, because the definition was biased toward a limited number of outpatient conditions. Third, we did not conduct patient interviews to determine whether there were other individual characteristics associated with ambulatory caresensitive conditions requiring intensive care. Bindman et al (13 ) showed that communities in which people perceive poor access to medical care have higher rates of hospitalization for chronic diseases. Such perceptions may further contribute to people's not seeking health care until hospitalization becomes necessary. In addition, mistrust of health care systems has been shown to occur in certain ethnic and racial communities, which may further impede access to needed medical care. (28,29) Also, this study's design was observational, and no attempt was made to prevent the ICU admissions.

Conclusions

We have shown that a substantial number of admissions to medical ICUs were associated with ambulatory care-sensitive conditions. Given the increasing future demands that will be placed on hospital and ICU resources, it seems prudent to develop health care strategies aimed at improving access to medical care for patients with these conditions. It seems, however, that the number or the proportion of patients with ambulatory care-sensitive conditions who require ICU admission is, at least in part, a function of access to ambulatory care. Additional studies are required to determine whether greater access to ambulatory medical care can reduce the need for hospital admission and reduce overall health care expenditures. (30) This task is daunting but has the potential to offer benefits for individual patients and society as a whole.

Table 1

Diagnoses in patients requiring medical intensive care unit admission

                                           No. of
Diagnosis                               patients (%)

Ambulatory care-sensitive
conditions (n = 627)
 Drug overdose                          141 (22.4%)
 Congestive heart failure               134 (21.4%)
 Pulmonary edema                        116 (18.5%)
 Diabetic ketoacidosis                  111 (17.7%)
 Asthma                                  58 (9.3%)
 Hypertension                            52 (8.3%)
 Epilepsy                                15 (2.4%)

Non-ambulatory-care-sensitive
conditions (n = 3,487)
 Sepsis                                 762 (21.9%)
 Gastrointestinal bleeding              481 (13.8%)
 Respiratory failure (nonasthma)        360 (10.3%)
 Infection                              272 (7.8%)
 Gastrointestinal disease               252 (7.2%)
 Chronic obstructive pulmonary disease  234 (6.7%)
 Neoplasm                               164 (4.7%)
 Postrespiratory arrest (nonasthma)     159 (4.6%)
 Renal failure                          150 (4.3%)
 Cardiovascular disease                 148 (4.2%)
 Post-cardiac arrest                    126 (3.6%)
 Coronary artery disease                 91 (2.6%)
 Other                                   70 (2.0%)
 Hemorrhagic shock/hypovolemia           64 (1.8%)
 Neurologic (nonseizure)                 61 (1.8%)
 Pulmonary embolism                      43 (1.2%)
 Rhythm disturbance                      22 (0.6%)
 Cardiogenic shock                       12 (0.4%)
 Dissecting aortic aneurysm               6 (0.2%)
 Intracranial/subarachnoid hemorrhage     6 (0.2%)
 Aspiration                               4 (0.1%)

Table 2

Baseline characteristics of the study cohort (a)

                                   Ambulatory care-sensitive
Characteristic                    condition present (n = 627)

Mean age (yr)                         50.0 [+ or -] 19.6
Sex, no. (%)
  Male                                    271 (43.2%)
  Female                                  356 (56.8%)
Race, no. (%)
  White                                   238 (38.0%)
  Black                                   385 (61.4%)
  Other                                    4 (0.6%)
Chronic obstructive pulmonary             67 (10.7%)
 disease, no. (%)
Underlying malignancy, no. (%)             61 (9.7%)
Human immunodeficiency virus               15 (2.4%)
 scropositive, no. (%)
Chronic organ insufficiency, No.          266 (42.4%)
 (%)
Chronic hemodialysis, no. (%)              57 (9.1%)
Immunodeficiency, no. (%)                  98 (15.6%)
Pa[O.sub.2]/Fi[O.sub.2] ratio           164 [+ or -] 149
APACHE II score                        14.8 [+ or -] 8.3
Predicted hospital mortality (%)       15.8 [+ or -] 19.8%
Creatinine (mg/dl)                      2.5 [+ or -] 3.1
Hematocrit (%)                         34.0 [+ or -] 6.9%
Health insurance status
  Private                                 121 (19.3%)
  Public                                  367 (58.5%)
  None                                    139 (22.2%)

                                   Ambulatory care-sensitive
Characteristic                    condition absent (n = 3,487)  P value

Mean age (yr)                          60.0 [+ or -] 17.7       <0.001
Sex, no. (%)
  Male                                    1703 (48.8%)           0.010
  Female                                  1784 (51.2%)
Race, no. (%)
  White                                   1944 (55.7%)          <0.001
  Black                                   1489 (42.7%)
 Other                                      54 (1.6%)
Chronic obstruetive pulmonary              720 (20.6%)          <0.001
 disease, no. (%)
Underlying malignancy, no. (%)             743 (21.3%)          <0.001
Human immunodeficiency virus                79 (2.3%)            0.845
 scropositive, no. (%)
Chronic organ insufficiency, No.          2254 (64.7%)          <0.001
 (%)
Chronic hemodialysis, no. (%)              259 (7.4%)            0.150
Immunodeficiency, no. (%)                  975 (28.0%)          <0.001
Pa[O.sub.2]/Fi[O.sub.2] ratio           194 [+ or -] 179        <0.001
APACHE II score                        21.4 [+ or -] 8.6        <0.001
Predicted hospital mortality (%)       41.8 [+ or -] 25.3%      <0.001
Creatinine (mg/dl)                      2.3 [+ or -] 2.5         0.175
Hematocrit (%)                         29.3 [+ or -] 7.2%       <0.001
Health insurance status
  Private                                 914 (26.2%)           <0.001
  Public                                 2280 (65.4%)
  None                                    293 (8.4%)

(a)Pa[O.sub.2]/Fi[O.sub.2], arterial blood oxygen tension to fractional
concentration of inspired oxygen; APACHE, Acute Physiology and Chronic
Health Evaluation.

Table 3

Process of medical care

                                   Ambulatory care-sensitive
Variable                          condition present (n = 627)

Mechanical ventilation, no. (%)           197 (31.4%)
Reintubation, no. (%)                      13 (2.1%)
Tracheostomy, no. (%)                      12 (1.9%)
Central venous catheter, no. (%)           63 (10.0%)
Enteral nutrition, no. (%)                 39 (6.2%)
Acute hemodialysis, no. (%)                13 (2.1%)
Sedative infusion, no. (%)                 30 (4.8%)
Corticosteroids, no. (%)                   47 (7.5%)
Histamine Type 2 antagonist, no.          265 (42.3%)
 (%)

                                   Ambulatory care-sensitive
Variable                          condition absent (n = 3,487)  P value

Mechanical ventilation, no. (%)          1,602 (45.9%)          <0.001
Reintubation, no. (%)                      149 (4.3%)            0.007
Tracheostomy, no. (%)                      192 (5.5%)           <0.001
Central venous catheter, no. (%)           792 (22.7%)          <0.001
Enteral nutrition, no. (%)                 726 (20.8%)          <0.001
Acute hemodialysis, no. (%)                 78 (2.2%)            0.884
Sedative infusion, no. (%)                 232 (6.7%)            0.091
Corticosteroids, no. (%)                   431 (12.4%)          <0.001
Histamine Type 2 antagonist, no.         2,277 (65.3%)          <0.001
 (%)

Table 4

Outcome and disposition profile (a)

                                    Ambulatory care-sensitive
                                        condition present
Variable                                    (n = 627)

Duration of mechanical ventilation      3.2 [+ or -] 4.3
 (d)
Intensive care unit days                4.0 [+ or -] 6.8
Hospital days                           8.6 [+ or -] 11.5
Hospital mortality, no. (%)                33 (5.3%)
Bacteremia, no. (%)                         9 (1.4%)
Hospital-associated pneumonia, no.          9 (1.4%)
 (%)
Acquisition of VRE, no. (%)                 3 (0.5%)
Acute renal failure, no. (%)              100 (15.9%)
Disposition of survivors, no. (%)
Home                                      560 (94.3%)
Nursing home                               34 (5.7%)
Discharged against medical advice,         21 (3.3%)
 no. (%)

                                    Ambulatory care-sensitive
                                        condition absent
Variable                                   (n = 3,487)         P value

Duration of mechanical ventilation      5.3 [+ or -] 6.9       <0.001
 (d)
Intensive care unit days                7.0 [+ or -] 9.0       <0.001
Hospital days                          17.3 [+ or -] 24.0      <0.001
Hospital mortality, no. (%)                905 (26.0%)         <0.001
Bacteremia, no. (%)                        156 (4.5%)          <0.001
Hospital-associated pneumonia, no.         157 (4.5%)          <0.001
 (%)
Acquisition of VRE, no. (%)                 71 (2.0%)           0.005
Acute renal failure, no. (%)               691 (19.8%)          0.023
Disposition of survivors, no. (%)
Home                                     2,239 (86.7%)         <0.001
Nursing home                               343 (13.3%)
Discharged against medical advice,          20 (0.6%)          <0.001
 no. (%)

(a)VRE, vancomycin-resistant enterococci.

Accepted July 25, 2002.

References

(1.) Angus DC, Kelley MA, Schmitz RJ, White A, Popovich J Jr; Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS). Caring for the critically ill patient: Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease--Can we meet the requirements of an aging population? JAMA 2000;284:2762-2770.

(2.) White J. Uses and abuses of long-term Medicare cost estimates. Health Aff (Millwood) 1999;18:63-79.

(3.) Iglehart JK. The American health care system: Medicare. N Engl J Med 1999;340:327-332.

(4.) Goldfrad C, Rowan K. Consequences of discharges from intensive care at night. Lancet 2000;355:1138-l142.

(5.) Darchy B, Le Miere E, Figueredo B, Bavoux E, Domart Y. latrogenic diseases as a reason for admission to the intensive care unit: Incidence, causes, and consequences. Arch Intern Med 1999;159:71-78.

(6.) Leape LL, Cullen DJ, Clapp MD, Burdick E, Demonaco HJ, Erickson JI, et al. Pharmacist participation on physician rounds and adverse drug events in the intensive care unit. JAMA 1999;282:267-270.

(7.) Jarvis WR. Selected aspects of the socioeconomic impact of nosocomial infections: Morbidity, mortality, cost, and prevention. Infect Control Hosp Epidemiol 1996;17:552-557.

(8.) Ordonez GA, Phelan PD, Olinsky A, Robertson CF. Preventable factors in hospital admissions for asthma. Arch Dis Child 1998;78:143-147.

(9.) Michalsen A, Konig G, Thimme W. Preventable causative factors leading to hospital admission with decompensated heart failure. Heart 1998;80:437-441.

(10.) Rauh RA, Schwabauer NJ, Eager EL, Moran JF. A community hospital-based congestive heart failure program: Impact on length of stay, admission and readmission rates, and cost. Am J Manag Care 1999;5:37-43.

(11.) Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: A severity of disease classification system. Crit Care Med 1955;13:818-829.

(12.) Sporer KA. Acute heroin overdose. Ann intern Med 1999;130:584-590.

(13.) Bindman AB, Grumbach K, Osmond D, Komaromy M, Vranizan K, Lurie N, et al. Preventable hospitalizations and access to health care. JAMA 1995;274:305-311.

(14.) Gage BF, Boechler M, Doggette AL, Fortune G, Flaker GC, Rich MW, et al. Adverse outcomes and predictors of underuse of antithrombotic therapy in Medicare beneficiaries with chronic atrial fibrillation. Stroke 2000;31:822-827.

(15.) Parchman ML, Culler SD. Preventable hospitalizations in primary care shortage areas: An analysis of vulnerable Medicare beneficiaries. Arch Fam Med 1999;8:487-491.

(16.) Hannan EL, van Ryn M, Burke J, Stone D, Kumar D, Arani D, et al. Access to coronary artery bypass surgery by race/ethnicity and gender among patients who are appropriate for surgery. Med Care 1999;37:68-77.

(17.) Carlisle DM, Leake BD, Shapiro MF. Racial and ethnic differences in the use of invasive cardiac procedures among cardiac patients in Los Angeles County, 1986 through 1988. Am J Public Health 1995;85:352-356.

(18.) Ayanian JZ, Kobler BA, Abe T, Epstein AM. The relation between health insurance coverage and clinical outcomes among women with breast cancer. N Engl J Med 1993;329:326-331.

(19.) 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.

(20.) Whittle J, Conigliaro J, Good CB, Lofgren RP. Racial differences in the use of invasive cardiovascular procedures in the Department of Veterans Affairs medical system. N Engl J Med l993;329:62l-627.

(21.) Shea S, Misra D, Ehrlich MH, Field L, Francis CK. Predisposing factors for severe, controlled hypertension in an inner-city minority population. N Engl J Med 1992;327:776-781.

(22.) Simon MS, Gimotty PA, Moncrease A, Dews P, Burack RC. The effect of patient reminders on the use of screening mammography in an urban health department primary care setting. Breast Cancer Res Treat 2001;65:63-70.

(23.) Escalante A, Espinosa-Morales R, del Rincon I, Arroyo RA, Older SA. Recipients of hip replacement for arthritis are less likely to he Hispanic, independent of access to health care and socioeconomic status. Arthritis Rheum 2000;43:390-399.

(24.) Eisenberg JM, Power EJ. Transforming insurance coverage into quality health care: Voltage drops from potential to delivered quality. JAMA 2000;284:2100-2107.

(25.) Katz SJ, Hofer TP. Socioeconomic disparities in preventive care persist despite universal coverage: Breast and cervical cancer screening in Ontario and the United States. JAMA 1994;272:530-534.

(26.) Hsia J, Kemper E, Kiefe C, Zapka J, Sofaer S, Pettinger M, et al. The importance of health insurance as a determinant of cancer screening: evidence from the Women's Health Initiative. Prey Med 2000;31:261-270.

(27.) Parker JD, Schoendorf KC. Variation in hospital discharges for ambulatory care-sensitive conditions among children. Pediatrics 2000;106(4 Suppl):942-948.

(28.) LaVeist TA, Nickerson KJ, Bowie JV. Attitudes about racism, medical mistrust, and satisfaction with care among African American and white cardiac patients. Med Care Res Rev 2000;57(Suppl l):146-161.

(29.) Galavotti C, Richter DL. Talking about hysterectomy: the experiences of women from four cultural groups. J Womens Health Gend Based Med 2000;9(Suppl 2):S63-S67.

(30.) Rothschild JM, Bates DW, Leape LL. Preventable medical injuries in older patients. Arch Intern Med 2000;160:2717-2728.

RELATED ARTICLE: Key Points

* Patients with ambulatory care-sensitive conditions comprise a significant proportion of intensive care unit admissions.

* Individuals in disadvantaged sectors of society are disproportionately represented among patients with ambulatory care-sensitive conditions.

* Lack of adequate access to outpatient medical care in the community may increase the number of patients with ambulatory care-sensitive conditions who require hospital admission.

From the Pulmonary and Critical Care Division and the Division of Infectious Diseases, Washington University School of Medicine, and the Department of Nursing, Barnes-Jewish Hospital, St. Louis, MO.

Supported in part by grants from the Centers for Disease Control and Prevention (No. U50/CCU717925-01) and the Barnes-Jewish Hospital Research Foundation.

Reprint requests to Mario H. Kollef, MD, Pulmonary and Critical Care Division, Washington University School of Medicine, Campus Box 8052, 660 S. Euclid Avenue, St. Louis, MO 63110. Email: [e-mail]

Copyright [c] 2003 by The Southern Medical Association 0038-4348/03/9602-0172