Alison Glendenning-Napoli, BSN, RN-BC, Beverly Dowling, CPA, CHFP, John Pulvino, PA-C, Gwen Baillargeon, MS, and Ben G. Raimer, MD

Transcription

1 CE Professional Case Management Vol. 17, No. 6, Copyright 2012 Wolters Kluwer Health Lippincott Williams & Wilkins Community-Based Case Management for Uninsured Patients With Chronic Diseases Effects on Acute Care Utilization and Costs Alison Glendenning-Napoli, BSN, RN-BC, Beverly Dowling, CPA, CHFP, John Pulvino, PA-C, Gwen Baillargeon, MS, and Ben G. Raimer, MD ABSTRACT Purpose of the Study: To examine the effects of a community-based case management program on acute health care utilization and associated costs in uninsured patients with 1 or more chronic diseases. Primary Practice Setting: Large regional academic medical center that provides health care services for the vast majority of indigent patients in the area. Methodology and Sample: This was a retrospective study of 83 patients who enrolled in a case management program between April 2007 and August 2008 on the basis of 1 or more emergency department visits or acute hospitalizations. Paired t tests were used to compare utilization and costs before and after enrollment. Results: Overall, acute outpatient encounters decreased by 62% and inpatient admissions by 53%, whereas primary care visits increased by 162%. Participation in the case management program was also associated with a 41% reduction in overall aggregate costs, from $16,208 preintervention to $9,541 postintervention ( p.004). Implications for Case management Practice: The results of this study suggest that intensive case management can reduce acute care utilization and costs and increase primary care follow-up among uninsured patients with certain chronic diseases. Key words: case management, chronic diseases, emergency care, medical indigency, primary health care The United States is the only industrialized nation that does not provide universal health coverage for its citizens. In 1987, when the Census Bureau began collecting data on health insurance coverage, an estimated 31 million U.S. residents (12.9% of the population) lacked health insurance (DeNavas-Walt, Proctor, & Smith, 2011). Although the proportion of uninsured Americans has gradually increased over the past two decades, the current economic recession has given rise to an unprecedented number of people without health insurance. In 2010, an estimated 49.9 million individuals (16.3% of the population) were uninsured ( DeNavas-Walt et al., 2011 ). Because about 60% of Americans obtain health insurance through their employers, the loss of a job can lead to potentially catastrophic health consequences. Between 2008 and 2010, an estimated 15 million adults lost both their jobs and their job-based health benefits. The majority of these individuals (57%) were unable to find another source of health care coverage and became uninsured (Collins, Doty, Robertson, & Garber, 2011). Many newly unemployed and uninsured Americans have turned to public insurance programs in the wake of what has been characterized as the worst financial meltdown since the Great Depression. Medicaid enrollment increased by 17.8% between the start of the recession in December 2007 and June By the end of June 2010, more than 50 million persons were enrolled in Medicaid the largest number of enrollees in the history of the program (Kaiser Commission on Medicaid and the Uninsured, 2011). Even though Medicaid has provided a safety This study was presented in part at the 139th Annual Meeting of the American Public Health Association, Washington, DC, November 1, Address correspondence to Alison Glendenning-Napoli, BSN, RN-BC, Community Health Program, University of Texas Medical Branch, 301 University Boulevard, Galveston, TX ( alglende@utmb.edu ). The authors report no conflicts of interest. DOI: /NCM.0b013e f2b Vol. 17/No. 6 Professional Case Management 267

2 net for millions of Americans who have lost their jobs and health insurance, millions more do not qualify because of the program s restrictive eligibility criteria. Most states now restrict Medicaid coverage to pregnant women, children, and parents with very low incomes; nondisabled adults without children remain ineligible for Medicaid in the vast majority of states (Heberlein, Brooks, Guyer, Artiga, & Stephens, 2011). The scheduled implementation of the Affordable Care Act in 2014 should expand eligibility for Medicaid benefits ( Collins et al., 2011 ). Until then, however, the nation s continuing economic problems have forced many states to either impose or propose cuts in Medicaid spending (Kaiser Commission on Medicaid and the Uninsured, 2012; Parisi, 2010 ). Compared with persons who have medical insurance, the uninsured face numerous barriers to health care, resulting in poorer overall health status, delays in accessing treatment, and increased morbidity and mortality (Ayanian, Weissman, Schneider, Ginsburg, & Zaslavsky, 2000; Freeman, Kadiyala, Bell, & Martin, 2008; McWilliams, 2009 ; Wilper et al., 2009 ). Historically, the United States has relied on a patchwork system of safety net providers to provide health care for the uninsured (Institute of Medicine, 2000). Unfortunately, the limited capacity of the system to provide adequate access to preventive, primary, and specialty care has forced many uninsured persons to rely on hospital emergency departments as a major source for their health care services (Grumbach, Keane, & Bindman, 1993; Richardson & Hwang, 2001). For example, an analysis of emergency department usage patterns of the major safety net hospitals in Houston, TX, showed that approximately one-half of all emergency department visits were for Unfortunately, the limited capacity of the system to provide adequate access to preventive, primary, and specialty care has forced many uninsured persons to rely on hospital emergency departments as a major source for their health care. For example, an analysis of emergency department usage patterns of the major safety net hospitals in Houston, TX, showed that approximately one-half of all emergency department visits were for nonemergent, primary care-related conditions and that the uninsured were responsible for 52% of such visits in 2002 and 54% in Such reliance on the emergency department for basic health care contributes to fragmented services and the growing problem of emergency department overcrowding and rising health care costs. nonemergent, primary care-related conditions and that the uninsured were responsible for 52% of such visits in 2002 and 54% in 2003 (Begley, Vojvodic, Seo, & Burau, 2006). Preliminary reports suggest that this problem has been exacerbated by the current economic crisis. According to the National Association of Public Hospitals and Health Systems (2010), emergency department visits at safety net health systems by uninsured patients increased by 10% 15% in the last 6 months of 2009, when compared with the beginning of the recession. Uninsured patients with chronic diseases are more likely to visit the emergency department for treatment of their conditions than those who are insured (Collins, Davis, Doty, Kriss, & Holmgren, 2006; Wilper et al., 2008 ). They also are less likely to have a primary care medical home ( Wilper et al., 2008 ) and more likely to be hospitalized for their conditions ( Collins et al., 2006 ). Such reliance on the emergency department for basic health care contributes to fragmented services and the growing problem of emergency department overcrowding and rising health care costs. Case management interventions to reduce inappropriate emergency department usage and its associated costs have met with some success among populations covered by Medicaid (Grossman, Rich, & Johnson, 1998; Hurley, Freund, & Taylor, 1989). As an example, analysis of claims utilization data from four Medicaid demonstration programs showed that implementation of primary care case management in a population of both children and adults was associated with substantial reductions in the proportion of persons making emergency department visits ( Hurley et al., 1989 ). Experience with case management programs targeted at uninsured populations is more limited, however, and varying degrees of success in reducing acute health care utilization and associated costs have been reported (see the Discussion section; Horwitz, Busch, Balestracci, Ellingson, & Rawlings, 2005; Wetta-Hall, 2007 ). To further examine the use of case management interventions among the uninsured, we analyzed the outcomes of an intensive case management program among a cohort of uninsured patients with a history of emergency department visits and/or hospitalizations for treatment of one or more chronic diseases 268 Professional Case Management Vol. 17/No. 6

3 over a 17-month study period. We hypothesized that patient participation in an intensive case management program would reduce the number of acute health care encounters and associated costs and increase the use of primary care services. M ETHODS Sample The study population consisted of a convenience sample (i.e., nonprobability sample; see the Limitations section) of 83 patients without medical insurance who volunteered to participate in an intensive case management program at the University of Texas Medical Branch (UTMB) between April 1, 2007, and August 31, Criteria for inclusion in the study included: (1) a diagnosis of diabetes mellitus, essential hypertension, congestive heart failure, or coronary artery disease (alone or in combination); (2) a history of at least one inpatient admission or acute outpatient encounter (i.e., emergency department visit, outpatient day surgery, or hospital observation but not admission) at the UTMB within 12 months before enrollment; and (3) participation in the case management program for a minimum of 6 months, including at least two encounters with a case manager. Patients with a co-occurring serious mental illness or substance abuse problem were not eligible for enrollment. Each patient s primary and secondary diagnoses were established at the time of the initial acute outpatient visit or inpatient admission and classified according to the International Classification of Diseases, Ninth Revision, Clinical Modification. During the 17-month study period, a total of 363 eligible patients were identified. In spite of multiple attempts, 266 of the patients could not be contacted for various reasons (e.g., did not return phone calls, disconnected phones). Another 14 patients who were contacted declined to enroll. This left a study sample of 83 patients. Recruitment of additional study subjects and a longer follow-up period were not possible because of Hurricane Ike, which made landfall in Galveston, TX, in September The storm inundated more than 100 buildings on the UTMB campus with up to 10 feet of water, necessitating the closure of all clinics and hospitals and the temporary suspension of most indigent care services, including the intensive case management program (Lozano, 2008; Ortolon, 2009 ). The study was reviewed and approved by the UTMB institutional review board. Intervention The UTMB Community Health Program implemented a case management program in 2007 with the goal of reducing potentially avoidable hospitalizations and acute care visits, while increasing access to primary care for uninsured patients with select chronic diseases. Patients who are eligible for the program are identified on the basis of a hospitalization, acute outpatient encounter, or recurring clinic visits. A case manager, who is a registered nurse, then contacts the patient by telephone, provides an overall description of the program, and invites the patient to enroll. If the patient agrees to participate, the case manager schedules a home visit to enroll the patient and conduct a needs assessment that focuses on identifying barriers to accessing health care and determining the patient s health literacy level, especially knowledge of his or her medical condition and how to manage it. The case manager then develops a preventive care regimen tailored to the specific needs of the patient. A master s-level social worker assists the case manager in identifying the patient s need for public health programs and social services (e.g., county indigent health care program, Medicaid, disability, housing, and food assistance). Additional home visits are scheduled on an as-needed basis to assist patients in the day-to-day management of their chronic condition, such as education about the correct use of glucometers and blood pressure monitors, dietary and lifestyle modifications, and strategies to enhance adherence to prescribed medication regimens. Case managers also provide medication education sessions and help patients apply for pharmacy assistance programs. In addition, case managers accompany patients to their provider/clinic visits in an effort to better engage the patients in their care and promote a positive, effective relationship with their primary care provider. The case manager then reinforces the interventions with telephonic followup or further home visits or both to check progress, answer questions, and provide support. Data Collection Two data sources were used: an administrative billing database maintained by the UTMB Department of Finance and an electronic clinical information system and registry (Patient Electronic Care System, Aristos Group, Inc., Austin, TX) maintained by the UTMB Community Health Program. The administrative database was used to identify all clinical encounters (i.e., inpatient admissions, acute outpatient encounters, and clinic visits) that occurred during the study period and were related to treatment for any of the four chronic diseases under study and also to identify the costs associated with these services. The clinical information system was used to determine the patients demographic characteristics (age, gender, and race/ethnicity) and the duration of active participation Vol. 17/No. 6 Professional Case Management 269

4 in the case management program. The two data sets were linked using medical record numbers. Statistical Analysis Health care utilization and associated costs before enrollment in the case management program were compared with an equal period of time after enrollment for each individual. For example, if a patient was enrolled for 10 months, we compared utilization and cost history for the period of 10 months prior to enrollment with the 10-month period during enrollment. Because parallel health care utilization and cost data were collected before and after enrollment in the program, analyses were conducted such that each patient served as his or her own historical control. Statistical analyses were conducted using SAS software, version (SAS Institute, Cary, NC). Paired t tests were used to compare the mean differences for both health care utilization and associated costs before and after the intervention. All of the analyses were stratified across gender, race/ethnicity, and age. Administrative costs of the intervention program itself (e.g., enrolling patients, providing intervention services) were not included in our cost analyses. Statistical significance was defined as a p value of less than.05. TABLE 1 Demographic and Clinical Characteristics of Sample Total Sample Males Females N (%) 83 (100) 33 (39.8) 50 (60.2) Dempographic characteristics Race/ethnicity Non-Hispanic White 43 (51.8) 15 (45.5) 28 (56.0) Hispanic 19 (22.9) 6 (18.2) 13 (26.0) African American 21 (25.3) 12 (36.4) 9 (18.0) Age range (years) (26.5) 10 (30.3) 12 (24.0) (73.5) 23 (69.7) 38 (76.0) Clinical characteristics Chronic diseases Diabetes 44 (53.0) 16 (48.5) 28 (56.0) Hypertension 73 (88.0) 29 (87.9) 44 (88.0) Congestive heart failure 25 (30.1) 14 (42.4) 11 (22.0) Coronary artery disease 35 (42.2) 13 (39.4) 22 (44.0) Comorbidity Only 1 chronic disease 13 (15.7) 4 (12.1) 9 (18.0) 2 chronic diseases 18 (21.7) 8 (24.2) 10 (20.0) 3 chronic diseases 28 (33.7) 13 (39.4) 15 (30.0) 4 chronic diseases 24 (28.9) 8 (24.2) 16 (32.0) R ESULTS Table 1 presents the demographic and clinical characteristics of the study cohort. Of the 83 patients in the sample, 60.2% were female, 51.8% were non- Hispanic White, and 73.5% were between 50 and 65 years of age. Patient age at the time of enrollment in the case management program ranged between 18 and 65 years. Of the four chronic diseases under study, hypertension was the most prevalent among the cohort (88.0%), followed by diabetes (53.0%), coronary artery disease (42.2%), and congestive heart failure (30.1%). The vast majority of the patients (84.3%) had more than one of the four chronic diseases. Most of the patients (67.5%) had participated in the intervention program for 12 months or longer ( see Table 2 ). Participation in the case management program was associated with statistically significant reductions in both acute outpatient encounters and inpatient admissions along with a concomitant increase in primary care clinic visits ( see Table 3 ). The mean number of acute outpatient encounters declined from 0.70 before enrollment in the program to 0.27 after enrollment ( p.0007); inpatient admissions declined from a mean of 1.24 to 0.58 ( p.0001). Conversely, clinic visits increased from a mean of 4.13 to ( p.0001). All demographically stratified analyses showed similar statistically significant decreases or increases, with the exception of acute outpatient encounters for non- Hispanic Whites ( see Table 3 ). The total number of acute outpatient encounters decreased by 62%, from 58 preenrollment to 22 postenrollment, whereas inpatient admissions decreased by 53%, from 103 to 48. In contrast, the total number of primary care clinic visits increased by 162%, from 343 to 898 ( see Table 4 ). Likewise, participation in the program was associated with statistically significant reductions in costs for acute outpatient encounters ( p.009) and inpatient hospitalizations ( p.002), and an increase in costs for primary care clinic visits ( p.02) ( see Table 5 ). Specifically, the mean per patient cost for acute outpatient TABLE 2 Duration of Participation in Community Health Program (Cohort of 83 Patients) Monthsn % 6 to to to to Professional Case Management Vol. 17/No. 6

5 TABLE 3 Comparison of Health Care Utilization Before and After Enrollment in Community Health Program Characteristics Acute Outpatient Encounters a Inpatient Admissions Clinic Visits p p p Total sample ( N 83) 0.70 (0.95) 0.27 (0.61) (1.09) 0.58 (1.00) (4.01) (9.08).0001 Gender Female ( n 50) 0.70 (1.05) 0.24 (0.59) (1.02) 0.60 (0.97) (4.29) (8.84).0001 Male ( n 33) 0.70 (0.77) 0.30 (0.64) (1.19) 0.55 (1.06) (3.44) (9.56).0001 Race/ethnicity Non-Hispanic White ( n 43) 0.60 (0.93) 0.33 (0.71) (1.13) 0.74 (1.16) (4.39) (9.53).0001 Hispanic ( n 19) 0.84 (1.12) 0.11 (0.46) (0.96) 0.32 (0.48) (3.44) 9.42 (5.94).004 African American ( n 21) 0.76 (0.83) 0.29 (0.46) (1.15) 0.48 (0.98) (3.75) (10.62).004 Age range (years) ( n 22) 1.09 (1.41) 0.32 (0.57) (1.41) 0.77 (1.07) (3.91) (10.58) ( n 61) 0.56 (0.67) 0.25 (0.62) (0.94) 0.51 (0.98) (4.06) (8.56).0001 Note. CHP Community Health Program. a Includes emergency department visit, outpatient day surgery, and hospital observation without admission. encounters decreased by 62%, from $1,830 preenrollment to $700 postenrollment; mean cost for inpatient admissions decreased by 53%, from $13,341 to $6,324; and mean cost for primary care visits increased by 143%, from $1,036 to $2,517. When stratified by gender, men showed a greater percentage decrease in costs than women for both acute outpatient encounters (69% vs. 57%; p.03) and inpatient admissions (58% vs. 49%; p.01). When stratified by race/ethnicity as well as age, statistically significant reductions in inpatient admission costs were observed among non-hispanic Whites ( p.01), African Americans ( p.03), and patients between 50 and 65 years of age ( p.01). However, a statistically significant reduction in costs for acute outpatient encounters was observed only among African American patients ( p.04). The increase in the postenrollment costs for clinic visits did not reach statistical significance when analyzed across any of the demographic strata. The mean aggregate health care utilization cost per patient decreased by TABLE 4 Comparison of Health Care Encounters Before and After Enrollment in Community Health Program Type of Encounter Number of Encounters Pre-CHP Post-CHP Percent Change Acute outpatient encounters a Inpatient admissions Primary care clinic visits Note. CHP Community Health Program. a Includes emergency department visit, outpatient day surgery, and hospital observation without admission. 41%, from $16,208 preintervention to $9,541 postintervention ( p.004) ( see Table 6 ). When stratified by demographic characteristics, statistically significant reductions in mean aggregate costs were observed among men ( p.03), non-hispanic Whites ( p.02), and patients between 50 and 65 years of age ( p.01). D ISCUSSION In this study of uninsured patients with one or more chronic diseases, the introduction of an intensive case management program was associated with statistically significant reductions in acute health care utilization and a concomitant increase in primary care clinic visits. Overall, acute outpatient encounters decreased by 62% and inpatient admissions by 53% whereas primary care visits increased by 162%. Participation in the case management program was also associated with a 41% reduction in overall aggregate costs, from $16,208 preintervention to $9,541 postintervention ( p.004). Only a few studies have evaluated the impact of case management interventions on health care utilization and costs among patients without medical insurance. A direct comparison of our results with these earlier studies is difficult because of differences in study designs, methods, and outcome measures. Wetta-Hall (2007) analyzed the outcomes of a convenience sample of 492 uninsured patients who enrolled in a case management program that was based on a model similar to ours. Half of the sample had a history of chronic illness. At 6 months postintervention, the total number of emergency department visits among the sample had declined by 48%. On the basis Vol. 17/No. 6 Professional Case Management 271

6 TABLE 5 Comparison of Health Care Utilization Costs a Before and After Enrollment in Community Health Program Characteristics Acute Outpatient Encounters b Inpatient Admissions Clinic Visits p p p Total sample ( N 83) 1,830 (3,087) 700 (2,224) ,341 (21,400) 6,324 (12,390).002 1,036 (1,059) 2,517 (5,456).02 Gender Female ( n 50) 1,801 (3,436) 776 (2,594).10 12,737 (24,056) 6,522 (12,291).05 1,178 (1,159) 2,206 (4,592).13 Male ( n 33) 1,875 (2,518) 584 (1,534).03 14,258 (16,911) 6,024 (12,724) (858) 2,998 (6,604).06 Race/ethnicity Non-Hispanic White ( n 43) 1,453 (2,860) 941 (2,829).40 16,655 (22,158) 8,483 (15,079) (1,143) 2,277 (4,751).07 Hispanic ( n 19) 2,397 (3,813) 331 (1,441).05 11,822 (27,911) 4,616 (8,880).27 1,235 (1,007) 1,367 (981).63 African American 2,090 (2,852) 541 (1,153).04 7,931 (9,431) 3,449 (7,709).03 1,052 (940) 4,049 (8,362).12 ( n 21) Age range (years) ( n 22) 1,909 (3,662) 438 (1,084).09 20,227 (31,238) 8,064 (14,002) (1,002) 2,700 (5,049) ( n 61) 1,802 (2,886) 794 (2,513).05 10,858 (16,146) 5,697 (11,818).01 1,099 (1,080) 2,450 (5,634).07 Note. CHP Community Health Program. a Costs are given in U.S. dollars. Administrative costs of the intensive case management program were not included in cost calculations. b Includes emergency department visit, outpatient day surgery, and hospital observation without admission. of the assumption that the rate of emergency department visits would remain the same in the absence of interventions, the author not only estimated that case management resulted in a charge avoidance of nearly $1.5 million but also cautioned that the cost savings might be overestimated because the costs of primary care services were not measured directly. Like those TABLE 6 Aggregate Costs of Health Care Utilization Before and After Enrollment in Community Health Program Characteristics Aggregate Costs a p Total sample ( N 83) 16,208 (21,265) 9,541 (13,681).004 Gender Female ( n 50) 15,715 (24,061) 9,504 (13,373).05 Male ( n 33) 16,954 (16,473) 9,597 (14,344).03 Race/ethnicity Non-Hispanic White ( n 43) 19,048 (22,254) 11,700 (15,994).02 Hispanic (n 19) 15,454 (27,423) 6,314 (9,180).16 African American 11,073 (9,573) 8,039 (11,492).25 ( n 21) Age range (years) ( n 22) 22,998 (30,948) 11,203 (14,969) ( n 61) 13,759 (16,118) 8,942 (13,266).01 Note. CHP Community Health Program. a Costs are given in U.S. dollars. Post-CHP aggregate costs do not include the administrative costs of the CHP program. of our study, the results reported by Wetta-Hall were based on a convenience sample whereby each subject served as his or her own control. In addition, approximately two thirds of those eligible for the case management programs in the two studies failed to respond to repeated attempts to contact them. Horwitz et al. (2005) evaluated a case management program designed to increase access to primary care and reduce emergency department usage among uninsured patients. Using a randomized design, 230 patients who were seen at a Level 1 urban trauma center were enrolled in the study and assigned to either an intervention group or a comparison group. Case managers used mail, telephone, and home visits to contact patients in the intervention group and help link them with a primary care provider. At 60 days after enrollment, the intervention subjects were significantly more likely to have established a primary care contact than the comparison subjects (51.2% vs. 13.8%, p.0001). In contrast, at 6 months after enrollment, there was no statistically significant difference between groups in either the number of postintervention emergency department visits or hospitalizations, although the average cost of a postenrollment emergency department visit was 31% lower in the intervention group than in the comparison group. The failure of the case management interventions to reduce emergency department visits in this study may have been a result of several factors. The interventions described by Horwitz et al. appear to have focused primarily on helping patients select a primary care provider and then scheduling an appointment 272 Professional Case Management Vol. 17/No. 6

7 In this study of uninsured patients with one or more chronic diseases, the introduction of an intensive case management program was associated with statistically significant reductions in acute health care utilization and a concomitant increase in primary care clinic visits. Overall, acute outpatient encounters decreased by 62% and inpatient admissions by 53% whereas primary care visits increased by 162%. Participation in the case management program was also associated with a 41% reduction in overall aggregate costs, from $16,208 preintervention to $9,541 postintervention (p.004). with the provider. Although primary care contacts were significantly increased with these interventions, they may not have been sufficient to change longstanding emergency department utilization patterns. In contrast, our study and that of Wetta-Hall (2007) employed additional interventions that included case managers making home visits to educate patients about self-management of their chronic illnesses as well as attending clinic visits with their patients. In addition, social workers were paired with the nurse case managers to help patients access an array of social services and support networks. The length of the intervention may also play a role in reducing emergency department and other acute health care usage; the vast majority (67.5%) of the patients in our sample had participated in the case management program for 12 months or longer. When stratified by age, gender, and race/ethnicity, our analyses showed that case management was equally effective in reducing acute health care utilization and increasing primary care follow-up among both men and women as well as among young and older adults. The only exception was for non-hispanic Whites, who failed to demonstrate a significant postintervention reduction in acute outpatient encounters despite a substantial increase in primary care contacts. Horwitz et al. (2005) found no association between these demographic characteristics (i.e., age, gender, and race/ethnicity) and the likelihood of accessing primary care among case management intervention subjects. They did, however, observe a marginal relationship with language, in that non- English-speaking subjects were more likely to visit a primary care provider. In terms of costs, we found that the reductions in postintervention expenditures for both acute outpatient encounters and hospitalizations were significantly greater for men than for women. Statistically significant reductions in hospitalization costs were observed among non-hispanic Whites, African Americans, and patients between 50 and 65 years of age. In contrast, a statistically significant reduction (74% decrease) in costs for emergency department visits and other acute outpatient encounters was observed only among African American patients. Although it is likely that the demographic differences observed in our study were affected by additional variables such as socioeconomic and behavioral characteristics, our database did not permit us to perform such an analysis. Further studies with larger and more representative samples are needed to further examine the potential effects of demographic characteristics on health care usage and costs in the uninsured population. L IMITATIONS The results of this study should be interpreted in the context of several limitations. Because our study utilized a nonprobability sample (i.e., a convenience sample) rather than a probability sample (e.g., a random sample), it is possible that our results were affected by selection bias and are not generalizable to the large group of eligible patients who did not enroll in the program. That is, the subgroup of patients who chose not to participate in the intervention may have been less motivated to improve their health or may have had more complex health or socioeconomic problems ( Wetta-Hall, 2007 ). Furthermore, because our study was limited to an uninsured population from a relatively small geographic area served by a single hospital, it is unclear whether our findings can be generalized to populations from other geographic regions. In addition, our results may not generalize to those patients who were ineligible for the intervention because of co-occurring psychiatric disorders. Another limitation was the small sample size in our study. Finally, because we did not factor in the administrative costs of the intervention program itself in our calculations, it is possible that the cost savings associated with the intervention were overestimated. CONCLUSIONS Clinical case management programs have become increasingly popular as a strategy to curb excessive and expensive health care services, particularly in persons with chronic diseases and those who are uninsured or underinsured. Nonetheless, only a few Vol. 17/No. 6 Professional Case Management 273

8 studies have evaluated the effectiveness of case management interventions in meeting this goal. In spite of its limitations, our study provides additional evidence that intensive clinical case management can reduce acute care utilization and costs and increase primary care follow-up in uninsured patients with chronic diseases. We believe that the results of our study may be applicable to other safety net institutions that are struggling to cope with escalating health care costs and the growing number of uninsured persons. Additional studies with randomized designs and large sample sizes are needed to validate our findings and to identify the most effective components of case management interventions. A CKNOWLEDGMENTS We thank Mary Lou Wallin for assistance in data compilation, Leonard Pechacek for editing and writing assistance, and Jacques Baillargeon, PhD, for critical review of the manuscript. R EFERENCES Ayanian J. Z., Weissman J. S., Schneider E. C., Ginsburg J. A., & Zaslavsky A. M. ( 2000 ). Unmet health needs of uninsured adults in the United States. JAMA, 284, Begley C. E., Vojvodic R. W., Seo M., & Burau K. ( 2006 ). Emergency room use and access to primary care: Evidence from Houston, Texas. Journal of Health Care for the Poor and Underserved, 17, Collins S. R., Davis K., Doty M. M., Kriss J. L., & Holmgren A. L. ( 2006 ). Gaps in health insurance: An all-american problem (Commonwealth Fund Publication No. 920). Washington, DC : Commonwealth Fund. Retrieved January 25, 2012, from commonwealthfund.org/content/publications/fund- Reports/2006/Apr/Gaps-in-Health-Insurance-An-All- American-Problem.aspx Collins S. R., Doty M. M., Robertson R., & Garber T. ( 2011 ). Help on the horizon: How the recession has left millions of workers without health insurance, and how health reform will bring relief findings from The Commonwealth Fund Biennial Health Insurance Survey of 2010 (Commonwealth Fund Publication No. 1486). Washington, DC : Commonwealth Fund. Retrieved June 4, 2012, from Mar/Help-on-the-Horizon.aspx?page all DeNavas-Walt C., Proctor B. D., & Smith J. C. ( 2011 ). Income, poverty, and health insurance coverage in the United States: Current Population Reports (U.S. Census Bureau Publication No. P60-239). Washington, DC : US Government Printing Office. Freeman J. D., Kadiyala S., Bell J. F., & Martin D. P. ( 2008 ). The causal effect of health insurance on utilization and outcomes in adults: A systematic review of US studies. Medical Care, 46, Grossman L. K., Rich L. N., & Johnson C. ( 1998 ). Decreasing nonurgent emergency department utilization by Medicaid children. Pediatrics, 102 ( 1, Pt 1 ), Grumbach K., Keane D., & Bindman A. ( 1993 ). Primary care and public emergency department overcrowding. American Journal of Public Health, 83, Heberlein M., Brooks T., Guyer J., Artiga S., & Stephens J. ( 2011 ). Holding steady, looking ahead: Annual findings of a 50-state survey of eligibility rules, enrollment and renewal procedures, and cost sharing practices in Medicaid and CHIP, (Publication No. 8130). Washington, DC : Henry J. Kaiser Family Foundation. Retrieved June 4, 2012, from org/medicaid/8130.cfm Horwitz S. M., Busch S. H., Balestracci K. M., Ellingson K. D., & Rawlings J. ( 2005 ). Intensive intervention improves primary care follow-up for uninsured emergency department patients. Academic Emergency Medicine, 12, Hurley R. E., Freund D. A., & Taylor D. E. ( 1989 ). Emergency room use and primary care case management: Evidence from four Medicaid demonstration programs. American Journal of Public Health, 79, Institute of Medicine. ( 2000 ). America s health care safety net: Intact but endangered. Washington, DC : National Academies Press. Kaiser Commission on Medicaid and the Uninsured. ( 2011 ). Medicaid enrollment: June 2010 data snapshot (Publication No ). Retrieved June 5, 2012, from Kaiser Commission on Medicaid and the Uninsured. ( 2012 ). A mid-year state budget update for FY 2012 and a look forward to FY 2013 (Publication No. 8277). Retrieved June 5, 2012, from org/medicaid/enrollmentreports.cfm Lozano J. A. ( 2008, December 1 ). UTMB s role in indigent care uncertain/since Ike, many who depended on the medical facility have no place to go. Houston Chronicle, p B4. McWilliams J. M. ( 2009 ). Health consequences of uninsurance among adults in the United States: Recent evidence and implications. Milbank Quarterly, 87, National Association of Public Hospitals and Health Systems. ( 2010 ). Safety net health systems: An essential resource during the economic recession. Washington, DC : Author. Retrieved June 5, 2012, from Safety-Net-Financing/Recession-Brief-Aug aspx?ft.pdf Ortolon K. ( 2009 ). Rising from the storm: UTMB struggles to recover after Hurricane Ike. Texas Medicine, 105 ( 12 ), Parisi L. ( 2010 ). States in need: Congress should extend temporary increase in Medicaid funding. Washington, DC : Families USA. Retrieved June 6, 2012, from Richardson L. D., & Hwang U. ( 2001 ). America s health care safety net: Intact or unraveling? Academic Emergency Medicine, 8, Wetta-Hall R. ( 2007 ). Impact of a collaborative community case management program on a low-income uninsured 274 Professional Case Management Vol. 17/No. 6

9 population in Sedgwick County, KS. Applied Nursing Research, 20, Wilper A. P., Woolhandler S., Lasser K. E., McCormick D., Bor D. H., & Himmelstein D. U. ( 2008 ). A national study of chronic disease prevalence and access to care in uninsured U.S. adults. Annals of Internal Medicine, 149, Wilper A. P., Woolhandler S., Lasser K. E., McCormick D., Bor D. H., & Himmelstein D. U. ( 2009 ). Health insurance and mortality in US adults. American Journal of Public Health, 99, Alison Glendenning-Napoli, BSN, RN-BC, is Director of outpatient care management and oversees the Community Health Program at the University of Texas Medical Branch. She has more than 16 years of experience in program development, case management, and nursing leadership. She is board certifi ed in case management and is currently pursuing chronic care professional certifi cation. Beverly Dowling, CPA, CHFP, is Assistant Vice President of the University of Texas Medical Branch Community Health Network. She has 20 years of experience in health care with a focus on implementing new health care delivery models for uninsured and underinsured populations. John Pulvino, PA-C, is Senior Director of quality and outcomes for the Correctional Managed Care Program at the University of Texas Medical Branch. He has more than 20 years of experience in developing, implementing, and refi ning quality improvement and case management processes in the health care fi eld. Gwen Baillargeon, MS, is a biostatistician for the Correctional Managed Care Program at the University of Texas Medical Branch. Gwen has more than 13 years of experience as a biostatistician and SAS programmer in the pharmaceutical industry and the academic setting and has coauthored a number of publications in peer-reviewed medical journals. Ben G. Raimer, MD, is Professor in the Departments of Pediatrics, Family Medicine, and Preventive Medicine and Community Health at the University of Texas Medical Branch and Senior Vice President of the Offi ce of Health Policy and Legislative Affairs. He has extensive experience in developing programs to improve the delivery and quality of health care for rural and underserved populations in Texas. Vol. 17/No. 6 Professional Case Management 275