A Comprehensive Case Management Program - Free Essay Example

JOURNAL OF PALLIATIVE MEDICINE Volume 12, Number 9, 2009 ? Mary Ann Liebert, Inc. DOI: 10. 1089=jpm. 2009. 0089 Original Article A Comprehensive Case Management Program To Improve Palliative Care 1 1 Claire M. Spettell, Ph. D. , Wayne S. Rawlins, M. D. , M. B. A. ,2 Randall Krakauer, M. D. ,3 Joaquim Fernandes, M. S. , 2 2 2 Mary E. S. Breton, B. S. , J. D. , Wayne Gowdy, B. S. , Sharon Brodeur, R. N. , B. S. , M. P. A. , Maureen MacCoy, B. S. N. , M. B. A. ,2 and Troyen A. Brennan, M. D. , M. P. H. 4 Abstract Objective: The objective of this study was to evaluate the impact of comprehensive case management (CM) and expanded insurance bene? s on use of hospice and acute health care services among enrollees in a national health plan. Study Design: Retrospective cohort design with three intervention groups, each matched to a historical control group. Methods: Intervention groups were health plan enrollees who died after 2004: 3491 commercial enrollees with CM; 387 commercial enrollees with CM and expanded hospice bene? ts; and 447 Medicare enrollees with CM. Control groups consisted of enrollees who died in 2004 prior to the start of the palliative care CM program. The main outcomes measured were the proportion using hospice, mean number of hospice days, and number of inpatient days measured through medical claims. Results: Hospice use increased for all groups receiving CM compared to the respective control groups: from 30. 8% to 71. 7% ( p 0. 0001) for commercial members with CM and from 27. 9% to 69. 8% ( p 0. 0001) for Commercial members with CM and enhanced hospice bene? ts. Mean hospice days increased from 15. 9 to 28. 6 days ( p . 0001) and from 21. 4 to 36. 7 days ( p 0. 0001) for these groups, respectively. Inpatient stays were lower for all groups receiving CM services compared to their respective control groups. Conclusions: Comprehensive health plan CM and more liberal hospice bene? t design may help to break down barriers to hospice use; ben e? ts might be liberalized within the context of such case management programs without adverse impact on total costs. Introduction ospice care helps to meet the needs of patients with advanced illness by providing effective pain and symptom management and support for the emotional and spiritual needs of patients and their caregivers. Such care allows patients to achieve a sense of control over dying, many of whom would prefer to die at home. Hospice utilization among Medicare decedents increased dramatically in the last decade, to approximately 40% in 2005. 1 However, the current rate is considered less than ideal to fully meet the needs of those with advanced illness, and there is substantial variation in the use of hospice by age, race, diagnosis and geographic location. 2–5 Many individuals enter hospice shortly before death, substantially limiting the bene? t they might obtain 1 2 H from hospice services. In 2006, the median length of stay in hospice was 20. 6 days, down from 26. 0 days in 2005, and little changed from the 2001 rate of 20. 5 days. 6 Among Medicare decedents, the median length of stay was 15 days in 2005. 1 Barriers to election of hospice care include preferences for aggressive curative treatment among patients, families, and physicians, physician’s discomfort and dif? culty in initiating conversations about advanced illness choices, Medicare regulations requiring the patient’s physician to certify that the patient has a life expectancy of 6 months or less, limits on hospice bene? s, and the need to forego curative medical treatment in order to qualify for hospice. 7,8 In 2004, a national health plan launched a comprehensive case management (CM) program targeted speci? cally to patients with advanced illness and their families. The health Aetna Informatics, Aetna, Blue Bell, Pennsylvania. Aetna Government Health Plan, Aetna, Hartford, Connecticut. 3 National Care Management, Aetna, Hartford, Connecticut. 4 CVS Car emark, Woonsocket, Rhode Island. 827 828 plan also piloted a bene? t design change among 13 large employers that liberalized hospice and respite bene? s for seriously ill patients and families. The purpose of this article is to describe the impacts of the case management program and the liberalization of bene? ts on use of hospice and acute health care services in commercially insured and Medicare Advantage populations. Methods Program description A comprehensive case management program termed the ‘‘Compassionate Care Program’’ was launched at the end of 2004 and included comprehensive case management services provided by health plan nurse case managers who received extensive training in palliative care. This specialized case management program supplemented the traditional case management services available to all health plan members. Members were identi? ed as candidates for the program through the health plan’s process of concurrent review of in patient admissions, physician referral, self-referral, and monthly use of a proprietary predictive model examining medical and pharmacy claims to identify individuals whose claims history suggested a terminal illness. Case management services were available to all eligible members and few individuals declined these services. Physicians in the health plan network were noti? ed of the program at the time it was implemented via an article in the physician newsletter sent out by the health plan. Case managers reached out by telephone to identi? ed members and conducted a comprehensive assessment of their needs and developed individual plans of care that addressed the members’ needs and preferences. The number and frequency of contacts with the member was established with the member=caregiver during the initial outreach. The case manager assisted the member and family by addressing issues such as the need for education of the disease process for member and family=caregiver, understanding of advanced directives and assistance with obtaining these documents, understanding their preferences for care, identifying community resources for member and caregiver support, social work support, pain control, medication management, and home or respite care. The case manager worked with the member’s physician to coordinate care and with the hospice agency if hospice was in place. The case manager handled an average caseload of 40–45 health plan members, all in various stages of need for support. Members with advanced illness made up a small percentage of that caseload at any given time. The internal cost for a nurse case manager to manage a member with advanced illness was approximately $400. In January 2005, a pilot program was launched for 13 large employers whereby, in addition to the provision for case management support, insurance bene? ts for hospice and respite were expanded. The expansion included extending the durational de? nition of te rminal illness from 6 months to 12 months; continued receipt of curative treatment while also receiving hospice services; removal of length of stay for inpatient hospice and maximum dollar limits for outpatient hospice; provision of 15 days per year of respite bene? ts for family members; and availability of bereavement services through employer assistance programs. Study design and population SPETTELL ET AL. The study was a retrospective cohort design using matched historical control groups. Data for the analysis came from the health plan’s eligibility, claims and utilization management systems. Members who died during 2005, 2006, and the ? rst quarter of 2007 were identi? ed through the health plan case management database. These members comprised three groups: 1. Case Management (CM) Group (n ? 3491): Commercially insured members with usual hospice bene? ts who received comprehensive case management (CM) services. 2. Enhanced Bene? ts CM Group (n ? 387): Commercially in sured members whose bene? s were provided by one of the 13 large employers participating in the pilot program for which hospice and respite bene? ts were liberalized. These members also received the comprehensive CM services. 3. Medicare CM Group (n ? 447): Medicare Advantage members with Centers for Medicare Medicaid Services (CMS)-de? ned hospice bene? ts who received comprehensive CM services. Control groups Historical control groups were created for each of the groups above. Health plan members who died in 2004 were identi? ed from the Social Security Death Index ? es by matching on Social Security Number and two of the following: date of birth, gender and full name. 9 Control group members had been eligible for the health plan’s usual case management services in place prior to the specialzed training program in palliative care. Each member receiving CM was matched to a control group member on age, severity of illness score, presence of health plan pharmacy bene? ts, and diagnosis using information available in the health plan’s claims and eligibility systems. Severity of illness of each member was quanti? d using the Ingenix Episode Risk GroupO (ERGO) Score software. 10 This score was derived from weights assigned from a normative insurance claims database for each diagnosis group found in medical episodes constructed from medical and pharmacy claims data. Study period The date of enrollment in the CM program was determined for each member and the number of days between this index date and the person’s death was calculated. The number of days prior to death was used as the observation period for each matched pair. Primary outcome measures The primary outcome measures were rates of hospice use and mean number of days in hospice, which were expected to be higher in the groups receiving case management and expanded hospice bene? ts compared to the control groups. Hospice use measures were calculated from health plan claims data for t he commercial members and included the proportion of members using hospice in both inpatient and outpatient settings and the length of service in hospice. For the Medicare CM Group for whom hospice claims were paid directly by CMS, hospice use was calculated based on an CASE MANAGEMENT TO IMPROVE PALLIATIVE CARE indicator ? ag on the CMS Monthly Member Eligibility Files. The number of days in hospice was not available from this source. The ? ag indicating hospice in the health plan utilization management system was not available for the Medicare control group, thus, the hospice use rate was not calculated for this group. Secondary outcome measures The acute care utilization measures were calculated from health plan claims data, and included the proportion of members with acute care hospital admissions, the rate of acute hospital inpatient days per 1000 members, proportion of members with an intensive care unit (ICU) stay during an acute hospitalization, proportion of members w ith emergency visits, the rate of emergency department visits per 1000 members, and rate of primary care and specialist vists per member. No directional hypotheses were made for these measures. Measures expressed as days per 1000 members were calculated as the number of days divided by the number of members in the CM Group multiplied by 1000. Statistical analysis Generalized linear models were used to compare outcome variables between groups with a subject effect variable to adjust for the paired nature of the data. McNemar’s test was used for comparing proportions. A generalized linear model assuming a two parameter Poisson probability distribution was employed for comparing rates represented as counts per thousand. The two-parameter Poisson was chosen for the response probability distribution so that the scale parameter 829 could model the overdispersion in the data. Kaplan-Meier methods were used to estimate the number of days between hospice enrollment and death, an d group differences were tested using a two-sided log rank test. All models included a variable for the geographical region where the member resided to adjust for regional differences in hospice use. Results of statistical tests yielding p values 0. 5 were considered statistically signi? cant. All analyses were done using SAS v. 9. 0 (SAS Institute, Cary, NC). Results Table 1 shows sociodemographic characteristics of each CM group compared to its control group. There were no statistically signi? cant differences on the variables used in the matching process. Table 2 lists the top 15 diagnoses for each group. Within each cohort, the CM and Control groups varied in the geographic distribution of members; therefore, geographic region was used as an adjustor in the analyses of outcomes. Table 3 presents the use of health care services by the Enhanced Bene? ts CM Group, the CM Group and the Medicare CM Groups compared to their respective control groups, adjusted for differences in ge ographic region. The average number of days in the CM program was 42. 3 days (Enhanced Bene? ts CM Group), 39. 6 days (CM), and 56. 7 days (Medicare CM). For each group receiving CM, the percentage of members using hospice more than doubled compared to its control group (Enhanced Bene? ts CM 69. 8% versus 27. 9%, p 0. 0001; CM 71. 7% versus 30. %, p 0. 0001). The mean number of days with hospice increased from 21. 4 days to 36. 7 days ( p 0. 0001) for the Enhanced Bene? ts CM group, and from Table 1. Characteristics of Case Management (CM) Groups Enhanced Bene? ts CM Study group 387 59. 47 18. 19 18. 1% 74. 4% 61. 5% 96. 6% 9. 8% 20. 9% 4. 1% 9. 8% 39. 3% 8. 3% 7. 8% Control group 387 59. 04 17. 76 18. 1% 74. 4% 55. 8% 98. 2% 10. 3% 22. 0% 9. 3% 9. 8% 19. 4% 8. 8% 20. 4% Study group 3491 56. 52 19. 79 62. 4% 80. 7% 49. 7% 65. 1% 20. 3% 16. 4% 12. 7% 24. 7% 10. 3% 9. 8% 5. % CM Control group 3491 56. 87 19. 65 62. 4% 80. 7% 48. 1% 74. 9% 14. 9% 16. 6% 14. 0% 14. 4% 12. 1% 10. 0% 1 7. 9% Study group 447 77. 14 24. 83 100% 57. 5% 44. 5% 0% 47. 9% . 2% 48. 5% . 2% 0% 3. 1% 0% Medicare CM Control group 447 77. 36 24. 17 100% 57. 5% 44. 5% 0% 43. 0% 0% 34. 7% 0% 0% 22. 4% 0% n Matching variables Mean age Comorbidity risk scorea Health plan pharmacy Bene? t % with cancer as terminal condition Descriptive variables % Female % PPO Health plan geographic Region Mid-Atlantic North Central Northeast Southeast Southwest West Unknown a value 0. 45 0. 5582 1. 00 1. 00 0. 1086 p value 0. 1266 0. 5824 1. 00 1. 00 0. 1880 p value 0. 6588 0. 4181 1. 00 1. 00 1. 00 Episode Risk GroupO Score. PPO, preferred provider organization. 830 Table 2. Top Fifteen Conditions by Case Management Group Enhanced case management Lung cancer Gastrointestinal cancer Colorectal cancer Neoplasms—other Brain cancer Breast cancer Gynecologic cancer Neurologic disorders Hodgkin’s lymphoma COPD Hepatobiliary disorders Head and neck cancer Heart failure Malignant melanoma Sepsis 15. % 10. 6% 9. 0% 7. 2% 6. 2% 6. 2% 5. 2% 3. 9% 3. 1% 2. 6% 1. 8% 1. 6% 1. 3% 1. 3% 1. 0% Commercial case management Lung cancer Gastrointestinal cancer Breast cancer Neoplasms—other Colorectal cancer Gynecologic cancer Brain cancer Hodgkin’s lymphoma Hematologic cancer Hepatobiliary disorders Head and neck cancer Prostate cancer COPD Respiratory failure Malignant melanoma 20. 1% 12. 7% 9. 2% 7. 9% 7. 5% 5. 0% 3. 8% 2. 2% 2. 1% 1. 8% 1. 5% 1. 5% 1. 4% 1. 3% 1. 2% SPETTELL ET AL. Medicare case management Lung cancer Gastrointestinal cancer Congestive heart failure Neoplasms—Other COPD Colorectal cancer Breast cancer Prostate cancer Chronic renal failure Diabetes mellitus Respiratory failure Cerebrovascular disease Hematologic cancer Pneumonia Hypertension 19. 5% 9. 6% 6. 7% 6. 5% 6. 0% 4. 9% 3. 4% 3. 1% 2. 9% 2. 9% 2. 9% 2. 2% 2. 2% 1. 6% 1. 6% 15. 9 days to 28. 6 days ( p 0. 0001) for the CM group. The rate of use of hospice in the Medicare CM Group was 62. 9%. The percentages of members with an acute inpatient stay after program enrollment were reduced for the Enhanced Bene? ts CM Group (16. % versus 40. 3%, p 0. 0001), CM group (22. 7% versus 42. 9%, p 0. 0001), and Medicare CM group (30. 0% versus 88. 4%, p 0. 0001) compared to their respective control groups. The number of acute inpatient days was reduced for the Enhanced Bene? ts CM group (1549 versus 3986 days per thousand members, p 0. 0001), CM Group (2311 versus 3858 days per thousand members, p 0. 0001), and Medicare CM Group (2309 versus 15,217 per thousand members, p 0. 0001) compared to their respective control groups. The proportion of members with ICU stays during an acute inpatient admission was signi? antly lower for all of the groups receiving CM compared to their respective control groups, as was ICU days per thousand member (Enhanced Bene? ts CM Group 899 versus 2542, p 0. 0001, CM Group 1356 versus 2162, p 0. 0001, Medicare CM Group; 1189 versus 9840, p 0. 0001) c ompared to the control groups. Table 3. Adjusted Utilization of Health Care Servicesa Enhanced Bene? ts CMb Pilot Group Study group Average days in 42. 3 CM program Percent Using 69. 8% Hospice Mean days from hospice 36. 7 claim and death Hospice inpatient 1,424. days=1000 Hospice outpatient 14,607. 0 days=1000 Percent with acute 16. 8% inpatient stay Average Length of 5. 84 Stay Inpatient Percent With Emergency Visit 9. 8% Percent With ICU Stay 9. 6% Acute inpatient days=1000 1,549. 4 Emergency visits=1000 94. 4 ICU days=1000 898. 8 Primary care physician 0. 53 visits per Member Specialist visits per Member 1. 44 a c b CM Group Study group 39. 6 Control group p value Medicare CM Group Study group 56. 7 Control group p value Control group p value 27. 9% 21. 4 601. 2 3,914. 5 40. 3% 6. 91 15. 2% 23. 0% 3,986. 4 159. 3 2,541. 6 1. 00 2. 09