190 stroke patients were randomized to the intervention group and 190 were randomized to the control group. N=380

Transcription

1 Appendix 1: Summary of Evidence on Transitional Care (n=21) 1 Allen, et al., 2009 To evaluate whether comprehensive care management after discharge for stroke survivors is superior to stroke unit care with enhanced discharge planning Single site RCT randomized patients being discharged to home from an acute stroke care unit in a large community teaching hospital. Patients discharged to a rehab facility for < 8 weeks were randomized prior to facility discharge. Randomization in blocks of 10. All patients on the stroke unit received interdisciplinary care and an individualized care plan. Study participants received baseline assessments prior to discharge. For both the intervention and control groups, the PCPs received a written summary of all inpatient findings, the patient s risk factor profile, discharge plans, discharge medications, and baseline assessment findings obtained for the research study. The control group received mailings every two months to remind them about the study and provide stroke related educational materials. The intervention group received care management 190 stroke patients were randomized to the intervention group and 190 were randomized to the control group. N=380 Intervention: 48% male, average age 68, 17% African-American, 47% married Control: 52% male, average age 69, 15% African-American, 46% married Inclusion: diagnosis of ischemic stroke, NIH Stroke Scale 1, discharged to home from the acute care hospital or discharged to home within 8 weeks from a SNF or acute rehab facility, live within 25 miles of hospital, have no other illnesses that would dominate poststroke care, speak English, and do not have an endarterectomy planned.

2 from an APN. One week after discharge, the intervention group received an in-home assessment. Education and interventions for common post-stroke complications were provided. Assessment findings were shared with an interdisciplinary team and care plans were developed based on the findings. The PCP received a copy of the care plans and written academic detailing specific to the patient. The APN worked with the PCP for 6-months to implement recommendations. Home visits by PT and other services were provided as needed. Patients were contacted by phone a minimum of once/week for the first month and once/month for the remainder of the 6 month intervention. Neuromotor function: NIH Stroke scale, Timed Up and Go test & Physical Performance Test Institution Time and Death: days spent hospitalized (i.e., rehospitalized) or in a nursing home during the 6-month follow-up period, death Quality of Life: Stroke- Specific Quality of Life scale Management of Risk: management of systolic blood pressure (<140 mmhg), diastolic blood pressure (<90 mmhg), depression (CED-D), medication appropriateness (investigatordeveloped tool), hemoglobin A1c (<6.5%), total cholesterol (<180), and self-reported falls and incontinence Stroke Knowledge and Lifestyle Modification: investigator-generated questionnaire Measurement intervals: Baseline- depression, NIH Stroke Scale, blood pressure, and Stroke Knowledge Test 6-months)- all outcomes measures 2

3 There was no treatment effect for neuromotor The authors cite one reason for a In this study, very function, institution time or death, or quality of lack of treatment effect may be little difference in life. The global test of the management domain that the sample was relatively high outcomes was found (management of risk and stroke knowledge/lifestyle modification) was significantly better for the intervention group (p=0.002). However, in testing management of risk and knowledge/lifestyle functioning at baseline (average NIH Stroke Scale was 1.7 for the intervention group and 2 for the control group, systolic blood between patients who received comprehensive interdisciplinary postdischarge management modification separately, only stroke pressure and cholesterol were close after care on an knowledge/lifestyle modification remained to target, depression, falls and organized stroke unit significant (p=0.62 for management of risk; p= incontinence were good at compared to those who for knowledge/lifestyle). baseline). The care provided on the received enhanced interdisciplinary stroke unit may discharge planning at In exploratory subgroup analysis, only those with a have effectively met discharge discharge from the prior event (stroke, transient ischemic attack, atrial fibrillation) benefited more from the needs. organized stroke unit. The only difference was intervention on the neuromotor domain (effect size 0.54 standard deviation in favor of prior stroke The neuromotor domain had quite a better Stroke Knowledge bit of missing data at six-months and Lifestyle patients, p=0.02). However, this was not a formal (19% usual care and 13% Modification among those test and no adjustment was made for multiple testing intervention group). However, the with comprehensive postdischarge when generating p values. other domains had much lower missing data (0% to 8%). The authors do not believe that the management. The authors conclude that enhanced discharge missing data greatly influenced the planning, including results. communication of PCPs regarding inpatient findings, may be sufficient to optimize post-stroke outcomes at six-months. However, more research is needed to determine if comprehensive postdischarge management may be better suited for patient with greater baseline need or those discharged from facilities without a stroke unit. 2 Brown, et al., 2006 To evaluate the effect of asthma education after Single site RCT with stratified randomization (children and adults randomized separately). A convenience sample of patients presenting on a range of times and days of the week was used. The total sample included 51 adults Primary outcome: first asthma relapse (either an asthma-related visit to the ED or an unscheduled 3

4 an ED visit Patients randomized to the comprehensive asthma education intervention received telephone contact from an asthma nurse-educator 3 to 5 days after the ED visit. The asthma nurseeducator helped arrange a follow-up visit with the PCP with a goal of completing the PCP visit within 3 weeks of the ED visit. The asthma nurse attended the PCP visit, and worked with the PCP to establish treatments based on current guidelines. An asthma management plan was developed and the patient was educated about further asthma exacerbations, including medication use, contact with PCP, and appropriate use of the ED. Six-weeks after the ED visit, the asthma nurse conducted a home visit to review current medications and inhaler, spacer, and PEF meter techniques. The asthma management plan was also reviewed with the patient. Basic asthma education was also provided to the patient. randomized to the intervention and 59 randomized to the control. This study also included children, with 66 randomized to the intervention and 63 randomized to the control. Sample: 46% adults, 46& male, 30% African-American, 56% had moderate to severe persistent asthma Inclusion: moderate to severe persistent asthma or had used ED services for asthma care at least once in the past year. urgent visit to a physician office during the 6 month follow-up period Secondary outcomes: total number of ED visits and hospitalizations for asthma during the 6-month follow-up period, selfreported compliance with spacer and PEF meter, use of an asthma management plan, self-reported actions to reduce exposure to asthma triggers, and missed days of school or work. measurement was done at baseline and 6-months after ED visit Many patients were non-compliant with the intervention. The study was powered to detect a large treatment effect with adults and Primary outcome: There was no difference between the intervention and control groups in time to asthma relapse. 23.1% of the intervention group and 31.3% of the A comprehensive asthma education intervention after ED discharge was not effective in delaying 4

5 control group had an urgent asthma visit during the follow-up period (Hazard Ratio, 0.79 (95% CI ), p=0.34). When the sample was stratified by child and adult, a stronger, yet not statistically significant, trend of benefit was seen in the pediatric group. Children: 22.7% in intervention and 38.1% in control had an urgent asthma visit in the past 6- months (Hazard Ratio, 0.62 (95% CI ), p=0.29). Adults: 23.5% in the intervention and 23.7% in the control had an urgent asthma visit in the past 6-months (Hazard Ratio, 1.08 (95% CI , p=0.85). children combined, not as separate subgroups. Because the survival curves for relapse begin to split at 4 months, a follow-up beyond 6 months may be necessary to detect an effect of the intervention. Because of the trend toward and effect in children, but not adults was seen, a different intervention approach may be needed for adults. another urgent asthma visit in adults or children, although there was more of a trend toward benefit among children who received the intervention. Treatment compliance: 39% of patients in the intervention group did not comply with any of the intervention. However, there was no difference in the per protocol analysis that excluded non-compliant patients (Hazard Ratio, 0.75 (95% CI , p=0.32). Secondary outcomes: The only statistically significant secondary outcome was self-reported actions taken to reduce exposure to asthma triggers (65.8% intervention group vs. 48.6% control group, p=0.02). There were no differences between intervention and control groups for total number of urgent asthma visits, number of asthma hospitalizations, use of asthma management plan, self-reported compliance with PEF meter or space, and missed days of work or school. 3 Carroll et Cardiac rehabilitation al., 2007 program participation To determine if a communitybased peeradvisor and APN intervention for unpartnered MI and CABS patients increased Multisite RCT with consenting participants randomized during hospitalization at one of 5 academic medical centers. The intervention began in the first 48 hours 184 subjects were enrolled at the east coast sites and 63 were enrolled at the west coast sites (total=247). They were randomized to 126 in the usual care group and 121 in the intervention group. Sample: avg. age 76.3, 66% female, 8% minority Cardiac-related rehospitalization Measurement points: 6 weeks, 3 months, 6 months, 12 months 5

6 participation in cardiac rehabilitation programs and reduced rehospitalizati on after discharge and lasted 12 weeks. The APN made one home visit and telephoned the intervention participants at least 3 times during the intervention period. Peer advisors telephoned the participants weekly for the 12-week period. Control group: 69% female, average age 76.2, 7% minority (NOS) Intervention group: 63% female, average age 76.4, 9% minority (NOS) Inclusion: Adults who had a diagnosis of MI or CABS, were older than 65 years, were unpartnered (single, widowed, divorced), were able to read and speak English, and had access to a telephone. Peer advisors were older than 60 years, had a history of MI and/or CABS, had successfully completed a cardiac rehabilitation program, and had a healthy lifestyle. APNs recruited and trained the peer advisors in a program outlined in a previous publication. The key strategies included verbal encouragement and support, active listening, peer advisors sharing their experiences, reinterpretation of symptoms, exercise promotion, energy management, and teaching about cardiac disease. Cardiac rehabilitation program participation Those in the intervention group had significantly higher odds of being in a cardiac rehabilitation program at 3, 6, and 12 months post index hospitalization in comparison to usual care group (3 months: OR 2.79 ( ), p<0.005; 6 months: OR 1.82 ( ), p<0.05; 12 months: OR 2.0 ( ), p<0.05). There was no difference between the The participants in this study agreed to the possibility of randomization into a group receiving social support. The intervention may not be successful in those who do not wish to have social support in the postdischarge period. The study also A collaborative peer advisor/apn postdischarge intervention appears to be successful in promoting active participation in cardiac rehabilitation program, although may not impact 6

7 two groups at 6 weeks post index hospitalization. Cardiac-related rehospitalization Overall, there was a low incidence of rehospitalizations (5.6%). There were fewer rehospitalizations in the intervention group compared to the usual care group between 3 and 6 months (p=0.036) Strong differences observed in the incidence of rehospitalizations between MI and CABS patient categories (higher for MI than CABS patients). included only those who are unpartnered and may not be generalizable to those who live with a spouse or partner. Although the timing of the peer support contacts for the intervention participants was similar, the content of the contacts was individualized and may have varied between participants. Almost 40% (38.8%; 158/[494-87]) of eligible subjects declined to participate. Resource use was self report only and may have been underreported. rehospitalization. 4 Coleman, et al., 2006 To evaluate a care transition model designed to encourage patients and their caregivers to assert a more active role during the transition period Single site RCT conducted in collaboration with a large not-for-profit capitated delivery system that cares for patients 65 years and older in Colorado. The delivery system contracts with one hospital, 8 skilled nursing facilities, and a home health agency. Eligible patients were approached at admission for participation. The coaching intervention was developed based on 4 pillars medication self-management, patient-centered record, follow-up, and red flags. The 4 pillars were operationalized into a personal health record and a series of visits and telephone 750 Medicare Advantage enrollees were randomized 379 to the intervention group and 371 to the control group. Sample: Intervention: average age 76.0, 48.3% female, 58.1% married, 31.0% live alone Control: average age 76.4, 52.3% female, 53.8% married, 30.8% live alone Inclusion: 65 years or older, admitted to the delivery system s contracting hospitals for a nonpsychiatric condition, community-dwelling, reside in a pre-defined geographic area to facilitate home visits, have a working telephone, be English speaking, show no documentation of dementia in the medical record, have no plans to enter hospice, not be participating in another research protocol, and have 1 of 11 diagnoses. Exclusion: Not passing a cognitive Rate of nonelective rehospitalization (including contracted hospital and any noncontracted hospitals) Rate of rehospitalization for the for the same condition as the index hospitalization Nonelective hospitalization costs Measurement points: 30, 90, and 180 days after index hospitalization. 7

8 calls with a transition coach (an APN). The coach met with the patient in the hospital prior to discharge to establish rapport and arrange a home visit within 48 to 72 hours after discharge. For those transferred to a SNF, the coach called at least weekly to maintain continuity, facilitate preparation for discharge, and arrange a visit once home. At the home visit, the coach worked with the patient and caregiver on medication reconciliation and education, teaching skills for effective communication with healthcare providers, and red flags that indicate a condition is worsening. The coach telephoned the patient three times during the 28-day posthospitalization period. Intervention participants had significantly lower rehospitalization within: 30-days 8.3% vs. 11.9% (adjusted p=0.048, OR 0.59, 95% CI 0.35,1.00) o unadjusted p= days 16.7% vs. 22.5% (adjusted p=0.04, OR 0.64, 95% CI 0.42, 0.99) o unadjusted p=0.05 function test, although those who did not pass could participate if they had a willing proxy. The study included older adults with one of 11 diagnoses. The findings may not be generalizable to those with complex medical conditions who are not older or those who are older, but do not have complex medical conditions. In addition, the study was not designed to determine if the The transition coach intervention appears to be successful at decreasing nonelective hospitalizations within 30 to 90 days after discharge and decreasing rehospitalization for the same reason as the index 8

9 Intervention participants had significantly lower rehospitalization for the same diagnosis as the index hospitalization within: 90-days 5.3% vs. 9.8% (adjusted p=0.04, OR 0.50, 95% CI 0.26, 0.96) o unadjusted p= days 8.6% vs. 13.9% (adjusted p=0.046, OR 0.55, 95% CI 0.30, 0.99) o unadjusted p=0.045 intervention was more successful for any one of the 11 diagnoses. hospitalization within 90 to 180 days after discharge. Intervention participants had lower nonelective hospital costs at: 90-days $1,519 vs. $2,016 mean USD (log transformed, p=0.02) 180-days $2,058 vs. $2,546 mean USD (log transformed, p=0.049) 5 Currier et al., 2010 To determine the impact of a mobile crisis team (MCT) intervention on linking psychiatric patients to outpatient mental health services after ED discharge Single site RCT with randomization at the patient-level MCT conducted communitybased clinical assessment within 48 hours of d/c at a location of the subject s choice (vs. usual care consisted of referral to outpatient mental health clinic [OPC]); both MCT and OPC appointments consisted of a review of presenting problems, a reexamination of psychiatric symptoms and attitudes toward treatments, and an assessment of need for further mental health, medical, or chemical dependency services or other forms of 120 adult patients (> 18 yoa; mean=32.7 yoa) who presented voluntarily or were brought by police ( mental hygiene arrest ) to an urban regional Comprehensive Psychiatry Emergency Program (CPEP), a component of the ED, for evaluation M-F, 8 AM 8 PM and discharged; self-rated presence of suicidal thoughts, plans, or behaviors was necessary for participation, as indicated for English- and Spanish speaking patients; people with active substance abuse disorders, personality disorders, psychosis, and a variety of other disorders were included Exclusions: < 18 yoa, not rated as suicidal, were deemed incapable of providing informed consent, already participated in outpatient mental health treatment; previously diagnosed or suspected mental retardation or dementia rate of linkage into outpatient care (i.e., completion of the appointment with either MCT or outpatient mental health clinic [OPC] scheduled before ED discharge); cumulative amount of outpatient mental health clinical contact in the 6 months post ED discharge; changes in depression rating scores and functional assessments over 2-week and 3-month intervals subsequent to enrollment 9

10 continuing care Interviews by research staff immediately before d/c from ED, and subsequently in the community or in an office adjacent to the ED at 2 weeks (±1 week) following discharge and again at 3 months (±2 weeks) post-d/c Study sample was heterogeneous in terms of diagnosis and personal/social characteristics and represented a subset of patients who presented for ED services with variable degrees of suicidal ideas, intent, or behavior, but who were all evaluated as sufficiently stable and safe for discharge; methodology differed from other ED based studies making it difficult to make comparisons; sample was relatively small and suicide is a rare event limiting generalizability; given the relatively high number of dropouts from care (i.e., 22% at 2 weeks; 37% at 3 months), the intention-totreat method may bias results (although the authors reanalyzed their data according to actual group assignment and found no significant differences) Patients in the intervention group: Higher rate of successful first contact with 69.6% linkage to care vs. 29.6% in the control group (p<0.001) No significant difference in the mean # of visits among patients who subsequently presented to the ED (1.62 vs visits) or mean number of outpatient mental health contacts in the first 6 months after enrollment (5.45 vs. 4.61) While patients in both groups demonstrated improvement at 2 weeks and 3 months post-d/c, there were no statistically significant differences in symptom or functional outcome measures between treatment and control groups No significant differences in clinical outcomes (depression scores, functional outcomes) between participants who did attend their first prescribed appointment via either MCT or OPC versus those who did not Many patients maintained clinically significant levels of dysfunction and continued to rely on ED services at a similar rate in the 6 months after study enrollment While MCT was an effective method of contacting suicidal patients who were discharged from the ED, initial post-d/c contact in the community versus in the clinic did not prove more effective at enhancing symptomatic or functional outcomes, nor did result in more successful linkage with outpatient psychiatric care 6 Daly, et Rehospitalization rate al To evaluate the effectiveness and cost- Single site RCT conducted at a large tertiary care hospital. 231 patients were randomized to the intervention and 103 were randomized to usual care. Time-to-first 10

11 savings of a disease management program in chronically critically ill (CCI) patients Participants were randomized to the disease management program or usual care. After 18 months of the trial, the block randomization scheme was changed from 1:1 to 1:4 control/intervention ratio for the remainder of the 26 month study period. Patients in the disease management program received case management from an APN who had access to a pulmonologist and a geriatrician. The APN met with the patient and caregiver several days prior to discharge to review the hospital course, perform baseline assessments of the patient and caregiver, discuss discharge plans, and establish a plan of care. A post-discharge care plan, patient goals, presence of advanced directives, and family coping assessment was sent to the patient s out-ofhospital health care providers. Sample: average age 62.9, mean APACHE III 75.9, mean Glasgow score 10.5, mean comorbidities 5.7, mean number of pre-admission medications 5.1 Inclusion: patients who required mechanical ventilation for more than 72 hours, understanding of English language, absence of ventilator dependency prior to index hospitalization, and hospital discharge location <80 miles of the hospital (patients >30 miles from the hospital received phone calls instead of in-person visits) Exclusion: hospice patients and patients who received organ transplants and were under the care of transplant case management hospitalization Duration of rehospitalization Mortality during rehospitalization Associated costs Measurement points: Two-months after discharge from index hospitalization The patient was visited by the APN within 48 hours of discharge and a second time later that same week. Visits could occur at home or an 11

12 extended care facility. The patient was visited weekly for the next 3 weeks and at least every other week for the following 4 weeks. The intervention occurred for a minimum of 8 visits. The patient was also visited whenever a transition in care setting occurred. During the 8-week intervention the APN performed case management activities such as team meetings, coordinating provider services, counseling the family, arranging for follow-up with specialists, and monitoring the patient s condition. Days of rehospitalization: Intervention patients had significantly fewer days of rehospitalization (mean 11.4 days; 95% CI ) compared to usual care (mean 16.7 days; 95% CI ), p=0.03 No significant difference between the intervention and usual care groups during the 2-month period for: Hospital readmission at least once Percentage of patients with more than one rehospitalization Death with readmission Readmission within the first 48 hours after index discharge Days to first rehospitalization Cost-effectiveness Limited to those who had been mechanically ventilated for at least 72 hours (chronically critically ill) Costs used in the costeffectiveness analysis may not be generalizable to other settings. Authors also state that many of those who refused to participate in the study were caregivers who felt overwhelmed by the patients illness. The effect of the intervention may be an underestimate because those who refused to participate may have been those who could have benefited most APN disease management post-discharge for the chronically critically ill may be helpful in reducing the length of rehospitalization and decrease associated costs, but further development is needed to determine if other outcomes can be impacted. 12

13 Using a reduction in days of rehospitalization of 6.31 days for the intervention, a cost-savings of $5,180 per patient was calculated from the intervention. Follow-up period was only 2 months 7 Dixon, et al., 2009 To assess the effectiveness of a Brief Critical Time Intervention (B-CTI) on continuity of psychiatric outpatient care for those with serious mental illness Multisite study conducted in a VA health network with four inpatient psychiatric units. Consenting veterans who were being discharged from the acute inpatient psychiatric unit were randomized to B-CTI or usual care. 3-month intervention begins before discharge when the B-CTI nurse or social worker meets with the patient to establish rapport and establish a case management plan. There are nine areas for clinical focus including system coordination, medication adherence, integration of medical care, and establishment of clinical linkages. 64 participants were randomized to the B-CTI group and 71 were randomized to the usual care group. Total sample: mean age 47.77, 90% male, 43% Caucasian, 37% diagnosis of schizophrenia Inclusion: years of age, live within 50 miles of the inpatient facility, diagnosis of schizophrenia, major depression, bipolar disorder, or psychotic disorder not otherwise specified Exclusion: homeless or currently receiving or eligible for services from the mental health intensive case management service (four or more psychiatric inpatient hospitalizations within the past year) VA performance measures: Continuity from inpatient care to outpatient care Continuity of outpatient care Outpatient treatment intensity Acute care service use: inpatient hospital days, partial hospital days, and emergency room visits over the six-month postdischarge period Satisfaction and help received Quality of Life Measurement time points: 30 and 180 days after discharge After discharge, the C- BTI clinician conducts home visits, accompanies patient to initial appointments, and provides support to the patient and family. The intervention group had significantly better outcomes compared to the control group for: Number of days until earliest visit after discharge (3.5 vs. 15.0, p<0.001) VA setting with almost all male veteran population Limited to those with serious mental illness The B-CTI appears to be successful in increasing continuity of care in a veteran population transitioning from 13

14 Any psych or substance abuse outpatient visit during the 30 days after discharge (97% vs. 69%, p<0.001) Any psych or substance abuse outpatient visit during the 180 days after discharge (100% vs. 86%, p<0.001) Number of psych or substance abuse outpatient visits in the first 30 days after discharge (6.67 ±3.64 vs ±2.59, p<0.001) Number of psych or substance abuse outpatient visits in the first 180 days after discharge (20.80 ±14.49 vs ±12.36, p<0.001) Number of 60-day periods with 2 or more services during the first 180 days after discharge (2.23 ±0.79 vs ±1.19, p<0.001) Unlike other interventions, there was no pre-specified home visit and patient contact schedule inpatient to outpatient psychiatric services. The intervention did not have an effect on the use of acute care services and had mixed effects on quality of life and help received with postdischarge needs. There was no difference between the two groups for: Relative odds of no service Relative rate of hospital days Relative rate of ED visits There was no difference between the two groups in overall satisfaction with mental health services. Intervention participants reported receiving significantly more help with some, but not all, aspects of post-discharge care needs Quality of life was significantly better for intervention patients in legal and safety issues and greater frequency of social contacts. There was no difference in quality of life measures for living situation, daily activities and functioning, family relations, finances, work and school, and health. 8 Easton et Patient and caregiver al., 1995 coping strategies To study the impact on coping strategies of a nurse-managed follow up program Single site RCT with randomization at the patient-level APRN-managed post-d/c follow up program deemed psych-educational support for 100 patients discharged from a hospital s inpatient rehabilitation unit and who would not be confined to their home (NOTE: Authors indicate that other outcomes complication rates, rehospitalizations, level of 14

15 rehabilitation patients; 3 meetings with the APRN and contact hours after d/c to discuss transition issues; the nature of the services provided by the APRN are not detailed by these authors (NOTE: The intervention is not sufficiently described to detail its components.) At the time of d/c, significant differences were found between the mean total, optimistic, and selfreliant coping styles (experimental group scores were higher); at 4-months, significant differences were found between the mean evasive, fatalistic, emotive, palliative, and supportant coping styles (control group scores were higher) satisfaction, anxiety, coping ability, number of post-d/c phone calls to the nursing unit, amount of nurse/social worker time used for patient f/u are reported elsewhere.) Subject homogeneity and the single site study limit the generalizability of these findings; without knowing more details regarding the intervention the impact of these findings is questionable Patients receiving follow-up from the APRN reported higher use of the more positive coping strategies 9 Fitzgerald, 333 in the intervention group and Frequency of VA and nonet al., 335 in the control group VA office visits 1994 To determine if a case manager intervention decreases readmission Single site RCT conducted in a VA facility with patients being discharged from the medical center to the general medicine outpatient clinic. Intervention: mean age 64.4, 82% white Control: mean age 64.6, 82% white ED visits, hospital admissions and days Nursing home days Within 3 days of discharge, the physician would alert the nurse case manager. The nurse case manager would complete baseline assessments At discharge patients were randomized to intervention or control within age Inclusion: age 45 or older, receive primary care services in the VA outpatient clinics, and live in the primary service area of the hospital Exclusion: patients believed to have < 60 days to live Mortality Measurement time point: 12 months 15

16 strata Intervention: Day after discharge, case manager mailed information packet to patients 5 th days after discharge, case manager telephoned the patient to discuss unmet needs, discuss warning signs, review follow-up appointments, and elicit new needs. Case manager met with patient at each physician office visit For hospital readmission, case manager evaluated patient and assisted with discharge planning For ED visits, case manager called patient the day after to provide counseling and reassess needs If no contact with case manager for 30 days, case manager called patient Closeout: Patients were eligible for closeout after 12 months. If no closeout occurred, an appointment was made at 15 months. 16

17 Intervention patients had better outcomes compared to control patients for: General medicine clinic kept visits (mean 0.30±0.23 vs. 0.26±0.22, p=0.02) Other clinic no-show visits (mean 0.24±0.30 vs. 0.31±0.41, p=0.02) Intervention patients had worse outcomes compared to control patients for: Number of nursing home days per patient per month (0.64±3.42 vs. 0.22±1.27, p=0.04) There was no difference between the two groups for: Outpatient utilization: General medicine no show visits, other clinics kept visits, total kept visits, ED visits, non-va visits, total unmet service needs, total service needs being provided Hospital readmission and nursing home placement: VA readmission and days of readmission per month, non- VA hospital readmission and days of readmission per month, total VA and non-va readmissions and days per patient per month, number of nursing home admissions per month, and mortality VA population Included all types of medical diagnoses not just those at high risk This post-discharge case management intervention was not effective in improving outpatient utilization or decreasing hospital readmission. 10 Jack, et 373 intervention group and 376 al., 2009 usual care group To test the effects of an intervention designed to minimize hospital utilization after discharge Single site RCT at a large, urban, safety-net hospital Block randomization with varying block sizes of 6 or 8, 2 participants enrolled each day Control group: Usual care Intervention: Nurse discharge advocate created an afterhospital care plan that contained medical provider contact Usual care: 47% male, mean age 49.6, 27% white non-hispanic, 52% black non-hispanic, 17% private insurance, 49% Medicaid, 13% Medicare Intervention: 52% male, mean age 50.1, 28% white non-hispanic, 51% black non-hispanic, 16% private insurance, 47% Medicaid, 14% Medicare Inclusion: have a telephone, be able to understand English, be discharged to a U.S. community Rate of resource use per patient within 30 days of discharge from index hospitalization ED visits Rehospitalization PCP visits Self-reported preparedness for discharge at 30 days of discharge from index hospitalization Knowledge of the discharge diagnosis at 30 days of discharge from index hospitalization 17

18 information, dates for appointments, an appointment calendar, color-coded medication schedule, list of tests with pending results at discharge, illustrated description of discharge diagnosis, and information about what to do if problems arise. On day of discharge, after-hospital care plan and discharge summary was faxed to the PCP 2-4 days after index discharge, a clinical pharmacist phoned the intervention participant. The discharge plan was reinforced using a scripted interview and medications were reviewed. Issues were communicated to the PCP or discharge advocate. Exclusion: discharged to a skilled nursing facility or other hospital, transferred to a different hospital service prior to enrollment, admitted for a planned hospitalization, on hospital precautions or suicide watch, deaf or blind, discharged to non-us community, unable to consent, sickle cell disease as admitting diagnosis, previously enrolled. Intervention group had significantly better performance than control group for: Hospital utilization(total number of emergency department visits and readmissions per patient per month) vs , p=0.009 ED utilization, visits per patient per month (0.165 vs , p=0.014) Ability to identify discharge diagnosis (79% vs. 70%, p=0.017) Able to identify PCP name (95% vs. 89%, 0.007) Visited PCP (62% vs. 44%, p<0.001) Self-reported discharged preparedness intervention participants significantly higher 30-day follow-up may limit ability to determine long term impact of intervention Costs cannot be generalized Utilization information not available on the electronic record had to be obtained through participant self-report Not targeted to a high-risk group Young sample (mean age 50 years) Assumption that if not able to reach patient at 60 days the patient was still alive. The discharge advocate intervention appears to improve overall postdischarge hospital utilization (hospital readmission + ED utilization). 18

19 percentage answered in the top two Likert categories for all 5 items (all 5 significant at p 0.05). Cost 33.9% lower observed costs for intervention group The intervention was more effective for patients with hospitalizations in the previous 6 months (p=0.014) There was no difference between the groups for: Readmissions alone, visits/participant/month (0.149 vs , p=0.090) but when combined with ED visits it was significant. 11 Laramee, et al., intervention and 146 control patients with CHF 90-day all-cause readmission To test the effect of a hospital-based nurse case management intervention readmission among CHF patients. Single site RCT conducted at an acute care hospital. After simple randomization of the first 42 patients, patients were randomized in blocks of 8. Intervention: Master s prepared CHF case manager While patient was in the hospital, CHF case manager coordinated care and facilitated discharge planning and needed consultations; submitted progress reports to the PCP; after discharge a letter was sent to the PCP that outlined the case management program Sample: mean age 70.7 years, even sex distribution, 71% had ischemic heart disease, half had severe left ventricular dysfunction, 20% had normal-to-mild dysfunction Inclusion: clinical signs and symptoms of CHF and either moderate-to-severe left ventricular dysfunction or radiographic evidence of pulmonary congestion and symptomatic improvement following diuresis; at-risk for early readmission Exclusion: discharge to long-term care facility, planned cardiac surgery, cognitive impairment, anticipated survival of fewer than 3 months, and long-term hemodialysis Adherence to the treatment plan Patient satisfaction ACEI, ARB, and BB dosages Overall cost of medical care Cause for readmission Length of stay Number of CHF readmissions Cumulative number of hospital days Number of days to first readmission 19

20 After discharge, patient and/or caregiver received telephone calls at 1 to 3 days after discharge and at weeks 1, 2, 3, 4, 6, 8, 10, and 12. There was no significant difference between the two groups in terms of readmission and length of stay. At 4-weeks and 12-weeks, the intervention group had significantly better adherence to weigh self-daily, check ankles and feet for swelling, follow fluid recommendations, and follow low-salt diet. In addition, at 12-weeks the intervention group also had significantly better adherence with taking medication compared to the control group. The intervention group was also significantly more satisfied with their hospital care, hospital discharge, care instructions, recovery at home, and total satisfaction score. There was no difference between the two groups for medication use and target dose advancement. There was also no significant difference in 90-day cost of care between the two groups. Timepoints: 4, 8, and 12 weeks After discharge, treatment adherence was assessed using patient self-report Researchers weren t blinded to group assignment The CHF case manager intervention was successful in improving discharge plan adherence and satisfaction. However, there was no improvement with readmission, length of stay, or medication use and target dose advancement. 12 Mion et al., 2003 To test the effects of using the transitional care model in the ED to reduce 30 and 120 day subsequent service use among Multisite (2-hospital) RCT using block randomization conducted at two EDs in teaching hospitals. Patients were assessed and stratified as high and low risk for service utilization, then randomized to the intervention or control Participants were 650 communityresiding individuals 65 years or older discharged after a visit to the ED. To be eligible patients needed to have a telephone, lived in the geographical service area of the ED, able to hear and understand in English, and, if cognitively impaired, have a caregiver/proxy. Intervention group mean age At 30 and 120 days: Service use - ED visits - Hospitalizations - Health care costs (both ED and readmissions) - Nursing home admission Physical & mental health 20

21 communitydwelling older ED patients. groups. The intervention group received a comprehensive assessment by an APN specialized in geriatrics. The APN collaborated with the physician, ED social worker, nurses, participant (and proxy if needed) to develop a discharge plan and community follow-up outreach, as needed. Short-term telephone follow-up by the APN was provided, as needed, until community agency staff contacted the participant. 74.4±6.5 years and 55% female; control group mean age 74.5 ± 7.3 years and 63% female. (SF-36) Satisfaction with discharge care Statistics: Chi-Square likelihood ratio tests, t-tests, Wilcoxon Ranksum tests The generalizabilty of these findings is limited due to: Location of service provision (i.e., teaching hospital) Population served (i.e., underserved populations impoverished & minority, including 60% non-white) Availability of services to make referrals to for post-dc followup Limited to having and using a telephone The intervention had no effect on overall service use rates at 30 or 120 days post discharge from the ED. There were no differences between the intervention and control groups in health care costs for subsequent ED visits or health care costs at either 30 or 120 days. The intervention was effective in lowering nursing home admissions at 30 days (0.7% versus 3%; odds ratio 0.21; 95% confidence interval [CI] 0.05 to 0.99) and in increasing patient satisfaction with ED discharge care (3.41 versus 3.03; mean difference 0.37; 95% CI 0.13 to 0.62). In subgroup analyses by risk group, low risk usual care participants were less likely to return to the ED within 30 days compared to those in the intervention group (10% versus 17%; OR 1.9; 95% CI 1.0 to 3.5). Participants in the high risk intervention group compared to those receiving usual An ED-based transitional care model reduced subsequent nursing home admissions and increased referrals to community agencies but did not decrease overall service use or costs for older ED patients. 21

22 care experienced fewer hospital days (0.6 }2.3 versus 1.6 }5.6; mean difference 1.0; 95% CI 2.0 to 0) and fewer nursing home admissions (2% versus 7%; OR 0.2; 95% CI 0.04 to 0.96) at 30 days. At 120 days, the intervention high risk group continued to have lower nursing home admissions (3% versus 10%; OR 0.3; 95% CI 0.07 to 0.94). There were no significant differences between groups or subgroups (risk level) based on SF-36 component subscale scores for physical and mental health. Participants in the intervention group were more satisfied with one item from a 6-item interview compared with participants in the usual care group (3.41 versus 3.03; mean difference 0.37; 95% CI 0.13 to 0.62). 13 Naylor et Multisite (2-hospital) al., 1999 RCT To test the effectiveness of an advanced practice nurse led discharge planning and home follow-up intervention for older adults at high risk for rehospitalizati on. Control group: standard care Intervention group: APNs assumed responsibility for discharge planning during hospitalization and substituted for visiting nurse during the first 4 weeks postdischarge. Standardized comprehensive discharge planning and home follow-up protocol individualized in collaboration with patient s PCP. Length/Dose: hospital admission (APN visits at least every 48 hours until discharge) to subjects, 65+ years admitted from home to one of two urban hospitals in an academic health care system for one of several medical or surgical diagnoses, speak English, be alert and oriented, live in the service area, and be contactable by phone post discharge. Plus subjects had to meet at least one of the following conditions: 80 or older Inadequate support system Multiple, active, chronic health problems History of depression Moderate-to-severe functional impairment Multiple hospitalizations in 6 months prior Hospitalization in the past 30 days Fair or poor self-rated health History of non-adherence to Time to first readmission for any reason, recurrence, or exacerbation of index DRG, comorbid conditions, or new health problems. Cumulative days of rehospitalization. Costs, functional status, depression and patient satisfaction. Kaplan-Meier survival curves compared control and intervention groups using log-rank statistic to compare the 2 cumulative readmissionfree rate curves. Proportional Hazards regression, used to adjust confounding variables, providing an adjusted hospital 22

23 weeks post-discharge (at least 2 home visits and weekly phone contact + plus 7 days/week phone availability) therapeutic regimen Enrolled: Control, n=186; Intervention group, n=177 Percent that completed the trial: 74% control, 70% intervention readmission rate ratio and 95% CI. Final multivariate model Time to first readmission for any reason was increased in the intervention group (log-rank χ 2 =11.1, p<0.001). Effect of intervention on time to first readmission for any reason remained significant (p<0.001) after adjusting for significant covariates. Mean days spent in the hospital per patient were significantly (p<0.001) less in the intervention patients. Average costs per patient were lower for the intervention group ($3630) than the control group ($6661; p<0.001). Intervention group: age (mean) years and 46% female; 44% African American Control group: age (mean) years and 54% female, 46% African American Not limited to just those patients who met guidelines for home care. An APN-centered discharge planning and home followup intervention for at risk hospitalized elders increased the length of time between hospital discharge and readmission and decreased costs at 6 months. 14 Naylor et Multisite RCT using al., 2004 block randomization To test the effectiveness of a 12-week comprehensive advanced practice nurse led discharge planning and home follow-up intervention for older adults with heart failure focused on rehospitalization, costs, quality of Control: standard of care Intervention group: Received a nurse led comprehensive care management intervention with a foundation in the Quality-Cost Model of APN Transitional Care. This included: patient and caregiver goals setting, individualized care plans, educational and behavioral 239 adults with heart failure ages 65+ from 6 urban academic and community hospital were randomized to receive the intervention or standard of care (control). Intervention, n=118 Control, n=121 Inclusion: 65 or older, diagnosis of HF (DRG #127), spoke English, alert and oriented, reachable by telephone after discharge, Time to first rehospitalization or death Cumulative days of rehospitalization, Readmission length of stay, Unscheduled resource utilization and Total costs Quality of life (Minnesota Living with Heart Failure Questionnaire) Functional status (Enforced Social Dependency Scale) 23

24 life, functional status, and patient satisfaction through 52 weeks. strategies to address learning needs, continuity of care and care coordination across settings, and clinical management of high risk patient groups. Length of intervention: index hospital admission through 3 months after the index hospital discharge reside within a 60-mile radius service area of the admitting hospital. Exclusion: end-stage renal disease Intervention group: age (mean) years and 60% female, 34% African American. Control group: age (mean) years and 56% female, 38% African American Patient satisfaction Tests: Group-specific Kaplan-Meier survival curves, proportional hazards regression, Wilcoxon ranksum tests (Analyzed as intent to treat.) Limited to Philadelphia/East coast of US. Rehospitalizations or deaths at 52 weeks were lower in the intervention group (56/118, 47.5%) compared to the control group (74/121, 61.2%; p=0.01). Distribution of times to first readmission or death was shifted toward longer time intervals in the intervention group than in the control group (Kaplan-Meier log rank χ 2 =5.0, p=0.026). Fewer total number of readmissions in the intervention compared to control group (104 vs. 162, p=0.047). Intervention group 52-week total costs ($7,636) were lower than the control group ($12,481, p=0.002). Significant differences (p<0.05) between the intervention and control groups for quality of life and satisfaction in the short-term but no differences in function. A comprehensive transitional care intervention for elders hospitalized with HF increased the length of time between hospital discharge and readmission or death and reduced total number of rehospitalizations at substantially reduced costs. The effect of the intervention on quality of life and patient satisfaction was only found for the intervention period (<3 months post-discharge) and had no effect on functional status. Results suggest the potential benefit of a comprehensive, multidisciplinary, individualized intervention directed by clinical nurse experts that spans the entire episode of acute illness and bridges the 24

25 transition from hospital to home. 15 Naylor et Single site RCT Service use and related al., 1994 costs To test the effects of a comprehensive discharge planning protocol on elderly patient and caregiver 2, 6, and 12 week outcomes and costs Control group: standard discharge planning. Intervention included: gerontologic clinical nurse specialists implementing a comprehensive, individualized discharge planning protocol developed specifically for elderly patients. 276 medical and surgical cardiac patients ages 70+ years (and 125 caregivers) admitted from their homes to an urban academic teaching hospital Eligible subjects: age 70+; DRGs: medical -congestive heart failure and angina/myocardial infarction, DRGs surgical: coronary artery bypass graft and cardiac valve replacement; had a telephone; and were assessed as alert and oriented when admitted. Statistics used: chisquares or Fisher exact test and independent t- test. The plan included: initial and ongoing assessment of the discharge planning needs; development of discharge plan with patient, caregiver, and hospital team; validation of patient and caregiver education; coordination of discharge plan throughout hospitalization to 2- weeks post-discharge; interdisciplinary communication regarding discharge status; ongoing evaluation of the discharge plan s effectiveness. Sample mean age 75.5 years. 25