Table 2.

Primary Research Studies—Characteristics, Settings, and Results of Effective Care Coordination Models

Author, year; study design*; N(I,C)	Intervention name; eligibility criteria	Setting characteristics	Description of patient contacts	Intervention effects
				Hospitalizations	ED visits
Transitional care interventions
Capp, 201751; cohort; I=406 C=3396	Bridges to care; adults with ≥ 2 ED visits and/or hospitalizations in past 180 days	Large urban academic medical center, Colorado	First home visit by community health worker within 24–72 h, second visit by PCP within 1 week of ED or hospital discharge; 8 visits over 60 days (community health worker, nurse, primary care provider, and/or behavioral health provider) depending on patient needs.	Average no. of admissions per person, 180 days before enrollment: I=1.04, C=1.15 180 days after 60-day intervention: I=0.75, C=1.02 Difference of differences = −0.16, P<0.1	Average no. of visits per person, 180 days before enrollment: I=5.12, C=4.93 180 days after 60-day intervention: I=2.79, C=3.60 Difference of differences = −1.01, P=<0.01
Hamar, 201654; cohort; I=560 C=3340	Care transition solution; adults admitted with ≥ 1 condition (COPD, heart failure, myocardial infarction, pneumonia)	14 community hospitals in north Texas	Initial visit in hospital with nurse before discharge, then 4 calls over 4 weeks	Proportion with ≥ 1 readmission at 30 days: AOR=0.56 (0.41–0.77) At 6 months: AOR=0.47 (0.35–0.65)	NR
Gardner, 201453; cohort; I=21 C=21	Care transitions intervention; adults participating in Medicare fee-for-service, admitted to hospital	6 community hospitals, Rhode Island	Initial visit in hospital by nurse, home visit “shortly after discharge,” 2–3 phone calls during 30-day post-discharge period	Propensity score–matched no. of readmissions at 6 months: I=0.65, C=0.93 P=0.01	Propensity score–matched no. of visits at 6 months: I=0.44, C=0.50 P=0.55
Coleman, 200638; RCT; I=379 C=371	Care transitions intervention; older adults (≥65) admitted with ≥ 1 condition (stroke, heart failure, diabetes mellitus, etc.)	Community health system, Colorado	Nurse met patients in hospital before discharge, home visit within 48–72 h of discharge, then 3 more times during the 28-day post-discharge period.	Proportion with ≥ 1 readmission at 30 days: I=0.08, C=0.12 (AOR 0.59 [0.35, 1.00], P=0.048) At 90 days: I=0.17, C=0.23 (AOR 0.64 [0.42, 0.99], P=0.04) At 180 days: I=0.26, C=0.31 (AOR 0.80 [0.54, 1.19], P=0.28)	NR
Naylor, 199943; RCT; I=177 C=186	Transitional care model; older adults (≥65) admitted with ≥ 1 condition (heart failure, respiratory infection, orthopedic procedure, etc.)	2 urban hospitals affiliated with the University of Pennsylvania	Initial nurse visit within 48 h of admission, visits at least every 48 h during admission, home visits after discharge (first within 48 h, second 7–10 days post-discharge, additional visits based on patients’ needs), weekly nurse-initiated phone contact	Proportion with ≥ 1 readmission at 24 weeks: I=0.20, C=0.37 P<0.01	NR
Outpatient care or case management
Shah, 201157; cohort; I=98 C=160	Care management program; adults aged 18–64, <200% federal poverty level, uninsured, “met frequent user criteria”	Public safety-net hospital and clinics in Kern County, CA	Care managers (social worker or medical office assistant) met with patients at least monthly in the home and/or clinic, for variable lengths of time (care manager decided when patient graduated program)	Adjusted ratio of no. of admissions per year (I:C) was 0.81, P=0.38	Adjusted ratio of no. of visits per year (I:C) was 0.67, P<0.001
Peikes, 200946; RCT; Mercy Medical Center (1 of 15 sites) I=669, C=467	Medicare coordinated care demonstration; adults participating in Medicare fee-for-service and with ≥ 1 condition (heart failure, COPD, etc.)	Mercy Medical Center—rural community hospital, Iowa	Nurse completed in-person evaluation within 2 weeks of enrollment, contacted patient at least monthly, 69% were in-person (either at home or during clinic visit)	Average no. of admissions per person per year: I= 1.15, C=0.98 P=0.02	NR
Shumway, 200848; RCT; I=167, C=85	Comprehensive case management; adults with ≥ 5 ED visits in past 12 months and had “psychosocial problems that could be addressed with case management”	Urban public hospital in San Francisco, CA	Social workers completed assessments, individual and group supportive therapy, assistance to a variety of community resources, and “assertive community outreach” (frequency and schedule of patient contacts NR)	Effect size NR, P=0.08 for treatment effect in adjusted model for visits over 2 years	Effect size NR, P<0.01 for treatment effect in adjusted model for visits over 2 years
Sommers, 200058; Cohort I=280 C=263	Senior care connections; adults ≥65 with difficulty in ≥1 instrumental activity of daily living and 2 ≥ chronic conditions	Primary care clinics in San Francisco Bay area, CA	Initial home visit with case manager (nurse or social worker), treatment plan drafted by care team (nurse, social worker, primary care provider), patients contacts via phone, home visits, small group sessions, or office/hospital visits at least once every 6 weeks	Number of admissions per person per year at baseline: I=0.35, C=0.06 during year 1: I=0.38, C= 0.34 during year 2: I=0.36, C=0.52 P=0.03	Proportion with ≥1 visit at baseline: I=0.09, C=0.06 during year 1: I=0.20, C=0.17 during year 2: I=0.21, C=0.17 P=0.77
Other intensive primary care models
Crane, 201252; cohort; I=34 C=36	Drop-in group medical appointments; uninsured, family income ≤ 200% federal poverty level, ≥ 6 ED visits in past year	Rural community hospital, North Carolina	Twice weekly groups sessions, short individual visit right after; direct phone access to nurse care manager; team included nurse, primary care, and behavioral health providers	NR	Median no. of visits per month during 1 year before: I=0.58, C=0.58 during 1 year after: I=0.23, C=0.42 Difference in differences: 0.23, P=0.005
Meret-Hanke, 201155; cohort; I=3889 C=3103	Program for all-inclusive care for the elderly; adults >65, with functional limitations or dementia, income <300% supplemental security income	National US program	Interdisciplinary care teams provided care management, clinical monitoring, and updated care plan in response to changes in enrollee’s health and functional status	Propensity score-matched any hospitalization at 6 months: AOR 0.35, P<0.01 At 2 years: AOR 0.16, P<0.01	NR

AOR, adjusted odds ratio; C, control group; COPD, chronic obstructive pulmonary disease; ED, emergency department; I, intervention group; RCT, randomized controlled trial

*Study designs were either RCT or observational cohorts with comparative controls