Beyond surviving: A scoping review of collaborative care models to inform the future of post-discharge trauma care

Abstract

Trauma centers demonstrate an impressive ability to save lives, as reflected by inpatient survival rates of over 95% in the United States. Nevertheless, we fail to allocate sufficient effort and resources to ensure that survivors and their families receive the necessary care and support after leaving the trauma center. The objective of this scoping review is to systematically map the research on collaborative care models (CCM) that have been put forward to improve trauma survivorship. Of 833 articles screened, we included 16 studies evaluating eight collaborative care programs, predominantly in the U.S. The majority of the programs offered care coordination and averaged 9-months in duration. Three-fourths of the programs incorporated a mental health provider within their primary team. Observed outcomes were diverse: some models showed increased engagement (e.g., Center for Trauma Survivorship, trauma quality-of-life follow-up clinic), while others presented mixed mental health outcomes and varied results on pain and healthcare utilization. The findings of this study indicate that collaborative interventions may be effective in mental health screening, PTSD and depression management, effective referrals, and improving patient satisfaction with care. A consensus on core elements and cost-effectiveness of CCMs is necessary to set the standard for comprehensive care in post-trauma recovery.

Traumatic injuries, ranging from road traffic injuries to falls, burns, and violence, can result in immediate life-threatening conditions. Patients with traumatic injuries tend to receive coordinated, high-quality care during hospitalization, but their post-discharge care is often fragmented and disorganized (1). Common concerns from patients and caregivers are a lack of coordination and continuity of post-discharge care, a lack of a consistent point of contact, delays or failure to receive necessary outpatient appointments, and limited time with clinicians (2).

There is empirical evidence that injury survivors are unlikely to fully return to their pre-injury physical, mental, occupational, and social status over time (3,4). Previous studies show that 20% of injury survivors screen positive for mental health conditions, 48% report daily pain, 37% have a new functional limitation for performing activities of daily living, 45% have social dysfunction, and 40% have not yet returned to work at 6–12 months post-injury (5–9). Further, there is evidence that these impairments could extend to 5 years and beyond for certain injuries (10,11).

Part of this challenge can be attributed to the lack of a standard approach for post-discharge care. After discharge, it is common for patients to have a single visit with the trauma team and then be referred back to their primary care providers (PCP) for ongoing care (12). Such directives are given without considering whether the patient has a PCP and, if so, whether that individual is capable of coordinating the necessary care (12). Oftentimes, PCPs lack experience and knowledge in caring for patients with traumatic injuries and may not have established connections with providers, such as physical therapists or behavioral health specialists, required for treating specific conditions. The sole visit with the trauma team also tends to be confined strictly to trauma surgery specialists, with few centers having additional providers, such as behavioral health providers or physiatrists, readily available in the clinic (12,13).

As an approach to providing holistic post-discharge care, collaborative care models (CCMs) have been developed and proven effective in the context of several chronic conditions, such as dementia, cancer, and psychiatric disorders (14–16). CCMs involve a team-based approach in which healthcare professionals from different disciplines work together to provide comprehensive and coordinated care. A well-known CCM within the traumatic injury community is the post-intensive care syndrome clinics (PICS), designed to address the physical, cognitive, and psychological challenges faced by survivors of critical illness following an intensive care unit stay. PICS employs a multidisciplinary approach involving pulmonologists, physical therapists, neuropsychologists, and social workers to provide specialized care(17).

In the context of traumatic injuries, programs like hospital-based violence intervention programs adopt some of those principles to prevent injury recidivism and mitigate violence-related health issues among violence survivors (18). However, only a select group of trauma centers have applied a comprehensive, trauma-specific CCM approach to address the biopsychosocial needs of injury survivors and improve outcomes. There is an ongoing dialogue within the trauma care community that suggests a transition in focus from merely ensuring survival to fostering a concept of survivorship, viewing traumatic injury as a chronic condition(19) and treating it as such. In parallel with these shifts, the integration of CCMs could be considered significant. As such, it is crucial to map out the available CCMs that have been tailored or adapted to address the unique biopsychosocial needs of the injured population to improve survivorship. This involves understanding the factors that lead to the success of these models, including their structure, processes, outcomes, and cost data. Accordingly, this scoping review aimed to systematically map the research on traumatic injury recovery CCMs, identify gaps in this area, and propose best practice recommendations.

MATERIALS AND METHODS

Protocol and Registration

We followed the process and recommendations developed by Arksey and O’Malley (2005) and Levac et al. for conducting this review. Then, we reported our findings based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA-2018) and the EQUATOR checklist (Table S1, http://links.lww.com/TA/D832) (20–22).

Information Source

We searched Ovid MEDLINE, Web of Science, and the Cochrane Library on November 18, 2022, from the first date available.

Search

We searched the combination of the search terms (Table S2, http://links.lww.com/TA/D833). Additionally, a medical librarian reviewed the terms and provided guidance. We identified additional studies from citation searching and relevant reviews. At the outset of the search, we did not impose any restriction on the study design, literature language, or any other restriction.

Selection of Sources of Evidence

Two reviewers independently conducted screenings as part of the first round of a two-stage screening process. The same two reviewers proceeded to the second step, reviewing the full texts. A third reviewer with expertise in the field resolved any disagreement. The criteria for full-text inclusion and exclusion were as follows:

Inclusion Criteria

• Adults (18 years or older) admitted for traumatic injury.

• Implementation of a multidisciplinary post-discharge intervention for more than three months. Multidisciplinary was defined as involving two or more health professionals addressing at least two of the health domains described in the National Trauma Research Action Plan (NTRAP) scoping review on long-term trauma outcomes (physical health, mental/emotional health, social health, cognitive health, and quality of life/global functioning)(23).

Exclusion Criteria

• Articles, grey literature, and reviews that did not present primary research findings, including those that were narrative, commentaries, or presented theories or frameworks.

• Books and conference proceedings

• Articles published before January 1, 2000

Data Charting Process

Two researchers developed the data extraction table, and two others cross-validated the spreadsheets.

Data Items

We extracted the data of interest as follows: (a) study-related variables (first author name, publication year, sample size, country of origin, and the presence/type of control group), (b) population characteristics (age, injury type), (c) intervention characteristics (program name, core staff, duration and contact points, resources, and general description) and (d) outcome measures (patient engagement, patient satisfaction, mental health outcomes, social and occupational outcomes, physical health outcomes, pain, cost and cost-effectiveness, healthcare utilization, financial outcomes, and quality of life)

Critical Appraisal of Individual Sources of Evidence

Two investigators assessed the quality of randomized clinical trials and non-randomized trials using the Cochrane Risk of Bias 2.0 (Rob2) and the Risk of Bias in Nonrandomized Studies of Interventions (ROBINS-I) tools, respectively (Figure S1, http://links.lww.com/TA/D831). Newcastle-Ottawa scale (NOS) was utilized for cohort studies (Table S3, http://links.lww.com/TA/D834) (24).

Ethics

No ethics approval was needed as this review used publicly available literature.

RESULTS

1. Selection of Sources of Evidence, Study Design, and Quality Appraisal

Our initial search identified 833 articles, of which 16 met the criteria for final inclusion (Figure 1). These 16 studies assessed eight distinct programs (1,12,13,25–37) (Table 1). Most of the included studies were trials (75%) (1,25–27,30–37) and conducted in the United States (75%) (1,13,25,27–34,37). Roughly half were multi-institutional (56%) (13,26,27,29–32,35,37). Two studies (13%) included only TBI patients (25,26). Detailed characteristics of the included programs are summarized in Table 2 (13,31,32,35,38). The brief report by Bradford et al. (39) contained in-depth information about the Trauma Collaborative Care (TCC), which was used to gain a better understanding, but it was not included due to the study type. The same scenario was adapted to the Fast Track study protocol (40). We also assessed two additional trials of stepped collaborative care interventions but did not include them in the review because their main focus was the treatment of PTSD and substance abuse (41,42). Table 3 presents a detailed overview of the program characteristics.

Figure 1.

Prisma flow diagram of included studies.

Table 1.

List of Programs and Associated Papers

Name of the Program	N. (Study reference)
Brain Injury Education, Training, and Therapy to Enhance Recovery (BETTER) – United States	1 (25)
The Center for Trauma Survivorship (CTS) – United States	2 (1,13)
Fast Track - Netherlands	2 (35, 37)
The Multidisciplinary Outpatient Follow-up Program - Norway	1 (26)
Multidisciplinary Screening and Coordinated Care Intervention - Australia	1 (37)
Stepped Collaborative Care Intervention – United States	3 (32, 33,34)
Trauma Collaborative Care/Trauma Survivors Network (TCC-TSN) – United States	5 (27, 28,29, 30, 31)
Trauma Quality of Life Clinic (TQoL) – United States	1 (12)

Table 2.

Program Characteristics^*

	Count (%)
Country of origin (n = 8)
United States	5 (63%)
Netherlands	1 (12%)
Australia	1 (12%)
Norway	1 (12%)
Year of publication (n = 8) **
2000–2005	1 (12%)
2005–2010	0 (0%)
2011–2015	2 (25%)
2016–2019	2 (25%)
2020–2023	3 (38%)
Intervention arm Sample size
<50	2 (25%)
50–100	3 (37%)
101–400	2 (25%)
>400	1 (12%)
Duration of follow-ups (Last follow-up)
6m	2 (25%)
12m	6 (75%)
Duration of intervention (Last intervention)
3m	1 (12%)
4m	1 (12%)
6m	1 (12%)
12m	4 (50%)
Not specified	1 (12%)
Outcomes (n=8)
Program engagement	2 (25%)
Patient satisfaction	3 (37%)
Caregiver burden	1 (12%)
Mental health outcomes	7 (87%)
Social/Occupational	3 (37%)
Physical health outcomes***	4 (50%)
Pain	3 (37%)
Cost/Cost-effectiveness of program	2 (25%)
Healthcare utilization	4 (50%)
Financial burden	1 (12%)
Quality of life	2 (25%)

^*To avoid showing redundant data, the numbers are determined for each program, rather than per study. If multiple studies were published from a single program, only the one with the highest quality assessment score was documented in the table.

^**According to the first paper published out of the program.

^***It includes physical function independence, activities of daily living, Perception of physical function ability, and physical disabilities.

AIS, Abbreviated Injury Scale; ISS, injury severity score; DRS, Disability Rating Scale,

Table 3.

In-depth Overview of Program Features and Associated Research Publications

Program Name/Country	Study	N* (total/intervention arm)	Type of study	Study population	Exclusion criteria	Evaluation(E), Intervention(I)	Touch Points (P), Number of visits(V), Duration (D) of visits	Outcome measures (tool)	Summary findings
BETTER/USA	Oyesanya (2022)	15/15	Non-randomized trial (Single arm)	TBI English/Spanish speakers Age 18–64	Discharged before approached	E: Psychological status Symptom tracking I: Education Behavioral intervention Social facilitation Goal setting, Workbook, Brain injury coping skills (BICS) sessions, Referral	P: Bedside 4 wks., 6wks 8wks 12 wks 16 wks 24 wks V: 7 visits in 16w D: Intro: 20–40 Support calls: 15 to 60 min	1. QoL 2. Self-efficacy, 3. Depression(PHQ- 9) 4. Readmission 5. Healthcare usage 6. PTSD (PCL) 7. Social outcomes	NOT REPORTED YET
CTS/USA	S1: Livingston (2020)	107/107	Non-randomized trial (Single arm)	Admitted trauma pts. ICU>2 d ISS≥ 16 Age 18–80	Mild trauma	E: Psychological status Physical status Behavioral health Medication usage Rehab needs I: Pain management Specialized rehab Physical therapy Subspecialty appointments Referrals, Uber health 24/7 Phone access Referral	P: ICU discharge Discharge 12 mos. V: 1,280 visits 1000 calls 1500 texts D: >1 hour	S1,S2: 1. PTSD(PCL) 2. Depression (BDI) S1,S2: 3. Patient engagement S1,S2: 4. ED usage 5. Reinterventions S1,S2: 6. Referrals	S1: 32% PTSD positive, 32% depression positive, 23% both. Inadequate pain management was detected. S1: 100% visited by nurse practitioner, 80% by psychiatrist and ~25% by trauma surgeons. S2: almost double follow up rate withing trauma department in CTS group. S2: significantly higher follow up with surgical specialties. S2: ED utilization was higher in the first 6 months, no difference after 6 months. More avoidable ED visit in control group. Screen positive patients referred to behavioral health service
	S2: Goldstein (2022)	236/119	Cohort
Fast Track/Netherlands	S1: Bauman (2017)	132 / 65	Non-randomized controlled trial	Poly trauma patients. (Complex extremity injuries or complex pelvic/aceta bulum fx) Age≥18 ISS≥ 16	Alcohol/drug abuse inadequate Dutch language skills Severe psychiatric problem	E: Symptom tracking Psychological status I: Physical therapy, Care coordination, Psychoeducation, Social work, support, Goal setting *Duration and intensity depended on injury severity & type	P: Baseline 3 mos. 6 mos. 9 mos 12 mos. V: Weekly f/u/monthly with trauma surg.	S1,S2: 1. Functional status (FIM) S1,S2: 2. quality of life (SF-36) S1: 3. Bodily pain S1,S2: 4. Anxiety and depression (HADS) S2: 5. Incremental Cost-Effectiveness Ratios (ICERs)- costs per unit of FIM improvement and per QALY	Both groups improved over time, but FIM and SF-36 scores differed slightly between the groups at any measured time points t recovery for Fast track group in 6 mos., no difference in 12 mos. More pain in favor of the control No sig. difference in HADS scores at any of the measured time points Mean total costs Є18,918 higher in Fast track group. Average incremental effects on the FIM were 3.7 in favor to Fast track. Cost-effectivness remains inconclusive
	S2: Wijnen (2019)
The multidisciplinary outpatient follow-up program/Norway	Vikane (2017)	151/81	RCT	TBI Age 16–55 Admitted≥48h persistent PCS symptoms	language barrier Not working Psychiatric condition	E: Psychological status Occupational status Symptom tracking Pain I: Psychoeducation memory aids Time management CBT Meeting with employer or NAV, Plan and monitor graded work return, Goal setting,	P: 6–8 wks 8–9 wks 9–20 wks 20–52 wks V: Individualized Two initial vists followed by Four work education sessions for everyone D: 2 hours	1. Return to work 2. Depression (HADS) 3. Anxiety (HADS) 4. Exacerbation of post-concussion symptoms 5. Patient global impression of change (PGIS) 6. Disability	Intervention did not improve return to work. No sig. difference between groups for anxiety and depression The number of post-concussion symptoms were fewer in the intervention. No difference in improvement reported on PGIS detected within groups No group difference for disability
MI/Australia	Browne (2013)	142/69	RCT	Sustainable Trauma Age 18–80	Moderate to severe TBI Post traumatic amnesi>24h GCS<13 in admission	E: Symptom tracking Psychological status Pain evaluation, I: Goal setting Scarring management, Physical therapy, Psychoeducation CBT, Treatment plans, Referral	P: 1 mos 3 mos 6 mos V: 2 D: 2–4 hours	1. Pain (BPI) 2. PTSD (PAS, PCL-C) 3. Depression (CES-D) 4. Functional capacity 5. Alcohol use (AUDIT) 6. Healthcare use 7. Financial burden 8. RTW	Not sig for pain symptom. severity No sig. difference in PTSD, Depression, functional disability and mobility and alcohol use behavior No sig. difference in healthcare use post discharge 45.2% of intervention group had no change in income versus 36.4% in control. (No p-value provided) Greater RTW in control group
Stepped collaborative care/USA	S1: Zatzick (2001)	34/16	RCT	Trauma survivors English speakers Age: 14–65 AIS≥5	Discharged before approach Cognitive impairment Self-inflicted injuries Psychosis	E: Psychological status Symptom tracking Recovery tracking Prior traumas Substance abuse I: Education Core case Mgmt., Psychopharmacology Motivational speech CBT	P: Discharge 1 Mos 4 Mos D: 1-h	S1, S2, S3: 1. PTSD (PCL) 2. Depression (CES-D) 3. Patient engagement S1, S2, S3: 4. Substance abuse (AUDIT-C) S1, S2, S3: 5. Physical function (PCS)	S1, S3: significantly decreased PTSD and depressive symptoms in the first follow up no difference in next follow-ups. S2: ↑ engagement of pts. No reductions in PTSD and depressive symptoms or in alcohol use in the intervention group compared to the control group S1, S3: No difference in functional outcomes and alcohol consumption S3: Patients with 3 or more baseline risk factors for enduring PTSD and patients treated at sites with good to excellent protocol implementation had greater treatment outcomes
	S2: Tsosie (2011)	30/15					P: 3 Mos 6 Mos D: 45–60 min
	S3: Zatzick (2021)	625/265	RCT				P: 1 Mos 3 Mos 6 Mos 12 Mos D: Mean: 122 min
TSN/USA	S1: Castillo (2013	251/126	Cohort	Major orthopedic injuries English speakers Age 18–69	TBI Immobile due to SCI Psychosis Dementia No access to phone or PC Brain injury with AIS>2 Not in jail or homeless	E: Psychological status Pain Mgmt. Rehab needs Recovery tracking I: Counseling (Coaching calls) Peer visitation Support group, Education, Crisis prevention, CBT, Self-mgmt. Classes, Personalized recovery plan, Handbook, Website Referral	P: 1w 2w 3w 4w 5w 6w/ 10w/ 6mos/ 12 mos	S1–3: 1 .Patient satisfaction S1–5: 2. Depression (PHQ-9) 3. Anxiety (BSI) 4. Psychologic distress (BSI) 5. Self-efficacy (PAM) 6.Social support (MSPSS) 7. Health status (SF-12), physical (PCS) and mental(MCS) 8. Return to work S2,S3: 9. Pain (BPI) 10. Functional outcome 11 .Recovery likelihood 12.Return to daily activity likelihood 12. Program use and engagement 13. Referral	S1, S2, S3: High satisfaction S1,S4: No difference in health status, anxiety, self-efficacy, and patient-activation Marginally sig. improvement in depression S4: Sig. improvement of TSN patients depression score in 6 mos., no difference in 12 mos. S1 S4: Small improvement in self-efficacy but not sig. S4: No sig. difference in functional outcomes S2: ↑ likelihood of recovery and return to daily activities for TSN patients S1–5: Lack of participation in program S1: 47% used at least one TSN resource. S5: 29% of pts received all the key resources of the program. Highly variable use across sites S2: Coordinators assisted pts access to different professionals
	S2: Wegner (2017)	896/481	RCT	Major orthopedic injuries Length of stay> 5 d or >3 d with planned readmissio n or reinterventi on
	S3 Simske (2019)	485/211	Cohort
	S4, S5: METRC (2019,2022)	896/481	RCT	English speakers Age 18–60
TQOL/USA	Trevino (2020)	36/18	Cohort	Admitted Trauma pts. Age≥18 Screen positive for depression or PTSD or chronic pain	Moderate to severe traumatic brain injury (TBI)	E: Psychological status Symptom tracking Pain tracking I: Pain management Physical therapy Social work support Psychoeducation Referral	P: 1 wk 6 mos (f/u appointments were made as needed)	1 .No show rate 2. Mental health visits engagement 3. ED visits 4. Readmission 5. Cost	↓ no-show rates in intervention group (22% vs. 40%) More behavioral health visits in intervention (23 v. 1) No difference in ED visits at 6-months No difference in readmission rates at 6-months On average, the social worker cost was US $25 per hour, and the physical therapist cost was US $50 per hour

Note:

^*Total number of participants / number of participant in intervention arm.

RCT, randomized clinical trial; MI, multi-institutional; SI, single-intuitional; TBI, traumatic brain injury; AIS, Abbreviated Injury Scale; TSN, trauma survivors network; SCI, spinal cord injury; PHQ, Patient Health Questionnaire; BSI, Brief Symptom Inventory; PAM, Patient Activation Measure; MSPSS, Multidimensional Scale of Perceived Social Support; PCS, Physical Components Summary; MCS, Mental Components Summary; BPI, Brief Pain Inventory; METRC, The Major Extremity Trauma Rehabilitation Consortium; TRRP, Trauma Resilience and Recovery Program; CTS, Center for Trauma Survivorship; BDI, Beck Depression Inventory; ED, emergency department; 3CM, Three Component Model; PCP, primary care physician; CBT, Cognitive-Behavioral Therapy; CIDI, Composite International Diagnostic Interview; PDEQ, peritraumatic dissociative experiences question; RA, research assistant, OT, occupational therapist; ISS, Injury Severity Score; CTS, Center for Trauma Survivorship; Surg, Surgeons, Rehabilitation; FIM, Functional Independence Measure; HADS, Hospital Anxiety and Depression Scale; Eval, Evaluation; TQOL, trauma quality-of-life follow-up clinic; GP, general practitioner; PAS, posttraumatic adjustment scale.

2. Collaborative Care Models (CCM)

2.1. Program Characteristics

Two of the eight included programs, the Center for Trauma Survivorship (CTS) and the Trauma Quality of Life Clinic (TQOL), are designated as structured trauma survivorship clinics. Here, patients receive comprehensive in-person follow-up care from a multidisciplinary team of trauma experts during a single visit. Three programs describe structured rehabilitation care plans delivered in inpatient or outpatient rehabilitation facilities (26,35–37). Every study from the TCC program indicated that they provided peer support and online networking(27–31). Of these, three studies also incorporated a care coordinator responsible for case management and linking patients to TCC resources (27,30,31). Finally, the Brain Injury Education, Training, and Therapy to Enhance Recovery (BETTER) program (25) and the stepped collaborative care program focused primarily on care coordination post-injury (32,33). The latter also included a step-up approach to the management of PTSD.

Despite heterogeneity in design, several elements were consistent across programs. All programs included some form of symptom tracking, behavioral health screening, and the development of an individualized recovery plan (1,12,13,25–32,34–37,41). Seven out of eight programs, including those without a designated care management role, provided care coordination services, including referral to subspecialty care (1,12,25–27,32,36). Half of the programs included in-person multidisciplinary follow-up visits (1,12,26,36), and 38% were predominantly delivered virtually (25,27,32). Finally, 25% of programs provided transportation assistance, (1,36) and only one program described 24/7 access to the care team through email and cell phone(1).

2.2. Injury Type and Mechanism

Two programs specialized in TBI patients (25,26), one in patients with major orthopedic injuries (43), and the remaining five on the broader trauma population. All programs except the TQOL clinic reported the injury mechanism for their study populations. However, not all programs utilized the same categories to report it, making it challenging to determine the overall breakdown of injury mechanism. Falls were reported in five programs(1,26,30,35,36), ranging from 7%−37%. Road traffic injuries were reported in five programs, ranging from 20–75%(1,25,26,30,36). All programs in the United States included participants who had suffered firearm injuries, with rates ranging from 3–23%.

2.3. Duration of Intervention and Follow-Up.

The average duration of the collaborative care intervention across studies and programs was nine months. Studies describing 6 of the programs had intervention periods of one year (1,13,26,27,30–32,35,37), and all the included programs assessed outcomes for at least six months after discharge (1,12,13,25,26,28,30–32,34,35,37).

2.4. Core staff

Table S4 illustrates the core staff of each program, http://links.lww.com/TA/D835. Most programs (87%) included a mental health provider as part of their core staff (1,12,13,26,32,33,35–37). Half of the programs involved a physical or occupational therapist(12,35–37), a social worker(1,12,13,26,35,37), and a designated team member providing case management or patient navigation(1,13,25,27,30–33). 38% involved a trauma surgeon or nurse practitioner (1,12,13,35,37) and a rehabilitation medicine specialist (26,35–37).

3. Outcomes

3.1. Patient Engagement

Studies evaluating the CTS and TQOL clinics found significantly higher rates of follow-up in the intervention arm(1,12,13). For example, 85% of patients in the CTS cohort were seen for at least one post-trauma appointment compared to 41% in the control group(1). No-show rates among patients seen in the TQOL clinic were 22%, compared to 40% in the control group (12). Attrition rates were reported for five programs, and the mean rate was 22%(26,28,30,31,33–37). The highest dropout rate was 47% for Browne et al.’s outpatient rehabilitation program.

3.2. Patient Satisfaction

Only studies of the TCC program had patient satisfaction as an outcome. Within five TCC program publications, three of them reported high satisfaction with the TCC resources overall, but none found higher satisfaction with care in the intervention group (28–30).

3.3. Mental Health Outcomes

Mental health outcomes were assessed for all programs, and most of them (87%) included a mental health provider as part of their core staff. In three programs, outcomes were not compared to those of a control group (CTS)(1), or results have not yet been published (TQoL clinic and BETTER)(25,44). Only one of four programs with mental health providers (Stepped Collaborative Care intervention) showed improved mental health outcomes. The remaining programs with mental health providers as part of their core staff did not significantly improve mental health outcomes. The one program that did not include mental health providers (the TCC program) but offered resources to help patients access mental health services found lower rates of depression in patients enrolled after TCC implementation.

3.5. Social and Occupational Outcomes

Social and occupational health outcomes were assessed for only three programs(26–30,36). None of these programs improved rates of return to work after traumatic injury. Simske et al. and Wegner et al. did find that engagement with the TCC increased return to daily activities, and Castillo et al. demonstrated that TCC participation increased reported social support.

3.6. Physical Health Outcomes

Physical health outcomes including functional status and disability, were assessed for half of the programs (26–29,31–34,36). Vikane et al.’s multidisciplinary outpatient rehabilitation program improved post-concussive symptoms in TBI patients. The remaining three programs that assessed physical health outcomes did not show improvement in these outcomes when compared to usual care (28–34,36).

3.7. Pain

Pain was an outcome of interest for three programs (27,30,36,37). One study of the TCC program found a modest reduction in the odds of experiencing pain in the intervention group (30) but the other found no effect (30). The Fast Track inpatient rehabilitation program described by Bouman et al. and Browne et al.’s outpatient rehabilitation regimen did not improve participants’ pain.

3.8. Intervention Costs and Cost-Effectiveness

Costs were described for only two of the eight programs described in this review(35,44). Trevino et al. described the cost incurred by the TQOL pilot beyond the standard of care, which included the hourly rate of the trauma psychologist, social worker ($25/hour), and physical therapist ($50/hour). Winjen et al. found that, compared to usual care, the Fast Track inpatient rehabilitation program was associated with higher cost-effectiveness per unit of functional status but not quality-adjusted life years (QALY).

The eight programs received funding from different sources, including the federal government, local governments, nonprofit organizations, charities, professional organizations, and their own institutions to conduct the research studies. However, no information was included on the financing of the programs and their sustainability beyond the study period.

3.9. Healthcare Utilization and Readmission

Healthcare utilization was a primary outcome in half of the described programs(1,13,25,36,44). TBI patients enrolled in Browne et al.’s program reported a lower frequency of general practitioner visits. The TQOL clinic patients did not experience decreased readmission rates or emergency department visits (12). Results were mixed for studies evaluating the CTS clinic(1,13). Livingston’s study found that the CTS cohort had lower ED utilization in the year following discharge. In contrast, in Goldstein’s study, the CTS cohort had higher ED utilization at six months post-injury and similar utilization to the control group at one year. Both studies found that receiving care in the CTS clinic did not significantly change readmission rates(1,13). On the other hand, both the CTS and the TQoL clinic showed increased follow-up and healthcare utilization in terms of additional provider visits.

3.10. Financial Outcomes

Browne et al. found similar degrees of financial strain in the control group and intervention groups. Financial outcomes were not assessed for the other seven programs(1,12,13,26–35,37).

3.11. Quality of Life (QOL)

The Fast Track program did not significantly improve QOL at 12 months post-injury (37). QOL was not a primary outcome in studies of the remaining six programs (1,12,13,26–34,36). The BETTER initiative had delineated QOL as a metric among their intended outcome measures. However, as mentioned, a substantial portion of their empirical results has not yet been published(25).

Overall, the outcomes assessed in the studies reviewed align with 76% (22/29) of the Core Outcome Measures for Research in Traumatic Injury Survivors recommended by the National Trauma Research Action Plan (NTRAP). Specifically, the studies covered 6/8 mental health outcomes, lacking only suicidal ideation and behavior change; 6/8 physical health outcomes, lacking only sleep and prescribed narcotic use; 8/9 social health and quality of life outcomes, lacking only trauma-specific quality of life; and 2/4 cognitive health outcomes, lacking only executive functioning and concentration/attention. While the studies employed some measures like PHQ-9, SF-12, GOS-E, and AUDIT-C as recommended by NTRAP, none incorporated PROMIS measures, despite their recommendation for assessing anxiety, depressive symptoms, mental health, pain, and physical mobility outcomes.

DISCUSSION

Our comprehensive review, drawn from data collected from the 16 included studies involving 1113 patients and 8 distinct programs, demonstrated that traumatic injury CCMs exhibit promising strides in certain mental health metrics and patient satisfaction. It is noteworthy that the programs aligned well in terms of tracking symptoms, behavioral health screenings, and developing personalized recovery plans. Most of these programs (87%) included a mental health expert in their core team. CTS and TQOL clinics recorded an uptick in follow-up rates. Figure 2 provides a concise review of what we know about CCMs in traumatic injury care, indicates major gaps in ongoing care, and underlines a future direction for specific initiatives to address these shortcomings and enhance patient outcomes.

Figure 2.

Knowledge, gaps, call to action.

Impact of Collaborative Care on Treatment Outcomes

Beginning with the domain of physical health outcomes, the findings present a nuanced picture. Vikane et al.’s program (26) exhibited positive results in post-concussive symptoms for TBI patients, while other programs did not demonstrate significant improvements in physical health parameters compared to usual care. Involvement with programs like TCC did boost the resumption of daily activities (27)

Results for mental health were more promising, although there was some nuance as well. Certain programs initially showed a reduction in PTSD symptoms, yet these effects were not sustained over time. Depression rates varied, with some studies showing reduced post-intervention rates, while others found no significant improvement. There was an increase in mental health visits and engagement after applying the intervention(1).

The observed lack of change in post-injury return-to-work rates across several treatment programs investigated is noteworthy, especially given that some programs promoted patient engagement and rehabilitation (28,32,33). A plausible rationale for this phenomenon might be derived from the research conducted by Finlay and colleagues, which brought attention to an unusual pattern among patients with severe burns (45). Some patients skip follow-up appointments even when they’re improving, hinting that current engagement and recovery metrics may be incomplete. They might feel fully recovered or “back to normal” and skip further care. This can skew how we view the success of rehabilitation programs due to missing data.

Necessary Components of CCMs for Traumatic Injury Survivors

Recovery from injury is a complex process that is impacted by the heterogeneity of the population, the injuries sustained, and the lack of any consensus on what constitutes recovery or even a “good outcome”. Table 4 presents a conceptual and idealized model for a post-injury recovery program. The levels of recommendation are subjectively derived from extensive literature reviews and cross-validation among our research team. These recommendations reflect the aggregate evidence and collaborative validation for optimal recovery pathways.

Table 4.

Recommended Components for Intervention and Their Subjective Levels of Recommendation in Designing the Ideal Collaborative Care Model

Characteristics		Strength of recommendation*
Setting	Comprehensive follow up clinic with same day multidisciplinary visit	2
Interventions	Symptom tracking Pain management Screening for risk and protective factors for poor outcomes Rehabilitation (physical therapy, occupational therapy) Goal setting Behavioral services (PTSD and depression screening, psychoeducation, cognitive behavioral therapy, motivational speech, pharmacotherapy) Medication and comorbidities management Referrals to clinical and community-based resources Evaluation of social determinants of health	3
Duration of intervention	6–12 months (booster sessions as needed)	1
Duration of follow up	At least 12 months	2
Modality	Hybrid: In-person and telehealth options	3
Staff	Mental health provider Advanced practice provider Surgeon Physiatrist Physical therapist Social worker Community health worker Patient navigator/Care coordinator	3
Additional Services	Transportation 24/7 clinic staff availability via phone Peer support Patient educational materials (handbooks/website)	2

^*Recommendation levels are determined subjectively based on our data review and have been validated by independent researchers. The levels are as follows: 3 - Highly Recommended, 2 - Medium Recommendation, 1 - Low Recommendation

At the heart of the approach is a comprehensive follow-up clinic that not only provides multidisciplinary visits but also emphasizes the importance of addressing social determinants in all aspects of patient care, marked by a recommendation level of 3.

Secondly, the model recognizes the marathon nature of recovery by including extended intervention and follow-up periods, as well as a hybrid of in-person and telemedicine modalities to keep patients engaged over time. Each team member—from patient navigators to surgeons—is essential to this ongoing endeavor. This is emphasized by strong support for a comprehensive team and additional services, such as transportation and 24-hour clinic access, to make the recovery road as smooth and supportive as feasible.

Looking ahead, the ideal model must consider patient-defined quality or healthcare coproduction, using user feedback and input to improve and refine services over time (46).

Barriers to CCMs Success for Traumatic Injury Survivors

We have identified three major barriers for these programs to achieve favorable outcomes: 1) Lack of sustained funding and support, 2) inconsistency in intervention implementation across institutions, and 3) low patient engagement.

Lack of time and dedicated funding were the largest barriers to intervention implementation, with another concern being the lack of institutional support (39,47). However, despite the scarcity of data on costs and cost-effectiveness, findings from the CTS and TQOL clinic indicated that implementing CCMs could result in financial advantages for healthcare institutions and the system. An in-depth assessment of the cost-effectiveness of these interventions will provide valuable insights for stakeholders, clinicians, and researchers and will be vital for their sustainability.

The use of CCM services varied notably across multicenter studies (28,30,32). In the largest included multi-institutional study (32), centers with good to excellent protocol implementation experienced significantly better PTSD outcomes. However, that was not the case for centers with poor implementation. Focusing on the rigor of implementation and monitoring implementation outcomes, such as feasibility, appropriateness, and acceptability, will be crucial for effectively translating the interventions into improved outcomes.

Dropout rates in studies varied widely, with some sites reporting less than 10% and others as high as 47%. Furthermore, there was a noticeable inconsistency in service utilization across various sites(28,30). Certain strategies, including case management, communication, and setting of delivery, have been shown to reduce the loss to follow-up rate in trauma centers to less than 7% (48). Early initiation of the CCMs and mental health interventions (either at the bedside or within 24–72 hours after the injury) can minimize loss of follow-up significantly (32,47).

Next Steps

Future research on traumatic injury CCMs should prioritize four key areas. Firstly, there is a need to define the target population requiring such care and develop a risk-stratification tool to identify individuals who would benefit most and determine the optimal duration of follow-up, which may vary on a risk-category level. Given that not every injured patient requires the intensity of care provided by these models, this will help guarantee the operability and sustainability of CCMs. Secondly, using mixed methods approaches and implementation science frameworks can provide valuable insights into the barriers and facilitators of CCM implementation. Addressing these factors is crucial for overcoming challenges associated with poor implementation and patient engagement, which have been linked to limited improvements in patient outcomes. Thirdly, conducting randomized controlled trials is essential to demonstrate the cost-effectiveness of traumatic injury CCMs, considering various healthcare utilization measures and outcomes across different health domains. Finally, when assessing long-term patient-reported outcomes, adherence to the recommendations of the NTRAP is crucial to standardize outcomes, enabling meaningful comparisons across different CCM models in diverse contexts. However, researchers should also explore novel methods for evaluating “successful injury recovery” that extend beyond traditional patient-reported outcome measures.

Limitations

We relied on published data, as no patient-level data was available for this study. This hindered our ability to conduct match analyses based on patient demographics. The heterogeneity of patient populations (TBI, orthopedic injuries, polytrauma) and outcome measures prevents us from providing a higher level of evidence via subgroup analysis and meta-analysis. Further, the diversity among the evaluated CCMs, each designed with distinct purposes and target populations, complicates direct comparisons between the interventions and their impact on different outcome domains. In addition, the lack of standardized outcome measures across all CCMs may impact the comparability and robustness of our results. Despite these limitations, this study contributes valuable insights into the diverse landscape of CCMs and their potential impact on patient outcomes.

CONCLUSION

If properly implemented, CCMs have the capacity to improve at least some health outcomes in the context of traumatic injury survivorship. Effective collaboration among healthcare providers can help ensure that injury survivors receive holistic care that addresses all their physical, emotional, and psychological needs. A consensus on cost-effective CCMs is necessary to set the standard for comprehensive care in traumatic injury recovery. Such models should consider patients’ risk and protective factors and enable healthcare providers to implement targeted interventions that improve long-term patient outcomes.