Evaluation of Psychosocial Factors Influencing Recovery From Adult Orthopaedic Trauma
This guideline was produced in collaboration with METRC, with funding provided by the US Department of Defense. Endorsed by: OTA, SCCM, AAPM&R
Factors Associated with Patient Outcomes
It is recommended that clinicians evaluate the following factors, as they are associated with increased biopsychosocial limitations after adult orthopaedic trauma: • Anxiety • PTSD • Depression • Premorbid psychiatric conditions • Smoking • Lower education level • Less social support • Resilience Issues (i.e. Limited self-efficacy, less effective coping strategies)
Moderate Recommendation Moderate Recommendation

This recommendation was derived from data regarding the association between psychosocial factors and patient outcomes. Clinicians should actively address the presence of these factors appropriately. However, this guideline did not evaluate effective treatment strategies for psychosocial factors.

Studies meeting criteria for this analysis which indicate anxiety is a factor associated with worsened biopsychosocial outcomes in orthopaedic trauma are few, with just one recent high quality retrospective observational study of 601 patients (Castillo, 2013) indicating that increased anxiety at six and twelve-months post-injury is associated with increased anxiety and pain at 18 and 24 months. Additional low quality studies and two moderate quality studies (Bosma, 2004, O’Toole, 2008) also support this association. Research studies have used the Brief Symptom Inventory (BSI) Anxiety Scale (Anxiety) and the Hospital Anxiety and Depression Scale (HADS) to screen and diagnose anxiety respectively.

One high quality study (Castillo, 2013) found that increased depression at 6 months leads to increased depression at 12 months, and increased depression at 12 months leads to increased depression at 24 months. Increased scores on the Brief Symptom Inventory (BSI) Depression Scale are associated with decreased functional outcomes (Wegener, 2011). Two moderate quality studies (Papadakaki, 2017, Bosma, 2004) suggested depressive symptoms at baseline with the Center for Epidemiologic Studies Depression Scale (CES – D) greater than or equal to 16 are associated with higher odds of depression at 6 months and with post-treatment depression at 1 year; both studies showed associated negative patient outcomes. Two moderate quality studies (Hou, 2013, O’Toole, 2008) suggested higher Brief Symptom Rating Scale (BSRS-5) scores are associated with lower EQ-5D Quality of Life (QOL) and decreased odds of satisfaction; both were associated with negative patient outcomes. However, two moderate quality studies (Nota, 2015, Rivara, 2008) also found depression measured by CES - D and depression before injury was found to be significantly associated with patient outcomes in bivariate analysis. Five low quality studies (Wegner, 2011, Schweininger, 2015 [HADS depression at 3 months and worse depression at 12 months]; Zatzick, 2007, Zatzick, 2008, Archer, 2015) found depression by a variety of measures is associated with negative patient outcomes, and there were no studies that found depression to be associated with positive outcomes.

Three low quality (Scweininger, 2015, Zatrzick, 2010, Liedl, 2010) and one moderate quality study (Papadakaki, 2017) found a significant association between PTSD and negative patient outcomes including pain, function, anxiety, depression, mental health, and return to activity/work. One study (Schweininger, 2015) showed a significant relationship with PTSD 3 months after the traumatic injury to be related to negative outcomes 12 months after injury. Only one moderate quality study (Nota, 2015) of 130 patients found a significant relationship between PTSD and patient outcomes in bivariate analyses, and there were no studies showing an association of PTSD to positive outcomes.

Pre-Morbid Psychiatric Conditions
In the literature there are existing studies examining the relationship between pre-morbid psychiatric conditions and negative patient outcomes in adult orthopaedic trauma. In these studies, pre-morbid psychiatric conditions included either a specific condition, including PTSD, or the presence of any psychiatric medical comorbidity. In a low quality study by Zatzick (2007), pre-injury psychiatric diagnoses abstracted from the medical chart were adjusted for in the statistical analysis (including alcohol/substance use, depression, etc.). PTSD and depression post-injury remained independently associated with elevated odds of impairment in activity of daily living (ADL or IADL), reduced physical and mental health, and lost productivity. In another low quality study by Shields (2015), the presence of any pre-injury psychiatric history was associated with lower odds of having a satisfactory Physical Components Summary Score and Mental Components Summary Score as measured by the SF-12. Additionally, the study found any pre-morbid psychiatric history was associated with lower odds of a satisfactory Simple Shoulder Test. However, the association between pre-injury psychiatric diagnosis and a satisfactory Disabilities of the Arm, Shoulder, and Hand (DASH) Score was not significant.

One low quality study examined pre-injury psychiatric diagnosis in a military population of 772 individuals and found the presence of a pre-injury psychiatric diagnosis is associated with higher odds of developing PTSD as well as higher odds of substance abuse (Melcer, 2013)54.

This assessment included three low quality (MacKenzie, 2005, Bosse, 2002, Castillo, 2011) and two moderate quality (MacKenzie, 2004/2006) articles that found a significant relationship between smoking and negative patient outcome including sickness impact profile (SIP), function, mental health, and return to work. The studies assessed smoking as currently smoking compared to non-smoker as well as continuously and categorically for number of cigarettes smoked. There was one low quality study of 154 patients (Shields, 2015) that did not find an association between smokers and patient outcomes. However, there were no studies that found smoking to be related to positive patient outcomes.

Lower Education Level
One high quality (Hou, 2012), six moderate quality (Bosma Hans, 2004, Holtslag, 2007, Kugelman, 2018, MacDermid, 2002, MacKenzie, 2004/2006) and nine low quality (Archer, 2015, Bosse, 2002, Castillo, 2011, Clay, 2010, MacKenzie, 1998/2005, Pezzin, 2000, Soberg, 2007/2012) articles found a significant association between higher education levels and improved patient outcomes. Outcomes included pain, quality of life, return to activity/work, mental health, function, anxiety, and overall SIP score. There were an additional five moderate quality (Hou, 2013, Nota, 2015, O'Toole, 2008, Papadakaki, 2017, Walsh, 2010) and six low quality (Andrew, 2008/2012, Hou, 2008, Ouellet, 2009, Ponsford, 2008, Soberg, 2011) articles that found no significant relationship. However, a majority of the articles favored higher education for improved outcomes, and none of the articles favored lower education levels for positive outcomes. Though the included literature was not entirely consistent in favor of higher education, there were no studies that showed lower education levels to be related to positive outcomes.

Less Social Support
A high quality study (Hou, 2012) compared return to work outcomes in married versus single, divorced or widowed patients. Return to work time was found to be slow in single, divorced or widowed patients and average to fast in married patients. One moderate strength study (Papadakaki, 2017) found divorced and widowed patients had higher odds of depression compared to single patients at six months.

There were also low quality studies (Bosse, 2002, Castillo, 2011) demonstrating an association of lower overall SIP scores and higher social support scores. Additionally, a low quality study (Soberg, 2012) found better SF-36 scores at the five-year mark for those with higher levels of societal participation, and another study (Ouellet, 2009) found greater social support decreases the risk of poor mental health in trauma patients. Compared to patients with low social functioning, those with higher social functioning had less time off from work according to a low quality study (Clay, 2010), and another study (Soberg, 2012) found higher social functioning is associated with higher probability of return to work.

Resilience Issues (i.e. Limited Self-Efficacy, Less Effective Coping Strategies)
Two moderate (Hou, 2013, Schnyder, 2001b) and four low quality (Ni, 2013, Soberg, 2010/2012, Tuncay, 2015) studies assessing varying coping strategies found positive patient outcomes significantly associated with more effective coping. Outcomes with significant association included quality of life, function, PTSD, mental health, and varying levels of post-traumatic growth.

Six moderate quality (Walsh, 2016, MacKenzie, 2006, Rusch, 2003, Bosma, 2004, Schnyder, 2001b, Schnyder, 2003) and ten low quality (Hou, 2008, Vranceanu, 2014, Clay, 2010, Steven, 2010, MacKenzie, 2004/2005, Archer, 2015, Bosse, 2002, Castillo, 2011, Bot, 2011) studies showed significant associations between negative patients outcomes and negative personal view (catastrophic thinking, low self-efficacy, low resiliency, etc.). Two additional studies (Hou, 2013, O'Toole, 2008) found significant associations at the bivariate level between personal view and patient outcomes, and no studies were found to favor negative personal view for positive patient outcomes.

Other Considerations:
Note the following factors may be associated with greater biopsychosocial symptom intensity, magnitude of limitations, and/or diminished health related quality of life:

  • Age
  • BMI
  • Race
  • Gender
  • Low income
  • Lack of employment
  • Comorbidities
  • Pre-injury exposure to combat related circumstances

Several low quality studies have demonstrated that increasing age at the time of injury leads to: higher disability (Mackenzie, 2005), lower SF-12 physical scores (Andrew, 2008), higher pain (Ponsford, 2008), and lower return to work rates (MacKenzie, 2006, Pezzin, 2000). One moderate quality study (MacKenzie, 2004) demonstrated better SIP scores in amputation patients older than 55 years old.

There is moderate evidence (Walsh, 2010) that increasing BMI in musculoskeletal injuries is related to increased pain. Race and gender may also be associated with greater biopsychosocial symptom intensity, magnitude of limitations, and/or diminished health related quality of life. There are abundant data that race disparities exist in the care of musculoskeletal injuries, however, no objective data exist to support a recommendation with regard to race or gender.
Two moderate quality studies (Kugelman, 2018, MacKenzie, 2004) and three low quality studies (Soberg, 2012, Ouellet, 2009, MacKenzie, 2006) failed to demonstrate a significant relationship between employment type (blue collar vs white collar), employment status (employed vs unemployed), or income at or below poverty level with physical or mental health symptoms after injury. One low quality study (Bosse, 2002) suggested that income at/below poverty level predicted lower % change in SIP score, and another low quality study (Hebert, 2006), demonstrated that pre-injury income of >$75,000 was associated with higher likelihood of return to work.

Three moderate quality studies (Holtslag, 2007, Abraham, 2014, MacKenzie, 2014) demonstrated that co-morbidity is an independent predictor of long-term functional consequences including higher overall SIP scores.

In military and veteran populations, one low quality study (Gunardena, 2007) demonstrated that war/combat exposures (i.e., being shot at, being threatened with arms, or witnessing war-related violence) predicted greater psychological distress after physical trauma.

Benefits/Harms of Implementation
There appears to be low risk of harm in evaluating the presence or absence of psychosocial factors. Support for how best to screen/evaluate for these factors and their effects is limited and requires further study.

Future Research
Current evidence regarding mental and social health influences on recovery from injury consists largely of correlations and associations with a few preliminary studies of treatment interventions. As documented in this report, there is consistent, compelling, and increasing evidence that mental and social health are associated with symptom intensity and magnitude of limitations after adult orthopedic trauma injuries. Studies of general traumatic injuries do not always stratify by the specific orthopaedic population, which leaves gaps in the evidence for future research to address.

Next steps and areas for additional investigation include:

1. The optimal strategies for identifying mental and social health opportunities

a. The following may have a role:

i. Questionnaires
ii. Monitoring for verbal and non-verbal signs
iii. Identification of greater symptoms and limitations than expected for a given injury and stage of recovery as measured on patient reported outcome measures (PROMs)
iv. Formal interviews.

b. What is the optimal timing and frequency of screening?
c. When is it appropriate to stop screening?

2. The relative benefits and harms of routine screening and tailored treatment
3. The effectiveness of various interventions
4. Economic analysis of the actual or potential benefits of screening and treatment with respect to optimal stewardship of resources
5. Assessment of the relationship between barriers to screening (condition, access, EHR, referral methods, etc.) and their impact on a patient’s ability to  respond to evaluation, seek treatment, or recover from injury
6. Determining to whom/when specialty referral is needed for trauma-associated factors

Barriers to investigation in this area include the following:

1. Altered consciousness or cognitive capacity.

2. Potential for problems arising from screening with no access or delayed access to treatment resources, or inadequate quality/training for mental and social health aspects of recovery from adult orthopedic trauma.

3. Ethical issues associated with no screening given that cognitive, emotional, and social aspects of recovery are to be expected based on human illness behavior.

4. Ethical issues associated with no treatment given the evidence that cognitive behavioral therapy and its derivatives (as well as psychotherapy and medication) are effective at alleviating stress and psychological distress and fostering optimal cognitive coping strategies.


Guideline Work Group:

  • Stephen Wegener, MA, PhD, Non-Military Co-Chair
  • Benjamin Keizer, PhD, Military Co-Chair
  • Erik Ensrud, MD
  • David Benedek, MD
  • Ann Marie Warren, PhD, ABPP 
  • Todd A. Swenning, MD. FAAOS
  • David C. Ring, MD, FAAOS
  • Kelly L. Cozza, MD
  • Wade T. Gordon, MD
  • Saloni Sharma, MD
  • Peggy L. Naas, MD, MBA, FAAOS

AAOS Oversight Chairs

  • Atul F. Kamath, MD, FAAOS
  • Julie B. Samora, MD, MPH, PhD, FAAOS


  • Jayson Murray, MA, Director, Department of Clinical Quality and Value
  • Kyle Mullen, MPH, Manager, Department of Clinical Quality and Value
  • Danielle Schulte, MS, Manager, Department of Clinical Quality and Value
  • Kaitlyn S. Sevarino, MBA, CAE, Senior Manager, Department of Clinical Quality and Value
  • Barbara Krause, Qualilty Improvement Specialist, Department of Clinical Qualilty and Value
  • Connor Riley, MPH, Research Analysty,Department of Clinical Quality and Value
  • Jennifer Rodriguez, Quality Development Assistant, Department of Clinical Quality and Value
  • Anne Woznica, MLIS, AHIP, Medical Research Librarian, Department of Clinical Quality and Value
  • Mary Demars, Quality and Value Coordinator, Department of Clinical Quality and Value
  • Ellen J. MacKenzie, PhD, METRC

Additional Contributor:

  • Mara Schenker, MD