Implementation and Utility of an Automated Text Messaging System to Facilitate Symptom Self-Monitoring and Identify Risk for Post-Traumatic Stress Disorder and Depression in Trauma Center Patients

Abstract

Background and Introduction:

Comprehensive monitoring and follow-up after traumatic injury is important for psychological recovery. However, scalable services to facilitate this are limited. Automated text message-based symptom self-monitoring (SSM) may be a feasible approach. This study examined its implementation and utility in identifying patients at risk for mental health difficulties after traumatic injury.

Materials and Methods:

Five hundred two patients admitted to a Level I trauma center between June 20, 2016 and July 31, 2017 were offered enrollment in a text message-based SSM service. Patients who enrolled received daily text message prompts over 30 days and most participated in a mental health screening 30 days postbaseline.

Results:

Approximately 67% of patients enrolled in the service; of these, 58% responded to the text messages, with an average response rate of 53%. Younger patients and those with elevated peritraumatic distress were more likely to enroll. Patients with higher levels of mental health stigma, who were White, or had been in a motor vehicle collision were more likely to enroll and respond to text messages once enrolled. Patients' daily ratings of distress detected clinically elevated 30-day mental health screens with high sensitivity (83%) and specificity (70%).

Discussion and Conclusions:

Text message-based SSM can be implemented as a clinical service in Level I trauma centers, and patient participation may increase engagement in mental health follow-up. Further, it can inform the use of risk assessments in practice, which can be used to identify patients with poor psychological recovery who require additional screening.

Introduction

Nearly 2.5 million (9%) of the 27 million injuries resulting in emergency department visits each year are severe enough to require urgent medical attention and hospitalization.^1,2 Between 20% and 40% of these patients develop post-traumatic stress disorder (PTSD) and/or depression within a year postinjury.^3
–5 Increasing emphasis is being placed on monitoring and screening of patients' psychological recovery. For example, the American College of Surgeons (ACS) explicitly recommends center-based screening and referral for treatment of PTSD and depression.⁶ However, systematic follow-up on patients' psychological recovery by U.S. trauma centers is rare and resource intensive, requiring roughly 30 min for each follow-up assessment,^3,7

–11 which equates to a minimum of 36,000 h each year for Level I trauma centers with high patient volumes (i.e., ≥1,200 patients annually). As such, efficient, scalable, and sustainable solutions are needed to address the psychological needs of trauma center patients postdischarge.

It is common for individuals to experience distress immediately after a traumatic event, and this often decreases naturally over time.¹² Symptom self-monitoring (SSM) is used for a variety of clinical conditions, particularly when natural psychological recovery is common. A provider using SSM typically provides patients with a log or journal and asks them to track their symptoms after a prescribed schedule (e.g., daily, multiple times per day, after each symptom occurs). This provides patients with better insight into the course of their condition and their psychological responses to symptoms. This also enables providers to obtain valuable data on patients' day-to-day symptom experience.¹³ As such, SSM may be an ideal, scalable approach to track and assess the psychological recovery of trauma center patients and identify those at risk for mental health problems.

Mobile phone and text message-based SSM programs have been studied for a variety of clinical conditions (e.g., cancer, diabetes, anxiety, depression).^14

–20 Text message-based SSM offers a potentially cost-effective and low-burden solution for trauma centers to assess the psychological recovery of their patients. There are several advantages to using this approach. First, it has wide reach and accessibility, as 95% of the U.S. population now owns a cellphone.²¹ Second, it is consistent with behavioral-health intervention preferences of trauma center patients.²² Third, scalable programs can automate text message delivery, reducing trauma center burden, and can also provide an efficient and effective means by which to estimate risk for mental health problems.²³ This could be especially useful for identifying patients in need of additional follow-up mental health services.

To date, two small pilot studies have demonstrated the feasibility of implementing text message-based SSM with trauma center patients.^24,25 The first study included 29 adult trauma center patients who received automated daily text messages asking them to respond with ratings on their emotional functioning for 15 days after their hospital discharge. The overall participation rate was 83%, the average daily response rate was 63%, and patient satisfaction was high (i.e., 61% of participants found the messages to be helpful).²⁴ The second study included 19 adolescent trauma center patients and their caregivers who received daily symptom tracking text messages for 30 days after hospital discharge. Overall response rates were 95% for adolescents and 100% for caregivers, and the average daily response rates were 88% and 97% for adolescents and caregivers, respectively. The majority of adolescents (81%) and more than half of caregivers (58%) found the messages to be helpful.²⁵ Given the promising levels of engagement and response in these pilot studies, we sought to examine (1) the implementation of automated text message-based SSM as a clinical service in a Level I trauma center, (2) factors associated with patient engagement in the clinical service, and (3) the utility of daily ratings of distress provided by patients via text messages in predicting clinically elevated PTSD and depression at 30-day follow-up.

Materials and Methods

Procedure

Data were obtained from Level I trauma center patients enrolled from June 20, 2016 to July 31, 2017 in the Trauma Resilience and Recovery Program (TRRP) at the Medical University of South Carolina (MUSC).²⁶ TRRP is a stepped-care clinical service that includes: (1) in-hospital psychoeducation about psychological recovery after traumatic injury, (2) enrollment in an automated text message-based SSM program, (3) completion of a brief peritrauma questionnaire battery, (4) completion of a mental health screening 30 days postinjury, and (5) referral for mental health services for patients with clinically elevated symptoms based on 30-day screens. Potential patients were identified daily by using a customized trauma registry, which excludes patients hospitalized for reasons related to self-harm and sexual abuse, as patients meeting these criteria are seen by other clinical services within the hospital. Patients were visited by TRRP staff who enrolled them in TRRP and informed them about the option of participating in the automated text message-based SSM program, and obtained verbal consent to enroll them in the program. Patients were informed that the text message program used HIPAA-compliant software and data storage and that responses would not be actively monitored on a daily basis.

Patients who enrolled in the text message service received a welcome message that read, “Welcome to the MUSC TRRP 30-day mood monitoring program, which can help your emotional recovery,” and then received one text per day over a period of 30 days. Messages were sent at 12:00 PM each day, and one reminder text at 6:00 PM if a response was not received by that time. Each message asked patients to rate the extent to which they had experienced a certain emotion or behavior during the previous week. A total of 10 questions were used (see Measures section) and administered sequentially over 3, repeating 10-day blocks. After 30 days, patients received a message that read, “Thank you! We will contact you shortly for your 30-day follow-up phone call.” Patients were then contacted by TRRP staff ∼30 days postbaseline (Median_Days = 39) to administer the telephone screening.

Automated Text Message-Based SSM System

TRRPs text message-based SSM system was developed by MUSCs Technology Applications Center for Healthful Lifestyles. The system uses Twilio's Programmable Short Message Service API to automatically send and receive text messages to and from patients, whose data are stored on a secured, HIPAA-compliant SQL Server database. The cost of using this program is approximately $0.05 per text message. The system includes tools for customizable survey development and text message delivery scheduling, data entry fields for staff to enter limited patient information (i.e., ID, name, phone number), and service-wide and patient-specific database export functions.

Measures

Baseline

The Peritraumatic Distress Inventory (PDI)²⁷ is a 13-item self-report measure of the level of distress experienced during and shortly after a traumatic event. Items are rated on a 5-point Likert scale (0–4). Total scores range from 0 to 52 and are calculated as the sum of all item ratings, with a total score ≥23 indicating clinically elevated distress.^28,29 The PDI demonstrated strong internal consistency (Cronbach's α = 0.84). Similar to prior studies,^30,31 social support was measured by using five items from the Medical Outcomes Study Social Support Survey,³² which measured emotional, instrumental, and appraisal aspects of social support. Items are rated on a 4-point Likert scale (1–4). Total scores range from 5 to 20 and are calculated as the sum of all item ratings, with a higher total score indicating higher levels of social support. The measure demonstrated strong internal consistency (Cronbach's α = 0.84). The Self-Stigma of Seeking Help Scale (SSOSH)³³ is a 10-item self-report questionnaire that measures stigma toward seeking mental health services. Items are rated on a 5-point Likert scale (1–5). Total scores range from 10 to 50, with higher scores indicating higher levels of stigma. The SSOSH demonstrated adequate internal consistency (Cronbach's α = 0.67).

SSM questions

A total of 10 questions were used for the text message-based SSM service (see Appendix 1). These included six questions from the Kessler-6,^34,35 which assess feelings of nervousness, hopelessness, restlessness/fidgetiness, depression, worthlessness, and perceived effort burden. Four additional questions were developed for the service to assess worrying, avoidance, distress related to reminders of their traumatic injury, and sleep quality, as these domains are relevant to this patient population but not addressed by the Kessler-6. Each question asked patients to indicate how much of the time they experienced the emotion/behavior during the past week by responding to the text via a 5-point Likert scale (1 = All to 5 = None; reverse coded for the present analyses).

30-Day postbaseline mental health screening

The PTSD Checklist for DSM-5 (PCL-5)^36,37 is a 20-item self-report questionnaire that measures PTSD severity during the past month. Items are rated on a 5-point Likert scale (0–4). Total scores range from 0 to 80 and are calculated as the sum of all item ratings, with a total score ≥33 indicating probable PTSD.³⁸ The PCL-5 demonstrated strong internal consistency (Cronbach's α = 0.96). The Patient Health Questionnaire (PHQ-9)^39,40 is a 9-item self-report measure of depression severity during the past 2 weeks. Items are rated on a 4-point Likert scale (0–3). Total scores range from 0 to 27 and are calculated as the sum of all item ratings, with a total score ≥10 indicating probable depression.⁴⁰ The PHQ-9 demonstrated strong internal consistency (Cronbach's α = 0.91).

Data Analysis

Factors associated with enrollment and engagement in the text message-based SSM service (i.e., declining enrollment, enrolling but not responding, enrolling and responding) were examined by using Univariate Analyses of Variance for continuous variables and chi-square tests of Independence for categorical variables. Associations among study variables were examined by using bivariate correlation analyses, and binary logistic regression was used to predict engagement in follow-up based on the number of patient responses to text messages. Receiver operating characteristic (ROC) curve analyses were conducted to assess whether daily ratings of distress provided by patients via text messages adequately detected clinical elevations in mental health screens 30 days postbaseline, to examine the sensitivity and specificity of ratings in predicting clinical elevations, and to establish optimal cutoff scores, maximizing sensitivity and specificity, for predicting clinical elevations at this time-point. All analyses were conducted by using SPSS^® 24.0 (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp.).¹⁴

Results

Sample Characteristics and Associations with Engagement

Sample characteristics are presented in Table 1. A total of 502 patients ages 18 to 94 years were offered enrollment in the text message-based SSM service and were included in this study. Of these, 192 (38.3%) enrolled and provided at least one response, 143 (28.5%) enrolled but did not respond, and 167 (33.3%) declined enrollment. Among these three groups, there were significant differences in age [F(2, 499) = 33.11, p < 0.001, η ² = 0.117], peritraumatic distress [F(2, 383) = 6.92, p = 0.001, η ² = 0.035], and stigma toward seeking mental health services [F(2, 355) = 3.32, p = 0.037, η ² = 0.018]. Specifically, those who enrolled were significantly younger (p < 0.001) and had higher peritraumatic distress (p < 0.005) than those who declined enrollment. Among enrollees, those who responded to the text messages had significantly lower levels of stigma than those who did not respond (p = 0.039). There also were significant associations between group membership and race (χ ²[8, 502] = 26.75, p = 0.001, Φ = 0.231), ethnicity (χ ²[4, 502] = 12.31, p = 0.015, Φ = 0.157), and trauma type (χ ²[14, 502] = 36.78, p = 0.001, Φ = 0.271). Once enrolled, White patients were more likely to respond to the text messages compared with Black patients (p < 0.001). In comparison to patients with other injuries, patients injured in motor vehicle collisions were more likely to respond once enrolled; patients injured in a fall were less likely to enroll; and patients injured by a gunshot wound or stabbing were more likely to enroll, but not respond (p's < 0.05). The three groups did not differ significantly in levels of social support [F(2, 379) = 1.42, p = 0.243, η ² = 0.007], and group membership was not significantly associated with engagement in the 30-day follow-up (χ ²[2, 502] = 1.29, p = 0.524, Φ = 0.051) or sex (χ ²[2, 502] = 5.25, p = 0.073, Φ = 0.102).

Table 1.

Sample Characteristics

	ENROLLED RESPONDED, N = 192	ENROLLED, NO RESPONSE, N = 143	DECLINED ENROLLMENT, N = 167	OVERALL, SAMPLE, N = 502
VARIABLE	M (SD)	M (SD)	M (SD)	M (SD)
Age	42.57 (16.61)^a	39.00 (16.76)^a	55.13 (22.19)^b	45.73 (19.85)
Peritraumatic distress (PDI)	21.70 (12.54)^a	21.69 (11.56)^a	16.71 (12.01)^b	2.20 (12.27)
Social support	17.32 (3.90)	16.96 (4.03)	16.50 (3.89)	16.96 (3.94)
Stigma (SSOSH)	2.49 (5.04)^a	22.18 (5.42)^b	21.73 (6.04)	21.37 (5.50)
PTSD severity (PCL-5)	25.59 (22.45)	19.87 (2.33)	14.57 (18.99)	2.43 (21.23)
Depression severity (PHQ-9)	8.33 (7.58)	6.38 (6.21)	5.50 (7.04)	6.87 (7.14)
	n (%)	n (%)	n (%)	n (%)
Engaged in 30-day follow-up	83 (43.2)	53 (37.1)	68 (4.7)	204 (4.6)
Sex
Male	133 (69.3)	108 (75.5)	106 (53.5)	347 (69.1)
Female	59 (3.7)	35 (24.5)	61 (36.5)	155 (3.9)
Race
Black	65 (33.9)^a	73 (51.0)^a	44 (26.3)	182 (36.3)
White	113 (58.9)^b	60 (42.0)^b	113 (67.7)	286 (57.0)
Asian	1 (0.5)	0 (0.0)	0 (.0)	1 (0.2)
Other/bi-racial	12 (6.3)	10 (7.0)	8 (4.8)	30 (6.0)
Unknown	1 (0.5)	0 (0.0)	2 (1.2)	3 (0.6)
Ethnicity
Non-Hispanic	182 (94.8)	137 (95.8)	161 (96.4)	480 (95.6)
Hispanic	9 (4.7)	6 (4.2)	1 (0.6)	16 (3.2)
Unknown	1 (0.5)	0 (0.0)	5 (3.0)	6 (1.2)
Trauma type
Motor vehicle collision	95 (49.5)^a	54 (37.8)^b	64 (38.3)	213 (42.4)
Motorcycle collision	26 (13.5)	21 (14.7)	22 (13.2)	69 (13.7)
Fall	28 (14.6)^a	21 (14.7)^a	45 (26.9)^b	94 (18.7)
Gunshot wound/stabbing	20 (1.4)^b	31 (21.7)^a	12 (7.2)^b	63 (12.5)
Pedestrian vs. auto	10 (5.2)	9 (6.3)	13 (7.8)	32 (6.4)
Assault/abuse	1 (0.5)	1 (0.7)	5 (3.0)	7 (1.4)
Medical injury	0 (0.0)	1 (0.7)	0 (0.0)	1 (0.2)
Other	12 (6.3)	5 (3.5)	6 (3.6)	23 (4.6)

a,b

Significant group differences.

PCL-5, PTSD Checklist for DSM-5; PDI, Peritraumatic Distress Inventory; PHQ-9, Patient Health Questionnaire; PTSD, post-traumatic stress disorder; SSOSH, Self-Stigma of Seeking Help Scale.

Response Patterns

Of the 335 (66.7%) patients who enrolled in the service, 206 (57.3%) responded to at least 1 message, 98 (29.3%) responded to at least 15 messages, and 24 (7.2%) responded to all 30 messages. The average number of responses was 15.86 (SD = 1.89; 52.9% response rate) across the 30-day administration, 5.66 (SD = 3.77; 56.6% response rate) during the first 10-day block, 5.33 (SD = 4.18; 53.3% response rate) during the second 10-day block, and 4.88 (SD = 4.43; 48.8% response rate) during the third 10-day block.

The average daily rating in response to SSM items was 2.55 (i.e., A Little to Some; SD = 1.05) for all questions across the 30-day administration, 2.60 (SD = 1.01) during the first 10-day block, 2.56 (SD = 1.08) during the second 10-day block, and 2.63 (SD = 1.18) during the third 10-day block.

Associations between Study Variables and Response Patterns

Correlations among study variables are presented in Table 2, and significant correlations are discussed here. Higher social support at baseline was associated with lower levels of daily distress via text messages. Higher peritraumatic distress was associated with higher levels of daily distress via text messages. Higher levels of stigma were associated with lower likelihood of responding to text messages. Daily distress ratings via text messages were positively correlated with levels of PTSD and depression severity at 30-day follow-up. Finally, patients who responded to the text messages more often were more likely to engage in the 30-day follow-up. The logistic regression predicting engagement in 30-day follow-up based on the number of responses was significant (β = 0.02, SE = 0.01, Wald = 4.28, p = 0.038, OR = 1.02, 95% CI [1.00–1.04], R ² _Nag = 0.017).

Table 2.

Correlations Among Study Variables

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1 Age	—
2 Peritraumatic distress	−0.19^*	—
3 Social support	−0.17^*	−0.06	—
4 Stigma	−0.01	0.05	−0.17^*	—
5 PTSD severity	−0.22^*	0.53^*	−0.18^*	0.03	—
6 Depression severity	−0.16^*	0.44^*	−0.15	0.00	0.90^*	—
7 Engaged in follow-up	−0.05	−0.01	−0.04	−0.04	^a	^a	—
8 Number of responses	0.09	−0.09	0.04	−0.20^*	0.06	0.12	0.11^*	—
9 Number of responses B1	−0.10^*	0.03	0.07	−0.15^*	0.14	0.15^*	0.04	0.90^*	—
10 Number of responses B2	−0.07	0.00	0.07	−0.16^*	0.10	0.12	0.09	0.97^*	0.86^*	—
11 Number of responses B3	−0.06	−0.01	0.05	−0.16^*	0.14^*	0.15^*	0.11^*	0.92^*	0.74^*	0.89^*	—
12 Average rating	−0.22^*	0.49^*	−0.23^*	0.13	0.59^*	0.54^*	0.10	0.00	0.06	0.00	−0.04	—
13 Average rating B1	−0.19^*	0.46^*	−0.19^*	0.15	0.45^*	0.40^*	0.17^*	0.03	0.04	0.04	0.01	0.91^*	—
14 Average rating B2	−0.15	0.42^*	−0.24^*	0.07	0.59^*	0.62^*	−0.03	−0.08	0.07	−0.12	−0.15	0.93^*	0.72^*	—
15 Average rating B3	−0.18^*	0.48^*	−0.27^*	0.16	0.64^*	0.56^*	0.08	−0.16	−0.09	−0.13	−0.20^*	0.92^*	0.70^*	0.81^*	—

p < 0.05.

Cannot be computed because at least one of the variables is constant.

B, 10-day block number.

Predicting Clinical Elevations at 30-Day Follow-Up based on SMS Ratings

The results of the ROC curve analysis predicting clinically elevated mental health screens at 30-day follow-up from daily ratings of distress provided via text messages was significant (p < 0.001). Examination of the coordinate points of the ROC curve suggested a cutoff value of 2.6 to optimize sensitivity (83%) and specificity (70%) in predicting clinically elevated screens. Given variability in responding to text messages over the full 30-day period, similar analyses were conducted for average daily ratings of distress from each 10-day block to examine the utility of using these values in predicting clinically elevated mental health screens. The results of these analyses were significant, with increased specificity and sensitivity for daily ratings occurring closer to the end of the 30-day period. The results of the ROC curve analyses are presented in Table 3.

Table 3.

Receiver Operating Characteristic Curve Analysis Results

	AUC (95% CI)	SE	SIGNIFICANCE	CUTOFF SCORE	SENSITIVITY, %	SPECIFICITY, %
Average daily rating of distress	0.801 (0.708–0.895)	0.048	<0.001	2.61	83	70
Average daily rating of distress B1	0.753 (0.643–0.862)	0.056	<0.001	2.68	72	64
Average daily rating of distress B2	0.778 (0.669–0.887)	0.056	<0.001	2.45	72	71
Average daily rating of distress B3	0.838 (0.741–0.935)	0.049	<0.001	2.55	76	74

B, 10-day block number.

Discussion and Conclusions

This study was the first to examine the implementation of text message-based SSM as a clinical service for traumatically injured patients admitted to a Level I trauma center. The results replicated and extended those found in previous pilot and feasibility studies with adult and adolescent trauma center patients.^24,25 The enrollment rate for this clinical sample was high (67%) and the response rate was moderate among the 38% of enrollees (53%). Unsurprisingly, the rates were lower than those reported in prior studies using mobile-based SSM in various clinical populations and settings where patients are formally recruited for this purpose and typically incentivized.^{15,17,19,20,24,25} Future work should examine ways to increase participation in text message-based SSM clinical services, particularly as the clinical benefits of these services are better understood.

Several patient characteristics were associated with enrollment and engagement in the service. Patients who were younger were more likely to enroll in the service. This was to be expected as, although increasing, technology adoption and confidence is still relatively poor among older adults.⁴¹ Patients with elevated peritraumatic distress also were more likely to enroll in the service. This was consistent with previous findings,²⁴ and it is understandable that patients experiencing more immediate symptoms might be more likely to find relevant and take advantage of assistance during the peritrauma time-period. Another notable finding was that patients with lower levels of stigma were more likely to respond to text messages once enrolled and responded more often throughout the course of the 30-day period. This is consistent with the literature suggesting that patients with elevated stigma toward seeking mental health services are less likely to engage in such services.⁴² White patients who were enrolled in the service were more likely to respond to text messages compared with Black patients, consistent with literature suggesting lower treatment utilization in Black patients with significant mental health symptoms.^43
–45 Further, patients injured in a fall were less likely to enroll in the service, and patients injured in motor vehicle collisions were more likely to respond to text messages once enrolled. The fact that older adults are more likely to be injured in a fall,⁴⁶ and that older adults are less likely to engage in the text messaging service, may explain this finding, in part. Moreover, patients injured in motor vehicle collisions are likely to experience high levels of functional impairment related to their injury and PTSD symptoms,⁴⁷ which may have played a role in their increased engagement in the service. A novel finding was that the number of responses to text message prompts predicted participation in follow-up 30 days postbaseline. This suggests the importance of continuity of communication during this time-period with respect to patients' engagement in future mental health screening.

A major advantage to using text message-based SSM is its propensity to facilitate the collection of clinically useful patient data.²⁴ This study demonstrated the utility of patients' daily ratings of distress in response to text messages to predict clinically elevated 30-day mental health screens. Further, this utility extends to situations where patients exhibit variable responding during the administration period. Specifically, average ratings of distress from the first, second, and third 10-day administration blocks effectively identified patients with clinically elevated 30-day mental health screens. This finding is important because some patients are unlikely to respond to all text message prompts, and some may be more likely to respond at different times during administration (e.g., during the first 10 days vs. the last 10 days).

This study has limitations that provide areas for future research. First, outcomes were measured by using self-report questionnaires, limiting the ability to make firm conclusions about specific mental health diagnoses at follow-up. Second, although the sample used in this study was diverse, clinical data were obtained from traumatically injured patients presenting for treatment at a trauma center. Replication of these results with survivors of other types of traumas (e.g., natural disasters) is warranted to better understand their generalizability. Third, although the Kessler-6 is a well-validated and widely used questionnaire, four additional items were added to assess domains relevant to this patient population. Further, SSM is prone to bias and inconsistency in response rates and ratings. Thus, the reliability and validity of the SSM questions used in this study are limited. Fourth, the potential clinical benefits of the SSM service were not assessed because group membership was not randomized and was highly dependent on patient preferences and characteristics. Several studies have shown positive effects of SSM on PTSD and depressive symptom severity.^{17,20,48
–50} Thus, future work should examine patient outcomes after administration of a similar program using a randomized clinical trial design. Fifth, the SSM program was not actively monitored on a daily basis, which would have provided a unique opportunity to monitor patient risk in real-time while potentially enhancing patient interaction and satisfaction with the program. Finally, reasons for declining enrollment in the text message-based SSM program were not collected from patients given the clinical nature of the program, which would have provided valuable data toward improving participation in future versions of the program.

Despite limitations, the results of this study are novel and should be of use to providers and researchers in trauma care settings. In addition to providing a cost-effective and low-burden approach to monitoring patients' psychological recovery postinjury, automated text message SSM may increase the likelihood of engagement in follow-up mental health services such as screening, possibly due to continuity of communication. In addition, this approach can be particularly useful in identifying traumatically injured patients whose psychological recovery is poor, and who will likely require additional mental health screening, in compliance with ACS recommendations. Moreover, because text-messaging systems can be automated with low maintenance costs, this approach has the potential to be efficiently integrated into stepped-care models to monitor and facilitate psychological recovery after traumatic injury.

Footnotes

Acknowledgments

This work was supported by funding from the Medical University of South Carolina (MUSC) Center for Telehealth and the South Carolina Telehealth Alliance. Dr. Bunnell was supported by the National Institute of Mental Health (grant number F32 MH108250). J.D. was supported by the National Institute on Drug Abuse (grant number K23 DA045766). The authors wish to acknowledge the MUSC Department of Surgery's support of the Trauma Resilience and Recovery Program (TRRP). The authors also wish to acknowledge TRRP staff for their tireless efforts toward improving the emotional recovery of their patients.

Disclosure Statement

No competing financial interests exist.

Appendix A1. SSM Questions

References

Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Web-based Injury Statistics Query and Reporting System (WISQARS) nonfatal injury data. 2016. Available at: www.cdc.gov/injury/wisqars/overview/key_data.html (last accessed June 18, 2018 ).

Florence

, Haegerich

, Simon

, Zhou

, Luo

. Estimated lifetime medical and work-loss costs of emergency department-treated nonfatal injuries-United States, 2013. MMWR Morb Mortal Wkly Rep, 2015; 64:1078–1082.

Fakhry

, Ferguson

, Olsen

, Haughney

, Resnick

, Ruggiero

. Continuing trauma: The unmet needs of trauma patients in the postacute care setting. Am Surg, 2017; 83:1308–1314.

O'Donnell

, Creamer

, Pattison

. Posttraumatic stress disorder and depression following trauma: Understanding comorbidity. Am J Psychiatry, 2004; 161:1390–1396.

Zatzick

, Kang

S-M

, Müller

H-G

, et al. Predicting posttraumatic distress in hospitalized trauma survivors with acute injuries. Am J Psychiatry, 2002; 159:941–946.

American College of Surgeons Committee on Trauma. Resources for optimal care of the injured patient. Chicago, IL: American College of Surgeons, 2014.

O'Donnell

, Creamer

, Elliott

, Atkin

, Kossmann

. Determinants of quality of life and role-related disability after injury: Impact of acute psychological responses. J Trauma, 2005; 59:1328–1334.

Van horn

. Loss and depression after traumatic injury: The importance of case management in the recovery process. Prof Case Manag, 2009; 14:66–73.

Wiseman

, Foster

, Curtis

. Mental health following traumatic physical injury: An integrative literature review. Injury, 2013; 44:1383–1390.

10.

Love

, Zatzick

. Screening and intervention for comorbid substance disorders, PTSD, depression, and suicide: A trauma center survey. Psychiatr Serv, 2014; 65:918–923.

11.

Gabbe

, Sutherland

, Hart

, Cameron

. Population-based capture of long-term functional and quality of life outcomes after major trauma: The experiences of the Victorian State Trauma Registry. J Trauma, 2010; 69:532–536.

12.

Rothbaum

, Foa

, Riggs

, Murdock

, Walsh

. A prospective examination of post-traumatic stress disorder in rape victims. J Trauma Stress, 1992; 5:455–475.

13.

Bakker

, Kazantzis

, Rickwood

, Rickard

. Mental health smartphone apps: Review and evidence-based recommendations for future developments. JMIR Ment Health, 2016; 3:e7.

14.

Wei

, Hollin

, Kachnowski

. A review of the use of mobile phone text messaging in clinical and healthy behaviour interventions. J Telemed Telecare, 2011; 17:41–48.

15.

Faurholt-Jepsen

, Vinberg

, Christensen

, Frost

, Bardram

, Kessing

. Daily electronic self-monitoring of subjective and objective symptoms in bipolar disorder—The MONARCA trial protocol (MONitoring, treAtment and pRediCtion of bipolAr disorder episodes): A randomised controlled single-blind trial. BMJ Open, 2013; 3:e003353.

16.

Donker

, Petrie

, Proudfoot

, Clarke

, Birch

M-R

, Christensen

. Smartphones for smarter delivery of mental health programs: A systematic review. JMIR, 2013; 15:e247.

17.

Dewey

, McDonald

, Brown

, Boyd

, Bunnell

, Schuldberg

. The impact of ecological momentary assessment on posttraumatic stress symptom trajectory. Psychiatry Res, 2015; 230:300–303.

18.

Morris

, Kathawala

, Leen

, et al. Mobile therapy: Case study evaluations of a cell phone application for emotional self-awareness. JMIR, 2010; 12:e10.

19.

Reid

, Kauer

, Hearps

SJC

, et al. A mobile phone application for the assessment and management of youth mental health problems in primary care: Health service outcomes from a randomised controlled trial of mobiletype. BMC Fam Pract, 2013; 14:84.

20.

Kauer

, Reid

, Crooke

AHD

, et al. Self-monitoring using mobile phones in the early stages of adolescent depression: Randomized controlled trial. JMIR, 2012; 14:e67.

21.

PewResearchCenter. Demographics of mobile device ownership and adoption in the United States. Internet & Technology Mobile Fact Sheet 2017. Available at: www.pewinternet.org/fact-sheet/mobile (last accessed June 18, 2018 ).

22.

Ranney

, Choo

, Wang

, Baum

, Clark

, Mello

. Emergency department patients' preferences for technology-based behavioral interventions. Ann Emerg Med, 2012; 60:218–227.e48.

23.

Cole-Lewis

, Kershaw

. Text messaging as a tool for behavior change in disease prevention and management. Epidemiol Rev, 2010; 32:56–69.

24.

Price

, Ruggiero

, Ferguson

, et al. A feasibility pilot study on the use of text messages to track PTSD symptoms after a traumatic injury. Gen Hosp Psychiatry, 2014; 36:249–254.

25.

Davidson

, Bunnell

, Ruggiero

. An automated text-messaging system to monitor emotional recovery after pediatric injury: Pilot feasibility study. Psychiatr Serv, 2017; 68:859–860.

26.

Ruggiero

, Davidson

, Bunnell

, Maples-Keller

, Fakhry

. Trauma Resilience and Recovery Program: A stepped care model to facilitate recovery after traumatic injury. Int J Emerg Ment Health, 2016; 18:25.

27.

Brunet

, Weiss

, Metzler

, et al. The Peritraumatic Distress Inventory: A proposed measure of PTSD criterion A2. Am J Psychiatry, 2001; 158:1480–1485.

28.

Nishi

, Matsuoka

, Yonemoto

, Noguchi

, Kim

, Kanba

. Peritraumatic Distress Inventory as a predictor of post-traumatic stress disorder after a severe motor vehicle accident. Psychiatry Clin Neurosci, 2010; 64:149–156.

29.

Bunnell

, Davidson

, Ruggiero

. The Peritraumatic Distress Inventory: Factor structure and predictive validity in traumatically injured patients admitted through a Level I trauma center. J Anxiety Disord, 2018; 55:8–13.

30.

Galea

, Ahern

, Resnick

, et al. Psychological sequelae of the September 11 terrorist attacks in New York City. N Engl J Med, 2002; 346:982–987.

31.

Acierno

, Ruggiero

, Galea

, et al. Psychological sequelae resulting from the 2004 Florida hurricanes: Implications for postdisaster intervention. Am J Public Health, 2007; 97(Suppl 1):S103–S108.

32.

Sherbourne

, Stewart

. The MOS social support survey. Soc Sci Med, 1991; 32:705–714.

33.

Vogel

, Wade

, Haake

. Measuring the self-stigma associated with seeking psychological help. J Couns Psychol, 2006; 53:325–337.

34.

Kessler

, Andrews

, Colpe

, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med, 2002; 32:959–976.

35.

Kessler

, Barker

, Colpe

, et al. Screening for serious mental illness in the general population. Arch Gen Psychiatry, 2003; 60:184–189.

36.

Blevins

, Weathers

, Davis

, Witte

, Domino

. The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and initial psychometric evaluation. J Trauma Stress, 2015; 28:489–498.

37.

Weathers

, Litz

, Keane

, Palmieri

, Marx

, Schnurr

. The PTSD Checklist for DSM-5 (PCL-5), 2013. Available at: www.ptsd.va.gov (last accessed June 18, 2018 ).

38.

Bovin

, Marx

, Weathers

, et al. Psychometric properties of the PTSD Checklist for Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition (PCL-5) in veterans. Psychol Assess, 2016; 28:1379–1391.

39.

Spitzer

, Kroenke

, Williams

. Validation and utility of a self-report version of PRIME-MD: The PHQ primary care study. Primary care evaluation of mental disorders. Patient Health Questionnaire. JAMA , 1999; 282:1737–1744.

40.

Kroenke

, Spitzer

, Williams

. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med, 2001; 16:606–613.

41.

Anderson

, Perrin

. Tech Adoption Climbs Among Older Adults. Pew Research Center. 2017. Available at: www.pewinternet.org/2017/05/17/tech-adoption-climbs-among-older-adults (last accessed June 18, 2018 ).

42.

Corrigan

. How stigma interferes with mental health care. Am Psychol, 2004; 59:614–625.

43.

Nobles

, Valentine

, Zepeda

, Ahles

, Shtasel

, Marques

. Usual course of treatment and predictors of treatment utilization for patients with posttraumatic stress disorder. J Clin Psychiatry, 2017; 78:e559–e566.

44.

Spoont

, Nelson

, Murdoch

, et al. Are there racial/ethnic disparities in VA PTSD treatment retention?. Depress Anxiety, 2015; 32:415–425.

45.

Hebenstreit

, Madden

, Koo

, Maguen

. Minimally adequate mental health care and latent classes of PTSD symptoms in female Iraq and Afghanistan veterans. Psychiatry Res, 2015; 230:90–95.

46.

O'loughlin

, Robitaille

, Boivin

, Suissa

. Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. Am J Epidemiol, 1993; 137:342–354.

47.

Beck

, Coffey

. Assessment and treatment of PTSD after a motor vehicle collision: Empirical findings and clinical observations. Prof Psychol Res Pr, 2007; 38:629–639.

48.

Tarrier

, Sommerfield

, Reynolds

, Pilgrim

. Symptom self-monitoring in the treatment of posttraumatic stress disorder. Behav Ther, 1999; 30:597–605.

49.

Ehlers

, Clark

, Hackmann

, et al. A randomized controlled trial of cognitive therapy, a self-help booklet, and repeated assessments as early interventions for posttraumatic stress disorder. Arch Gen Psychiatry, 2003; 60:1024–1032.

50.

Foa

, Zoellner

, Feeny

. An evaluation of three brief programs for facilitating recovery after assault. J Trauma Stress, 2006; 19:29–43.