Identifying Populations at Risk for Difficulty Managing Emotions and Social Interactions After Burn Injury

Abstract

There has been conflicting data on the relationship between burn severity and psychological outcomes. The present study aims to characterize the baseline psychosocial disposition of adults attending outpatient burn clinic at a large urban safety net hospital, as well as the impact of clinical course on self-reported psychosocial well-being. Adult patients attending outpatient burn clinic completed survey questions from the National Institutes of Health Patient-Reported Outcomes Measurement Information System Managing Chronic Conditions: Self-Efficacy for Managing Social Interactions (SEMSI-4) and Managing Emotions (SEME). Sociodemographic variables were collected from surveys and retrospective chart review. Clinical variables included total body surface area burned, initial hospital length of stay, surgical history, and days since injury. Poverty level was estimated by U.S. census data using patient’s home ZIP code. Scores on SEME-4 and SEMSI-4 were compared to the population mean by one-sample T-test, and independent variables evaluated for associations with managing emotions and social interactions by Tobit regression while adjusting for demographic variables. The 71 burn patients surveyed had lower scores in SEMSI-4 (mean = 48.0, P = .041) but not SEME-4 (mean = 50.9, P = .394) versus the general population. Marital status and neighborhood poverty level were associated with SEMSI-4, while length of stay and % total body surface area burned were associated with SEME-4. Patients who are single or from poorer neighborhoods may have difficulty interacting with their environment after burn injury and need extra social support. Prolonged hospitalization and increased severity of burn injury may have more impact on emotional regulation; these patients may benefit from psychotherapy during recovery.

With improving survivorship and functional status after significant burn injury, increased focus has been placed on optimizing psychological recovery. Prior investigations of psychosocial outcomes after burn have largely emphasized the relationship of burn size and length of stay to psychological symptoms, self-esteem, and social reintegration. In general, worse outcomes are seen in patients with larger burns^1–3 and prolonged hospitalizations,^3,4 however some studies did not observe such a relationship between burn severity and psychological symptoms.⁵

There has been comparatively less work done in exploring the relationship of non-clinical factors to psychosocial outcomes. With regard to racial or ethnic minority background, patients from racial or ethnic minority groups have reported less community reintegration^6,7 and satisfaction with appearance⁷ compared to white burn survivors. The limited literature on the relationship of socioeconomic status and psychosocial outcomes after burn demonstrates mixed strength of associations between socioeconomic status and mental health after burn injury; one study reported greater psychosocial needs in patients who are unemployed or whose spending exceeds income,⁸ while another study found a measure of economic well-being was associated with physical, but not mental, health outcomes after burn injury.⁹ With regards to companionship and psychosocial outcomes, patients who were living alone¹⁰ or unmarried¹¹ may have poorer social reintegration and activity after burn injury.

Psychologist Arthur Bandura first popularized the term self-efficacy, or the belief of an individual in their ability to interact with their environment in order to achieve a desired outcome.¹² Bandura makes the distinction between expectations of outcome and expectations of efficacy, with the latter referring to a persons belief in their ability to execute the behavior necessary to generate the desired outcome.¹² Bandura theorized that self-efficacy is the strongest predictor of behavioral change, asserting that a person’s estimation of the efficacy of a behavior affects both initiation and persistence of said behavior.¹² For example, a patient undergoing rehabilitation after burn injury must believe they can physically perform range of motion exercises before said actions can provide the therapeutic benefit of increased functionality of the injured extremity. Self-efficacy has been significantly associated with disability,¹³ depression,¹⁴ and functional ability¹⁵ in patients with chronic medical conditions. For burn patients, their self-efficacy may impact progress in not only physical and occupational therapy, but also in social activities and emotional regulation after significant injury.

We hypothesized that patients with increased severity of burn injury, economic disadvantage, or from a racial or ethnic minority group would report lower self-efficacy in managing emotions and managing social interactions after receiving treatment for burns at a regional burn center and at a safety net hospital.

METHODS

Sample Population

Patients were approached for enrollment at the outpatient burn clinic of a regional burn center and county-safety-net hospital between June and September 2021. Only patients speaking English or Spanish were approached for participation. Exclusion criteria included patients under 18 years of age, patients without self-consent capacity, and patients who did not speak English or Spanish. All study procedures were approved and overseen by the Institutional Review Board.

Data Collection

To measure the primary outcome, self-efficacy for managing emotions and social interactions, patients completed two 4-question short forms versions of two sub-domains of the National Institute of Health (NIH) Patient-Reported Outcomes Measurement Information System (PROMIS) Self-Efficacy for Managing Chronic Conditions: Self-Efficacy for Managing Emotions (SEME-4) and Self-Efficacy for Managing Social Interactions (SEMSI-4), the full 27- and 23-item versions of which have been previously validated.¹⁶ SEME-4 was developed to assess confidence in managing or controlling feelings such as stress, disappointment, and anxiety; SEMSI-4 aims to asses confidence in participating in social activities, especially asking for help and communication about medical conditions.¹⁶Figure 1 shows the questions included in the 4-item short forms of SEME-4 and SEMSI-4. The official Spanish translation of the survey was provided to Spanish speaking patients, and patients completed surveys independently or with verbal assistance from the first or second author. Patients were also asked about current marital and employment status when completing the short forms.

Figure 1.

(a) Four-item short form PROMIS^® Self-Efficacy for Managing Emotions. (b) Four-item short form PROMIS^® Self-Efficacy for Managing Social Interactions.

Chart review was performed to collect patient age, race, ethnicity, primary language, ZIP code, percent of total body surface area burned (TBSA), hospital length of stay, surgical history, and days from injury to survey collection. To approximate patient socioeconomic status, home ZIP code was used to determine percent of neighborhood living below the U.S. federal poverty level, based on U.S. census data.¹⁷

Statistical Analysis

Scores on SEME-4 and SEMSI-4 were converted to standardized scores for comparison with the general population mean of 50 by one sample t-test.

The relationship between raw scores on both PROMIS measures and clinical course (TBSA, hospital length of stay, surgical intervention, and days since injury), and sociodemographic factors (age, race, ethnicity, primary language, percent of home ZIP code below federal poverty level, and marital and employment status) were assessed by Tobit regression analysis. Histograms of the PROMIS scores had two peaks, one of which was at the upper limit of the PROMIS scale; consequently, we assumed there was a right-censored underlying variable, for which Tobit regression was appropriate.¹⁸ Tobit regression produced an estimate of association with the latent variable, not with the original PROMIS measure score, but was intended to detect an association with the original PROMIS measure score. The PROMIS score was assumed to be equal to the latent variable where the latent variable was below a cutoff (the upper limit of the PROMIS scale), and equal to the cutoff otherwise. The model for each independent variable of interest was adjusted for patient age, employment status, marital status, and percent of home ZIP code below federal poverty level. Employment statuses were grouped by not employed versus employed, student, retired, or other. Marital statuses were grouped by married or domestic partnership versus single, divorced, separated, or widowed. Where the independent variable of interest was one of the four sociodemographic variables, the three other sociodemographic variables were used as the only adjustment variables.

Statistical analysis were done using R version 4.1.2 (Vienna, Austria). A significance level of .05 was used to evaluate P-values.

RESULTS

One hundred two patients were approached for participation, and 71 (69.6%) were enrolled in the study. Compared to the United States general population, the burn patients in this study had lower mean scores on SEMSI-4 (mean 48.0, P = .041) but not SEME-4 (mean 50.9, P = .394, Table 1). Patients were majority male (64.8%, 46/71), Hispanic or Latino (76.1%, 54/71), and English speaking (60.6%, 43/71) with a median age of 40 years. The median percent of the cohort’s home ZIP code living below the federal poverty line was 17.8%, compared to 13.4% in the county and 11.8% in the country at the time of enrollment. The median TBSA was 2% (range 1–60), and about one-quarter of patients underwent surgery.

Table 1.

Comparison of standardized SEME-4 and SEMSI-4 scores of burn patients with adult U.S. population.

	Mean standard score	Population mean	P-value
Managing emotions	50.9	50.0	.394
Managing social interactions	48.0	50.0	.041

Among sociodemographic variables assessed, being married or in a domestic partnership was significantly associated with higher latent SEMSI-4 score (36.7%, P < .001, Table 2). There was a significant negative association between percent of home ZIP code below the federal poverty level and latent SEMSI-4 score (P = .017).

Table 2.

Multivariable Tobit regression analysis of sociodemographic factors and latent four-item PROMIS Self-Efficacy for Managing Emotions (SEME4) and Self-Efficacy for Managing Social Interactions (SEMSI4) scores.

Factor	n	% or Median (range)	SEME-4			SEMSI-4
			β	95% CI	P-value	β	95% CI	P-value
Age (years)	71	40 (18–83)	0.016	−0.068, 0.100	.715	−0.004	−0.078, 0.071	.924
Female (%)	71	35.2	−1.642	−3.764, 0.479	.129	1.103	−0.811, 3.016	.259
Black or African American (%)	71	8.5	0.897	−2.844, 4.638	.639	0.217	−3.008, 3.442	.895
White (%)	71	5.6	−3.826	−8.179, 0.528	.085	0.496	−3.473, 4.465	.806
Hispanic or Latino (%)	71	76.1	0.045	−2.341, 2.432	.970	−0.739	−2.822, 1.344	.487
English-speaking (%)	71	60.6	−0.686	−2.918, 1.547	.547	0.392	−1.575, 2.360	.696
Not employed (%)	70	37.1	−1.522	−3.798, 0.754	.190	0.370	−1.652, 2.393	.720
Married or domestic partner (%)	71	36.7	1.923	−0.575, 4.421	.131	4.487	2.189, 6.785	< 0.001*
ZIP code below poverty line (%)	69	17.8 (6.4–31.4)	−0.152	−0.321, 0.016	.076	−0.183	−0.333, −0.032	.017*

^†A P-value ≤ 0.05 was considered statistically significant.

Among clinical factors assessed, greater percentage TBSA (median 2%, P = .044, Table 3) and increased hospital length of stay (median 1, P = .024) were associated with lower scores on the SEME-4. No clinical factors were significantly associated with latent SEMSI-4 scores.

Table 3.

Multivariable Tobit regression analysis of clinical factors and latent four-item PROMIS Self-Efficacy for Managing Emotions (SEME4) and Self-Efficacy for Managing Social Interactions (SEMSI4) scores.

Factor	n	% or Median (range)	SEME-4			SEMSI-4
			β	95% CI	P-value	β	95% CI	P-value
TBSA (%)	71	2 (0.25–60)	−0.090	−0.178, −0.003	.044*	−0.053	−0.130, 0.024	.178
Hospital LOS (days)	71	1 (1–126)	−0.065	−0.122, −0.009	.024*	−0.046	−0.096, 0.004	.069
Surgery (% yes)	71	25.4	−1.564	−3.861, 0.733	.182	−0.720	−2.754, 1.315	.488
Time since injury (days)	64	24 (4–1712)	−0.004	−0.020, 0.011	.572	0.002	−0.010, 0.015	.719

^†A P-value ≤ 0.05 was considered statistically significant.

DISCUSSION

In this cross-sectional study, patients who were married or in a domestic partnership had higher scores on Managing Social Interactions. Previous studies have found a similar relationship between companionship and outcomes after burn injury, with living alone associated with lower scores on the Burn Specific Health Scale,¹⁰ and marriage associated with higher ratings of social interactions and social integration.¹¹ Self-rated degree of social and emotional support has been associated with poorer health related quality of life.¹⁹ This relationship is not unique to burn survivors; poorer psychological outcomes among single patients has been reported in other trauma populations,²⁰ and a recent meta-analysis found that loneliness was an indicator of all-cause mortality in the general population.²¹ Patients without a partner may be especially vulnerable to deterioration of their social support system after injury.

Previous work has found mixed strength of associations between socioeconomic status and mental health after burn injury. Worse mental health after burn injury has been associated with spending that exceeds income,⁸ and self-rated degree of financial distress,²² however the level of home ZIP-code level community distress was found to be associated with physical, but not mental, health outcomes.⁹ In the present study, percent of home ZIP code living below the federal poverty line was associated with lower scores on the Managing Social Interactions sub-scale, and was not significantly associated with Managing Emotions. A low rate of participation in peer support groups in patients treated for burns at an urban safety-net hospital has been previously reported;²³ patients from lower income neighborhoods may have less access to educational resources such as support groups, possibly contributing to worse self-rated ability to interact with peers compared to patients from less disadvantaged neighborhoods.

Higher TBSA and increased hospital length of stay were significantly associated with lower scores on Managing Emotions. The findings here are consistent with previous research indicating a relationship between higher TBSA and both anxiety and depression¹ as well as increased overall psychosocial needs.⁸ TBSA and hospital length of stay have also been correlated with social functioning,^3,24 however here there was not a significant relationship between these measures of injury severity and self-efficacy of managing social interactions here.

Existing literature on the trajectory of psychological recovery after burn injury reports mixed relationships between time since injury and psychological symptoms.^1,8,11,25 In the present study, there was no significant association between time since injury and scores on Managing Emotions or Managing Social Interactions. Because patients were enrolled when attending outpatient burn clinic appointments, only patients with more severe injury were seen at higher intervals from time of injury, limiting interpretation of this null finding.

CONCLUSIONS

The present study demonstrates a possible relationship between scores on PROMIS Self-Efficacy for Managing Emotions, which assesses a patient’s self-perceived ability to effectively regulate their own internal state, and two measures of injury severity. In contrast, scores on Self-Efficacy for Managing Social Interactions, which assesses a patient’s self-perceived ability to effectively interact with peers, were associated with two sociodemographic factors. These findings may represent a pattern in the impact of burn injury on survivor’s livelihood, where severity of injury has greater effects on emotional state, and personal sociodemographic factors play a greater role in shaping a patient’s interactions with their environment after injury. To interpret our findings in the clinical context, patients with more severe injuries may benefit most from psychotherapy after injury, and patients from less advantaged neighborhoods or without a dedicated companion may benefit from peer support groups.

Funding:

The contents of this manuscript were developed under a grant from the National Institute on Disability, Independent Living, and rehabilitation Research (NIDILRR grant number 90DPBU0007). NIDILRR is a Center within the Administration for Community Living (ACL), Department of health and Human Services (HHS). The contents of this manuscript do not necessarily represent the policy of NIDILRR, ACL, or HHS, and you should not assume endorsement by the Federal Government. This work was supported by grants UL1TR001855 and UL1TR00130 from the National Center for Advancing Translational Science (NCATS) of the U.S. National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.