Acupuncture use for pain after traumatic brain injury: a NIDILRR Traumatic Brain Injury Model Systems cohort study

Abstract

Background

Pain after traumatic brain injury (TBI) is common and can become chronic. Acupuncture is an increasingly popular non-pharmacologic option in the United States and is commonly used for pain.

Objective

We explored demographics, injury characteristics, and pain characteristics of individuals who reported using acupuncture for chronic pain after TBI.

Methods

We analyzed a subset of data collected as part of the Pain After Traumatic Brain Injury collaborative study and identified individuals reporting a history of acupuncture as part of management for chronic pain after TBI. We characterized and compared basic demographic data, pain treatment engagements, pain severity, pain interference, functional independence, and pain locations using descriptive and inferential statistics.

Results

Our sample included 1,064 individuals. Acupuncture use (n=208) was lower proportionally among females, Blacks/African Americans, Asians, less educated, and non-military service members. Insurance type varied between acupuncture and non-acupuncture users. Functional and pain outcomes were similar, but acupuncture users reported a higher number of pain sites.

Discussion

Acupuncture is one treatment utilized by individuals with TBI and chronic pain. Further investigation would be helpful to understand the barriers and facilitators of acupuncture use to inform clinical trials to examine the potential benefit of acupuncture on pain outcomes after TBI.

Background

Traumatic brain injury (TBI) remains a significant public health concern globally and is considered a leading cause of disability. An estimated 69 million individuals worldwide sustain a TBI annually (1), of which approximately 2.8 million individuals are in the United States (2). The clinical presentation for TBI is complex, with many interwoven factors contributing to the complexity, such as sleep disturbance, depression, anxiety, cognitive difficulties, and pain (3–5). However, these cognitive, behavioral, and somatic sequelae are common and diminish health-related quality of life and interfere with day-to-day activities (6–9). In the United States alone, over three million Americans are estimated to live with a disability related to a TBI (6, 9, 10).

Chronic pain is one of the most common medical complaints after TBI (11–15). Pain after TBI can include headache (16), musculoskeletal pain (3, 11, 14), central pain syndrome (3, 17) and be directly related to the TBI or other comorbidities. While pain is common in the acute stage after injury, pain also commonly becomes chronic, lasting three months or more (3).

Treatments for TBI rehabilitation are varied and can include medications and non-pharmacologic approaches (6, 13, 18). And despite pain being a common issue after TBI, there are no uniform guidelines for treatment. However, clinical practice guidelines in non-TBI specific populations have recommended a variety of non-pharmacologic approaches, including acupuncture (19).

Acupuncture is a non-pharmacologic treatment that involves the insertion of thin needles through the skin (20). This therapy is distinguished from other modalities involving needle insertion, such as dry needling. The points stimulated with acupuncture are based on practices in Traditional Chinese Medicine (20), and dry needling chooses points to stimulate based on the location of myofascial trigger points (21). Acupuncture is steadily gaining popularity in the United States (22–24), partly driven by suboptimal results from conventional treatments (25). Acupuncture is beneficial for many types of chronic pain, including musculoskeletal pain and headache (26), but it has not been well studied in individuals with TBI. Many mechanisms have been proposed to explain the therapeutic effects. For pain, neuroimaging studies demonstrate that acupuncture modulates areas of the brain associated with pain (27) and affects pain-related functional connectivity (28). However, small sample sizes, methodological issues, and a lack of controlled trials limit the research on acupuncture in TBI-specific populations. Nevertheless, these limited results suggest benefits for headaches (29), sleep disturbance (30), and disorders of consciousness (31). Additionally, some studies have indicated that acupuncture decreases the risk of developing dementia (32), stroke (33), or utilization of emergency care and hospitalization (34) after TBI.

Currently, there is little research on the use of acupuncture in the setting of chronic TBI, and fewer still on pain with chronic TBI (35). And despite some evidence for the potential benefit of acupuncture and inclusion in clinical practice guidelines, there is little known about who uses acupuncture for pain after TBI in the United States. Therefore, the objective of this study was to examine the demographics, injury characteristics, and pain characteristics of individuals with TBI who report the use of acupuncture for pain after TBI and compare them to those with chronic pain who did not use acupuncture to identify whether differences existed between groups.

Methods

Study design

We analyzed a subset of data collected as part of the Pain After Traumatic Brain Injury collaborative study funded by the National Institute on Disability, Independent Living, and Rehabilitation Research (pain.tbindsc.org). The collaborative pain study is a cross-sectional, multi-site, observational cohort study that included individuals who were enrolled and followed as part of their participation in the TBI Model Systems (TBIMS) study (36). The goal of the collaborative pain study was to learn about the experience of chronic pain in individuals with TBI, including determining the frequency and characterization of pain and, for those who had experienced or were currently experiencing chronic pain, what treatments they had received.

TBIMS participants are individuals hospitalized for TBI enrolled during acute inpatient rehabilitation and are subsequently followed telephonically or in-person at post-injury timepoints of years one, two, five, ten, and every five years thereafter. Upon concluding the TBIMS follow-up interview, participants were invited to take part in the collaborative pain study (Pain Questionnaire). Those who were eligible (able to speak English and to respond by themselves) and interested completed the informed consent process for this interview and were offered the option to complete it online, via paper survey, or by telephone interview. Participants were defined as having current chronic pain if they endorsed experiencing persistent or recurring pain, including headaches or pain anywhere in the body that occurs more than half the days over three months. Those not endorsing current pain were asked if they had experienced previous chronic pain now resolved since their TBI as defined above (past pain). Those not endorsing either chronic pain scenario were categorized as “no pain.”

For the current analysis, we selected a sample of individuals who provided data from March 2019 to April 2021. Individuals categorized as having “no pain” were excluded from the current analysis. TBIMS data collected at baseline and the follow-up interview preceding the survey were also included for analysis. The University of Washington Human Subjects Division Institutional Review Board (IRB) determined that the current study is not human subjects research and is exempt from IRB review.

Study dataset

The data for the current study included all participants who completed the survey by April 1, 2021, from 13 civilian TBIMS sites and one Veterans Health Administration (VHA) TBIMS site. In addition, demographic data already recorded for respondents by the TBIMS were included.

Identification of acupuncture use

Within the sample of participants reporting current or past pain, we identified all respondents who reported a history of acupuncture as part of the management of chronic pain after TBI.

Sociodemographic data

We included data on age, gender, race, ethnicity, level of educational attainment, marital status, annual household income, year of survey completion since injury, military service status, and insurance coverage. Regional geographic residence was presented in accord with the United States census regions and divisions (37).

Pain characterization and treatment

To explore the experience of pain, we included data on past or current pain, pain intensity and interference, number of pain locations, and endorsement of specific pain locations. In addition, we provided counts of engagement in different approaches to the treatment of pain summarized as biomedical (e.g., injections, prescription medication, surgery), biopsychosocial (e.g., exercise, physical therapy, psychotherapy, biofeedback, relaxation), and non-acupuncture complementary and integrative health (e.g., chiropractic, guided imagery, hypnotherapy, massage, mindfulness therapy, tai chi, yoga). All free-text entries under “other” were categorized according to their nature, such as classifying the self-reports of having trigger point injections or arthroscopy as biomedical. All remaining entries from the free-text field inappropriate for grouping elsewhere were categorized as “other.”

Selected measures

We abstracted data from the Pain Questionnaire including two instruments to report pain severity and interference (Brief Pain Inventory-Short Form, BPI-SF) (38) and functional independence (Functional Independence Measure, FIM^™) (39).

Pain severity and interference

The Pain Questionnaire included data on pain severity and interference from the BPI-SF (38). Pain severity was recorded on an 11-point scale (0–10), with higher numbers indicating more pain. The Brief Pain Inventory uses four domains, “worst,” “least,” “average,” and “now,” for pain severity over the last seven days. We summarized information from the single BPI “average” pain item for pain intensity. Pain interference is also recorded on an 11-point scale (0–10), with higher numbers indicating more interference. Seven items are used for pain interference with daily activities, “general activity,” “mood,” “walking,” “working,” “relations with others,” “sleep,” and “enjoyment of life.” A composite score calculated from the mean of all seven pain interference subscales reflects the pain interference.

Functional independence

The current function was assessed from the FIM^™, which has 18 items of seven-level choices to measure outcomes of rehabilitative care (39). Scoring includes a total composite score from 18 to 126 calculated from the summed two subtotals for motor and cognitive independence scales. Higher scores reflect higher functional independence.

Analysis

We used descriptive statistics to characterize sociodemographic information and pain-related responses in acupuncture and no acupuncture users. We provided counts and proportions for categorical variables and the mean and standard deviation for continuous variables. In addition, we compared the responses for individuals reporting acupuncture use and individuals not reporting acupuncture use for pain with the Mann-Whitney U test for continuous variables to account for a non-normal distribution of the data, or an unpaired t-test for continuous variables with data arising from a normal distribution. Because of the exploratory nature of our study aims, we did not specify a reference group and compared the values for each group collectively with chi-squared for all categorical variables that contained all values of 10 or more or Fisher’s Exact test for all categorical variables that contained all or only some values that were less than 10. We conducted our analyses with R version 4.0.3 statistical software (40).

Results

From the initial 1,857 respondents who provided data to the Pain Questionnaire from March 2019 to April 2021 we excluded 756 respondents who reported “no pain” and another 37 respondents who did not provide a response to having acupuncture. This process left a sample of 1,064 (Table 1). The majority (77.1%) of respondents endorsed experiencing current pain since TBI. Two hundred and eight (19.5%) reported using acupuncture for pain after TBI. Seven individuals reported having dry needling without specifically reporting acupuncture and were grouped with the non-acupuncture group. Respondents reported a higher proportion of acupuncture use at two- or ten-years post-injury than in other follow-up years.

Table 1.

Characteristics of Survey Respondents with Current or Past Pain (N=1064)

	Acupuncture (n = 208)		No Acupuncture (n = 856)
	Response Rate (n, %)	Value	Response Rate (n, %)	Value	P a
Age (years) [mean (SD)]	207 (99.5)	46.3 (13.6)	854 (99.8)	45.6 (16.0)	0.33
Sex	208 (100.0)		856 (100.0)		0.66
Female		56 (26.9)		246 (28.7)
Male		152 (73.1)		610 (71.3)
Race	196 (94.2)		821 (95.9)		< .001
American Indian or Alaska Native		10 (5.1)		22 (2.7)
Asian		3 (1.5)		19 (2.3)
Black or African American		12 (6.1)		167 (20.3)
Hawaiian or Other Pacific Islander		1 (0.5)		5 (0.6)
White		168 (85.7)		604 (73.6)
More than one race		2 (1.0)		4 (0.5)
Ethnicity	208 (100.0)		851 (99.4)		0.18
Hispanic or Latino		26 (12.5)		96 (11.3)
Not Hispanic or Latino		181 (87.0)		755 (88.7)
Not disclosed (refused or unknown)		1 (0.5)		0 (0.0)
Education	120 (57.7)		792 (92.5)		< .001
Less than High School		5 (4.2)		143 (18.1)
High School or GED		25 (20.8)		210 (26.5)
Some college		43 (35.8)		276 (34.8)
Bachelor’s or higher		47 (39.2)		163 (20.6)
Years since injury at follow-up	208 (100.0)		856 (100.0)		< .001
1 year		25 (12.0)		186 (21.7)
2 years		41 (19.7)		151 (17.6)
5 years		35 (16.8)		178 (20.8)
10 years		53 (25.5)		129 (15.1)
15 years		34 (16.3)		124 (14.5)
20 years or more		20 (9.6)		88 (10.3)
Marital status	120 (57.7)		791 (92.4)		0.21
Married		50 (41.7)		258 (32.6)
Divorced or separated		29 (24.2)		184 (23.3)
Widowed		1 (0.8)		22 (2.8)
Never married		40 (33.3)		326 (41.2)
Other		0 (0.0)		1 (0.1)
Household Income	116 (55.8)		756 (88.3)		0.11
< $25,000		31 (26.7)		253 (33.5)
$25,000 – $49,999		24 (20.7)		188 (24.9)
$50,000 – $99,999		30 (25.9)		184 (24.3)
$100,000 or more		27 (23.3)		106 (14.0)
Not disclosed (refused or unknown)		4 (3.4)		25 (3.3)
Regional United States	119 (57.2)		788 (92.1)		< .001
Pacific		40 (33.6)		117 (14.8)
Mountain		32 (26.9)		100 (12.7)
West North Central		7 (5.9)		69 (8.8)
West South Central		14 (11.8)		163 (20.7)
East North Central		7 (5.9)		107 (13.6)
East South Central		1 (0.8)		32 (4.1)
Middle Atlantic		4 (3.4)		65 (8.2)
South Atlantic		10 (8.4)		99 (12.6)
New England		4 (3.4)		36 (4.6)
Military service memberb	208 (100.0)		855 (99.9)		< .001
Yes		99 (47.6)		125 (14.6)
No		90 (43.3)		640 (74.9)
Not disclosed (refused or unknown)		19 (9.1)		90 (10.5)
Insurancec	208 (100.0)		856 (100.0)		< .001
Private		137 (65.9)		418 (48.8)
Medicare		41 (19.7)		258 (30.1)
Medicaid		30 (14.4)		241 (28.2)
VA		67 (32.2)		71 (8.3)
Casualty		13 (6.2)		67 (7.8)
Other		13 (6.2)		83 (9.7)
None		23 (11.1)		108 (12.6)
Pain	208 (100.0)		856 (100.0)		< .001
Current Pain		185 (88.9)		635 (74.2)
Past Pain		23 (11.1)		221 (25.8)
Therapy engagement for paind	208 (100.0)		856 (100.0)
Biomedical (n=988) [mean (SD)]		2.9 (1.3)		1.7 (0.9)	< .001
Biopsychosocial (n=900) [mean (SD)]		4.3 (1.9)		2.8 (1.4)	< .001
Non-acupuncture CIH (n=661) [mean (SD)]		3.2 (1.6)		1.9 (1.2)	< .001
Other (n=125) [mean (SD)]		1.1 (0.3)		1.1 (0.3)	0.95
Counts of therapy engagement for paine	208 (100.0)		856 (100.0)		< .001
Biomedical		200 (96.2)		788 (92.1)
Biopsychosocial		197 (94.7)		703 (82.1)
Non-acupuncture CIH		194 (93.3)		467 (54.6)
Other		46 (22.1)		79 (9.2)

Note: Values are presented as number (%) of the total number of respondents displayed in the response rate for categorical variables and mean (standard deviation) for continuous variables where noted.

Abbreviations. CIH = Complementary and Integrative Health; SD = standard deviation; VA = Department of Veterans Affairs.

^aStatistical significance by Mann-Whitney, Fisher’s exact, or chi-squared as appropriate.

^bMilitary service status can be active or previous.

^cRespondents may have more than one type of insurance. It is possible that some respondents did not provide a response to all questions on individual insurance types.

^dRespondents may have multiple types of therapies. It is possible that some respondents did not provide a response to all questions on unique treatment modalities.

^eCounts of therapies are indicators only and respondents may have more than one therapy within each category of therapy. It is possible that some respondents did not provide a response to all questions on unique therapy.

The acupuncture and non-acupuncture groups were similar in age, primarily identified as male, and primarily were not of Hispanic ethnicity. Among individuals reporting being of Hispanic ethnicity, over half of the acupuncture and non-acupuncture users also reported identifying as White (Table 2). Among respondents reporting acupuncture use, a higher proportion were married (41.7%) and had an annual household income of $50,000 or more (49.1%). Among the non-acupuncture group, a higher proportion of respondents were never married (41.2%) and had an annual household income of less than $50,000 (58.3%). However, the annual household income and marital status differences were not statistically significant.

Table 2.

Hispanic or Latino Ethnicity by Race for Acupuncture and Non-Acupuncture Users (N=122)

	Acupuncture (n = 26)		No Acupuncture (n = 96)
	Response Rate (n, %)	Value (n, %)	Response Rate (n, %)	Value (n, %)	P a
Race	18 (69.2)		68 (70.8)		0.72
American Indian or Alaska Native		0 (0.0)		5 (7.4)
Asian		0 (0.0)		0 (0.0)
Black or African American		3 (16.7)		7 (10.3)
Hawaiian or Other Pacific Islander		0 (0.0)		2 (2.9)
White		15 (83.3)		53 (77.9)
More than one race		0 (0.0)		1 (1.5)

Note: All values are presented as number (%) of the total number of respondents displayed in the response rate.

^aStatistical significance by Fisher’s exact.

The two groups differed on race, education, regional residence, military service status, and insurance use (all p < .001, see Table 1). Compared to the non-acupuncture group, the acupuncture group had a higher proportion of respondents that were White (85.7% vs. 73.6% in the non-acupuncture group) and a smaller proportion of Black/African American (6.1% vs. 20.3%) and Asian individuals (1.5% vs. 2.3%). A higher proportion of individuals in the acupuncture group had a college degree or higher than the non-acupuncture group. Compared to the non-acupuncture group, a higher proportion of respondents reporting acupuncture use lived in the Pacific or the Mountain regions. Individuals in the acupuncture group had a higher proportion of military service members than individuals in the non-acupuncture group. A higher proportion of acupuncture users had private health insurance or health care coverage through the VHA. Medicaid and Medicare use was reported less in the acupuncture group compared to the non-acupuncture group. Acupuncture users had higher utilization of non-acupuncture complementary and integrative health treatment than non-acupuncture users.

Pain outcomes were similar between the groups (Table 3). Both acupuncture and non-acupuncture users reported similar pain intensity, pain interference, and functional independence. However, the acupuncture group reported a higher average number of 5.1 pain sites compared to 3.5 in the non-acupuncture group (p < .001). While the pain locations were similar between the groups overall, over half of the acupuncture group reported pain in the back, legs or feet, neck, shoulder, or headache compared to the non-acupuncture group where only half of the respondents in this group reported having back or leg or foot pain.

Table 3.

Pain Experience (N=1064)

	Acupuncture (n = 208)		No Acupuncture (n = 856)
	Response Rate (n, %)	Value	Response Rate (n, %)	Value	P a
Number of pain sites [mean (SD)]	208 (100.0)	5.1 (2.8)	852 (99.5)	3.6 (2.5)	< .001
Pain Sitesb	208 (100.0)		852 (99.5)		0.11
Abdomen		18 (8.7)		88 (10.3)
Arms and/or hands		102 (49.0)		270 (31.5)
Back		165 (79.3)		464 (54.2)
Buttocks		34 (16.3)		87 (10.2)
Chest		23 (11.1)		93 (10.9)
Face or jaw		48 (23.1)		133 (15.5)
Head, such as headache or migraine		121 (58.2)		388 (45.3)
Hips		97 (46.6)		244 (28.5)
Legs and/or feet		129 (62.0)		448 (52.3)
Neck		125 (60.1)		307 (35.9)
Pelvic area or groin		47 (22.6)		105 (12.3)
Shoulder		124 (59.6)		342 (40.0)
Widespread pain or Fibromyalgia		24 (11.5)		56 (6.5)
Other		5 (2.4)		11 (1.3)
BPI- SF
Average Pain Intensity [mean (SD)]	208 (100.0)	4.6 (1.7)	852 (99.5)	4.9 ( 2.0)	0.08c
Pain Interference [mean (SD)]	203 (97.6)	4.4 (2.4)	849 (99.2)	4.3 (2.5)	0.63
FIM Total [mean (SD)]	201 (96.6)	117.8 (7.4)	846 (98.8)	117.7 (10.1)	0.18
FIM Motor [mean (SD)]	203 (97.6)	86.8 (5.1)	847 (98.9)	86.1 (8.3)	0.50
FIM Cognitive [mean (SD)]	201 (96.6)	30.9 (3.8)	848 (99.1)	31.5 (3.3)	0.06

Note: Values are presented as number (%) of the total number of respondents displayed in the response rate for categorical variables and mean (standard deviation) for continuous variables where noted.

Abbreviations: BPI-SF = Brief Pain Inventory- Short Form; FIM = Functional Independence Measure; SD = standard deviation.

^aStatistical significance by Mann-Whitney, unpaired t-test, Fisher’s exact, or chi-squared as appropriate.

^bRespondents can have multiple pain sites. It is possible that some respondents did not provide a response to all questions on unique pain sites.

^cStatistical significance by unpaired t-test.

Discussion

Our results add to the non-pharmacologic pain management knowledge base by characterizing individuals with chronic TBI who use acupuncture for chronic pain and their self-reported pain experience. This study set out to examine the characteristics of acupuncture users and pain qualities reported among individuals with chronic TBI, and not to examine the effectiveness of acupuncture on pain outcomes in the context of other reported pain rehabilitation modalities. The limitations of the data for this study do not allow us to either support or not support the use of acupuncture as part of a comprehensive pain management program after TBI. They neither indicate a superiority or inferiority of acupuncture to other pain management strategies nor a contribution or detriment to other pain management strategies. Rather, our results indicate that individuals with chronic TBI use acupuncture for pain management.

Few studies examine the characteristics specifically of acupuncture users in the United States. A secondary analysis of the National Health Interview Survey (41) found that the average acupuncture user is 65 years of age and older, identifies as female, not Black or African American, has a bachelor’s degree or higher, reports an annual income of at least $75,000, lives regionally in the Western United States, and either reports a good health status or reports having a nervous system disorder, or arthritic condition (25). Despite differences in study samples, our findings are largely consistent with this study by Austin and colleagues. However, in our study sample, males comprised the largest proportion of reported gender, and the average age was less than 65 years. The contribution of at least two factors may explain these discrepancies. First, our study sample focused on a specific clinical population, and the distribution for age and gender are consistent with the data from the TBIMS (42). Second, the VHA provides health care coverage for acupuncture, facilitating access to more Veterans and service members (43), and our study sample had a significant number of military service members in the acupuncture use group. A recent examination of acupuncture use among VHA patients from records through the Veterans Informatics and Computing Infrastructure by Redd and colleagues showed that more males than females reported using acupuncture, and frequently for pain management (44).

The use of acupuncture varied with insurance type, especially with having private insurance or health care coverage through the VHA, and overall most of our study sample had some form of insurance. Acupuncture can be costly for self-pay individuals. The median cost for an acupuncture treatment in the United States is estimated to be approximately $80 per visit, including the initial and follow-up visits and the generally lower-cost community-style group treatments (45). Some private health insurers provide coverage for at least partial benefits (46), and state and federal insurance recently began offering coverage under benefits from Medicaid (47), Medicare (48), and as previously noted, through the VHA (43). Even with limited benefits available, at least 50% of acupuncture visits do not involve health insurance (46, 49). Having health insurance coverage can facilitate access to this treatment for individuals with TBI who otherwise might have limited financial resources (49, 50).

Acupuncture users also reported more utilization of conventional and non-conventional pain management strategies. This finding aligns with prior reports that acupuncture use typically stems from suboptimal results from conventional medicine (25) and that acupuncture is a last treatment resort for many conditions (51). Acupuncture users in our study sample possibly tried multiple non-acupuncture modalities before employing acupuncture, which might explain the higher utilization rate. The areas of pain reported by acupuncture users are consistent with the primary areas most reported for pain in general for individuals after TBI (3, 11). While it is possible that the acupuncture users in our study sample had more pain areas and therefore elected to try acupuncture, it is also possible that physician referrals, especially those within the VHA, contributed to the higher proportion of acupuncture utilization for these pain sites (25).

Pain outcomes were similar between acupuncture and non-acupuncture users, which is not surprising, given the limitations of the data. Chronic pain and inflammatory conditions require multiple acupuncture treatments more than one day weekly over many weeks for a clinically meaningful effect – at least 12 to 24 treatments over six to twelve weeks (52). This frequency and course of treatment exceed the average course of acupuncture for the United States consumer, with or without insurance coverage (25, 53). The Pain Questionnaire from the primary study did not collect information on the temporality of acupuncture treatments for pain after TBI. While some participants were only one-year post-injury, a majority were at least five years after injury, and over half of the acupuncture users were ten years or more post-injury.

Strengths and limitations

The surveillance for both acupuncture use and chronic TBI in the United States is limited, even with the treatment’s growing popularity and the high TBI prevalence. The TBIMS database is representative of patients receiving inpatient rehabilitation in the United States for TBI (54). A strength of this study is that the data are from TBIMS individuals nationwide, all with a moderate to severe chronic TBI, who report pain management strategies.

However, there are several limitations to this study. The most significant limitation is the cross-sectional study design which limits our ability to make conclusions about acupuncture effects on pain outcomes among individuals with TBI. As noted, the details of acupuncture treatment are absent from the data, including temporal aspects of the acupuncture treatments, the frequency, and duration of treatments, patient satisfaction, or whether it was helpful for these individuals. As such, we do not have information on how pain and functional outcomes changed with acupuncture, and possibly the acupuncture users had more pain before treatments. Additionally, there are limitations to our study stemming from the primary data collection itself. The study sample only included individuals able to respond independently without proxy responses. As such, any cognitive impairment might have affected a respondent’s recall for item accuracy or recall whether specific treatments were received, including acupuncture. Some respondents may not know the difference between acupuncture and dry needling, and the research staff did not provide specific definitions for these modalities. Misclassification of having acupuncture might have affected our results. Last, this study examined a subset of the Pain Questionnaire, and individuals agreed to participate. In light of possible selection bias, the responses provided by our study sample may not be generalizable to all individuals with pain after TBI.

The results of our study suggest that future research should focus on the specific barriers and facilitators to access acupuncture, including the availability of acupuncture services, the scope of insurance benefits, sociodemographic factors, and other factors predisposing an ability to afford treatments in the absence of insurance coverage. Future research is also needed to characterize factors motivating an individual to seek acupuncture based on medical symptomatology and to characterize recommendations for its use by healthcare providers. Finally, our results suggest a higher utilization among military service members, and future research should focus on the treatment guidelines used by the VHA and those clinical outcomes and how this may differ from acupuncture treatment plans and clinical outcomes for non-service members, for whom access and insurance benefits are likely to be less.

Conclusions

Acupuncture is one treatment utilized by individuals with TBI and chronic pain. Identified users of acupuncture include those typically found in other literature on acupuncture, but the current study also shows the use of acupuncture in military service members where acupuncture has become more available in the VHA system. While we cannot determine the clinical effect of acupuncture in this population, these results suggest that further investigation would be helpful to understand the barriers and facilitators of acupuncture better to inform clinical trials to examine the effects on pain outcomes after TBI.