Anoxic Brain Injury: A Subtle and Often Overlooked Finding in Non-Fatal Intimate Partner Strangulation

Abstract

BACKGROUND:

A paucity of literature exists dedicated to the identification of anoxic brain injury in patients that survive non-fatal intimate partner strangulation (NF-IPS). While some individuals report experiencing symptoms of brain hypoxia followed by a loss of consciousness, other individuals report symptoms of brain hypoxia prior to amnesia, rendering some unable to recall loss of consciousness (LOC).

OBJECTIVE:

Using a standardized clinical assessment tool, the purpose of this retrospective analysis is to describe anoxic brain injury symptom prevalence in a sample of patients reporting NF-IPS.

METHODS:

191 unique patients, reporting a total of 267 strangulation events, were assessed by a member of the Shasta Community Forensic Care Team utilizing the Strangulation Hypoxia Anoxia Symptom TBI Assessment (SHASTA) tool. The sample is 98% female and includes adult patients ages 18–68. Examination records were categorized based on the presence or absence of hypoxia and anoxia symptoms. This manuscript utilizes the STROBE checklist.

RESULTS:

Amnesia was reported in 145 of the 267 strangulations (54.3%). Of those, 74 reported LOC (51.0%) while 71 did not recall LOC (49.0%).

CONCLUSIONS:

Within our sample, 49% of patients with amnesia did not recall losing consciousness, demonstrating that LOC is an imperfect measure of anoxia for patients following NF-IPS. Healthcare providers examining NF-IPS patients should inquire about additional symptoms of hypoxia and amnesia, which can be captured on the SHASTA tool.

Background

“Time is brain” (1,2), scorched into the minds of emergency department clinicians for the better part of two decades, underscores the importance of rapid anoxic brain injury identification in patients presenting for emergency care. Anoxic brain injury results in the death of 1.9 million neurons and 14 billion synapses per minute (1,2). Cells located within the hippocampus and layers 3, 5 and 6 of the prefrontal cortex are particularly susceptible due to their location and high metabolic demand. As these neurons play a vital role in the formation of working (i.e., short-term) and episodic (i.e., long-term) memory, symptoms of brain anoxia may manifest as amnesia resulting in denial of loss of consciousness (3–11).

While literature dedicated to the assessment of anoxic brain injury in stroke victims remains robust, there is scant research focused on the identification of anoxic brain injury in patients that survive non-fatal intimate partner strangulation (NF-IPS). Emphasis on loss of consciousness (LOC) as a measure of anoxia often fails to recognize the importance of amnesia as a separate marker of anoxic brain injury. Considering ED utilization by patients experiencing IPV is four times higher than the non-IPV population (12), and up to half report strangulation (13), an obvious gap in the medical literature exists. The detection of amnesia among patients who have experienced IPV and present for emergency care remains an important consideration (14). This study seeks to understand this dichotomy using a standardized assessment tool to identify symptoms of anoxic brain injury, including amnesia, in NF-IPS patients seeking emergency care. Detection of amnesia by the ED clinician is crucial to the diagnosis of anoxic brain injury and referral to neurologists, physiatrists, and occupational therapists for the treatment of neurocognitive deficits, including memory impairment.

Objective

Using a standardized clinical assessment tool, the purpose of this retrospective analysis is to describe anoxic brain injury symptom prevalence in a sample of patients reporting NF-IPS.

Methods

Design

This study utilized a retrospective, cross-sectional secondary data analysis adhering to the 22 item STROBE checklist (15).

Setting

The Shasta Community Forensic Care Team consists of physicians, advanced practitioners, and registered nurses conducting forensic medical examinations on adult and pediatric patients. Consultations performed in multiple counties of northern California include four Emergency Departments embedded within three regional trauma centers ranging from Level II – IV. The program sees approximately 175–200 forensic medical patients per year, inclusive of but not limited to IPV, child abuse, sexual assault, elder abuse, and human trafficking.

Data Collection

Electronic health records of patients aged 18 and older who reported one or more strangulations that were assessed by a member of the Shasta Community Forensic Care Team utilizing the Strangulation Hypoxia Anoxia Symptom TBI Assessment (SHASTA; available as Online Supplement A) between May 1^st, 2019 and April 30^th, 2022 were reviewed. The SHASTA uses a trauma-informed approach in both sequencing and assessment prompts (e.g., “Did your position change during the strangulation (standing to laying, etc)? If yes, do you remember?”) to aid clinicians in distinguishing symptoms of brain hypoxia and anoxia from other differential diagnoses such as blunt head trauma, dissociative amnesia, post-traumatic stress disorder (PTSD), and substance use. Providers are prompted to assess for neurological symptoms, including LOC and amnesia, associated with blunt head injury. Evidence of brain injury excludes the diagnosis of dissociative amnesia according to the DSM-V (16). Diagnosis of PTSD and brain injury are not mutually exclusive in patients reporting amnesia. PTSD may be caused by the memory of the assault while amnesia may be the result of a brain injury (17). Distinguishing symptoms from substance use is based on worsening or new symptoms occurring after application of force to the neck.

In addition to information from the SHASTA, study-specific variables were abstracted, including age, race, ethnicity, relationship to perpetrator, documentation of neck imaging, and documentation of co-occurring blunt head injury. Data from 204 unique patients’ records representing 280 distinct strangulation events were included in the dataset. Data was abstracted by a physician member of the clinical and research team, entered into a structured Excel spreadsheet, and checked for accuracy by a second physician team member. Given this study’s purpose was to examine IPV-specific strangulation characteristics, data from 13 non-IPV-related, unique patient strangulation events were excluded from analyses. This resulted in a final analytic sample of 191 unique patients, and 267 unique strangulation events. Study protocols were reviewed by the Institutional Review Board of Dignity Health and determined not to meet the criteria for human subjects’ research.

The SHASTA Form

To minimize diagnostic error and standardize the assessment of NF-IPS, the clinical team developed the Strangulation Hypoxia Anoxia Symptom TBI Assessment (SHASTA; Online Supplement A). The SHASTA tool uses a trauma-informed approach in both sequencing and assessment prompts. The tool first assesses hypoxic symptoms followed by anoxic symptoms.

Anoxia Symptoms

For analyses, we created a category for anoxic symptoms that included items from the SHASTA that indicated loss of consciousness consistent with loss of consciousness during the event (see Table 1). The SHASTA also includes a prompt about memory problems after the event, which were not included in our anoxia variable.

Table 1:

Hypoxia and Anoxia Symptoms Reported During Strangulation Events

	All Events			With any hypoxia symptoms			With any anoxia symptoms			With recalled loss of consciousness
	n=267	%	95%CI	n=206	%	95%CI	n=149	%	95%CI	n=78	%	95%CI
Hypoxia symptoms
Feeling faint	136	50.9	45.1-57.1	136	66	59.4-72.6	87	58.4	50.6-66.0	42	53.8	42.7-65.3
Dizziness	125	46.8	40.8-53.2	125	60.7	53.7-67.6	75	50.3	42.8-58.2	43	55.1	43.6-66.2
Disorientation	112	42.1	36.1-7.9	112	54.4	47.8-61.4	81	54.4	45.9-61.9	43	55.1	42.9-66.3
Vision changes	154	57.7	51.3-63.7	154	74.8	68.4-80.4	108	72.5	65.1-79.1	54	69.2	58.7-79.5
Hearing changes	86	32.2	26.7-37.7	86	41.7	35.3-8.5	63	42.3	34.4-50.7	31	39.7	28.6-0.7
Anoxia Symptoms
Amnesia	145	54.3	48.5-60.3	129	62.6	55.9-69.3	145	97.3	94.7-100	74	94.9	89.2-98.9
Loss of consciousness	78	29.2	24.0-34.3	65	31.6	24.5-38.3	78	52.3	43.9-60.4	78	100	-

Notes: Confidence intervals not calculated for cells with a 100% occurrence.

Hypoxia Symptoms

The category of hypoxic symptoms was created during the analytic process to collapse data across the symptoms of dizziness, feeling faint, disorientation, changes in vision, and changes in hearing during the event (see Table 1). Again, we elected to use only the during strangulation documentation for this variable, to minimize measurement bias resulting from varying event-to-assessment time intervals and lack of clarity regarding whether a symptom may have been due to hypoxia (e.g. a delayed vision changes from vascular injury) or from another cause (e.g. chronic vision changes from co-occurring facial injuries). While the SHASTA includes a prompt for headache in the hypoxia category, without additional clinical data for analysis (e.g. was the headache pre-existing, patient history of headaches, resolution of headache), we elected not to include this symptom in our collapsed hypoxia category.

Data Analysis

Data were analyzed in Microsoft Excel and SPSS Version 28 (18). Descriptive statistics were used to summarize the sample. Bootstrapping was used to estimate 95% confidence intervals for symptom frequencies. Sensitivity analyses were used to assess for differences in symptom reporting among patients who sought care within 1) the first 7 days; 2) the first 30 days; 3) all patients. A second set of sensitivity analyses were run to compare inclusion vs exclusion of patients reporting headache during the event in the hypoxia variable. As this was a descriptive analysis of clinical data, no hypothesis testing was completed. Missing data for any variable was less than 2%. Given this low rate of missingess, data were not imputed. Missing symptom data were treated as “0” (not present) when calculating summary symptoms and individually missing data points were excluded from totals and summary statistics.

Results

Patients in this sample were primarily white (84.8%, n=162), 18–30 years only (41.9%, n=80), female (97.9%, n=187), and assaulted by male intimate partners (97.9%, n=187). Further demographic data are presented in Table 2 (Online Supplement B). Of the 191 unique patients, 40 (20.9%) reported more than one strangulation event, the number of strangulation events for any individual patient ranged from 1–6. The nature of the clinic and its partnership with local intimate partner violence advocacy and law enforcement agencies allowed patients to be seen at any time following a strangulation event for an examination. This resulted in a wide range of exam timeframes. The mean number of days between the strangulation event and the exam was 120, the median was 39, and the range was 0 days to nearly 10 years (3,421 days). With approximately 19.1% of events (n=51) being seen on days 0–6, 27.0% of events (n=71) being seen on days 7–29, 46.4% of events (n=124) being seen on days 30–364, 6.7% of events (n=18) being seen on days 365–3,421. Imaging of the neck was ordered for 82.7% of patients (n=158), with 74.7% (n=118) of those with ordered imaging completing the recommended imaging examination. Of the 40 patients that did not complete imaging, 4 did not complete imaging due to denial by insurance. Positive neck findings were noted in 14 of 118 (11.9%) completed imaging examinations. Positive findings included: carotid artery dissection, apposition or displacement of the arytenoid cartilage, hyoid bone fracture, soft tissue or muscular findings, and anterolisthesis. CT head is not routinely ordered on patients presenting for a forensic medical examination after strangulation, even if blunt head injury is disclosed. Blunt head injury in intimate partner violence often involves a slap or a punch to the head and often does not warrant a CT head. In addition, only 51 of the 191 patients (26.7%) presented within 7 days of the assault. The number of patients that received a head CT within this sample was low and did not have enough power for statistical analysis.

Hypoxia and Anoxia Symptoms

Of the 267 total strangulation events, 206 events (77.2%) included a reported hypoxia symptom. Hypoxic symptom reporting ranged from 32.2% for hearing changes to 57.7% for vision changes in the entire sample, 41.7% for hearing changes to 74.8% for vision changes in the group reporting at least one hypoxia symptom (Tables 1 and 3). Loss of consciousness (LOC) was reported during 74 of 267 (27.7%) strangulation events, while amnesia was reported during 145 of 267 events (54.3%), as depicted in Figure 1^*.52 Seventy (94.9%) of the strangulations in which LOC was reported also reported amnesia during the event, an unsurprising nearly perfect overlap. This also leaves 49.0% of anoxic events (n=71) in which amnesia was reported without LOC. Nearly all events in which anoxia was reported (n=149) also included at least one hypoxia symptom (n=131, 87.9%), and 18 events (12.1%) only included reported anoxia without any associated hypoxia symptoms.

Figure 1.

Venn diagram depicting the percentage of patients with LOC only, amnesia and LOC, amnesia only, and neither amnesia or LOC.

Comparing symptom reporting between those with anoxic and hypoxic symptoms

During the 267 strangulation events, symptom frequency reporting varied from 0.4% (95%CI: 0–1.5%; n=1) for bowel incontinence to 85.8% (95%CI: 81.8–89.7%; n=229) for difficulty breathing at the time of the event (see Table 3, Online Supplement C). Given the descriptive nature of this study and overlap in individuals in each group noted above we did not use hypothesis testing statistics to compare proportions across groups but have instead highlighted the highest proportion in each symptom reporting category in Table 3 (Online Supplement C) for illustration. In the hypoxia category there are 10 symptoms with the highest reporting rates, 12 in the anoxia category, and 9 in the LOC group. The highest difference in reporting rate between groups is urinary incontinence and coughing without blood.

Sensitivity Analyses

We completed three sets of sensitivity analysis. Two sets compared for timing of exam: one including only incidents where patients were seen in <7 days from the strangulation event (Online Supplement D) and one including only incidents where patients were seen in <30 days from the strangulation event (Online Supplement D). A third set included headache during the strangulation in the hypoxia symptom category (Online Supplement D). None of these analyses provided results that changed our overall description or interpretation of the sample.

Discussion

Our study is the first to describe the large proportion of strangulation events (49%) in which patients were unable to recall LOC yet describe symptoms of brain hypoxia followed by amnesia. The overall minimal variation in symptom reporting across groups (i.e. hypoxia, anoxia, LOC) in our sample demonstrates that LOC is an imperfect measure of anoxia for patients following strangulation and should not be the single marker of anoxia used in evaluating these patients.

Emergency Medicine providers often serve as the initial point of contact for patients seeking medical care after strangulation (19). Screening patients reporting strangulation for amnesia in addition to LOC will enhance detection of anoxic brain injury, increase considerations for imaging, and improve the treatment of acute and chronic symptoms by way of referral to specialists (neurologists, physiatrists, and occupational therapists) (20–27).

In addition, patients with a range of hypoxic and anoxic symptoms also reported co-occurring neck, respiratory, EENT, and gastrointestinal/genitourinary symptoms, emphasizing the importance of obtaining a complete and thorough history to avoid missing important symptoms and opportunities for additional diagnostics (laryngoscopy, dysphagia evaluation, and electrodiagnostic studies), and referral (ENT, GI, speech therapy, and physical therapy) (19–20).

Next Steps for Using Strangulation Assessments in Practice

• Future research to explore the prevalence of amnesia among patients reporting NF-IPS.

• Development of a standardized approach to the evaluation of strangulation in the emergency department through assessments like the SHASTA.

• Use of clinical data obtained to support appropriate consultations and referrals to neurology, physiatry, occupational therapy, and other brain injury specialists.

• Work with healthcare providers, systems, and patients to ensure the standardized approach is culturally relevant.

Limitations

The care-seeking convenience sample was limited with respect to geography, race, ethnicity, gender, and sexual orientation; it may not be generalizable to other populations. Given the exploratory descriptive nature of this clinical dataset, we decided hypothesis testing was premature. This limits any conclusions or generalizations that can be drawn from our results.

Several confounding factors (e.g. blunt head injury, recency of event), though measured were not able to be examined statistically given the descriptive nature of the data and analyses. Future prospective work should ensure that in addition to standardized measures and preset hypotheses, an appropriate range of confounding factors is collected for analysis. As patients were asked about past events, recall bias is possible. Given the context of the original data source (medical records) under reporting and documentation of symptoms would be the most common error (28). We attempted to address self-report and recall bias by abstracting data from a standardized clinical assessment tool (i.e. the SHASTA). In addition, symptoms were initially self-reported for the purposes of medical care instead of research (29).

The SHASTA was used on all patients reporting strangulation referred for an examination by a Shasta Community Forensic Care Team member. Our sample represents the minority of patients nationwide who are referred for forensic subspecialist evaluation. As the SHASTA is a clinical tool, as opposed to a research tool, it was not static through the time data were reviewed. For example, lightheadedness was added to the SHASTA in November 2020 after several patients reported feeling lightheaded when asked about “other” symptoms experienced. Other confounding factors, such as other causes of brain injury (including blunt head injury), were asked for each strangulation event. Prior to their inclusion on the tool, these factors (e.g. substance use, blunt head injuries) were still commonly asked about and included in patient history and physical completed by the examiner but lacked a specific documentation location within the SHASTA form.

Conclusions

49% (71/145) of the patients reporting symptoms of brain hypoxia followed by amnesia did not recall losing consciousness. Our data demonstrates that LOC is an imperfect measure of anoxic brain injury following strangulation. Healthcare providers that examine NF-IPS patients in the emergency care setting should inquire about additional symptoms of hypoxia and anoxia, including amnesia, which can be captured on the SHASTA tool. Future research on the prevalence of amnesia in NF-IPS and the integration of a standardized approach to NF-IPS patients is recommended.

Highlights.

• Non-fatal intimate partner strangulation is a common condition presenting to the ED

• Anoxic brain injury due to non-fatal strangulation can result in LOC with amnesia

• Nearly half of events denied LOC but reported hypoxia symptoms followed by amnesia

• ED Providers should screen for amnesia using a standardized assessment