Case Presentation

A 48-year-old African American man with a previous history of poly-substance abuse, traumatic brain injury, epilepsy, and behavioral issues with impulsivity and disinhibition, was admitted to the Neuroscience ICU with an aneurysmal subarachnoid hemorrhage. His hospital course was complicated by obstructive hydrocephalus and cerebral vasospasm as well as intermittent behavioral outbursts and disinhibition. Healthcare providers noted his previous history of behavioral issues, and he was often subjected to both physical and chemical measures of restraint. Additional security staff were frequently called to his bedside for further attempts at de-escalation. Though no direct harm to the patient was readily identifiable during the hospitalization, it led to a poor patient and family experience.

Introduction

Workplace violence is increasing across all types of healthcare settings. Two-thirds of physicians and nurses in the emergency department have reported some level of workplace violence up to and including physical assault, with 97% of episodes reported to be at the hands of patients. In reaction to increasing incidents, a common intervention has been the addition of an Electronic Behavior Alert to the electronic medical record (EMR). (1)

Studies have since shown that electronic behavior alerts may negatively impact future patient relationships as well as lead to disparities in the use of restraints and other security measures, with an especially negative impact on patients identifying as black or African American and/or male. Stigmatizing language, though often unintentional, has been shown to have a racial bias in the EMR. (1) Despite not being intended as a diagnosis, electronic behavioral alerts are often interpreted as such and can impact patient care, often to the detriment of patients from certain demographics. (2)

However, some healthcare providers feel that these alerts provide an early warning about potential issues in a patient that a provider may not be familiar with. Here, we examine some of the data regarding Electronic Behavioral Alerts (EBA), demographics, examples of stigmatizing language, and strategies that may help mitigate the negative effects of a tool that may otherwise be beneficial.

The Impact of Demographics

One large cohort study across three urban hospitals within a single large academic health care system found that behavioral flags were not widely utilized. However, when EBAs were used, there was a disparity between Black and White patients, with Black patients receiving more flags. In this study a tendency toward overall decreased laboratory and imaging was noted as were increased treatment times for those with flags. (2) In a larger study reviewing about 3 million ED visits, younger patients who identified as Black, non-Hispanic, publicly insured, and male were more likely to have an EBA. (1)

Urban and rural healthcare systems serve different patient populations, and each environment has a unique set of patients who may have higher or lower thresholds for receiving physical or chemical restraints. For example, a retrospective study at a large urban safety net hospital in Minneapolis that represents many rural Indigenous American communities noted that restraints were more likely to be placed on Indigenous American patients, especially those with a history of severe alcohol use disorder. The same study found that other racial minorities were less likely to have restraints. Those patients who received restraints were also more likely to have an EBA for a history of violent behavior. (3)

Though EBAs have not been researched in correlation with Security Emergency Responses (SERs), a retrospective study including 24,000 patients at a large tertiary care center in a non-psychiatric inpatient setting found that Black patients had higher SER utilization than White patients, while patients who identified as Hispanic did not have significantly higher rates of SER or restraint use. (4)

The Risks of Stigmatizing Language

Many health care providers are unaware of their own unintentional use of stigmatizing language in the EMR and its potential impact in transmitting bias, alienating patients, and leading to negative outcomes in some cases.

Stigmatizing language, even when unintentional and otherwise factual, can suggest that a patient may somehow be less worthwhile to engage with and thus merit less treatment. There are three forms of stigmatizing language, including 1) marking or labeling a patient, 2) assigning responsibility or blame, and 3) invoking danger or risk. For example, marking or labeling a patient as non-adherent with their medication regimen assigns a sense of blame and can lead to health care providers potentially considering that patient less likely to benefit from a medication that would otherwise be indicated. Though healthcare providers can justifiably fear the risks of injury from a combative patient, an excessive response to combativeness, whether through chemical or physical restraints, can negatively impact patient care and potentially lead to patient harm.

Examples of common stigmas include patients with substance use disorders being labeled as substance abusers or using terms such as non-compliant or poorly controlled, which all assign patient responsibility to medical, social, or educational factors, and patients in distress or otherwise agitated as part of their disease process (e.g., post-ictal states or certain brain injuries) being labeled as belligerent and combative. (5) Even if they are factual and are not used with an intent to bias, such phrases can be poorly perceived by others who may encounter them in documentation and eventually lead to an EBA being created in a patient’s medical record.

A retrospective study reviewing nearly 30,000 patients noted that 2.5% of admission notes contained stigmatizing language. Admission notes are often the first notes read in a patient’s chart and their verbiage is often carried forward multiple times by different healthcare providers. This phenomenon may lead to a false sense of accuracy and mislabeling or otherwise misrepresenting a patient over time. The same study found that stigmatizing language was most commonly used in patients with diabetes, substance use disorders, and chronic pain. Specifically, a one-point increase in the diabetes severity index was associated with a 1.23% greater probability of stigmatizing language in the chart. (5)

Maintaining Potential Benefits while Minimizing Unexpected Consequences

Though the advent of EBAs arose from the honest intent to combat an increase in patient violence over the last several years, an unexpected consequence has been a risk for inadvertent mislabeling of patients and demographically driven disparities in EBA utilization. EBAs may benefit healthcare providers by alerting them to suicide risk, which may positively impact how a patient is triaged and managed.

Though most studies don’t suggest an excessive use of EBAs yet, there is a potential for minority patient populations (most often young Black, Non-Hispanic males) to be further marginalized and at risk for substandard care. Further research needs to continue across all health care disciplines to investigate the causation and correlation between EBA and restraint use, health care delivery, and racial and gender disparities.

In this hypothetical case presented, several factors may have contributed to a potential for adverse outcome or poor patient/family experience. Race and gender, in addition to a previous history of traumatic brain injury with resulting behavioral changes would predispose him to higher risk of an EBA and potentially higher risk for more security emergency responses.

Healthcare providers should indeed receive training on stigmatizing language and the impact it may create for patient care. The long-term effects of EBA’s on individual patients and populations have not been studied, though it is reasonable to hypothesize some patients/populations may feel more vulnerable to seeking out healthcare. Individual hospital systems should have regular reviews of their EBA system and the benefits to healthcare providers as well as possible harm to patients and the standards of care that we provide.