Influence of patient race on administration of analgesia by student paramedics

Disparities in the provision of health care have been described in the literature, and these include disparities associated with age and gender as well as race or ethnicity. A landmark report on racial and ethnic disparities in health care from the US Institute of Medicine found that African American individuals receive fewer procedures and poorer-quality medical care than white individuals [1]. The cause of racial disparities is complex, and encompasses patient, provider and health system issues. Provider factors include prejudice and stereotyping based on patient characteristics that may influence clinical judgements and treatment decisions [2]. Disparities in the management of pain have been shown to be associated with race, with a study of the prescribing of opioids in the emergency department finding that white patients reporting pain were more likely to receive an opioid than Black, Hispanic, or Asian/other patients [3]. Evidence from other treatment settings has revealed racial and ethnic disparities in acute, chronic and palliative pain care across the lifespan, and this has led to calls for reform of health care agendas and public health policy to reduce and eliminate disparities in care [4, 5].

Pain is a commonly encountered complaint in paramedic practice, and despite having guidelines for the assessment and management of pain and effective analgesics to manage this symptom, inadequate analgesia has been frequently reported, with paramedic attitudes identified as a potential barrier to appropriate care due to the effect of affective bias on clinical judgements [6]. Despite the evidence of racial disparities arising from other health settings, there is limited research describing associations between patient race and the paramedic provision of analgesia. This study therefore aims to examine associations between patient race and paramedic student records of pain management completed during field internships where the veracity of the record was confirmed by the supervising paramedic.

The original dataset comprised 290,670 unique cases. After applying inclusion criteria, 59,915 cases were available for analysis. Race was recorded in 41,071 of these cases, with 31.1% of cases missing race data. Pain severity was recorded in 26,272 cases (43.8%). Median age was 50 years (IQR 39 years), with 50.1% female (n = 30,040). The analytic sample used in the logistic regression models discussed below and presented in Table 2 comprised 17,729 unique cases.

The most common cause of injury was falls. The cause of injury variable included 23 separate categories, and these were collapsed into 8 categories for the analysis. Categories combined included: drowning, lightning, radiation exposure, excessive cold and excessive heat into an ‘environmental’ category. Aircraft related accident, water transport accident, bicycle accident and motor vehicle accident were combined into a ‘transport’ category. A toxicological category included drug poisoning, chemical poisoning, chemical poisoning and venomous stings. Lastly, a ‘trauma other’ category was generated comprising mechanical suffocation, electrocution (non-lightning), fire or flames, rape, bites, machinery accident and smoke inhalation. Patient characteristics are shown in Table 1.

The unadjusted odds ratio for African American patients receiving analgesia is statistically significantly (p < 0.001) at 0.62 times that of Caucasian patients. The adjusted odds ratio for Hispanic patients receiving analgesia is statistically significant (p < 0.01) at 0.81 times that of Caucasian patients. The odds ratios for other racial categories included in the model (Asian, Native American and Other) were not statistically significant (p > 0.05). The Pseudo R-squared statistic suggests that 0.4% of variance in the outcome variable is explained by this model.

In model two, gender is added. The odds ratio for female, however, is not statistically significant (p > 0.05), suggesting that gender does not account for analgesia administration.

In model three, when patient age is added with the reference age category 18–40 the patients aged 61–80 have an odds of receiving analgesia 1.14 times that of patients age 18–40 years. The odds ratio of the African American category remains highly statistically significant, though increases slightly to an odds of 0.64. The Pseudo R-squared increases to 0.5% of variance explained in the model.

Model four adds pain score category to the analysis. Patients in severe pain had a statistically significant odds of receiving analgesia 2.45 times that of patients in moderate pain. This suggests that patients in severe pain have an increased odds of receiving analgesia in comparison to those in moderate pain. The Pseudo R-squared increased in this model to 0.104, indicating that this model explain 10.4% of variance in the outcome of receiving analgesia.

Model five adds cause of injury. Cause categories of family violence, firearm, stabbing/cutting and transport were all highly statistically significant (p < 0.05). The odds of an African American patient receiving an analgesic, with cause of injury included in the model, increases slightly from model four, but it is still lower than model one, suggesting that even when gender, age, pain score and cause of injury are accounted for, African American patients still have lower odds that Caucasian patients of receiving analgesia. The same conclusion can be applied to Hispanic patients as well, though they have a higher odds of receiving analgesia than African American patients. Table 2 presents a logistic regression analysis, including all independent variables, of the odds of a patient receiving any analgesic.

Previous studies have shown disparities in patient management for minority patients. These results have held true in multiple disciplines, but almost all studies have examined patients who are in a healthcare facility, not in the prehospital environment. Previous prehospital research has described an association between patient race and paramedic management of suspected cardiac chest pain. However, that study was limited by poor reporting of race and a very small number of non-white patients in the dataset [11]. One recent study reported a low frequency of paramedic administration of analgesics for painful injury and a reduced frequency of analgesia for Black patients. However, this study did not control for variables that may be associated with a reduced odds of analgesia administration [8].

This study contributes to the understanding of disparities in EMS by describing the association between race and analgesia administration by paramedic students, and reaffirms the pattern previously reported of suboptimal care being provided to minority patients.

How racial disparities are mitigated is a complicated question. Disparities affecting pain management decisions may arise from complex interactions between patient race, racial bias of the care provider, context and clinical ambiguity [12]. Professional associations and educational programs that train EMTs and paramedics should increase their conversations around healthcare disparities. Education should include guided reflective strategies to help the student understand pain from the patient’s perspective, as these strategies have been shown to induce empathy and reduce racial bias associated with the management of pain [13]. EMS training programs should ensure that educational design aims to increase their students’ exposure to multi-racial patients in all settings and across the lifespan. Introducing minority patients during simulation, using cognitive debriefing to carefully evaluate differences in treatment from similar cases involving white patients, and teaching that helps students understand the effect that subconscious bias often has on the quality of care may help to address disparities based on race. Currently, many EMS training manikins have ‘white’ skin, which means students may not have opportunities to treat non-white patients. As manufacturers now supply different skin tones for their training manikins, training programs should make concerted efforts to purchase manikins that better represent the patients their students will ultimately treat. At a national level, EMT education standards need to address the lack of detail regarding the development of cultural competency, as current standards fail to adequately address this need [7].

While education influences the practice of new graduates, these disparities are unlikely to be addressed solely through education. If EMT and Paramedic graduates begin work in a system which fails to asses for and mitigate racially based disparities in treatment, those graduates may inevitably conform to system norms. It is critically important, therefore, for those systems to actively assess for such disparities and design interventions that aim to reduce disparities [14]. EMS and Fire Department Quality Assurance and Quality Improvement processes (QA/QI) are typically designed to evaluate compliance with treatment protocols, but this process may not have the capacity to identify disparities in the management of specific case types across their system. Previous research has shown that clinician behavior can change when that behavior is specifically measured and reported [15]. QA/QI programs must consistently look for, measure and evaluate disparities such as those associated with race. Without robust investigations for racially based disparities in treatment, those disparities will likely remain.

This study also found a low (14%) rate of analgesia administered to patients with a complaint of pain. While informed patient refusal of analgesia and clinical guideline or protocol constraints may account for some of this finding, the low frequency of analgesia administration replicates previous research and requires further study and actions to improve pain management in this setting [8]. Patient disposition or lack of expected behavioral response to pain may be associated with cultural differences. Some EMS providers may believe that a patient not requesting pain management means that they do not require analgesia. If true, this belief is contrary to our responsibility to advocate on behalf of patients and to provide advice regarding the risks and benefits of treatment options.

Caucasian patients have significantly higher odds of receiving analgesia from a student paramedic in this study than non-Caucasian patients. When ethnicity is considered more closely, it is found that African Americans have the lowest odds of receiving any analgesia when compared to Caucasian patients. This adds to the evidence for disparities in the management of pain based on patient race and reinforces the need for clinical audit by EMS to identify and mitigate disparities in the quality of care based on patient characteristics such as race. Evidence arising from studies that aim to change health provider behavior to ensure equitable delivery of care should inform EMT education programs, in-service quality improvement education, and system design to eliminate racial disparities associated with pain management.