Healthcare disparities and inequities: Definitions for the current era

The Centers for Disease Control and Prevention (CDC) defines health equity as the state in which everyone has a fair and just opportunity to attain their highest level of health. ¹ A disparity is a difference in health, services, or outcomes by some variable (eg, age, race, or insurance), which may or may not be clinically justifiable. These variables have often been collected in the medical record for reasons unrelated to their influence on health measures. Contemporary definitions and descriptions for health equity, disparity, inequity, and determinants of health are presented in Table 1. Multiple interconnected determinants of health, structural racism, and provider and system biases contribute to disparities and inequities across the healthcare spectrum. ^{2–Reference Yehia, Winegar and Fogel10}

Table 1.

Contemporary Health Equity, Disparity, and Inequity Definitions

Note. Use and meaning of these terms continues to evolve. ^{1–Reference Braveman, Orleans, Proctor and Plough3} Determinants of health encompass many, often interconnected, categories.

A test case for evaluating health inequities: Antibiotic prescribing for respiratory conditions in urgent care

Intermountain Health is a nonprofit, integrated, healthcare delivery system in the Mountain West that operates 38 urgent-care clinics. We limited our exploratory analysis to encounters from July 1, 2018, to June 30, 2019, among adults aged ≥18 years. We selected patient characteristics for assessment based on data availability in our electronic health record (EHR). These included age group (18–64 years and ≥65 years), race, ethnicity, preferred language, clinician–patient sex combination, and clinician type (physician or advanced practice clinician [APC]). Patient race, ethnicity, preferred language, and sex were self-reported. Additionally, we included body mass index (BMI) ≥25 and <25 (overweight/obese and nonobese). Individual respiratory encounters were identified using a validated methodology based on International Classification of Diseases, Tenth Edition codes. ^{Reference Stenehjem, Wallin and Fleming-Dutra21} Antibiotic prescribing rates for respiratory conditions overall and rates for tier 3 respiratory conditions (ie, conditions in which antibiotics are not indicated such as acute uncomplicated bronchitis) were assessed. ²² Because no standard definition exists to identify a disparity or inequity, we considered an absolute percentage difference between groups within a characteristic group of ≥5.0% to represent a potential disparity or inequity for which further evaluation was merited.

We identified 122,930 (88.1%) respiratory urgent-care encounters in adults aged 18–65 years. Among them, 98% of patients were White, 89.6% were non-Hispanic, 98.3% preferred English, 60.4% were female, and 85.3% were seen by physicians. Of non-English speakers, Spanish was the preferred language (77.1%). Most clinicians were male (77.0%); 59,442 (42.6%) of all respiratory encounters were for tier 3 conditions; and 68.3% of encounters occurred in overweight adults (BMI ≥ 25).

Patient groups in which notable differences were identified included age, race, ethnicity, and preferred language (Table 2). Adults aged ≥65 years received more prescriptions for tier 3 conditions than patients aged 18–65 years (30.2% vs 20.8%). White patients received antibiotic prescriptions more frequently than their nonwhite counterparts overall (50.1% vs 39.5%) and for tier 3 encounters (22.8% vs 14.9%). Non-Hispanic patients received antibiotic prescriptions overall more frequently than Hispanic patients (50.0% vs 43.9%). Patients who preferred English received more overall antibiotic prescriptions compared to non-English preferred speakers (49.6% vs 43.2%). Among clinician groups, female clinicians prescribed antibiotics for tier 3 conditions less frequently than their male colleagues (18.5% vs 23.6%) and this distinction was preserved regardless of patient gender. APCs prescribed antibiotics overall and for tier 3 conditions similarly to physicians.

Table 2.

Antibiotic Prescribing for Urgent-Care Respiratory Encounters in Patients Aged ≥18 Years Between July 1, 2018, and June 30, 2019

Note. Abx, antibiotics prescription; BMO, body mass index; MD/DO, medical doctor/doctor of osteopathy. Tier 3 codes are those where antibiotics are not indicated (eg, acute uncomplicated bronchitis).

^a Absolute differences of ≥5.0% between groups within each category are indicated in bold.

Future directions: Auditing (and resolving) disparities and inequities

Our EHR-based evaluation of health disparities and inequities in antibiotic prescribing can be used by health systems to examine other clinical conditions, therapeutics, and patient outcomes. Healthcare systems should create diverse teams of researchers, clinicians, and community members to evaluate observed differences and identify actions to achieve equity in healthcare delivery in the unique context in which any healthcare organization operates. Modification of system-level factors (eg, language support services, clinician language fluency, operational hours, telehealth accessibility, and a diverse system leadership team) should be considered in addition to evaluating and addressing clinician bias. This approach may also highlight and reinforce equitable care when observed. Although this exploratory analysis has limitations, the value of our approach using EHR data, and working to improve its collection and quality, lies in presenting a sustainable and adaptable mechanism to identify and monitor interventions that aim to reduce health inequities in healthcare delivery.