Minutes Matter: An Audit of the Timeliness of Opioid Administration for Sickle Cell Disease Pain in the Emergency Department and Association With Disposition

Recurrent pain episodes are a frequent complication of sickle cell disease (SCD), leading to frequent emergency department (ED) visits [1, 2]. These pain episodes significantly impair quality of life, contribute to substantial healthcare utilization, and increase the risk of other serious complications [2, 3]. Clinical guidelines from the American Society of Hematology (ASH) [4] and the National Heart, Lung, and Blood Institute (NHLBI) [5] recommend that the first dose of pain medication be administered within 60 min of ED arrival, with subsequent doses given at 30- to 60-min intervals. Despite these recommendations, many patients do not receive timely pain management in the ED [6-9]. A significant limitation in existing research is the lack of comprehensive data on pain management timeliness among adults with SCD, as most studies focus on pediatric populations [6, 7]. Empirical evidence suggests that adults experience longer ED delays for pain relief compared to children [9] but robust, multisite data specifically for adults are lacking. Additionally, previous studies in pediatric SCD patients show that timely opioid administration was associated with decreased hospitalizations, but it remains unclear if this is generalizable to adults [10]. To address this gap, we analyzed de-identified electronic health records (EHRs) from over 260 U.S. health systems using the EPIC Cosmos research platform [11, 12]. Datapoints in EPIC Cosmos include patient-level demographic information, encounter-level diagnosis codes, ED arrival times, ED disposition, medication types, routes of administration, and medication administration times [12]. We defined the study cohort as patients ≥ 19 years who visited the ED between 1/1/2019 and 12/31/2024 with a primary diagnosis of SCD with pain crisis, received at least one opioid dose, and did not have a fever above 38.5°C or any other additional SCD complication. The primary outcome was the proportion of ED visits where opioid pain medication was administered according to guidelines-recommended timings for the first and second doses. The secondary outcome was ED disposition (hospitalization vs. discharge). We conducted a visit-level analysis, allowing patients to have multiple ED visits. We measured the median time and interquartile range (IQR) from ED arrival to first opioid administration and the median time interval and IQR between the first and second doses. Guideline adherence for the first dose was calculated as a percentage of overall visits where the first opioid was given within 60 min of ED arrival. The second-dose guideline adherence was calculated in two parts: the percentage of visits in which the second dose was administered within a 30-min interval of the first dose (consistent with NHLBI guidelines), and the percentage of visits in which the second dose was administered within a 60-min interval of the first dose (consistent with ASH guidelines). We calculated guideline adherence and 95% Confidence Intervals (CIs) for the first two opioid doses, stratified by route of administration. We evaluated the association between timeliness and demographic factors, including age categories, sex, insurance status, ED arrival time, and frequency of ED use. ED utilization was categorized according to previous research [13] by the highest frequency of visits in any study year, ranging from low (≤ 1 visit) to super high (≥ 33 visits). For statistical analyses, we used multilevel logistic regression models, adjusting for clustering within patients and sites and controlling for covariates such as age, sex, insurance, ED utilization, arrival time, site, and Emergency Severity Index (ESI) levels. We further examined whether timely opioid administration according to the guidelines was associated with lower hospitalization rates. We dichotomized ED disposition into hospitalized (admitted or observed) vs. not hospitalized, excluding visits where patients left without being seen, left against medical advice, died, or were transferred. Multi-level logistic regression models estimated the odds of hospitalization by guideline adherence category, adjusting for age in years, sex, ED utilization, acuity level, and site variation. We obtained ethical approval from Nemours Children's Health Institutional Review Board (IRB) and performed statistical analyses using Python. Statistical significance was determined by 95% CIs. The final analytic sample included 347 404 ED visits for uncomplicated SCD pain by 31 661 patients across 230 ED sites. Table 1 summarizes the study's ED visit characteristics and demographics. All visits included at least one opioid dose, with 78% involving multiple doses. Patients had a median age of 33 years, 56% were female, and 75% used public insurance. High- and super-high ED utilizers accounted for 14% of patients but 70% of visits. Only 29% of SCD pain visits were classified as ESI levels 1 or 2, and overnight visits were least frequent (25%). Across the whole sample, the median time from ED arrival to the administration of the first opioid dose was 90 (IQR: 55–157) minutes, while the median time to the second dose was 74 (IQR: 52–109) minutes. Guideline adherence for the timeliness of the first opioid dose was achieved in 29.6% [95% CI: 29.4%–29.8%] of visits. Parenteral administration had slightly higher adherence (30.1% [95% CI: 29.9%–30.2%]) than oral (26.1% [95% CI: 25.7%–26.5%]), and there was slight annual variation in first-dose adherence rates (27%–33%). Among visits with multiple doses, only 6.9% [95% CI: 6.8%–7.0%] were within 30 min, and 34.6% [95% CI: 34.4%–34.8%] were within 60 min of the first. Oral second doses were more likely to be timely than parenteral when given within 30 min, but adherence was similar between routes within 60 min. Demographic and visit characteristics were associated with guideline adherence (Table 2). Males had a higher rate of timely first doses 33.7% [95% CI: 33.5%–34.0%] than females 26.3% [95% CI: 26.1%–26.5%]. Guideline adherence was higher for young adults aged 19–30 years (31.4% [95% CI: 31.2%–31.6%]) compared to older age groups; visits by patients ≥ 65 years had the lowest guideline adherence (22.4% [95% CI: 21.3%–23.6%]). Overnight visits had the highest adherence of 39.2% [95% CI: 38.9%–39.6%] compared to evening hours with adherence of 23.5% [95% CI: 23.2%–23.7%]. Adjusted analyses confirmed these trends: compared to females, males had 36% higher odds of receiving timely opioids [aOR = 1.36: 95% CI (1.34–1.38)]; patients ≥ 65 years had 25% lower odds than young adults [aOR = 0.75: 95% CI (0.70–0.81)]; and overnight arrivals had the best adherence odds, with evening visits the lowest. Across all visits included in this analysis, receiving the first opioid dose within 60 min of ED arrival was associated with 27% lower odds of hospitalization [aOR = 0.73; 95% CI (0.70, 0.76)]. Table 3 displays the adjusted odds of hospitalization for 78% of the overall visits with at least two opioid doses. Compared to visits where the timeliness of both the first and the second opioid doses were not guideline adherent, receiving only a timely guideline-adherent first dose within 1 h of arrival was associated with 27% lower odds of hospitalization [aOR = 0.73; 95% CI (0.70, 0.77)], while having only a timely guideline-adherent second dose (within 60 min interval of the first) was associated with 25% lower odds of hospitalization [aOR = 0.75; 95% CI (0.71, 0.78)]. The lowest odds of hospitalization were observed when the timeliness of both the first and second opioid doses were guideline adherent [aOR = 0.62; 95% CI (0.57, 0.66)]. This analysis demonstrates that only one-third of uncomplicated SCD pain episodes received their first opioid dose within the recommended 1 h of ED arrival. Adherence to the guidelines for a second opioid dose within 30 min was even lower, occurring in just 7% of visits. These low rates of adherence remained consistent even during the COVID-19 pandemic, highlighting persistent challenges in delivering timely pain management for adults with SCD. Comparing these results with studies from pediatric EDs suggests better timeliness at pediatric EDs [6, 7]. Possible explanations include more robust protocols, lower suspicion of opioid-seeking behavior, greater empathy towards children, and lower patient acuity in pediatric EDs [14-18]. The observed sex disparity, with male patients more likely to receive guideline-adherent care, suggests the need for further exploration of sex-based differences in presentation and treatment. Analysis of ED utilization patterns revealed that frequent ED use did not correlate with receiving guideline-adherent pain management. Timeliness also varied by time of day, with the lowest adherence during evening hours, possibly due to ED crowding or staffing issues. Importantly, the timely administration of opioids according to guidelines was associated with significantly lower odds of hospitalization. Patients who received both the first and second opioid doses according to the guidelines had up to a 38% lower hospitalization rate. This association had not been previously demonstrated in adults, reinforcing the importance of prompt pain relief and supporting the implementation of standardized protocols, clinical decision support tools, and targeted staff education in EDs that care for adults. Despite leveraging a large, multisite, de-identified database, our analysis was limited by the inability to account for certain ED and patient characteristics, pain severity, and pre-ED analgesia use. Nevertheless, our study provides the most comprehensive evidence to date on the relationship between timely opioid administration and ED outcomes for adults with SCD pain, emphasizing the urgent need for quality improvement in ED pain management practices. The authors confirm contribution to the paper as follows: study conception and design: David C. Brousseau, Ibrahim Gwarzo, Harish chandra Dega; data analysis and interpretation of results: Harish chandra Dega, David C. Brousseau, Ibrahim Gwarzo; manuscript preparation: Ibrahim Gwarzo; manuscript and data review for quality: David C. Brousseau; study supervision: David C. Brousseau. All authors reviewed the results and approved the final version of the manuscript. We gratefully acknowledge the use of de-identified data from Cosmos (Epic Systems Corporation), a collaborative dataset aggregating electronic health records from participating health systems across the United States. This project was supported by the Delaware Sickle Cell Disease Center for Biomedical Research Excellence (COBRE) with funding from the National Institute of Health (NIH). Project Number: 5P20GM109021-07. This study was approved by the Nemours Children's Health institutional review board. The authors have nothing to report. The authors declare no conflicts of interest. The de-identified data used in this analysis are available from the Epic Cosmos Research Platform (cosmos.epic.com/about). Restrictions apply to the availability of these data.