Abstract
This cross-sectional analysis of 86,111 visits for sickle cell disease and vaso-occlusive episodes (VOE) in U.S. pediatric emergency departments between 2013 and 2023 shows increased use of NSAIDs, ketamine, and acetaminophen, with unchanged opioid use. Hospitals with a higher volume of VOE visits more frequently administered opioids.
Background:
Vaso-occlusive episodes (VOEs) are a hallmark of sickle cell disease (SCD), leading to frequent emergency department (ED) visits. Effective pain management is crucial, with guidelines recommending routine use of non-steroidal anti-inflammatory drugs (NSAIDs) with opioids, and emerging evidence supporting ketamine use. However, these recommendations are based on low-certainty evidence, and the impact of these guidelines on analgesia use over time remains unclear.
Objective:
This study aimed to analyze trends in analgesia use over an 11-year period in pediatric SCD patients presenting to U.S. EDs with VOE and assess variations in treatment across hospitals.
Methods:
A cross-sectional study was conducted using data from the Pediatric Health Information System covering 34 U.S. children’s hospitals from January 1, 2013, to December 31, 2023. The primary outcomes were the proportions of visits where opioids, NSAIDs, acetaminophen, and/or ketamine were administered on the first calendar day of the initial visit. Secondary outcomes included the co-administration of NSAIDs with opioids. Logistic and linear regression models were used to assess trends and hospital-level variations.
Results:
A total of 86,111 ED visits for VOE were analyzed. Opioids were administered in 82 % of encounters, NSAIDs in 72 %, acetaminophen in 17 %, and ketamine in 1 %. Co-administration of NSAIDs with opioids occurred in 59 % of the visits. Among discharged patients, there was a positive trend for NSAID use (slope: 1.68 %/year, 95 % CI: 0.91 %, 2.45 %) and NSAID-opioid co-administration (slope: 1.03 %/year, 95 % CI: 0.37 %, 1.69 %) over time. Acetaminophen use also increased over the study period (slope: 0.99 %/year, 95 % CI: 0.80 %, 1.17 %). In hospitalized patients, there was a significant upward trend for acetaminophen (slope: 1.29 %/year, 95 % CI: 0.69 %, 1.89 %) and ketamine (slope: 0.36 %/year, 95 % CI: 0.27 %, 0.45 %), while opioid use remained unchanged. Significant hospital-level variations were observed, with larger hospitals more likely to administer opioids but less likely to co-administer NSAIDs with opioids compared to medium-volume hospitals.
Conclusion:
Over the past decade, the use of NSAIDs, acetaminophen, and ketamine has increased in the management of VOE in pediatric SCD patients, while opioid use remains consistent. The co-administration of NSAIDs and opioids has also increased, reflecting guideline adherence. Variations in analgesia practices across hospitals underscore the need for standardizing pain management strategies in this population.
Keywords: Pediatric emergency medicine, Pain management, Hospital practice patterns, Vaso-occlusive crisis, Analgesic trends, Multimodal analgesia, Opioid use, Non-steroidal anti-inflammatory drugs (NSAIDs)
1. Background
Vaso-occlusive episodes (VOEs) are a defining feature of sickle cell disease (SCD), accounting for 75 % of all SCD-related emergency department (ED) visits [1]. Effective pain management is critical for patients experiencing VOE in SCD, as these episodes are associated with significant morbidity [2]. The American Society of Hematology (ASH) [3] a nd the National Heart, Lung, and Blood Institute (NHLBI) [4] recommend use of non-steroidal anti-inflammatory drugs (NSAIDs) for all patients with VOE pain, and rapid administration of opioids, taking into account factors such as prior response to specific potencies and doses of opioids. Furthermore, the use of ketamine has been recommended for hospitalized children, supported by few trials showing its non-inferiority to morphine [3,5,6]. However, all these analgesia recommendations in the guidelines were based on low-certainty evidence, with several systematic reviews demonstrating mixed and inconclusive results [7,8], raising questions about whether analgesia use over time has been influenced by these guidelines.
Our objectives were to: 1) analyze trends over an 11-year period ending in 2023 for commonly used analgesia in children with SCD presenting to a U.S. ED with VOE, and 2) assess variations in hospital treatment of VOE. By understanding these trends and variations, we aim to gain valuable insights into the effectiveness of current guidelines and identify gaps in care. This knowledge has the potential to inform future research and guideline development, ultimately leading to improved pain management strategies and better outcomes for these patients.
2. Methods
We performed a cross-sectional study of ED encounters for patients with a VOE from 34 US children’s hospitals who provide data to the Pediatric Health Information System (PHIS). PHIS utilizes de-identified administrative data on ED and inpatient visits and is managed by the Children’s Hospital Association (Overland Park, Kansas) [9]. The study followed STROBE guidelines [10] and was exempt from local ethics review.
2.1. Cohort definition
We included all visits of patients aged 0 to 19 years to a PHIS hospital between January 1, 2013, and December 31, 2023, with a primary or non-primary diagnosis of a VOE using the International Classification of Diseases, 9th and 10th revisions diagnosis codes (supplement). We excluded encounters of patients transferred from another hospital or those with in-hospital mortality.
2.2. Outcomes
The primary outcomes were the proportions of visits in which opioids, NSAIDs, acetaminophen and/or ketamine were administered on the first calendar day of the initial visit. The secondary outcome was the proportion of children who were co-administered NSAIDs and opioids (i.e., patient visits in which NSAIDs were given, when opioids were used for pain relief). This outcome was chosen since both 2014 NHLBI guidelines and the 2018 ASH guidelines recommend NSAIDs in addition to opioids for VOEs [3]. The PHIS database provides a comprehensive list of Multum Lexicon drug codes corresponding to medications administered during the visit (supplement). We included all relevant codes for opioids, NSAIDs, acetaminophen, and ketamine across all routes of administration. Additionally, we abstracted demographic data (age, sex, race, primary payer), other diagnoses related to SCD (i.e., acute chest syndrome, splenic sequestration, cerebral vascular involvement) and non-hematologic chronic complex conditions [11].
2.3. Analysis
We stratified our analysis by disposition and by hospital. We used a linear regression model for each medication type to assess time trends, with the year of the visit as the independent variable and the proportion of visits receiving the medication as the dependent variable. Hospitals were categorized into small, medium, and large volume based on tertiles of the number of VOEs treated during each study period, rather than the overall hospital size. This approach reflects the specific experience of each hospital in managing sickle cell disease, where the volume of VOEs treated is a more relevant indicator of experience than the hospital’s total number of patients seen. Using percentiles, hospitals were divided into three groups: 11 hospitals were classified as small-volume (209 to 999 VOEs), 11 as medium-volume (1176 to 2507 VOEs), and 12 as large-volume (2637 to 15,251 VOEs). We calculated odds ratios (OR) with 95 % confidence intervals (CI) using logistic regression to compare the likelihood of receiving opioids and combination opioids and NSAIDs among hospital size groups. All statistical tests were 2 sided, and P values <0.05 were considered statistically significant. Analyses were performed by using SAS software version 9.4 (Statistical Analysis System Institute, Inc., Cary, NC), and trends were visualized using line plots generated with the matplotlib library in Python (version 3.8).
3. Results
There were 86,111 ED visits for VOE during the study period, and adolescents 15 years and older accounted for 37,294 (43 %) of the visits. Overall, 44,037 (51.1 %) were males, and 78,897 (92 %) were non-Hispanic Black. Visit numbers were stable over time (p = 0.540) and 48,149 (56 %) of visits resulted in hospitalization (Table 1).
Table 1.
Study cohort (N = 86,111): any diagnosis of vaso-occlusive episode.
| Characteristic | ED visits N (%) |
Discharged N (%) |
Hospitalized N (%) |
|---|---|---|---|
|
| |||
| N | 86,111 | 37,962 (44.1) | 48,149 (55.9) |
| Age (years) <1 |
1445 (1.7) |
631 (1.7) |
814 (1.7) |
| 1–4 | 9639 (11.2) | 5687 (15.0) | 3952 (8.2) |
| 5–9 | 16,216 (18.8) | 9120 (24.0) | 7096 (14.7) |
| 10–14 | 21,517 (25.0) | 8788 (23.1) | 12,729 (26.4) |
| 15–18 | 26,056 (30.3) | 9344 (24.6) | 16,712 (34.7) |
| 19–20 | 11,238 (13.1) | 4392 (11.6) | 6846 (14.2) |
| Sex: Male | 44,037 (51.1) | 19,947 (52.5) | 24,090 (50.0) |
| Race and Ethnicity Non-Hispanic White |
1093 (1.3) |
481 (1.3) |
612 (1.3) |
| Non-Hispanic Black | 78,897 (91.6) | 34,680 (91.4) | 44,217 (91.8) |
| Hispanic | 4079 (4.7) | 1647 (4.3) | 2432 (5.1) |
| Non-Hispanic Asian | 171 (0.2) | 83 (0.2) | 88 (0.2) |
| Other | 1871 (2.2) | 1071 (2.8) | 800 (1.7) |
| Payor Government |
63,263 (73.5) |
27,452 (72.3) |
35,811 (74.4) |
| Private | 18,955 (22.0) | 8412 (22.2) | 10,543 (21.9) |
| Other | 3893 (4.5) | 2098 (5.5) | 1795 (3.7) |
| Any Non-Hematologic Complex Chronic Condition |
8713 (10.1) | 915 (2.4) | 7798 (16.2) |
| Weekend ED encounter | 22,190 (25.8) | 9886 (26.0) | 12,304 (25.6) |
| Acute chest syndrome | 4333 (5.0) | 153 (0.4) | 4180 (8.7) |
| Splenic sequestration | 809 (0.9) | 28 (0.1) | 781 (1.6) |
| Cerebral vascular involvement |
8 (0.0) | 0 (0.0) | 8 (0.0) |
| Year 2013 |
7753 (9.0) |
2829 (7.5) |
4924 (10.2) |
| 2014 | 8410 (9.8) | 3125 (8.2) | 5285 (11.0) |
| 2015 | 8480 (9.8) | 3544 (9.3) | 4936 (10.3) |
| 2016 | 7691 (8.9) | 3729 (9.8) | 3962 (8.2) |
| 2017 | 7942 (9.2) | 3813 (10) | 4129 (8.6) |
| 2018 | 7700 (8.9) | 3477 (9.2) | 4223 (8.8) |
| 2019 | 8034 (9.3) | 3607 (9.5) | 4427 (9.2) |
| 2020 | 6273 (7.3) | 2844 (7.5) | 3429 (7.1) |
| 2021 | 7377 (8.6) | 3403 (9.0) | 3974 (8.3) |
| 2022 | 8067 (9.4) | 3758 (9.9) | 4309 (8.9) |
| 2023 | 8384 (9.7) | 3833 (10.1) | 4551 (9.5) |
| Drugs Administered Opioids |
70,283 (81.6) |
27,706 (73.0) |
42,577 (88.4) |
| NSAIDs | 62,204 (72.2) | 23,070 (60.8) | 39,134 (81.3) |
| Opioids and NSAIDs combination |
56,085 (65.1) | 19,391 (51.1) | 36,694 (76.2) |
| Acetaminophen | 14,871 (17.3) | 3185 (8.4) | 11,686 (24.3) |
| Ketamine | 990 (1.1) | 109 (0.3) | 881 (1.8) |
3.1. Trends in the use of analgesics
The most frequently used analgesics were opioids, which were administered in 70,283 (82 %) encounters, followed by NSAIDs in 62,204 (72 %), acetaminophen in 14,871 (17 %), and ketamine in 990 (1 %). Among discharged patients, there was a positive trend for NSAIDs use (slope: 1.68 %/year, 95 % CI: 0.91 % to 2.45 %), NSAID co-administration with opioids (slope: 1.03 %/year, 95 % CI: 0.37 % to 1.69 %), and acetaminophen use (slope: 0.99 %/year, 95 % CI: 0.80 % to 1.17 %) over time. Among admitted patients, there was a significant upward trend for acetaminophen (slope: 1.29 %/year 95 % CI: 0.69 % to 1.89 %) and ketamine (slope: 0.36 %/year 95 % CI: 0.27 %, 0.45 %), but there were no significant changes for NSAIDs or NSAIDs co-administration with opioids. Among admitted and discharged patients, there was no change in the use of opioids over time (Fig. 1).
Fig. 1.
Trend lines for opioids, Non-Steroidal-Anti Inflammatory-Drugs (NSAIDs), opioids-NSAID combination therapy, acetaminophen and ketamine.
3.2. Hospital-level variation
The frequency of opioid administration varied across hospitals, ranging from 52.3 % to 90.3 % with a median of 81.8 % (interquartile range [IQR] 79.0 %, 86.2 %). Similarly, the frequency of NSAID use among patients who received opioids also showed considerable variation, ranging from 14.7 % to 93.5 % with a median of 82 % (IQR 77.8 %, 89.2 %).
Large-volume hospitals were more likely to administer opioids compared to small-volume hospitals (OR: 1.20, 95 % CI: 1.13 to 1.28) and medium-volume hospitals (OR: 1.20, 95 % CI: 1.15 to 1.24). There was no difference in the frequency of opioid administration between small-volume and medium-volume hospitals (OR: 1.01, 95 % CI: 0.94 to 1.08). Large-volume hospitals were less likely to co-administer NSAIDs with opioids than medium-volume hospitals (OR: 0.74, 95 % CI: 0.71 to 0.78) but not compared to small volume hospitals (OR: 0.97, 95 % CI: 0.91 to 1.04). Medium hospitals were more likely to co-administer NSAIDs with opioids than small-volume hospitals (OR: 1.31, 95 % CI: 1.21 to 1.41) (Fig. 2).
Fig. 2.
Scatter plot for Hospital Variation in opioid and non-steroidal anti-inflammatory drugs (NSAIDs) and opioid co-administration.
4. Discussion
In this 11-year cross-sectional study of 34 children’s hospitals, the use of NSAIDs alone or in combination with opioids, acetaminophen, and ketamine all had an upward trend in use, whereas opioid use was unchanged. The use of ketamine, though rare, increased 9fold during the study period, primarily among hospitalized children, indicating a growing recognition of its use in managing severe pain. The variability in opioid administration across hospitals, with large-volume hospitals being more likely to use opioids and less likely to use NSAIDs in combination, highlights differences in clinical practice and resource availability and underscores opportunities to standardize care.
Our analysis revealed that patients discharged from the ED were more likely to receive NSAIDs alone or in combination with opioids, which is consistent with recent ASH and NHLBI guideline updates [3,4]. These emphasize the importance of multimodal analgesia, aiming to minimize opioid use while ensuring effective pain management. The stable use of opioids, despite the upward trend in NSAID use, indicates that opioids remain a cornerstone of VOE pain management. This may reflect their established role in providing immediate relief for acute pain episodes in SCD.
The higher utilization of opioids in larger-volume hospitals could be attributed to greater familiarity and confidence in managing severe VOE pain in these settings. Larger-volume hospitals may also have more resources and protocols in place to standardize pain treatment, while providing adequate monitoring and management of potential opioid-related complications. Smaller-volume hospitals may be impeded by lack of those resources and some of the known barriers to using opioids in patients with SCD, such as fear of drug abuse, and skepticism in patients’ report of pain severity among clinicians and nurses, [12] and beliefs that opioid addiction frequently develops in patients with SCD [13]. This has not been supported by empirical evidence, with a recent study showing that SCD is not a risk factor for opioid use disorders relative to other chronic conditions [14].
Our multicenter study builds upon previous local evidence evaluating long-term trends of pharmacologic treatment for patients with SCD. Prior single-center retrospective chart reviews have suggested that opioid use ranges from 70 to 80 % for treatment of VOE [15,16]. After publication of the NHLBI guidelines, a cross sectional analysis of data from 7 Pediatric Emergency Care Applied Research Network sites over a 3-year period including 4578 visits was published [17]. This analysis focused on the time to opioid administration but did not provide data on the frequencies of different analgesics used [17]. They found poor adherence to timely management of pain with opioids with only half of the visits classified as guideline congruent (receiving opioid within 60-min of arrival) [17]. These findings, highlight the need for ongoing efforts to standardize pain management practices across pediatric EDs, ensuring that all children with SCD receive equitable care regardless of the hospital they visit.
4.1. Limitations
The retrospective design and reliance on an administrative database restrict our ability to adjust for treatment predictors such as pain severity, pre-hospital (EMS or home) treatments administered, socioeconomic status, or specific clinical protocols in different hospitals. Additionally, the de-identified nature of the PHIS database limits our capacity to conduct detailed patient-level analyses, which could provide deeper insights into individual treatment responses and outcomes. Our definition of first day of treatment (calendar day) may have resulted into an under-estimation of the proportions of patients receiving analgesia if patients presented late in the evening. All hospitals included in this study are tertiary care pediatric hospitals, and thus the findings may not reflect practice in general EDs. Furthermore, the exclusion of patients transferred from other hospitals or those with in-hospital mortality might introduce selection bias, potentially underestimating the severity of cases managed in smaller hospitals.
5. Conclusion
Our study provides valuable insights into the evolving practice patterns of analgesia use for VOE in children with SCD across U.S. children’s hospitals. While the use of NSAIDs, acetaminophen, and ketamine has increased, opioids use has not changed and remain a critical component of VOE pain management. The observed variations in analgesia practices highlight the need for ongoing efforts to standardize and optimize pain management strategies in this vulnerable population.
Supplementary Material
Abbreviations:
- VOE
Vaso-occlusive episodes
- SCD
Sickle cell disease
- NSAIDs
Non-steroidal anti-inflammatory drugs
- ED
Emergency department
- ASH
American Society of Hematology
- NHLBI
National Heart, Lung, and Blood Institute
- PHIS
Pediatric Health Information System
- STROBE
Strengthening the Reporting of Observational Studies in Epidemiology
- OR
Odds ratio
- CI
Confidence interval
- IQR
Interquartile range
- SAS
Statistical Analysis System (software)
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ajem.2024.10.028.
Footnotes
CRediT authorship contribution statement
Mohamed Eltorki: Writing – review & editing, Writing – original draft, Visualization, Software, Project administration, Methodology, Formal analysis, Data curation, Conceptualization. Matt Hall: Writing – review & editing, Methodology, Formal analysis, Data curation, Conceptualization. Sriram Ramgopal: Writing – review & editing, Visualization, Methodology, Investigation, Conceptualization. Pradip P. Chaudhari: Writing – review & editing, Methodology, Investigation, Conceptualization. Oluwakemi Badaki-Makun: Writing – review & editing, Methodology, Conceptualization. Chris A. Rees: Writing – review & editing, Investigation, Conceptualization. Kelly R. Bergmann: Writing – review & editing, Methodology, Conceptualization. Daniel J. Shapiro: Writing – review & editing, Methodology, Investigation, Conceptualization. Frank Gonzalez: Writing – review & editing, Methodology, Conceptualization. Timothy Phamduy: Writing – review & editing, Methodology, Investigation, Conceptualization. Mark I. Neuman: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
- [1].Lanzkron S, Carroll CP, Haywood C Jr. The burden of emergency department use for sickle-cell disease: an analysis of the national emergency department sample database. Am J Hematol. 2010;85(10):797–9. 10.1002/ajh.21807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Bailey M, Abioye A, Morgan G, Burke T, Disher T, Brown S, et al. Relationship between Vaso-occlusive crises and important complications in sickle cell disease patients. Blood. 2019;134(Supplement_1):2167. 10.1182/blood-2019131721. [DOI] [Google Scholar]
- [3].Brandow AM, Carroll CP, Creary S, Edwards-Elliott R, Glassberg J, Hurley RW, et al. American Society of Hematology 2020 guidelines for sickle cell disease: management of acute and chronic pain. Blood Adv. 2020;4(12):2656–701. 10.1182/bloodadvances.2020001851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Yawn BP, Buchanan GR, Afenyi-Annan AN, Ballas SK, Hassell KL, James AH, et al. Management of sickle cell disease: summary of the 2014 evidence-based report by expert panel members. JAMA. 2014;312(10):1033–48. 10.1001/jama.2014.10517. [DOI] [PubMed] [Google Scholar]
- [5].Lubega FA, DeSilva MS, Munube D, Nkwine R, Tumukunde J, Agaba PK, et al. Low dose ketamine versus morphine for acute severe vaso occlusive pain in children: a randomized controlled trial. Scand J Pain. 2018;18(1):19–27. 10.1515/sjpain-2017-0140. [DOI] [PubMed] [Google Scholar]
- [6].Alshahrani MS, AlSulaibikh AH, ElTahan MR, AlFaraj SZ, Asonto LP, AlMulhim AA, et al. Ketamine administration for acute painful sickle cell crisis: a randomized controlled trial. Acad Emerg Med. 2022;29(2):150–8. 10.1111/acem.14382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Lowe M, Bambhroliya Z, Patel H, Patel VJ, Vudugula SA, Cheruvu NP, et al. Emerging therapies for the Management of Pain and Vaso-Occlusive Crises in patients with sickle cell disease: a systematic review of randomized controlled trials. Cureus. 2023;15(4):e38014. 10.7759/cureus.38014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Cooper TE, Hambleton IR, Ballas SK, Johnston BA, Wiffen PJ. Pharmacological interventions for painful sickle cell vaso-occlusive crises in adults. Cochrane Database Syst Rev. 2019;2019(11). 10.1002/14651858.CD012187.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Pediatric Health Information System. Children’s hospital association. Available at. https://www.childrenshospitals.org/; 2024 August 20.
- [10].von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008; 61(4):344–9. 10.1016/j.jclinepi.2007.11.008. [DOI] [PubMed] [Google Scholar]
- [11].Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr. 2014;14:199. 10.1186/1471-2431-14-199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Pack-Mabien A, Labbe E, Herbert D, Haynes J Jr. Nurses’ attitudes and practices in sickle cell pain management. Appl Nurs Res. 2001;14(4):187–92. 10.1053/apnr.2001.26783. [DOI] [PubMed] [Google Scholar]
- [13].Zempsky WT. Evaluation and treatment of sickle cell pain in the emergency department: paths to a better future. Clin Pediatr Emerg Med. 2010;11(4):265–73. 10.1016/j.cpem.2010.09.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Jonassaint CR, O’Brien J, Nardo E, Feldman R, Stanton M, DeCastro L, et al. Prevalence of substance use disorders in sickle cell disease compared to other chronic conditions: a population-based study of black American adults. J Gen Intern Med. 2023; 38(5):1214–23. 10.1007/s11606-022-07786-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Zempsky WT, Loiselle KA, McKay K, Lee BH, Hagstrom JN, Schechter NL. Do children with sickle cell disease receive disparate care for pain in the emergency department? J Emerg Med. 2010;39(5):691–5. 10.1016/j.jemermed.2009.06.003. [DOI] [PubMed] [Google Scholar]
- [16].Frei-Jones MJ, Baxter AL, Rogers ZR, Buchanan GR. Vaso-occlusive episodes in older children with sickle cell disease: emergency department management and pain assessment. J Pediatr. 2008;152(2):281–5. 10.1016/j.jpeds.2007.06.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Brousseau DC, Alpern ER, Chamberlain JM, Ellison AM, Bajaj L, Cohen DM, et al. A multiyear cross-sectional study of guideline adherence for the timeliness of opioid administration in children with sickle cell pain crisis. Ann Emerg Med. 2020;76 (3S):S6–S11. 10.1016/j.annemergmed.2020.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.