Patient-reported outcomes in children with sickle cell disease at presentation for an acute pain episode

TO THE EDITOR:

Patient reported outcomes (PROs) provide multidimensional perspectives on the impact of pain^1-5 and have been assessed using the Patient Reported Outcomes Measurement Information System (PROMIS) in children with sickle cell disease (SCD) during outpatient visits,^3,4 acute care utilization for pain,^4,6 and in clinical trials to measure the effect of interventions.^7-9 Here, we report the feasibility of collection of PROs of pain interference and fatigue at enrollment (baseline) from participants 3 to 21 years of age enrolled in a phase 2 randomized double-blind placebo-controlled trial of intravenous arginine therapy for acute SCD pain (NCT02536170),¹⁰ and examine the impact of an acute pain episode on these PROs.

Study design has been previously published.¹⁰ At, or soon after enrollment, study participants and/or their parents completed the self-report pediatric or parent-proxy version of the 8-item PROMIS Pediatric Pain Interference v1.0 Short Form (SF) 8a and the 10-item PROMIS Pediatric Fatigue v1.0 SF 10a,^2-4,6 which generated domain-specific standard T-scores.¹¹ These questionnaires have a 7-day recall and are valid, reliable, and responsive to change in pain in SCD.^2-4,6 Demographic, clinical, and laboratory data were as collected during the clinical trial, and prior health care utilization (HCU) for pain represented the sum of emergency department visits and hospitalizations for pain in the 12-month period before enrollment. We defined high HCU for pain as ≥3 episodes of HCU for pain, given that this threshold is a marker of severity in SCD.^12,13 This study was approved by the Institutional Review Board at Emory University and was conducted in accordance with the Declaration of Helsinki.

Ninety-eight (mean age, 12.8; standard deviation [SD], 3.7 years; 53.1% female) of the 108 enrolled participants (90.7%) had at least 1 completed PRO (n = 94, for pain interference; n = 98, for fatigue), either by child self-report or by parent-proxy report (for n = 4, we retained self-report versions because both self-report and parent-proxy forms were completed). Both pain interference and fatigue domains were completed for 94 participants (87%), and neither of the domains for 10 participants (9.3%). For 64 participants (mean age, 14.1; SD, 3 years), at least 1 self-report pediatric PROMIS questionnaire was available, and for 35 participants, at least 1 parent-proxy PROMIS questionnaire was available (mean age, 10.5; SD, 3.6 years). Our experience confirms that collection of pain-related PROs is feasible in a clinical trial of an acute pain intervention in SCD, consistent with previous experience in observational and intervention studies of acute pain^4,6,14,15 in SCD.

Participant characteristics are reported in Table 1. Domain scores are reported in Table 2 and supplemental Figure 1. To determine the severity of patient reported symptoms, we used thresholds for self-report pediatric pain interference in SCD, in which a T-score of ≤48 represents mild symptoms, 49 to 64 represents moderate symptoms, and >64 represents severe symptoms.¹ Because severity thresholds are not described for self-report pediatric fatigue in SCD, we used thresholds established in juvenile idiopathic arthritis and systemic lupus erythematosus,¹⁶ in which a T-score of ≥57.5 represents moderate fatigue, and ≥67.5 represents severe fatigue. Most self-report pediatric pain interference scores (n = 53, 88.3%) met the threshold for moderate (n = 26) or severe (n = 27) pain interference, and most self-report pediatric fatigue scores (n = 41, 64%) met the threshold for moderate (n = 25) or severe (n = 16) fatigue. Our findings indicate that the majority experience at least moderate pain interference and fatigue when they experience pain that results in an acute care visit that requires parenteral opioids. As compared to those with mild pain interference, those with moderate to severe pain interference were more likely to have been prescribed NSAIDs (P = .034). We found that females were more likely to have moderate to severe fatigue compared with males (78.1% females vs 50% males; P = .037), consistent with published reports of higher fatigue in female children³ and adults with SCD.^17,18 We also found that those that were prescribed hydroxyurea (HU) were more likely to have at least moderate fatigue than those that were not (73.3% vs 42.1%; P = .036). However, the prescription of HU was associated with both a severe sickle cell genotype and high HCU for pain, so the relationship between fatigue and HU may be confounded by disease severity or may reflect the prescription of HU to those with more severe disease, as has been previously reported.¹⁹ No other differences in demographic, clinical, or laboratory characteristics between those who had at least moderate symptoms as compared with those who had mild symptoms were observed, including lack of differences in hemoglobin levels or mean corpuscular volume.

We then examined the effects of age (<12 years and ≥12 years), sex, HU, and history of high HCU for pain (independent variables) on PROs (dependent variables) in bivariable and multivariable models using general linear regression (Table 2). We found no differences in fatigue and pain interference by age in this study, different from reports of higher pain interference during routine clinic visits in children >12 years compared with younger children.³ In bivariable models, we observed a moderate effect size (ES, 0.45; P = .070) for female sex on self-reported fatigue. In multivariable models adjusted for all clinical covariates, we observed moderate effect sizes for female sex on self-report fatigue (ES, 0.48; P = .072) and pain interference (ES, 0.47; P = .095) scores, but these were not statistically significant. Higher pain interference and fatigue have been reported in female children and adults with SCD,^3,17,18 so the lack of significant associations during an acute pain episode must be examined in future studies.

We did not observe an association between genotype and PROs, consistent with reports from outpatient visits.³ We also did not observe an association of PROs with high HCU for pain, different from reports in which pain interference and fatigue are correlated with HCU for pain.²⁰ We did not find an association between HU and pain interference, consistent with lack of significant effects of HU on change in pain interference and other pain-related PROs after an acute pain episode.⁶ Contrary to findings in children with chronic pain conditions including SCD during outpatient visits,²¹ we did not observe a significant correlation between pain interference or fatigue with the highest pain score reported at baseline (all Pearson product moment correlations, −0.2 < ρ < 0.2; P > .1). Similarly, we did not observe significant correlations between pain interference or fatigue with hemoglobin or mean corpuscular volume values, contrary to reported associations^18,22 in SCD during clinic and hospital visits.

The lack of association between PROs and clinical correlates in this study may have several possible explanations: effects may be minimized during an acute pain episode, may not have been observed after accounting for the effect of other clinical covariates, or have lacked adequate power to detect an association. The presence of chronic pain was not collected in this study. Because chronic pain impacts PROs,¹⁷ it is crucial to study the impact of acute pain on PROs among those with chronic pain, as is being done in a subsequent randomized study of arginine in SCD (NCT04839354).²³ 

In summary, we report that PROs complement clinical outcomes and assessment of the impact of acute pain in SCD, concordant with the recommendations from the American Society of Hematology to include PROs as endpoints in clinical trials.²⁴ This study also indicates the need to examine the effects of clinical and biological correlates on PROs in adequately powered studies, specifically in the context of an acute painful episode.

Acknowledgments: The authors thank the patients and families who participated in the study. The authors acknowledge the Emory and Children’s Healthcare of Atlanta Pediatric Emergency Medicine Research Coordinator team for their assistance with this study. They acknowledge the support and contributions from Children’s Healthcare of Atlanta and Emory University Pediatric Biostatistics Core. They also thank Srikant Rangaraju for assistance with figures.

This study was supported by the Food and Drug Administration (R01FD004814-01A2) and in part, by the National Center for Complementary and Integrative Health K24AT009893-01 (C.R.M.). N.B. received funding from the National Heart, Lung, and Blood Institute of the National Institutes of Health under award numbers 1K23HL140142 and 1K23HL140142-03S1, from the Program for Retaining, Supporting, and EleVating Early-career Researchers at Emory (PeRSEVERE) from the Emory School of Medicine, a gift from the Doris Duke Charitable Foundation, and through the Georgia Clinical and Translational Science Alliance under award UL1-TR002378. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Contribution: N.B. and Z.L. designed and performed data analysis; N.B., Z.L., and C.R.M. wrote the manuscript; N.B., Z.L., S.G., R.K., C.A.R., C.D., and C.R.M. interpreted the data; C.R.M., S.G., and R.K. were responsible for data analysis; D.L., P.K., and K.K. performed data collection; C.D. and C.R.M. were responsible for the study design; C.R.M. was responsible for acquisition of funding and data collection; and all authors revised the manuscript.

Conflict-of-interest disclosure: C.D. has received consulting fees from CLS Behring, Cyclerion, Epizyme, Glycomimetics, Hilton Publishing Company, Micelle BioPharma, Modus Therapeutics, Novartis, and Pfizer Inc, and research grants from Merck, Micelle BioPharma, Novartis, and Pfizer Inc. C.R.M. is the inventor or co-inventor of several UCSF-Benioff Children’s Hospital Oakland patents/patent-pending applications that include nutritional supplements (receiving royalties), and biomarkers of cardiovascular disease related to arginine bioavailability; is an inventor or co-inventor of several Emory University School of Medicine patent application for nutritional supplements for autism and coronaviruses; is a consultant for Roche and CSL Behring; and is an editor for the sickle cell disease-fever reference for UpToDate. The remaining authors declare no competing financial interests.

Correspondence: Nitya Bakshi, 2015 Uppergate Dr, Atlanta, GA 30322; e-mail: nitya.bakshi@emory.edu.