Abstract
OBJECTIVE
The aim of this study was to review the use of patient-controlled analgesia (PCA) in sickle cell disease (SCD) for pediatric patients with vaso-occlusive crisis (VOC) in our institution and to compare the effect of early vs late PCA start on pain relief and LOS.
METHODS
This retrospective study included all pediatric patients treated with PCA for a severe VOC from 2010 to 2016. “Early-PCA” was defined as start of PCA within 48 hours of arrival. Time to reach adequate analgesia was defined as the time to reach 2 consecutive pain scores less than 5/10 at 4-hour interval.
RESULTS
During the study period, 46 patients presented 87 episodes of VOC treated with PCA. Sixty-three patients with VOC were treated with Early-PCA and 24 with Late-PCA. Both groups were comparable except for median pain score at admission; the Early-PCA group had higher scores: 9.0/10 vs 7.0/10. Time to reach adequate analgesia could be evaluated only in a subset of patients (n = 32) but was shorter in the Early-PCA group with a median difference of 41.0 hours (95% CI −82.0 to −6.0). Early-PCA was associated with a median reduction in LOS of 3.4 days (95% CI −4.9 to −1.9). There was no difference between the 2 groups in terms of side effects and occurrence of acute chest syndrome during hospitalization.
CONCLUSIONS
In this study, a reduced time to reach adequate analgesia and LOS was noted in the Early-PCA group for severe VOC. A prospective study is required to confirm these results.
Keywords: children, length of stay, patient-controlled analgesia, sickle cell disease
Introduction
Sickle cell disease (SCD) is a group of inherited diseases of the β-globin gene, with an autosomal recessive transmission. Patients with SCD suffer from recurrent painful episodes called vaso-occlusive crisis (VOC), which represents the most frequent reason for consultation in the pediatric EDs. Vaso-occlusive crises are associated with decreased quality of life and significant morbidity and mortality.1,2 In infants, VOC usually presents as dactylitis (painful swelling of the fingers and toes) and can involve all parts of the body later in life.2
The National Institutes of Health recommends pain assessment in adults and children within 30 minutes of triage or 60 minutes of registration.3 While such rapid assessment and interventions are crucial, emergency care of VOC may be challenging. It requires adequate recognition of a painful VOC, and a thorough evaluation using adequate tools such as pain scales or other pain scores.4,5 The distinct manifestations of pain in this population and the broad differential diagnosis associated to some pain locations may indeed delay pain recognition and treatment.2,6 Furthermore, the features of VOC pain in SCD are complex, because these patients often have recurrent and chronic pain in addition to their acute episodes. The intensity of pain is often underestimated by care providers, which may also hinder adequate pain management.2
Recently, our institution reported that hospitalization for moderate to severe painful VOC episodes might be avoided by improved ambulatory treatment with oral hydration and a combination of non-opioid analgesics and intranasal fentanyl followed by oral opioids.7,8 This new approach also increased pain scale's use on arrival in our institution.7,8 Using such approach, about half of our patients with moderate to severe VOC could be discharged home on oral medication without requiring hospitalization.7,8
Despite these measures, some patients suffer persistent severe pain and require hospitalization for aggressive pain management with IV opioids and scheduled doses of non-opioid medications. In such severe and refractory pain crises, aggressive pain management requires IV opioids followed by regular opioid administration until pain control is achieved.3
Patient-controlled analgesia provides the management of analgesics' dosing according to the patient's need through opioid self-administration. The impact on pain management of early use of PCA in patients admitted with a VOC has been poorly reported, particularly in children. In adult studies of patients with SCD, while PCA was found to have the same analgesic effect as bolus doses of IV morphine, PCA was associated with less treatment failure, as well as reduced opiate consumption and opiate side effects, compared with continuous morphine infusion.9–11 In a recent retrospective observational study on adult SCD patients admitted for acute VOC treated with parenteral opioids in an urban academic health system, the authors12 compared LOS, 30-day read-mission, and discharges against medical advice between admissions treated with PCA vs standard therapy.12 Of 823 admissions for VOC included, 536 (65.1%) were treated with PCA and 287 (34.9%) were treated with standard nurse-administered opioid therapy. Treatment with PCA was associated with significantly shorter LOS in the unadjusted analyses (7.46 vs 9.42 days, p = 0.001), but the difference was not significant after adjustment for demographic and clinical characteristics (adjusted difference: 1.47 days, p = 0.06). Initiation of PCA for SCD-related pain in a busy ED can also be possible and was associated, in another adult study, with faster PCA delivery and fewer opioid boluses following decision to admit.13 Moreover, in a non-randomized pilot cohort, PCA initiated in a pediatric ED shortened the time of initiation. Patient records were reviewed, and a brief patient/parent satisfaction survey was collected. Forty-eight of 50 patient surveys indicated preference for starting PCA at the ED compared with PCA introduced later; 2 did not have a preference.14
In our institution, PCA has been used in a subset of patients admitted with severe VOC, prescribed and managed (or adjusted) by the pain management team at the request of the attending physicians. Our institution's pain management team is composed of anesthesiologists and specialized nurses who manage severe or refractory pain and PCA by using either IV morphine or IV hydromorphone and IV ketamine as co-analgesic. In a few cases, nurse-controlled analgesia (NCA) (nurse-administered opioid bolus based on pain scores for younger patients or patients enabled to use the PCA machine themselves) can be prescribed.
The objective of this study was to describe the use of PCA in patients presenting with severe VOC during a 6-year period in our tertiary care pediatric center and to evaluate the PCA characteristics as well as the impact of “Early” vs “Late'' use of PCA on time to reach adequate analgesia, on LOS, and on occurrence of adverse effects.
Methods
Study Design and Setting. This retrospective single center study was conducted in a pediatric tertiary care hospital where a comprehensive team is dedicated to the management of patients with SCD and a multidisciplinary team, DrépaNoPain, is committed to pain management improvement throughout the hospital. This sickle cell cohort is growing steadily (from 270 pediatric patients with SCD in 2010 to 400 patients in 2019). A provincial screening has been available since 201615.
“Early-PCA” was defined as the beginning of PCA within 48 hours of arrival in the ED or the hematology clinic and “Late-PCA” was initiated after 48 hours of arrival. In our institution, patients with SCD are admitted to the hospital through either the ED or the hematology clinic. The cutoff of 48 hours corresponds to the time generally reported between arrival to the hospital (ED or hematology clinic) and consultation with the pain management team.
Population and Data Collection. All patients <18 years of age with a proven diagnosis of SCD admitted for a VOC treated with PCA from January 1, 2010, to August 12, 2016, were included. Patients were identified through the computerized drug system of the pharmacy and the pain team records. Patients with SCD treated with PCA for pain control related to conditions other than VOC and/or acute chest syndrome (ACS), transferred patients from another center, and patients with missing data from the specific admission were excluded.
Medical charts were reviewed by a single author (CA). The following data were extracted: PCA characteristics including the number of hours from hospital admission (to ED or clinic) to the initiation of PCA, drug used, amount of bolus dose and baseline infusion, duration of administration and modality of administration (PCA or NCA), pain scores, LOS, and adverse effects. Independent variables were also collected including demographic information (age, sex) and information specific to the disease (i.e., hemoglobin phenotype, hemoglobin level, and use of hydroxyurea).
Primary Outcomes. Outcomes were compared between Early-PCA and Late-PCA. Three primary outcomes were evaluated. The first, time to reach adequate analgesia, was defined as the time to achieve 2 consecutive pain scores <5/10 at a 4-hour interval obtained by using either the auto-evaluation of pain (using the verbal numeric scale [VNS], visual analog scale [VAS], Faces Pain scales [including the OUCHER scale]) or hetero-evaluation of pain (using Face, Legs, Activity, Cry, Consolability [FLACC] or EValuation ENfant DOuLeur [EVENDOL]). Pain scores were retrieved in the patient's charts in the dedicated section for nursing evaluation. Since 2012, pain scales have been included in the VOC management protocols (OUCHER for young African American individuals, EVENDOL for others). However, VNS, VAS, and FLACC scales are the most frequently used scales by the nursing staff for other pathologies. Therefore, those scales are also frequently used by nurses for their patients' pain evaluation. We chose the 5/10 threshold for pain scores because it is conventionally used for treatment decision to add opiate in our protocols.15 We chose a 4-hour interval because it is the minimal frequency for pains scores' monitoring in our protocols and is a hospital requirement. The second primary outcome was the duration between the arrival in the ED or the hematology clinic and the last administered bolus. The final primary outcome was LOS, which was defined as the duration from admission (in the ED or clinic) to discharge.
Secondary Outcomes. These involved side effects associated with PCA and were divided into 3 categories by severity and were compared between Early-PCA and Late-PCA. The first was “mild and moderate” adverse effects and included nausea/vomiting, pruritus, and constipation. The second subcategory included “significant” adverse effects such as apnea, desaturation, urinary retention, hallucination or visual disorders, and excessive sedation (requiring or not intervention). The final subcategory was “serious” adverse drug reaction and was defined as a reaction leading to a persistent disability (i.e., permanent change, impairment, damage to patient's body function, physical activities, or quality of life), prolonged hospitalization or death, and any life-threatening event as per Health Canada Definition.17
Occurrence of ACS. Acute chest syndrome at admission and during hospitalization was defined as a new pulmonary infiltrate on the chest x-ray accompanied by fever, chest pain, or respiratory symptoms.
Data Analysis. A first subanalysis was completed to evaluate the same outcomes with early-PCA defined as 24 hours from arrival. A second subanalysis was conducted including only 1 hospitalization per patient. The latest hospitalization was chosen because it was more representative of our current approach to pain management. Time to reach adequate analgesia, time between arrival to the hospital (ED or the hematology clinic) and the last self-administered bolus, and LOS were compared between Early-PCA and Late-PCA groups with Mann-Whitney test. Adverse effects were compared with Fisher tests.
Interobserver reliability of data extraction for LOS and delay to the last bolus were evaluated by a second author (EDT) for 30% of the patients' files. A priori, we considered that data between the 2 reviewers matched if the difference was <0.2 day for the LOS and <2 hours for the delay to the last bolus. The percentage of matching values between the 2 reviewers was calculated.
Data were entered in an Excel database (Microsoft Inc, Richmond, WA) and analyzed with SPSS version 21 software (IBM Software Group Inc) and confidence intervals were calculated by using CIA (Confidence Interval Analysis) 2.2 (Trevor Bryant, University of Southampton). All statistical tests were 2-sided and used a significance level of 5%.
Results
Population Selection. Over a 6-year period, 175 PCA treatments were prescribed. Of these, 88 did not meet the inclusion criteria for the following reasons: patients >18 years of age (n = 38), patients without SCD (n = 31), patients with SCD with conditions other than VOC (n = 16), transferred from another hospital (n = 1), and missing data (n = 2). Overall, 87 PCA prescriptions were included in our study. These 87 episodes of VOC treated with PCA occurred in 46 patients (median 1, average of 1.78 episodes per patient [IQR, 1, 2]). Median age was 13.2 years (IQR: 10 years 4 months, 15 years) at admission. Sixty-nine (74%) patients had hemoglobin SS phenotype. Overall, the population of SCD patients admitted and treated with PCA over the study period represents 10.2% of the patients admitted for a VOC (853 hospital admissions for 274 SCD patients occurred during the study period).
Early- and Late-PCA Groups' Characteristics. Sixty-three Early-PCA treatments (72.4%) were administered to 35 patients, while 24 (27.6%) Late-PCA treatments were administered to 17 patients. Six patients received Early- and Late-PCA on different admissions and were therefore included in both groups.
Both groups were comparable in terms of age, sex, use of hydroxyurea, and hemoglobin phenotype (Table 1). There was no difference in patients presenting with fever or ACS on admission. However, median pain score on admission was higher in the Early-PCA group (9.0/10.0) than the Late-PCA group (7.0/10.0) with a median difference of 1 (95% CI 0 to 2) (Table 1).
Table 1.
Demographic Data and Patient Characteristics
| Early-PCA (n = 63) | Late-PCA (n = 24) | Difference (95% CI) | |
|---|---|---|---|
| Sex, n (%) | |||
| Female | 34 (54) | 15 (62) | −8.5% (−28.9 to 14.5)* |
| Male | 29 (46) | 9 (38) | |
| Age, mean ± SD, yr | 12.03 ± 4.08 | 13.20 ± 3.30 | −1.30 (−3.16 to 0.55)* |
| Phenotype, n (%) | |||
| SS | 50 (79.4) | 19 (79.2) | 0.2% (−16.2 to 21.5)* |
| Sβ | 1 (1.6) | 1 (4.1) | |
| SC | 12 (19) | 4 (16.7) | |
| Chronic hydroxyurea treatment, n (%) | 39 (62) | 13 (54) | 7.7% (−14.0 to 29.8)* |
| Hb level at admission, mean ± SD, g/dL | 83.6 ± 13.3 | 88.3 ± 15.4 | −4.9 (−11.3 to 2.0)* |
| Fever on admission, n (%) | 9 (14.3) | 3 (12.5) | 1.8% (−17.9 to 15.2)* |
| ACS on admission, n (%) | 6 (9.5) | 2 (8.3) | 1.2% (−17.0 to 12.6)* |
| Pain score on admission, median (IQR) | 9 (8, 9) | 7 (6.2, 8) | 1 (0 to 2) |
ACS, acute chest syndrome; Hb, hemoglobin; PCA, patient-controlled analgesia
* Not statistically significant at p ≤ 0.05.
There was no difference in terms of the characteristics of PCA used between the 2 groups. In 76 (87.4%) cases, PCA was patient controlled and in 11 (12.6%) it was an NCA. Fifty-one (58.6%) PCA treatments included continuous infusion of opioid and ketamine. Hydromorphone was the most frequently used opioid in 51 PCA treatments (58.6%). There was no difference between the 2 groups regarding PCA regimen (total opioid doses and continuous perfusion) (Table 2).
Table 2.
Type of PCA Treatments Received in Both Groups
| Early-PCA (n = 63) | Late-PCA (n = 24) | Median Difference (95% CI) | |
|---|---|---|---|
| Administration, n (%) | |||
| PCA | 53 (84.1) | 23 (95.8) | 11.7% (−5.8 to 23.2)* |
| NCA | 10 (15.9) | 1 (4.2) | |
| Continuous infusion of opioid, n (%) | 50 (79.4) | 22 (91.6) | 9.7% (−10.6 to 24.0)* |
| Type of opioid, n (%) | |||
| Hydromorphone | 35 (55.6) | 16 (66.7) | 11.1% (−12.0 to 30.8)* |
| Morphine | 28 (44.4) | 8 (33.3) | |
| Total dose of hydromorphone, median, mg/kg/day | 0.186 (n = 35) | 0.157 (n = 17) | 0.020 (−0.071 to 0.095)* |
| Continuous perfusion dose of hydromorphone, median, mg/kg/day | 0.11 (n = 29) | 0.091 (n = 15) | 0.007 (−0.029 to 0.038)* |
| Total dose of morphine, median, mg/kg/day | 0.968 (n = 28) | 1.285 (n = 7) | −0.306 (−0.871 to 0.240)* |
| Continuous perfusion dose of morphine, median, mg/kg/day | 0.498 (n = 23) | 0.769 (n = 6) | −0.208 (−0.503 to 0,108)* |
| Ketamine, n (%) | 35 (55.6) | 16 (66.7) | 11.1% (−12.0 to 30.8)* |
| With hydromorphone, n | 21 | 11 | — |
| With morphine, n | 14 | 5 | — |
NCA, nurse-controlled analgesia; PCA, patient-controlled analgesia
* Not statistically significant at p ≤ 0.05.
Primary Outcomes.
Time to Reach Adequate Analgesia. Thirty-two patients admitted over 87 hospitalizations (36.7%) had sufficient data (auto-evaluation and hetero-evaluation) to determine the time to reach adequate analgesia: 21 in the Early-PCA group and 11 in the Late-PCA group (Table 3). Median time to reach adequate analgesia was statistically shorter in the Early-PCA group: 102.3 hours (IQR 60.5, 119) vs 127 hours in Late-PCA group (IQR 113.5, 176), with a median difference of −41.0 (95% CI −82.0 to −6.0) (Table 4).
Table 3.
Type of Pain Evaluation Observed in Patient Files
| Early-PCA (n = 63) | Late-PCA (n = 24) | Median Difference (95% CI) | |
|---|---|---|---|
| Pain hetero-evaluation, n (%) | |||
| FLACC | 30 (47.6) | 17 (70.8) | −23.2% (−41.7 to 0.2)* |
| FLACC + EVENDOL | 2 (3.2) | 0 | 3.2% (−10.8 to 10.9)* |
| Pain auto-evaluation, n (%) | |||
| VAS | 31 (49.2) | 14 (58.3) | −9.1% (−30.1 to 13.8)* |
| Faces scale | 3 (4.8) | 0 | 4.8% (−9.4 to 13.1)* |
| VAS + Faces scale | 2 (3.2) | 0 | 3.2% (−10.8 to 10.9)* |
| Pain evaluations per day, median (range) | 11.21 (1.31–29.25) | 11.1 (2.71–1916) | 0.05 (−2.06 to 2.16)* |
| Number of auto-evaluations per day | |||
| Specific | 0.14 | 0.36 | 0.00 (−0.35 to 0.11)* |
| Non-specific | 7.58 | 7.27 | 0.06 (−1.60 to 1.74)* |
| Number of hetero-evaluations per day | |||
| Specific | 0 | 0.86 | −0.23 (−0.86 to 0.1)* |
| Non-specific | 0.8 | 0.72 | 0.11 (−0.23 to 0.48)* |
EVENDOL, EValuation ENfant DouLeur (specific pain evaluation scale); FLACC, Face, Legs, Activity, Cry, Consolability (specific pain evaluation scale); PCA, patient-controlled analgesia; VAS, visual analog scale
* Not statistically significant at p ≤ 0.05.
Table 4.
Early- Versus Late-PCA and Pain Management of VOC
| Early-PCA, median (IQR) (n = 63) | Late-PCA, median (IQR) (n = 24) | Median Difference (95% CI) | |
|---|---|---|---|
| Delay to obtain PCA, hr | 25.1 (17.7, 30.6) | 79.0 (65.3, 94.2) | −53.9 (−65.3 to −46.8)* |
| Time to reach adequate analgesia, hr | 102.3 (60.5, 119) | 127 (113.5, 176) | −41.0 (−82.0 to −6.0)* |
| Delay to last bolus, hr | 103.6 (48.1, 161.7) | 177.5 (141.5, 256.4) | −72.9 (−106.9 to −38.8)* |
| Length of stay, days | 6.4 (4.6, 8.7) | 9.6 (7.1, 12.0) | −3.4 (−4.9 to −1.9)* |
PCA, patient-controlled analgesia; VOC: vaso-occlusive crisis
* Statistically significant at p ≤ 0.05.
Time to Last Bolus. Early-PCA was associated with a shorter time to last bolus administration: 103.6 hours in the Early-PCA (IQR 78.1, 161.7) vs 177.5 hours in the Late PCA (IQR 141.5, 256.4) (Table 4) for a median difference of −72.9 hours (95% CI −106.9 to −38.8).
Length of Stay. Early-PCA was associated with a shorter LOS of 3.4 days (95% CI −4.9 to −1.9): median Early-PCA, 6.4 days (IQR 4.6, 8.7) vs Late-PCA, 9.6 days (IQR 7.1, 12.0) (Table 4).
Secondary Outcomes. Mild and Moderate Adverse Effects. Nineteen prescribed PCA treatments (21.8%) had no side effects, and 37 (42.5%) had adverse effects such as nausea or constipation (Figures 1 and 2).
Figure 1.
Occurrence of side effects during the administration of patient-controlled analgesia.
Figure 2.
Occurrence of side effects during different PCA regimens.
Significant Adverse Effects. They were comparable between the Early- and Late-PCA groups (Figures 1 and 2). Overall, they occurred during 28 (32.2%) PCA treatments: 21/63 (33.3%) episodes in Early-PCA and 7/24 (29.2%) in Late-PCA group. The most frequent significant adverse effect was excessive sedation, in 20 PCA administrations (23.0%). All required intervention: decreasing dose of opioid in 4 patients, decreasing dose of opioid and ketamine in 4 patients, and change of opioid (morphine to hydromorphone) in 12 patients. The other significant adverse effects (16 PCA administrations [18.4%]) required immediate interventions and included 2 desaturations treated with oxygen administration, 8 neurologic side effects (hallucinations, visual abnormalities) that prompted ketamine discontinuation, and 6 urinary retentions treated with naloxone bolus or catheterization. No apnea was observed.
Serious Adverse Drug Reactions. None were noted.
Acute Chest Syndrome. Acute chest syndrome occurred after admission during 19 (21.8%) hospitalizations with 11 (17.5%) in the Early-PCA group and 8 (33.3%) in the Late-PCA group. This difference was not statistically significant (difference -15.9%; 95% CI −37.2 to 3.1).
First Subanalysis: Early-PCA Defined as Started Before 24 Hours. When early-PCA was defined as started <24 hours from arrival, both groups were similar in terms of baseline characteristics at admission (30 patients in the ≤24-hour PCA group, 57 patients in the >24-hour PCA group). There was no statistical difference in the time to reach adequate analgesia (median difference, 27.0 hours [95% CI −7.0 to 74]). However, time to last administered bolus was significantly reduced by −60.3 hours (95% CI −93.4 to −27.1) in the ≤24-hour PCA group: 91.8 hours (IQR 70, 124.4) vs 157.3 hours (IQR 115.8, 208). The LOS was shorter in the group with PCA introduced ≤24 hours from arrival with a median difference of −2.1 days (95% CI −3.7 to −0.5) from 5.8 days in the ≤24-hour PCA group (IQR 4.6, 8.0) to 7.5 days in the >24-hour PCA group (IQR 6.1, 9.8).
Second Subanalysis: One Hospitalization per Patient. When only the latest hospitalization for each patient was used in the analysis, both groups were similar in terms of baseline characteristics at admission (35 patients in the Early-PCA group and 17 in the Late-PCA group). Time to reach adequate analgesia was shorter in the Early-PCA group (median difference, -61.7 hours [95% CI −95.6 to −27.9]), as was the time to last auto-administered bolus (median difference, -70.3 hours [95% CI −111.7 to −29.0]). The LOS was shorter in the Early-PCA group (≤48 hours) with a median reduction of −3.6 days (95% CI −5.5 to −1.6): 6.14 days (IQR 4.7, 8.7) vs 9.42 days (IQR 7.2, 11.9).
Interobserver Reliability. Interobserver reliability for the delay to the last bolus and LOS was calculated by a second reviewer for 28 (32.1%) patients' files with an 89.3% match for the delay to the last bolus and an 82.1% match for the LOS.
Discussion
Our retrospective data show an association between Early-PCA (defined as started 48 hours from arrival) and decreased LOS, despite the higher median pain score on admission in this group. This association was also observed when early-PCA was defined as started 24 hours from arrival. Some patients had multiple episodes of VOC. To better represent the overall population, the analysis was restricted to 1 admission per patient, with similar results. Therefore, the result in the overall study is less likely to have been skewed by an imbalance due to a few patients requiring frequently an admission. The association between early-PCA and decreased LOS is also in keeping with the work of Payne et al18 who found that amongst patients with frequent pain admissions, those who receive earlier maximal therapy require shorter hospitalization than patients who require longer titration to maximum opioid dose. However, causality cannot be assessed. Interestingly, Averbukh et al12 have recently studied LOS in adult patients with SCD admitted with a VOC, comparing PCA with standard nurse-administrated opiate therapy. They concluded that PCA may be associated with shorter LOS in those patients. However, the timing of the beginning of the PCA was not considered in this study. Another small adult study reported a non-significant reduction in LOS in patients receiving PCA compared with patients receiving a continuous infusion of morphine.11 Shah et al10 retrospectively reviewed LOS in adult SCD patients with VOC, comparing intermittent injections to PCA, but did not find a difference between treatments. However, patients treated with PCA had less treatment failure than patients treated with intermittent injections (64% of patients with intermittent injection vs 14% in the PCA group [p < .0001]).10 Another adult study by Santos et al13 compared PCA initiation in the ED with PCA initiated after hospitalization in patients with SCD presenting with pain (PCA initiated at 4.5 hours vs 8.6 hours) and found no significant difference in the LOS. However, the short difference in time to initiate the PCA could explain this lack of difference.
In our study, early- and late-PCA were associated with similar but frequent adverse effects without serious adverse drug reactions. None of the adverse effects appeared to prolong hospital stay, or to cause significant disability or death.16, 19 However, the number of adverse effects appears to be high compared with that of other studies, for example, the pilot study of Jacob et al.19 Most frequent adverse effects were excessive sedation, pruritus, and nausea/vomiting. The incidence of excessive sedation and neurologic problems (hallucinations and visual disorders) might have been related to the concomitant use of adjunct medication such as ketamine instead of the opioid dose.20 Constant monitoring and readjustments are expected with PCA use for opioid administration. This was confirmed in our study with about a third of the episodes treated by PCA resulting in significant adverse effects requiring interventions. Anticipating these side effects with the prescription of monitoring and symptomatic treatments through pre-written orders can help standardize the care of these patients. While significant adverse events such as severe bradypnea or respiratory arrest are unlikely to have been omitted in our record, some other more subtle adverse reaction might have been unnoticed. Further studies on the effects of various additive treatments like ketamine on adverse effects are required.
Some authors have suggested that the use of opiates might increase the risks of developing an ACS20 but this association has not yet been confirmed.21, 22, 23 In our study, there was no difference in ACS occurrence between the 2 groups at admission or during hospitalization but more studies are needed to evaluate the influence of the type of opiate used, the dose, or the use of continuous perfusion on occurrence of ACS.
Although our study has not found a difference between various PCA regimens, it is limited by the low number of patients that received PCA during the study period. In a preliminary study, the combination of a high-dose infusion rate and low intermittent push dose of opiate suggested improved response and shorter recovery from the painful episode, compared with a regimen with lower background infusion rate and high intermittent push dose in adults, but the 2 regimens were quite similar in pediatric patients.24
The retrospective methodology of our study affected the quality of pain evaluation, as well as robustness on the record of potential adverse effects. Indeed, almost half of the patients did not have specific pain scores recorded in their medical charts. Underevaluation and underassessment of pain is common in children and has been reported in various settings.25 Furthermore, pain evaluation has a high impact in disposition of the patient during a VOC; in a pediatric study, a high pain score in the ED at triage was associated with a higher probability of admission and prolonged LOS.26 In another study, education improved pain evaluation with the use of a pain scale concomitant with some treatment improvement (a trend in reducing time to analgesic's administration).27 Adding pain scales at the end of the prewritten orders and offering training on pain evaluation are initiatives that have been developed in our setting to improve pain assessment of SCD patients in the ED and in the hematology clinics.7,8 However, further interventions will be required to improve pain reevaluation on the ward and to better evaluate readiness to discharge.
Over the last few years, new protocols in our departments have decreased the number of hospitalizations, decreased the number of installed IV, and decreased the time to first opiate by improved oral hydration and IN/oral opioid administration.7,8 However, patients with severe and persistent pain still need to be admitted and treated with IV opioids. Those patients with refractory pain might then represent a subgroup of patients with pain even more difficult to relieve, having failed to improve with maximal initial therapies. Improving the care of those patients is essential. In our study, higher pain score in the early-PCA group might have triggered early used of PCA through an earlier pain team consultation. This group seemed to have had an earlier success of pain control and a reduced LOS. In our institution, an earlier consultation of the expert group in pain management may have been one of the factors leading to a reduced LOS. Therefore, the administration of a PCA might be considered earlier if the patient requires an admission for severe pain despite optimal ambulatory management. Reducing time to adequate analgesia and LOS could improve the quality of life of these patients.
Conclusion
In conclusion, in this retrospective study, early use of PCA for severe VOC in children was associated with shorter time to reach adequate analgesia and a reduced LOS despite higher pain score on admission. There was no difference in adverse effects between the early- and late-PCA groups. A prospective study comparing early vs late PCA is required to confirm better pain control with early PCA start and to assess the impact on LOS and safety.
ABBREVIATIONS
- ACS
acute chest syndrome
- ED
emergency department
- EVENDOL
EValuation ENfant DOuLeur
- FLACC
Face, Legs, Activity, Cry, Consolability
- Hb
Hemoglobin
- IN
intranasal
- IV
intravenous
- LOS
length of stay
- NCA
nurse-controlled analgesia
- PCA
patient-controlled analgesia
- SCD
sickle cell disease
- VAS
visual analog scale
- VNS
verbal numeric scale
- VOC:
vaso-occlusive crisis
Footnotes
Disclosures. Dr Arbitre reports grants from Pfizer SDAC 2018 grant, outside the submitted work. Dr Yves Pastore has received financial compensation as consultant from Pfizer and Novartis. The other authors declare no conflicts or financial interest in any product or service mentioned in the manuscript, including grants, equipment, medications, employment, gifts, and honoraria. All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Ethical Approval and Informed Consent. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guidelines on human experimentation. The project was approved by the institution review board/ethics committee. Given the nature of the study it was exempt from informed consent.
References
- 1.Platt OS, Thorington BD, Brambilla DJ et al. Pain in sickle cell disease: rates and risk factors. N Engl J Med. 1991;325(1):11–16.1. doi: 10.1056/NEJM199107043250103. [DOI] [PubMed] [Google Scholar]
- 2.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):265–273. doi: 10.1016/j.cpem.2010.09.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.US Department of Health and Human Services Evidence-based management of sickle cell disease: expert panel report, 2014: guide to recommendations. September 2014. Accessed may 28th 2021. https://www.nhlbi.nih.gov/sites/default/files/media/docs/sickle-cell-disease-report%20020816_0.pdf (noending period)
- 4.Morrissey LK, Shea JO, Kalish LA et al. Clinical practice guideline improves the treatment of sickle cell disease vasoocclusive pain. Pediatr Blood Cancer. 2009;52(3):369–372. doi: 10.1002/pbc.21847. [DOI] [PubMed] [Google Scholar]
- 5.Dunlop RJ, Bennett KCLB. Cochrane Database Syst Rev. 2. 2006. Pain management for sickle cell disease; p. CD003350. [DOI] [PubMed] [Google Scholar]
- 6.Dampier C, Jaeger B, Gross HE et al. Responsiveness of PROMIS® pediatric measures to hospitalizations for sickle pain and subsequent recovery. Pediatr Blood Cancer. 2016;63(6):1038–1045. doi: 10.1002/pbc.25931. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Paquin H, Trottier ED, Robitaille N et al. Oral morphine protocol evaluation for the treatment of vaso-occlusive crisis in paediatric sickle cell patients. Paediatr Child Health. 2019;24(1):e45–e50. doi: 10.1093/pch/pxy074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Paquin H, Trottier ED, Pastore Y et al. Evaluation of a clinical protocol using intranasal fentanyl for treatment of vaso-occlusive crisis in sickle cell patients in the emergency department. Paediatr Child Health. 2020;25(5):293–299. doi: 10.1093/pch/pxz022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gonzalez ER, Bahal N, Hansen LA et al. Intermittent injection vs patient-controlled analgesia for sickle cell crisis pain: comparison in patients in the emergency department. Arch Intern Med. 1991;151(7):137–138. [PubMed] [Google Scholar]
- 10.Shah SP, Twilla JD, Kemp L et al. Comparison of parenteral opioid dosing in adult sickle cell disease patients with vaso-occlusive crisis. J Pain Palliat Care Pharmacother. 2018;32(4):201–207. doi: 10.1080/15360288.2019.1577938. [DOI] [PubMed] [Google Scholar]
- 11.van Beers EJ, van Tuijn CFJ, Nieuwkerk PT et al. Patient-controlled analgesia versus continuous infusion of morphine during vaso-occlusive crisis in sickle cell disease, a randomized controlled trial. Am J Hematol. 2007;82(11):955–960. doi: 10.1002/ajh.20944. [DOI] [PubMed] [Google Scholar]
- 12.Averbukh Y, Porrovecchio A, Southern WN. Patient-controlled analgesia for vaso-occlusive crisis: a cohort study. Clin J Pain. 2019;35(8):686–690. doi: 10.1097/AJP.0000000000000726. [DOI] [PubMed] [Google Scholar]
- 13.Santos J, Jones S, Wakefield D Pain Res Manag. 2016. 2016. Patient controlled analgesia for adults with sickle cell disease awaiting admission from the emergency department. 3218186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Melzer-Lange MD, Walsh-Kelly CM, Lea G et al. Patient-controlled analgesia for sickle cell pain crisis in a pediatric emergency department. Pediatr Emerg Care. 2004;20(1):2–4. doi: 10.1097/01.pec.0000106235.72265.29. [DOI] [PubMed] [Google Scholar]
- 15.Urgence CHU Sainte Justine Guide clinique Anémie falciforme – Drépanocytose. Accessed may 28th 2021. http://www.urgencehsj.ca/protocoles/anemie-falciforme-drepanocytose/ (no ending period)
- 16.Zempsky WT, Loiselle KA, Corsi JM et al. Use of low-dose ketamine infusion for pediatric patients with sickle cell disease-related pain: a case series. Clin J Pain. 2010;26(2):163–167. doi: 10.1097/AJP.0b013e3181b511ab. [DOI] [PubMed] [Google Scholar]
- 17.Government of Canada – Health Canada Side effect reporting form. Accessed may 28th 2021. www.canada.ca/en/health-canada/services/drugs-health-products/medeffect-canada/adverse-reaction-reporting/consumer-side-effect-reporting-form.html (no ending period)
- 18.Payne J, Aban I, Hilliard LM et al. Impact of early analgesia on hospitalization outcomes for sickle cell pain crisis. Pediatr Blood Cancer. 2018;65(12):e27420. doi: 10.1002/pbc.27420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Jacob E, Hockenberry M, Mueller BU et al. analgesic response to morphine in children with sickle cell disease: a pilot study. J Pain Manag. 2008;2(1):179–190. [PMC free article] [PubMed] [Google Scholar]
- 20.Birken CS, Khambalia A, Dupuis A et al. Morphine is associated with acute chest syndrome in children hospitalized with sickle cell disease. Hosp Pediatr. 2013;3(2):149–155. doi: 10.1542/hpeds.2012-0067. [DOI] [PubMed] [Google Scholar]
- 21.Buchanan ID, Woodward M, Reed GW. Opioid selection during sickle cell pain crisis and its impact on the development of acute chest syndrome. Pediatr Blood Cancer. 2005;45(5):716–724. doi: 10.1002/pbc.20403. [DOI] [PubMed] [Google Scholar]
- 22.Kavanagh PL, Sprinz PG, Wolfgang TL et al. Improving the management of vaso-occlusive episodes in the pediatric emergency department. Pediatrics. 2015;136(4):e1016–e1025. doi: 10.1542/peds.2014-3470. [DOI] [PubMed] [Google Scholar]
- 23.Finkelstein Y, Schechter T, Garcia-Bournissen F et al. Is morphine exposure associated with acute chest syndrome in children with vaso-occlusive crisis of sickle cell disease: a 6-year case-crossover study. Clin Ther. 2007;29(12):2738–2743. doi: 10.1016/j.clinthera.2007.12.016. [DOI] [PubMed] [Google Scholar]
- 24.Dampier CD, Smith WR, Wager CG et al. IMPROVE trial: a randomized controlled trial of patient-controlled anal-gesia for sickle cell painful episodes: rationale, design challenges, initial experience, and recommendations for future studies. Clin Trials. 2013;10(2):319–331. doi: 10.1177/1740774513475850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ceelie I, de Wildt SN, de Jong M et al. Protocolized postoperative pain management in infants; do we stick to it? Eur J Pain. 2012;16(5):760–766. doi: 10.1002/j.1532-2149.2011.00056.x. [DOI] [PubMed] [Google Scholar]
- 26.Rogovik AL, Li Y, Kirby MA et al. Admission and length of stay due to painful vasoocclusive crisis in children. Am J Emerg Med. 2009;27(7):797–801. doi: 10.1016/j.ajem.2008.06.022. [DOI] [PubMed] [Google Scholar]
- 27.Po' C, Colombatti R, Cirigliano A et al. The management of sickle cell pain in the emergency department: a priority for health systems. Clin J Pain. 2013;29(1):60–63. doi: 10.1097/AJP.0b013e318245764b. [DOI] [PubMed] [Google Scholar]