Ketamine for Pain in Sickle Cell Disease Reduces Opioid Usage

Abstract

Context

Pain attributable to sickle cell disease (SCD) is often unpredictable, recurrent, and requires complex treatments. Subanesthetic ketamine infusion has been studied in other diseases and disorders, but there is still limited data on its efficacy in pain management for SCD.

Objectives

The primary objective is to determine if subanesthetic ketamine infusion reduces pain scores and opioid requirements in hospitalized pediatric patients with SCD.

Results

Forty-six admissions among 22 patients between February 2018 and December 2019 were analyzed. We observed decrease in pain scores within 24 hours of ketamine initiation in 34 of 46 admissions (mean pain score per patient before ketamine initiation: 2.2–9.7, mean pain score per patient after ketamine initiation: 0–9.7; P<0.05). We observed a decrease in pain scores in the remaining 12 admissions after greater than 24 hours of ketamine initiation. Opioid usage declined after ketamine infusion, with a difference of means in oral morphine equivalents before and after ketamine of 122.8 mg/day. The side effects observed with ketamine infusion included hallucinations in 11 (23.9%) admissions. Only 4 (8.7%) admissions required cessation of the infusion due to side effects. The readmission rate at 2 weeks and 4 weeks after first ketamine infusion was the same (12.5%) at both time points. For all patients in the cohort, the introduction of ketamine into pain regimens did not reduce the number of admissions in the year following ketamine initiation relative to the year prior.

Conclusion

In pediatric patients with SCD, subanesthetic ketamine was safe as a continuous infusion and effectively reduced both pain scores and opioid requirements.

INTRODUCTION

The most common presenting symptom of a sickle cell disease (SCD) vaso-occlusive episode (VOE) is severe pain. The pain is often unpredictable, recurrent, difficult to prevent, and requires complex treatments that often fail to meet the patient’s needs in a timely fashion. Currently, standard of care for acute VOEs involves the use of NSAIDs and opioids, but these medications are associated with several adverse effects and potential long-term complications (1). During a VOE, persons with SCD commonly attempt to either manage their pain at home or present to the emergency department for further management. In the acute care setting, there is often discrimination against these patients, with an inaccurate perception from the medical community that they are drug-seeking (2). This common misconception is enhanced by the fact that persons with SCD usually have familiarity with which medications provide adequate pain relief in their individual experience (3, 4). This knowledge represents active patient involvement in their healthcare, rather than inappropriate medication use. A non-opioid alternative for pain management in SCD has potential to reduce stigma associated with seeking pain control.

Ketamine, when administered as a continuous infusion with subanesthetic dosing, has been studied in several diseases. There are limited data establishing its efficacy in pain management in SCD. Subanesthetic ketamine has been used in pediatric oncology in patients with chronic pain or advanced cancer (5). There are extensive data on its use in psychiatric disorders, most commonly major depressive disorder (6, 7). In the 2020 guidelines for the management of acute and chronic pain in SCD, the American Society of Hematology suggested subanesthetic ketamine infusion in adults and children for pain that is refractory or inadequately treated with opioids, though without strong evidence (1). This recommendation assumes that institutions with appropriate protocols for drug administration begin infusion at rates of 0.1 to 0.3 mg/kg/hr with an hourly maximum of 1 mg/kg (1). Institutional guidelines vary widely across the United States. Some institutions limit the use of subanesthetic ketamine to an intensive care unit (ICU), with patients who require infusions being transferred to higher acuity care than was otherwise unnecessary (1, 8). Other institutions have protocols that require an acute pain treatment service or Anesthesiologists to manage infusions within limited hours of operation. In 2018, our institution created an evidence-based clinical algorithm for the use of subanesthetic ketamine for acute pain episodes in hospitalized, non-ICU pediatric patients with SCD (9). A goal of this algorithm was to standardize the use of subanesthetic ketamine and provide a reference for clinicians.

The purpose of this study was to determine whether subanesthetic ketamine infusion reduced pain scores and opioid requirements in hospitalized patients with SCD following the implementation of an evidence-based clinical algorithm.

METHODS

A retrospective, cross-sectional study of all patients with SCD who received subanesthetic ketamine infusion for the management of VOE from February 2018 to December 2019 was conducted. The electronic medical record was queried for patients who received subanesthetic ketamine during that period using ICD-10 diagnostic codes for SCD and for ketamine administration. Admissions for those receiving ketamine for VOE due to SCD were selected for review and included if the patient was greater than 3 years old and did not require mechanical ventilation. Patient demographics and past medical history, including history of surgery with two weeks of admission, were collected. Pain was defined as acute-on-chronic or recurrent acute pain according to statements in the medical record. Medication data, including a patient’s home pain regimen and hydroxyurea use, was collected. Pain scores were recorded every four hours throughout the admission by nursing staff using a 10-point Likert scale or age-appropriate equivalent. Per institutional protocol, patients using opioid patient-controlled analgesia (PCA) are strongly encouraged to discontinue the continuous infusion and continue interval dosing only once ketamine infusion is initiated. The interval dosing is allowed to remain for breakthrough pain management or may be transitioned to an oral equivalent. At our institution, ketamine infusions were ordered and managed uniquely by the Acute Pain Treatment Servicestaffed by the Department of Anesthesiology. Ketamine infusion was started at 0.05–0.15 mg/kg/hr with a maximum dose of 0.5 mg/kg/hr on the orderset, or 1 mg/kg/hr with the approval of the Acute Pain Treatment Service. Approval for higher doses was granted based upon review of the patient’s history of ketamine use and response (e.g. tolerance and adverse effects) by the Acute Pain Treatment Service. Periodic infusion rate adjustments were made following consultation with the patient, assessment of pain response, and screening for adverse effects. Study data were collected and recorded with REDCap electronic data capture tools (10, 11). Descriptive statistics were analyzed using Microsoft Excel 2016 and GraphPad Prism V9.5.0 (San Diego, CA). Statistical significance of associations between collected parameters were determined using Fisher’s exact test. To examine the trend in pain scores across the cohort at each time point before and after ketamine, a mixed-model was fit using restricted maximum likelihood methodology. This study was approved by the local Institutional Review Board.

RESULTS

Forty-six admissions representing 22 unique patients were included (Table 1). The mean age at admission was 15.1 years (SD: 3.0). The median number of a repeat admissions in the year following ketamine infusion per patient was 4 (range: 0–10). The cohort consisted of 22 patients, 19 (86.4%) who identified as Black or African American and 3 (13.6%) who identified as Caucasian/White. The cohort included 3 (13.6%) subjects who identified as Hispanic and 19 (86.4%) non-Hispanic. Among all admissions 38 of 46 (82.6%) were amongst patients who were adherent to hydroxyurea therapy.

Table 1.

Pain Scores and Opioid Use Before and After Ketamine

Variable		N (%)
Sex
	Female	7 (31.8%)
	Male	15 (68.2%)
Age at admission	Average	15.13
	Minimum	9.42
	Maximum	20.42
Race/ethnicity
	African American	19 (86.4%)
	Hispanic	3 (13.6%)
Genotype
	HbSS/HbSbeta0	21 (95.5%)
	HbSC	1 (4.55%)
Pain diagnosis
	Recurrent acute pain	16 (34.8%)
	Chronic pain	30 (65.2%)
Length of admission in days
	Average	6.5
	Minimum	2
	Maximum	15
Previous ketamine infusion
	Yes	22 (47.8%)
	No	24 (52.2%)
Additional medications
	Hydroxyurea	38 (82.6%)
	NSAIDs (ibuprofen, ketorolac)	46 (95.7%)
	Acetaminophen	37 (80.4%)
	Morphine	28 (60.9%)
	Hydromorphone	25 (54.3%)
	Acetaminophen/hydrocodone	5 (10.9%)
	Oxycodone	16 (34.8%)
	Gabapentin	14 (30.4%)
	Methocarbamol	17 (37%)
Side effects *
	Nausea	8 (17.4%)
	Vomiting	2 (4.3%)
	Sedation	3 (6.5%)
	Hallucinations	11 (23.9%)
	Dizziness	8 (17.4%)
	Dyspnea	0 (0%)
	Respiratory Depression	0 (0%)
	Delirium	0 (0%)
	Hypersensitivity Reaction	0 (0%)

^*Patients may have experienced more than one side effect.

Ketamine infusion rates were adjusted for side effects and analgesic effects. Across all admissions, the starting ketamine infusion rate ranged from 0.02–0.15 mg/kg/hr (mean 0.09 ± 0.03 mg/hr) and the maximum infusion rate ranged from 0.03–0.21 mg/kg/hr (mean 0.11 ± 0.04 mg/hr). The infusions continued for an average 52.7 ± 25.30 hours (range: 4.4–109.28 hours). In this cohort, patients required less than one dose adjustment on average during an admission, including dose escalations and dose reductions (mean=0. 8, range: 0–4).

Ketamine infusion was associated with reduced pain scores and opioid usage (Figure 1A). For each patient, mean pain score over 48 hours before and 24 hours after initiation of ketamine was calculated using all available scores. The mean pain score per subject before ketamine ranged from 2.2–9.7. After ketamine, the mean pain score per subject ranged from 0 to 9.7. After initiation of ketamine, 34 of the 46 admissions had improved pain scores, including 12 admissions among 9 patients with statistically significantly lower pain scores after initiation of the infusion (P<0.05). Among these 9 patients, 4 patients (44.4%) were admitted for recurrent acute pain, and 7 patients (55.6%) were admitted for acute-on-chronic pain. A linear regression model for the entire cohort suggests that average pain scores at each time point were significantly different following initiation of ketamine than before (before R²=0.48; after R²=0.53; difference between slopes, P=0.02). Opioid use declined by more than 3-fold after ketamine initiation (Figure 1B). The mean morphine milligram equivalents (MME) use per day per patient was 173.2 mg/day (IQR: 33.9). After ketamine initiation, the mean MME use per day per patient was 50.4 mg/day (IQR: 11.0).

Figure 1.

Pain Scores and Opioid Use Before and After Ketamine. A: All patient data aligned to time of ketamine initiation (time 0, dashed line). Pain scores were collected as discrete (not continuous) measurements every four hours but are represented as lines to demonstrate trends for each patient. Linear regression line is shown. B: Oral and intravenous opioids were included. Oral morphine equivalents were tallied as discrete (not continuous) measurements every 24 hours but are represented as lines to demonstrate trends for each patient.

Hallucinations were the most common side effect reported. Among 46 admissions, 11 (23.9%) experienced hallucinations at any point throughout the infusion. These were exclusively visual hallucinations, with no reports of auditory or tactile hallucinations noted. Despite this effect, only 4 admissions (8.7%) required immediate discontinuation of the infusion due to reported hallucinations. Other adverse effects recorded included nausea, sedation, dizziness, and difficulty breathing (Table 1). Numerous medications prescribed for the treatment of acute and chronic pain, included some with similar side effect profiles, were used during admission.

We investigated the effect of subanesthetic ketamine infusion on short-term or long-term readmission rates. In this cohort, 24 (52.2%) of admissions were for first-time recipients of subanesthetic ketamine and 22 (47.8%) had received ketamine in a previous admission. The number of readmissions among patients were receiving ketamine infusions for the first time was 3 (12.5%) at two weeks and 3 (12.5%) at four weeks post-discharge. For first-time recipients of subanesthetic ketamine, the average number of admissions one year before and one year after first ketamine infusion were 3.9 and 4.2 hospital admissions, respectively.

DISCUSSION

Adequate treatment of VOE in persons with SCD has remained difficult, especially in patients with insufficient response to opioids. In our cohort, ketamine has shown to be beneficial as an adjunct to opioid therapy in the treatment of VOE when used as an infusion at subanesthetic dosing. Ketamine may also aid in reducing opioid use by enabling pain relief with lower doses of opioid medications. Through these effects, ketamine may be a useful component of a strategy to reduce stigma and misconception in opioid-tolerant patients.

Although pain scores did not significantly decrease during every admission where ketamine was used, pain scores remained stable despite a significant reduction in opioid usage for all admissions. A large portion of the opioid reduction is likely due to discontinuation of the continuous infusion. However, had there not been some benefit from the ketamine or a lack of efficacy of the opioid, we would have expected interval opioid use to increase after the change. The impact of subanesthetic ketamine infusion on pain scores in our study was variable, as seen in other studies. One study of 12 adults with SCD did not report any change in pain score or reduction in length of stay with ketamine infusion (12). In another study examining the use of subanesthetic ketamine in the treatment of children and adolescents with chronic pain in the outpatient setting, subanesthetic ketamine improved pain scores but had no effect on opioid consumption (13). These studies and our work suggest that there may be some variability in response for pain reduction and opioid use. Variability between institutions in ketamine protocols may also contribute to the difference in outcomes.

The side effects reported during admissions were generally tolerable as most patients did not require cessation of ketamine infusion due to adverse effects. Hallucinations represented an important adverse effect across admissions. Our findings were similar to those observed in a 2014 study of 33 hospitalized pediatric patients, where 3 patients had to discontinue ketamine infusion due to transient psychotomimetic effects (14). In a retrospective analysis of 9 adult patients with sickle cell pain, ketamine used in combination with benzodiazepines was associated with reduced pain scores and reduced frequency of hallucinations and dissociation (15). Sheehy et al correlated the psychotropic effects of ketamine directly to the patient’s age, with more adults reporting psychotropic effects than pediatric patients. They postulated this finding due to a lower sensitivity to ketamine in pediatric patients and meticulous dosing based on weight per unit of time in pediatric patients (13).

One limitation of this study is that due to the retrospective nature of the study some pain scores were not available. Missing pain scores could be mitigated in a prospective investigation. Additionally, between patients, there was variable duration of the hospital stay before initiation of ketamine. This could have influenced the average pain scores or could introduce a bias based on who received ketamine infusion earliest in their admission. The number of readmissions following ketamine infusion was too small to definitively suggest a relationship between ketamine and readmission. While the overall sample size was small, generalizability of our data may be limited due to the specifics of our institutional protocol. Future investigations should address the timing of initiation of ketamine and optimal management with patient-controlled analgesia. Nonetheless, our findings lend support to the recommendation to consider subanesthetic ketamine infusion for patients with SCD hospitalized with pain.

KEY MESSAGE.

• This retrospective study describes the effect of subanesthetic ketamine on pain control in children with sickle cell disease hospitalized for vaso-occlusive episodes. Improvement in pain scores and reduction in opioid consumption were observed with few side effects. These findings support the use of subanesthetic ketamine for pain control in children with SCD.