Lower Continuous Infusion, Higher Bolus Dose Patient-Controlled Analgesia Results in Shorter Hospitalization in Children with Sickle Cell Vaso-Occlusive Pain Crisis

Background:

Children with sickle cell disease (SCD) and vaso-occlusive pain crisis (VOC) require opioid analgesia, but there is little evidence to determine the most effective strategy for opioid dosing via patient-controlled analgesia (PCA) for this patient population. In January 2013, the opioid PCA dosing strategy for VOC at St. Louis Children's Hospital was changed from high continuous infusion/low demand dose (regimen A, Table) to lower continuous infusion/higher demand dose (regimen B, Table). In this retrospective cohort study, we tested the hypothesis that hospital length of stay (LOS), intravenous opioid utilization, and opioid side effects would be reduced in regimen B compared with regimen A.

Methods:

All hospitalizations for VOC management during January 1-July 31, 2012, and January 1-July 31, 2013, were identified from pharmacy records and patients' electronic medical records. Opioid doses, ibuprofen, and supportive care were part of standardized VOC order sets for both time periods. In both regimens, PCA opioid was weaned by 20-25% daily after achieving pain control, until equivalent to home opioid dose. Inclusion criteria were diagnosis of any SCD phenotype; admission for VOC; and treatment with morphine or hydromorphone PCA. Subjects were excluded for age 18 years or older; opioid continuous infusion ordered without patient-controlled demand dose; intensive care unit admission; opioid PCA ordered > 24 hours after hospital admission; post-operative PCA; participation in a clinical trial; or discharge diagnosis other than VOC. Data were abstracted from subjects' medical records into a REDCap database. Opioid administration was recorded as morphine equivalents (1 mg hydromorphone = 6.667 mg morphine). Data were analyzed in SPSS version 21 (IBM, Armonk, NY). Continuous variables were compared with the independent-samples Mann-Whitney U-test, and categorical variables with Fisher's exact test. The level of significance was specified as p<0.05.

Results:

Of 303 hospitalizations identified during the study period, 101 met all criteria for analysis: 45 in 2012 (all regimen A) and 56 in 2013 (2 regimen A, 54 regimen B). The median number of hospitalizations per subject was 2 (range 1-8). Age, sex, SCD type, acute chest syndrome diagnosis, and hydroxyurea use were not significantly different between regimens.

The median LOS was significantly shorter for regimen B hospitalizations at 5 days [interquartile range (IQR) 4, 6.25] versus 6 days (IQR 4, 9) for regimen A hospitalizations (p=0.004). The median total morphine equivalents administered per admission was significantly less in regimen B (150 mg, IQR 74.8-349) than in regimen A (246 mg, IQR 129-601; p=0.03). This was due to fewer morphine equivalents given by continuous infusion in regimen B hospitalizations (median 90.3 mg, IQR 52-158) than in regimen A (median 191 mg, IQR 102-361; p<0.001). Total morphine equivalents delivered by demand dose were not significantly different between groups (p=0.077), but a greater fraction of opioid was delivered by demand dose in regimen B (median 0.38, IQR 0.21, 0.48) compared to regimen A (median 0.09, IQR 0.021, 0.22) (p<0.001). Ondansetron was prescribed for nausea in 33.3% of regimen B versus 55.3% of regimen A admissions [odds ratio 0.65 for regimen B (95% confidence interval 0.43, 0.97, p=0.029)].

Conclusion:

In this cohort of children with SCD treated for VOC with two different opioid PCA regimens, lower continuous infusion/higher demand dose opioid resulted in significantly shorter LOS, less opioid exposure, and less use of antiemetics. Pain control may be improved when patients control a greater fraction of the opioid they receive. These results should guide future prospective studies of VOC management.