METHODS
We designed a protocol as a quality improvement initiative for
introduction of LDKI for refractory or chronic pain (Supporting
information) in 2014 for our pediatric and young adult hematology and
oncology patients at UCSF Benioff Children’s Hospital Oakland. Protocol
development occurred with involvement of the pain anesthesiologist,
palliative care team, hematology/oncology, nursing and pharmacy to
provide LDKI at a dose of 1-10 mcg/kg/min (0.06-0.6 mg/kg/h), with
patients generally being prescribed 1-5 mcg/kg/min unless with chronic
refractory pain or at end of life (Supplemental material). We
subsequently obtained institutional review board (IRB) approval to
retrospectively review our initial experience from protocol initiation
in March 2014 through October 2017, when IRB approval was obtained for
hematology/oncology patients through age 21 years prescribed LDKI.
Patients were included during their first admission with LDKI. There
were no exclusion criteria.
For these patients, their primary diagnosis and reason for admission
were recorded in addition to all concomitant scheduled and as-needed
medications (opioid and non-opioid) with a focus on use on
benzodiazepines for LDKI side effects. Cumulative daily (24 hour) opioid
doses were converted to morphine equivalents. Vital signs and pain
scores from up to five days before and after initiation of LDKI were
recorded and averaged in 24-hour increments. The day of initiation of
LDKI was designated day 0 and compared with the day prior to LDKI
initiation as well as day 1 and 2 after LDKI initiation to determine if
there was a significant daily decrease in opioid utilization, pain
scores or heart rate. Given the inter-patient variability and the fact
that most all patients with sickle cell disease (SCD) were initiated on
LDKI at time of admission, these patients were compared with prior
admissions for a vaso-occlusive event (VOE), with a maximum of three
admissions recorded within one year prior to the admission in which they
received LDKI. These admissions were reviewed to determine if there was
a significant decrease in pain scores, opioid utilization or hospital
length of stay in the LDKI admission as compared with the prior
admissions. Daily physician and nursing progress notes were reviewed for
subjective descriptions of each patient’s pain and any reported side
effects to ketamine. Given the small sample size, data were compared
using medians and Mann Whitney U testing. Friedman test for repeated
measures was utilized when there were repeated days with the same number
of data points.