PDSA #2: Implementation phase (February 1, 2016 to June 30,
2016).
We conducted an educational intervention focused on Vitamin D deficiency
and insufficiency testing and supplementation for oncology providers
(physicians and advance practice providers), nurses, case managers, and
pharmacists via direct one-on-one in-person communication, structured
in-person lectures and secure e-mails. The decision-making tree was
distributed as hardcopy and was made widely available and posted in the
inpatient unit, outpatient clinic working space, and the secure shared
oncology folder online. EMR triggers and Vitamin D automated testing
options were incorporated. A Vitamin D laboratory order was added into
the pre-existing oncology admission order set as well as in Beacon
chemotherapy treatment plans, an Epic® oncology module
for physicians and pharmacists to create treatment and supportive care
plans based on standardized protocols that also allows for easy
modification. A BPA alert was created that triggered if the patient had
a chemotherapy treatment plan in place but did not have recent Vitamin D
testing, or if Vitamin D testing was done but the patient was not taking
Vitamin D.
In order to assess end-user value, we created a 6-question secure online
de-identified survey utilizing the SurveyMonkey®platform. The survey was administered to oncology providers after
interventions were executed. The survey included the following
questions: 1) What are the barriers preventing Vitamin D initial testing
at the time of cancer diagnosis? (Free text response); 2) Do you use the
Vitamin D supplementation decision-making tree? (Likert scale); 3) Do
you find the Vitamin D supplementation decision-making tree helpful?
(Likert scale); 4) What are the barriers preventing repeat Vitamin D
testing at 6-8 weeks after Vitamin D supplementation initiation? (Free
text response); 5) Do you use the Vitamin D BPA? (Likert scale); 6) Do
you find Vitamin D BPA reminders helpful? (Likert scale). For questions
2, 3, 5 and 6, a traditional Likert scale with 1-5 ratings was used to
assess value/usefulness, with 1 = none, 2 = little, 3 = some, 4 =
moderately high, and 5 = high. The percentage of results for
“moderately high“ or “high” for each question were reported.