Transcript Prescribing Errors on Discharge Prescriptions

ELECTRONIC DISCHARGE
PRESCRIPTIONS AT PRINCESS OF
WALES HOSPITAL
Rowena Duffield
ELECTRONIC TRANSFER OF CARE (ETOC)
 First
ward in May 2007
 Now on over 440 beds
 Provides GP with a detailed list of
medicines on day of discharge
Pre-existing medicines
 Dose changes to pre-existing medicines
 New medicines and reason for starting
 Stopped medicines and reason for stopping

ELECTRONIC TRANSFER OF CARE
 Junior







doctor fills in the
initial review
test results
diagnosis
procedures
medication
management plan
follow up arrangements
 Pharmacist
verifies and signs off
medication
 Consultant or senior registrar signs off
clinical summary
TRANSFER TO GP
 On
the day of discharge the GP gets:
diagnosis, medication list, management
plan, follow up arrangements
 Once
the whole form is signed off the GP
gets the complete summary
COMMENTS FROM GPS
 “Useful
and timely”
 “Much clearer which medicines are preexisting, new or stopped”
 “Much easier to map information onto
patient medication record”
 “Has reduced errors due to poor
transcription and illegibility”
IN SUMMARY
 We
have an electronic discharge
prescription and discharge summary sent
directly to the GP
 We send the medicines list on the day of
discharge for 97% of patients
INVESTIGATION INTO INCIDENCE,
TYPES AND CAUSES OF PRESCRIBING
ERRORS ON HANDWRITTEN AND
ELECTRONIC DISCHARGE
PRESCRIPTIONS.
MSc Dissertation
BACKGROUND TO PROJECT
 Prescribing
errors are a risk to patient
safety
 Electronic prescriptions have been shown to
increase patient safety
 Electronic discharge prescriptions
implemented in Princess Of Wales
 Investigate prescribing error rate on paper
(TTH) and electronic (eTOC) discharge
prescriptions
OBJECTIVES OF PROJECT
 To
determine the proportion of prescribing
interventions on TTHs and eTOCs
 To categorise the types of prescribing
interventions on TTHs and eTOCs
 To compare the severity of errors made on
a sample of TTHs and eTOCs
 To identify prescribers’ ideas or reasons to
the causes of errors on eTOCs and potential
solutions.
METHODS
 Retrospective
collection of completed TTHs
and eTOCs
 Sample of TTH and eTOC errors graded for
severity by nurses/doctors/pharmacists
 15 prescribers interviewed, doctors that
made most errors and all four nurse
prescribers
EXAMPLES OF ERRORS FOUND ON ETOCS AND
TTHS
Type of Error
Example of Error
Name of drug
Rivaroxoban 20mg at night prescribed instead of
Rosuvastatin 20mg at night
Form of drug
Nifedipine (Adalat Retard) modified release tablet 10mg
three times daily prescribed instead of Nifedipine normal
release capsule 10mg three times daily
Strength of drug
Digoxin 250micrograms daily prescribed instead of digoxin
125micrograms
Dosage instructions
Clonidine 25micrograms daily prescribed instead of clonidine
25micrograms twice daily
Duration
Flucloxacillin 500mg four times daily for 6 days instead of 6
weeks
Omission
Diltiazem (Tildiem Retard) modified release tablet 90mg
twice daily omitted
Stopped medicine
listed as current
Lactulose 10ml twice daily stopped during admission listed
as current on discharge prescription
EXAMPLES OF ERRORS FOUND ON ETOCS ONLY
Date of initiation
Doxazosin 2mg twice daily listed as “prescribed prior to
admission” instead of “new medication”
Stopped medicine not Spironolactone 25mg daily stopped during admission not
listed on eTOC
listed on discharge prescription
Stopped medicine on
eTOC that is
current on
prescription
Glyceryl trinitrate spray listed on eTOC as “stopped” when
it was a current medicine
Medicine change not
communicated
Isosorbide mononitrate 20mg twice daily listed as
“unchanged” when increased from isosorbide
mononitrate 10mg twice daily in hospital
Medicine on eTOC
that is not current
Carbocisteine liquid 250mg/5ml 15ml twice daily listed as
current medicine when patient not prescribed it on
medicine chart
RESULTS FROM TTHS AND ETOCS
 342 eTOCs analysed ↔
336 TTHs analysed
 3491 items in total ↑
 2028 items in total
 23% items had intervention  22% items had intervention ↔
 71% TTHs had intervention  76% eTOCs had intervention ↔
 16% item had error ↑
 12% items had error
 60% eTOCs had ≥1 error ↑
 44% TTHs had ≥1 error

No difference in rates of interventions
Odds of eTOC having an error were 1.95 times
higher than a TTH
COMPARING LIKE-FOR-LIKE ERRORS
ONLY
 The
extra types of error on eTOCs
contributed ~35% of all errors
 Comparing like-for-like errors only
results in significantly more errors on
TTHs than eTOCs
 Odds of item having error on TTH is now
1.62 times higher than on eTOC
Overall intervention rate now significantly
higher on TTHs than eTOCs
ASSOCIATION BETWEEN TYPES OF ERROR
AND TYPES OF PRESCRIPTION
Paper
Prescription
Electronic
Prescription
Odds
ratio
Significant
Difference
Duration
3.3%
0.7%
5.13
Yes
Strength of Drug
4.8%
2.6%
2.11
Yes
Stopped medicine listed as current
4.6%
3.7%
1.37
Yes
Omission
8.9%
11.3%
1.20
Yes
Form of Drug
1.9%
5.6%
2.94
No
Name of Drug
1.0%
2.0%
1.87
No
Dosage Instructions
5.8%
8.2%
1.34
No
Date of initiation
n/a
11.2%
n/a
n/a
Medicine on eTOC that is not
current
n/a
8.8%
n/a
n/a
Stopped medicine not listed on
eTOC
n/a
8.7%
n/a
n/a
Medicine change not communicated
n/a
6.3%
n/a
n/a
Stopped medicine on eTOC that is
current on prescription
n/a
0.5%
n/a
n/a
MAJORITY OF PRESCRIBING AT DISCHARGE
WAS BY MORE JUNIOR DOCTORS
(DATA ONLY AVAILABLE FOR ETOCS)
45
Number of
Prescribers
40
35
Percent
30
Number of Items
Prescribed
25
20
15
10
5
0
F1
F2
CT1
CT2
Grade of Prescriber
Registrar
Nurse
Prescriber
COMPARISON OF PRESCRIBING SHOWED A TREND
OF INCREASING ERROR RATE WITH NUMBER OF
YEARS EXPERIENCE AS A DOCTOR
Prescriptions
prepared
Items
prescribed
Error rate
Clarification
rate
Intervention
rate
Registrar
2.0%
2.1%
24.0%
6.7%
29.3%
CT1
20.2%
21.1%
19.0%
9.4%
26.6%
CT2
9.1%
8.8%
18.6%
3.6%
21.9%
F2
20.8%
22.2%
17.2%
5.6%
20.8%
F1
42.1%
39.4%
13.1%
6.3%
18.9%
5.8%
6.4%
13.9%
4.5%
18.4%
Nurse
Prescriber
OTHER COMMENTS ABOUT ERRORS
 There
were significantly more errors on
Medicine and Elderly Care prescriptions
than General Surgery and Trauma and
Orthopaedics
 Significantly more errors on TTHs for
respiratory medicines
RESEARCHERS’ ESTIMATIONS OF SEVERITY OF
ERRORS SHOWED SIGNIFICANT DIFFERENCES
100%
Percentage of errors
80%
60%
Paper
Electronic
40%
20%
0%
Minor
Moderate
Severity of Error
Major
THE SEVERITY OF ERRORS MADE BY F1S AND NPS TENDED
TO BE OF LOWER SEVERITY THAN FROM DOCTORS WITH
GREATER NUMBER OF YEARS’ EXPERIENCE
SEVERITY JUDGING
 Sample
of eTOC and TTH errors
 Judged by nurses, doctors and pharmacists
 Range of specialties (including 2 GPs)
 Visual analogue scale 0-10cm
 Mean scores calculated for each error
 Researchers’ estimation confirmed in 87.5%
cases
Mean scores for eTOC errors were lower
than TTH errors
INTERVIEWS WITH PRESCRIBERS
 15
interviews
 Range of grades and specialties
 Audio-taped and transcribed verbatim
 Transcripts analysed looking for causes of
errors
CAUSES OF PRESCRIBING ERRORS
 Documentation





-EPC
Many errors blamed on poor documentation
Prescribers couldn’t find information in the notes
Decisions made without entries in notes
Changes made to chart with no explanation
But prescribers didn’t always look back through
all the notes because it was too time-consuming
CAUSES OF PRESCRIBING ERRORS
 Legibility

“Sometimes it is difficult because you have to
work out what the hell the drug is, there is
poor handwriting on drug charts” (NP)
 Time

in notes and on charts -EPC
Pressure – Error producing condition
“I think the reason why errors occur is due to
the time on the ward. I don’t have time to do
them properly” (CT1)
CAUSES OF PRESCRIBING ERRORS
 Distractions

on the ward
“when you’re on the ward, you’re doing a few
things at a time, and you’ve got to sit down and
do this properly, and sometimes that’s not
always possible, you’re called to do different
things” (F1)
CAUSES OF PRESCRIBING ERRORS
 Unfamiliar

“Those of us who do it more regularly just get on with
it, those who do it infrequently aren’t going to be
familiar with it” (CT2)
 Unfamiliar

with eTOC -EPC
with medication
“a major error on the Epo, it’s on as one three times a
week, and it should have been twice a week. Epo isn’t
a medicine I’m familiar with, and normally when I
prescribe things I know in my head whether it’s
clearly wrong and it wouldn’t have been an error I’d
pick up on when inputting it onto the computer” (F2)
CAUSES OF PRESCRIBING ERRORS
 Unfamiliar
with patient
Being absent during ward round
 Prescription for patient under a different team
 Cover prescriber on nurse-led rehab ward


“It is more difficult to do it if you don’t know the
patient, because normally we are not seeing the
patient and then you have to go to rehab ward and
do an eTOC for the patient” (F2)
CAUSES OF PRESCRIBING ERRORS
 More
than one person involved in eTOC -
EPC

“This is an example of where the nurse
practitioners look after the patients
throughout their journey in the rehab ward
including the eTOC and they put in all these
details for the patients. I am aware of them,
but I haven’t checked every detail on it which
is why these errors have occurred” (Registrar)
CAUSES OF PRESCRIBING ERRORS
 Length
of stay of patient - EPC
Errors more likely to occur when chart
rewritten
 “when I do the eTOCs I only use current
charts, that is something to be aware of” (F1)

 Other
causes:
“pressure on the doctors” (F2)
 “an oversight” (CT2)
 “not taking due care and attention” (NP)

KNOWLEDGE OF ERRORS
 Awareness
of mistakes
“With things like inhalers I was pretty sure I
was making mistakes with those” (F2)
 “If I would have known I wouldn’t have done it!
(laughs)” (F2)

 Change

of practice
“You only learn from knowing about your
mistakes, so now that I’m more aware, I think
I’ll be more careful next time” (CT1)
KNOWLEDGE OF ERRORS
 Feedback

“You are never going to able to improve if you
don’t know what you’re doing wrong, you’re just
going to keep doing it” (F2)
 Consequences

“GPs have to know from this exactly what
happened during the admission, and if it’s not
precise the GP doesn’t know what’s going on, and
that can be highly detrimental” (CT2)
PRESCRIPTION PREFERENCE
MORE PREFERRED TTH TO ETOC BUT MANY
FELT ETOC IS SUPERIOR COMMUNICATION
Advantages of TTHs
 Speed of handwriting over typing
 Takes less time out of their day
 Satisfies nursing requests faster
 Only needs to be list of medicines
Disadvantages of TTHs
 Legibility problems due to poor handwriting
 Less detail about medicines given to GP
 Limited transfer of information onto carbon
copies if prescribers pressed lightly
Advantages of eTOCs
 Short codes for dosing instructions
 Instant access to a perfectly legible copy of
prescription in a clinic or if patient re-admitted
 Benefits of improved information transfer to the
GP outweighed the downsides of the eTOC
Disadvantages of eTOCs
 The requirement to enter clinical details such as
diagnosis
 One prescriber thought that the increased
demand for information may lead to more errors
on eTOCs
 Inconvenience of eTOC lay with computers not
the form itself or level of detail required
E.g. slow computers, password problems or not
enough computers

TYPES OF ERRORS
 Name

of drug
“I know it’s not good to make errors, but creams
and eye drops is stuff that I consider to be a bit
‘over-the-counter’…” (CT2)
 Form
of drug
“If it’s not written on the chart, I just guess!”
(CT1)
 “I know there are lots of different types of
inhalers and it’s beyond me sometimes” (F1)

TYPES OF ERRORS
 Date

of initiation
“Sometimes you just put ‘prescribed before
admission’ for everything because it’s the easy
option” (CT2)
 Omission

“Umm… not sure why I’ve missed them,
perhaps the patient had a second chart?!” (F1)
OTHER COMMENTS
 Computer


related problems
Speed of computers and availability on wards
“I think once you’ve got the hang of doing it, you can do it
quite quickly - computer speed dependent” (F1)
 Pharmacist

annotations
“Yeah because otherwise we wouldn’t have a clue, cos
often, even if you look in the clerking notes, the doctor
could have missed stuff off, or we’re not sure whether it
has just been started. The annotations from the
pharmacist I rely on a lot” (F1)
OTHER COMMENTS
 Differences
between grades of prescriber
“Nurse prescribers may be better, especially with
inhalers, cos they open the locker and give it to
the patient, whereas we never clap eyes on it.
They may be better at trade names as well” (F1)
 “I think you probably get worse as you get more
experienced, cos I think registrars will be far less
familiar with eTOC” (F2)

OTHER COMMENTS
 Suggestions

to improve eTOC
“I’m not sure anything can be done, it’s just more time
consuming, but these things do need to be communicated
to the GP” (F2)
 Positives
about eTOC
“I often use them if we’ve got someone coming into A&E
and I haven’t got a clue about their history, I open [the
eTOC], they are really useful to know what meds they
have been on in the past” (F2)
 “We don’t have discharge summaries anymore because
they took an eternity to do, this is a combined process, it’s
much easier” (CT2)

SUGGESTIONS TO REDUCE ERRORS
 Attend
training asap and ask for help
 Start eTOC early in admission
 Allow time to do eTOC without rushing
 Write clearly on charts and in notes
 Fully document all decisions in notes/on
chart
 Justify prescribing choices in notes
 Use original medication history
 Extra care with long admission/many
SUGGESTIONS TO REDUCE ERRORS
 Be
mindful of distractions on the ward
 Don’t allow undue interruptions
 Be conscious of unfamiliar medicines
 Be aware anyone can make mistakes
 Responsibility lies with final prescriber
 Ask for feedback from pharmacist
IN SUMMARY…
 Significantly
more errors on eTOCs when
looking at all possible error types
 Omission most common type of error
 When looking at like-for-like error types
only, eTOCs had significantly fewer overall
interventions and errors
 F1s and NPs had lowest error rates
IN SUMMARY…
 Significantly
more minor and fewer major
errors on eTOCs
 F1s and NPs made more minor and fewer
major errors than doctors with more years’
experience
 Mean severity scores for eTOC errors were
consistently lower than for TTHs
IN SUMMARY…
 Interviews
highlighted many causes of
errors
 Time pressure, poor documentation,
distractions on the ward
 Being unfamiliar with the medicine,
patient, or eTOC system increased chances
of error
SINCE THE PROJECT FINISHED
 Met
with pharmacists to discuss what
we could do to help
Decided to fully endorse all types of
inhalers
 Ensure endorsements transcribed onto
rewritten charts

 Met
with Consultants to share results
and discuss how we use this data to
improve prescribing
SINCE THE PROJECT FINISHED
 Consultants
interested in findings and
surprised by some interview quotes
 Felt junior doctors had to take more
responsibility for providing good
communication to GPs
 Gave me permission to present to junior
doctors
 Junior doctors felt only extra time and less
pressure would reduce error rates