Transcript Schwarz Epic UGM 2014 - University of Utah School of Medicine

Improving Women’s
Reproductive Health Care
with Health IT
Eleanor Bimla Schwarz, MD, MS
[email protected]
Objectives
• Actionable data on pregnancy intentions
• Decision support for safe prescribing
to women of reproductive age
• Contraceptive e-Visits
Acknowledgements
● AHRQ R18HS017093
● Veterans Affairs
● Society of Family Planning
● Food and Drug Administration
● UPMC clinics and EpicCare data team
Each year in the US…
62 million women of reproductive age
6 million pregnancies
Maternal
Illness
Medication
Use
4 million births
150,000 babies with birth defects
March of Dimes Perinatal Data Center, 2001
Alan Guttmacher Institute, 2004
Safe prescribing to women of
childbearing potential
 Risks of teratogenic exposure
greatest in early pregnancy
 49% of US pregnancies
unintended
 6% of US pregnancies exposed to
potentially teratogenic medication
Andrade SE et al. Am J Obstet Gynecol 2004;191:398–407.
Information needed prior to conception
Perception (or misperception) of
risks impact behavior
– Discontinue needed medication
– Abortion of otherwise desired
pregnancy
Koren G et al Can J Pub Health 1991 May-Jun;82(3):S11-4, S33-7
Jasper JD, et al Lancet. 2001 Oct 13;358(9289):1237-8
“Scares” can have long-term effects
Is my child normal?
Fyro K et al
Acta Paediatr Scand. 1987 Jan;76(1):107-14
What do we know
about medications that can cause
birth defects?
Thalidomide,
Isotretinoin,
Phenytoin,
Coumadin,
Methotrexate,
Lithium…
FDA Classification System
• Class A: Fetal harm appears remote
• Class B: Animal studies revealed no evidence
of fetal harm
• Class C: No adequate studies in women.
• Class D: Evidence of human fetal risk
– Use in pregnant women may be acceptable
for serious disease when no safer drugs exist
• Class X: Contraindicated in women who are or
may become pregnant
Effective July 2015
• Previously used pregnancy risk
categories (i.e., A, B, C, D, X) will be
replaced with narrative summaries of
relevant studies
• Applies to all newly approved drugs
• Labelling for prescription drugs approved since
July 2001 will be gradually updated.
Effective July 2015
Organized into 3 sections:
• “Pregnancy”
• “Lactation”
• “Females and Males of Reproductive
Potential” (this is NEW)
Effective July 2015
Must be updated “when new information
becomes available that causes the
labeling to become inaccurate, false, or
misleading”
• Information on when data informing
the label was last reviewed is not
required.
Are women of reproductive age
using class D and X Medications?
How often are potentially teratogenic meds Rx’d?
Annually:
11.7 million teratogenic Rx = 1 of every 25 Rx
283 million prescriptions
for women age 14-44 years who were not pregnant
11.7 million teratogenic Rx
147 million outpatient visits = 1 of every 13 visits
for women age 14-44 years who were not pregnant
Schwarz EB,et al. Am J Med 2005
In a large HMO…
1/6 women filled
a class D or X Rx
Schwarz EB,et al. Ann Intern Med. 2007
Female Vets…
49% of female Vets who
received a Rx from a VA
pharmacy received a
potential teratogen
Schwarz EB, et al Med Care 2010
Most are chronic meds
Women who filled
>=1 Rx
N=38,205
N
%
Days supply per
woman/year
Average
Antihypertensive
8115
21.2
195.5
Warfarin
547
1.4
164.5
Statin
8790
23.0
153.5
Neurologic
4687
12.3
134.0
Psychiatric
20155
52.8
133.5
Tetracycline
6235
16.3
28.0
75
0.2
65.5
7562
19.8
34.5
Retinoid
Other
Schwarz EB, et al Med Care 2010
Does FDA label affect
contraceptive counseling?
Safe Rx
Class A or B
(95% CI)
Risky Rx
Class D or X
(95% CI)
p-value
Nationally, visits with
contraceptive counseling or Rx
5.4%
4.1%
p=0.24
In Northern CA, Rx with
contraceptive counseling or use
39.4%
37.0%
p<0.001
Apparently
not
Schwarz EB, et al. Am J Med 2005
Schwarz EB, et al. Ann Intern Med. 2007
Does contraceptive method affect
rates of positive pregnancy tests?
Women 15-44 yo with + pregnancy test
<3 months after filling a potentially teratogenic Rx
Overall
1.0
Women with any contraceptive documentation
0.8
Women with a most effective contraceptive
0.2
%
Schwarz EB, et al Ann Intern Med 2007
Nationally…
 <20% of US women receiving
potentially teratogenic medications
had documented receipt of family
planning services
Schwarz EB et al. Am J Med 2005;118:1240–9.
Limited PCP visit time
Electronic decision
support tools
Lack of
reimbursement
∆ reimbursement
structure
Difficulty identifying
teratogenic risk
Education resources for
providers and patients
Facilitators
Barriers
Barriers & Facilitators
Schwarz EB et al. Birth Defects Res A Clin Mol Teratol. 2009 85(10): 858–863.
Eisenberg DL et al. J Gen Intern Med. 2010;25(4):291-7.
What affects rates of safe prescribing to
women of reproductive age?
• Patient characteristics?
– Age=
• at KP: 25-34 yrs > 15-24 yrs > 35+ yrs
• at VA: 18-25 yrs > 26-34 yrs > 35+ yrs
• at UPMC: 18-29 yrs > 30+ yrs
• Provider characteristics?
– Gyn > non-gyn; NPs/PAs/CNMs > MDs
– Not gender, age, race, or years in practice
Schwarz EB, et al. Am J Med 2005
Schwarz EB et al. Med Care 2010
What affects rates of safe medication use?
• Medication class/Clinical indication
– Isotretinoin better than others
– Coumadin worse than others
• Practice location
– Local culture one of the most significant factors
• At the VA, OR=1.6
• At KP, OR=2.0
• At UPMC, OR=2.0
Schwarz EB, et al. Am J Med 2005
Schwarz EB et al. Med Care 2010
Objectives
• Actionable data on pregnancy intentions
• Decision support for safe prescribing
to women of reproductive age
• Contraceptive e-Visits
Electronic intake system
Contraceptive Vital Sign
Ideal Implementation
Patient greeted
& prompted to
use electronic
health
screening
system
Wirelessly
networked
tablet
computer
system
Patient reports
pregnancy
intentions,
contraceptive
use
Transferred to
EMR & informs
clinical decision
making
Specific Questions...
“Are you currently pregnant or trying to become
pregnant?”
• No
• I wouldn’t mind being pregnant
• I am trying to get pregnant
• I am currently pregnant
• I have been through menopause
• Prefer not to answer
Specific Questions...
“There are many ways that people try to avoid
becoming pregnant. Which are you using?”
•
•
•
•
•
•
•
•
•
Tubal ligation
Hysterectomy
Vasectomy
Intrauterine device
Implanon
Depo-provera shot
Progestin only pill
Birth control pill
Patch
•
•
•
•
•
•
•
•
•
Vaginal ring
Condoms
Diaphragm
Spermicide
No method
Withdrawal
Other
No sex with a man in the last 3 months
Prefer not to answer
Specific Questions...
“Some of the medications that your doctor may
prescribe can be harmful to babies who are
breastfeeding. Are you currently breastfeeding
or nursing a baby?”
• No
• Yes
Acceptability
• 93% answered questions on pregnancy
intentions and contraceptive use
• 7% either skipped the question
or ran out of time
Pregnancy
• 4% Pregnant or trying to conceive
• 3% Wouldn’t mind pregnancy
Contraception
• 8% No method
• 22% Behavioral or barrier methods
Lactation
• 1% Currently breastfeeding
Patients reporting information that
may affect prescribing decisions…
35%
Prescribing decisions
may be affected
Prescribing decisions NOT
affected
Useful, but…
• New family planning services increased
only minimally
• Even for women prescribed teratogens
– Only 7% with documented nonuse of
contraception prescribed potential teratogens,
were provided family planning services.
Effect of a contraceptive vital sign
on documented contraception
Challenges
• New family planning services increased
only minimally
• When women with documented nonuse of
contraception were prescribed potential
teratogens, only 7% were provided family
planning services.
Conclusions
Systematic collection of patient-reported
reproductive health information is…
Feasible
Acceptable to women
Helpful to PCPs
The coming STANDARD OF CARE
Going live Summer 2015
√
Going live Summer 2015
Some medicine can be harmful during pregnancy. To understand each
patient’s plans and risks for pregnancy, please ask the following
questions:
“Are you currently pregnant?”
 Yes
 No
 Don’t know  Prompt to order pregnancy test
“Do you want to become pregnant within the next year?”
 Yes Template for preconception care
 No
 Don’t know
“There are many ways that people try to avoid becoming pregnant.”
“What are you doing to prevent pregnancy?”
Extremely effective methods (Typically <1 pregnancy/100 woman-years)
 Intrauterine Device (IUD – Paragard, Mirena, Skyla)
 Implant (e.g., Nexplanon)
 Partner has a vasectomy
 Female partners only
 Life-long abstinence
Highly effective methods (Typically 3 pregnancies/100 woman-years)
 Contraceptive Shot (e.g., Depo-Provera)
Very effective methods (Typically 8 pregnancies/100 woman-years)
 Oral contraceptive (estrogen/progesterone)
 Progestin-only oral contraceptive
 Contraceptive vaginal ring (e.g., NuvaRing)
 Contraceptive patch (e.g. Evra)
Effective and less effective methods (Typically 14-29 pregnancies/100 woman-years)
 Male condom
 Female condom
 Diaphragm/cervical cap with spermicide
 Withdrawal
 Nothing
 No current partner
Objectives
• Actionable data on pregnancy intentions
• Decision support for safe prescribing
to women of reproductive age
• Contraceptive e-Visits
Design
41 primary care physicians
randomized
Simple Alerts
Multifaceted Alerts
“Concern has been raised about
use of this med during pregnancy.”
“Concern has been raised about
use of this med in pregnancy.”
Plus links to:
•Info on meds
•Order for pregnancy testing
•Orders for contraception
•Refer to specialist
Hypotheses
 Simple alert would be good
 Multifaceted alert would be better
Outcomes
 Data source = electronic medical record
= patient surveys
 Visits with Rx of a potential teratogen
 Visits with Rx of a potential teratogen
& documented “family planning services”:
• Pregnancy testing
• Contraceptive Rx
• Contraceptive counseling
• Referral to a gynecologist or
family planning specialist
Data Collection
T0: No alerts (10 months)
T1: Double Intervention (8.75 months)
- Simple vs. Multifaceted alerts
T2: Simple Intervention (8.75 months)
- Simple vs. No alerts
Physician characteristics
Physicians
Simple alert
n=17
Multifaceted
n=24
40 (10)
40 (10)
Gender, % Female
47%
50%
Training, % Family Med
29%
29%
214 (353)
163 (334)
Age, mean (SD)
Visits, median (IQR)*
* p < 0.05
Visits Characteristics
Simple alert
n=5,433
Multifaceted
n=7,243
34 (9)
34 (10)
% Married patients*
44%
47%
% White patients*
89%
86%
% New patient*
9%
11%
% Visit with regular MD*
56%
66%
Patient age, mean(SD)*
* p < 0.05
Results
Clinical Decision Support
Schwarz EB, et al
JGIM 2012
Clinical Decision Support
Schwarz EB, et al
JGIM 2012
Visits with Teratogenic Rx &
Family Planning Services
* p = 0.04 for improvement over time
+3.0%
+2.0%
+1.5%
+2.6%
Results
● 3 types of patients:
 Contraception on all visits = 26%
 NO contraception on any visits = 69%
 No contraception before intervention,
with contraception after alert = 5%
Challenges
● Doctors who received multifaceted alerts
accessed links only 16% of time.
● Alerts fired only once per visit:
 14% doctors who received alert cancelled Rx,
AND unknowingly Rx’d another teratogen!
 Limits visible effect of intervention.
Limitations
 Electronic medical record data incomplete
 Vasectomy, sterilization, condoms
 Contraceptive counseling ICD-9 codes rarely used
because no reimbursement provided
 Should not differ between groups
Not all counseling is documented
• Surveyed 801 women
• 5-30 days after clinic visit
• 27% received a potentially teratogenic Rx
Schwarz EB, et al J Womens Health 2013
Not just poor documentation
Counseling about risk of birth defects and/or
contraception was reported by Women
• 41% if no potentially teratogenic Rx
• 45% if pot’ly teratogenic Rx, no CDS
• 57% if pot’ly teratogenic Rx and CDS
Schwarz EB, et al J Womens Health 2013
Conclusions
 Computerized alerts can be helpful,
but refinement is needed.
 Alerts need to re-fire if Rx another teratogen.
 Routine documentation of fertility and
contraception is important for safe Rx to women
of childbearing age.
Version 2.0
includes Link to
Regularly Updated Drug Database
Objectives
• Actionable data on pregnancy intentions
• Decision support for safe prescribing
to women of reproductive age
• Contraceptive e-Visits
Challenges
• 11% of women (15-44yo) who do not desire
pregnancy are NOT using contraception.
• About 30% of all contraceptive users
experienced a gap in contraceptive use during
the previous year.
Frost et al., 2007; Grossman et al., 2010;
Potter et al. 2011
Some of these contraceptives
are NOT like the others
• Prescription methods are
at least 2x as effective as
barrier/behavioral methods
Some of these things
are NOT like the others
• Some are 20x more effective
than pills
– Subdermal
• Nexplanon
– Intrauterine
• Mirena, Paragard, Skyla
– Vasectomy
– Tubal ligation
Computer-assisted
contraceptive counseling
Schwarz EB et al JGIM 2008: 794-9
Schwarz EB et al Contraception 2013
Can we do more?
• Increase knowledge of contraceptive options
• Screen for contraindications to estrogen
• Facilitate Rx for hormonal contraception
Pregnancy
Intentions?
Trying to conceive
or Pregnant
Counseling
about
FOLATE
Trying to
avoid pregnancy
INFO about
contraceptives
CHC
Pill, Patch, Ring
Progestin-only
Pills, DMPA
HER-C
IUDs, implanon
EC
Offer of Rx
Screen for
contraindications
Rx printed
with summary
MD signs
Information on each method
Acceptability of
Family Planning Module
Easy to Use
Trust Information
No
Recommend
Yes
Learn Something
New
0
0.2
0.4
0.6
0.8
1
Rx requests
• 19% of all those in intervention group
• 48% of women who used no method
• 60% of women who used withdrawal
• Less educated & non-white women
MORE likely to request Rx
• 57% of women who received Rx reported
1+ episodes of unprotected sex in prior month
Schwarz EB et al Contraception 2013
Important points
It worked!
• 16% vs. 1% controls, p=0.001 received a
contraceptive Rx the day of clinic visit
• New Rx were NOT more likely than refills to
have potential contraindication to estrogen
identified by computerized screening
• 75% of refills vs. 52% new Rx, p=0.23
Schwarz EB et al Contraception 2013
Can we do this thru a patient
portal?
YES
Check out our
“Contraceptive eVisit”
https://myupmc.upmc.com/anywhere-careapp#!/tmnt/t1/3
Thank You!
Questions?
[email protected]