CME
Evidence-Based Practice
A Peer-Reviewed Publication of the Family Physicians Inquiries Network
LETTER TO THE EDITOR
2A question about strength
of recommendation
EDITORIAL
3
The $1,000 pill
DIVING FOR PURLs
4
Honey for acute cough in children
Brief intervention for medication overuse
headaches
5
Atrial fibrillation in cryptogenic stroke
Shoe advice for chronic foot pain
EBPEDIATRICS
6Office-based literacy interventions
to promote language development
in children
TOPICS IN MATERNITY CARE
7Safety, efficacy, and patient satisfaction
of uterine aspiration for miscarriage
HELPDESK ANSWERS
8Nonsurgical therapy for carpal tunnel syndrome
Panic disorder treatment
9BUN-to-creatinine ratio for predicting
acute kidney injury
10
Injection for trigger points
11
Nonpharmacologic treatment for IBS
12
Fennel tea for infantile colic
Treatment for alopecia areata
E1Microalbuminuria screening
in hypertensive patients
Routine screening for depression
in patients with asthma
E2Cannabinoids for treatment
of rheumatoid arthritis
E3Effectiveness of patient-controlled
analgesia for patients with sickle cell
vaso-occlusive crisis
E4
Routine testing for tuberculosis
in low-risk areas
E5
Biofeedback for essential hypertension
SPOTLIGHT ON PHARMACY
14
Serum magnesium levels for migraine sufferers
CME TEST
15
May 2015
V O L U M E 18 N UM BER 5 M AY 2015
IN DEPTH
What is the best way to determine if a patient
has type 1 or type 2 diabetes mellitus?
Evidence-based answer
The only way to differentiate various types of diabetes is by
assessing a patient’s history, physical, and laboratory workup. There
is no single test or finding that differentiates the 2 types (SOR: B,
cross-sectional studies).
Evidence summary
A multicenter, cross-sectional study of 2,435 newly diagnosed
youth with diabetes (age <20 years) classified incidence of diabetes
based on race/ethnicity and diabetes type.1 In the 0–9 year-old
group, type 1 diabetes (based on healthcare provider assessment)
was present in 99% of children and type 2 was present in 1%.
However, the incidence of type 2 diabetes increased in the 10–19
year-old group to 23%.
Further analysis of this age group revealed that GAD65
autoantibody was present in 21% of patients with type 2 diabetes
and 66% of patients with type 1 diabetes. This result suggested
that GAD65 antibody positivity was not unique to either type of
diabetes.1
A cross-sectional analysis of data from a multicenter RCT with
1,206 patients aged 10–17 years with type 2 diabetes, previously
diagnosed by an endocrinologist based on their phenotypic
presentation, evaluated the frequency of islet autoantibodies (ie,
GAD65 and IA2) and described associated clinical and laboratory
findings.2 Autoantibody positivity for either GAD65 or IA2, or
both, was noted in 9.8% of children who (based on comprehensive
clinical evaluation) had type 2 diabetes.
When various other clinical parameters were compared (BMI,
blood pressure, acanthosis nigricans, insulin use, HbA1C, c-peptide,
lipid panels), the authors found that regardless of antibody status,
clinical findings overlapped in antibody-positive and antibodynegative individuals. As an example, acanthosis nigricans, a typical
sign of insulin resistance, was present at a somewhat higher rate
Evidence-Based Practice / Vol. 18, No. 5
1
In Depth
in antibody-negative than antibody-positive individuals
(84% vs 68%, respectively; P<.01). The median BMI
in antibody-negative individuals was higher than in
antibody-positive individuals (34.9 vs 21.9 kg/m2;
P<.01).2
Ivan Zubkov, MD
Stacie Cruz, MD
Kaiser Permanente Fontana FMR
Fontana, CA
REFERENCES
1.Writing Group for the SEARCH for Diabetes in Youth Study Group, Dabelea D, Bell RA,
D’Agostino RB Jr, et al. Incidence of diabetes in youth in the United States. JAMA. 2007;
297(24):2716–2724. [STEP 3]
Recommendations
The American Association of Clinical Endocrinologists
recommends a combination of clinical presentation,
family history, and laboratory workup (including insulin
level, c-peptide, islet autoantibodies) to help distinguish
EBP
type of diabetes, particularly in younger people. 3
2.Klingensmith G, Pyle L, Arslanian S, et al; TODAY Study Group. The presence of GAD and
IA-2 antibodies in youth with a type 2 diabetes phenotype: results from the TODAY study.
Diabetes Care. 2010; 33(9):1970–1975. [STEP 3]
3.Handelsman Y, Bloomgarden ZT, Grunberger G, et al. American Association of Clinical
Endocrinologists and American College of Endocrinology— Clinical Practice Guidelines for
developing a diabetes mellitus comprehensive care plan—2015. Endocr Pract. 2015; 21
(suppl1):1–87. [STEP 1]
Letter to the Editor
A question about strength of recommendation
The Help Desk Answer (HDA) entitled, “What is the most
effective medication for weight loss?”1 provided a clear
summary of recent articles, all of which were STEP 2 level
of evidence. The “Evidenced-Based Answer” was graded
Strength of Recommendation (SOR) B. Yet the write-up reads
as if the SOR should have been C.
Weight is a factor similar to blood pressure or cholesterol
level. It is a measure of disease. Improving the number is
of patient outcome value if it is linked with improvement in
morbidity, mortality, quality of life, or cost. Hence, without
clear clinical outcomes linked to modest weight loss achieved
by the weight loss drugs in the HDA—phentermine with
topiramate (Qsymia®), lorcaserin, phentermine—SOR C would
have been more accurate.
Editor’s reply
Thank you for writing! You make an excellent point:
since weight is a risk factor for many noxious outcomes
like stroke and heart attack, one might reasonably
treat weight as an intermediate outcome like blood
pressure or cholesterol level. Yet weight alone is also
an outcome that matters to many patients, societal
pressures being what they are.
I think the authors are okay to use the B rating
and we can admit there will be some disagreement on
that point.
Stephen A. Wilson, MD, MPH, FAAFP
University of Pittsburgh UPMC St. Margaret
Pittsburgh, PA
REFERENCE
1.Milliron SK, Shimman J. What is the most effective medication for weight loss? Evid Based
Pract. 2014; 17(5):9–10.
2
Evidence-Based Practice / May 2015
INTERESTED IN SUBMITTING A LETTER TO THE EDITOR?
Visit www.fpin.org/letters or email
ebp@fpin.org for more information.
Evidence-Based Practice
Editorial
EDITOR-IN-CHIEF
Jon O. Neher, MD, FAAFP
University of Washington
EXECUTIVE EDITOR/FOUNDING EDITOR-IN-CHIEF
Bernard Ewigman, MD, MSPH, FAAFP
The University of Chicago
The $1,000 pill
DEPUTY EDITORS
Clinical Inquiries
Gary Kelsberg, MD, FAAFP
E. Chris Vincent, MD
University of Washington
HelpDesk Answers
Corey Lyon, DO
University of Colorado
Diving for PURLs
Goutham Rao, MD
The University of Chicago
Rick Guthmann, MD, MPH
University of Illinois
eMedRef
Robert Marshall, MD, MPH
Valley Medical Center
Kate Rowland, MD
Rush–Copley Medical Center
SECTION EDITORS
Behavioral Health Matters
Vanessa Rollins, PhD
University of Colorado
Geriatrics
Irene Hamrick, MD
University of Wisconsin
Musculoskeletal Health
Andrew W. Gottschalk, MD
Cleveland Clinic
EBM on the Wards
Corey Lyon, DO
University of Colorado
Integrative Medicine
David Rakel, MD, FAAFP
University of Wisconsin
Pharmacy HDAs
Connie Kraus, PharmD, BCACP
University of Wisconsin
EBPediatrics
Jonas A. Lee, MD
A. Ildiko Martonffy, MD
University of Wisconsin
Maternity Care
Lee Dresang, MD
University of Wisconsin
PRODUCTION
Medical Copy Editor
Melissa L. Bogen, ELS
Chester, NY
Managing Editor
Kerri Reynolds, BJ
Columbia, MO
Layout and Design
Robert Thatcher
New York, NY
Statement of Purpose
Evidence-Based Practice (EBP) addresses important patient care questions asked by
practicing family physicians, using the best sources of evidence in a brief, clinically
useful format. Our goal is to instruct our authors on how to write peer-reviewed
scholarly research for the medical and scientific community.
Disclosure
It is the policy of the University of Colorado School of Medicine to require the
disclosure of the existence of any relevant financial interest or any other relationship
a faculty member or a provider has with the manufacturer(s) of any commercial
product(s) discussed in an educational presentation. In meeting the requirements
of full disclosure and in compliance with the ACCME Essentials, Standards for
Commercial Support, and Guidelines, the following information has been provided
by the editors regarding potential conflicts of interest: Jon O. Neher, M.D. and John
Saultz, M.D. have disclosed no relationships with commercial supporters.
The PURLs Surveillance System is supported in part by Grant Number UL1RR024999
from the National Center for Research Resources, a Clinical Translational Science
Award to the University of Chicago. The content is solely the responsibility of the
authors and does not necessarily represent the official views of the National Center
for Research Resources or the National Institutes of Health.
EBP CME
Evidence-Based Practice (ISSN 1095-4120) is published monthly to family clinicians
by the Family Physicians Inquiries Network, Inc. FPIN is a nonprofit 501(C)3
educational and research consortium.
Accreditation Statement
This activity has been planned and implemented in accordance with the Essential
Areas and Policies of the Accreditation Council for Continuing Medical Education
through the joint sponsorship of the University of Colorado School of Medicine and
Family Physicians Inquiries Network. The University of Colorado School of Medicine
is accredited by the ACCME to provide continuing medical education for physicians.
Credit Designation
The University of Colorado School of Medicine designates this enduring material for
a maximum of 48 AMA PRA Category 1 CreditsTM (4 AMA PRA Category 1 CreditsTM
per issue). Physicians should only claim credit commensurate with the extent of their
participation in the activity. Credit must be claimed by March 31, 2016.
2014 Subscription Rates
Personal Subscriptions:
FPIN Member (includes 48 AMA PRA Category 1 Credits™)
Non-member (includes 48 AMA PRA Category 1 Credits™)
International (outside of the US or Canada)
Institutional Subscriptions:
$59
$119
$179
$209
$259
My parents were born during the Great Depression and
internalized the values of thrift and frugality required during
those desolate times. As young parents, they passed those
values along to me. Now I am always looking for deals and
silently bemoaning the price of milk, gasoline, and other staples
the prices of which fluctuate daily.
But price shock at the gas pump and grocery store are
nothing like price shocks in the medical world. In medicine, for
example, we have drugs that cost crazy amounts—like $1,000
a pill.
That pill is sofosbuvir, which, like simeprevir, is used to cure
hepatitis C. A 12-week course of sofosbuvir costs $84,000.
Simeprevir is cheaper, but still costs more than $66,000 for
12 weeks. The numbers are staggering, especially when you
consider that 12 weeks of sofosbuvir only costs between $70
and $140 to manufacture. A course of simeprevir costs between
$130 and $270 to manufacture, making simeprevir’s price tag
a slightly less egregious form of grand larceny.1
Let’s do some math. There are about 3.2 million people
in the United States with hepatitis C. If all were treated
with sofosbuvir, the makers would gross $269 billion (with
a “b”). Subtracting manufacturing costs and $2.4 billion in
development still leaves $266 billion in profit. Also consider
that there are about 140 million people in the world with
chronic hepatitis C. If someone managed to sell sofosbuvir
to all of them at the US price, the makers would reap over
$11 trillion. (Note: I had to do this math long hand because my
calculator does not have that many zeros.)
How much money is this? Well, according to the US Treasury,
the national debt expanded by $28 billion during the depression
and Roosevelt’s New Deal (1933–1941). Currently, 30 brandnew US Virginia-class nuclear attack submarines ($2.7 billion
each) will set you back $81 billion. This means the makers of
sofosbuvir could theoretically—after curing hepatitis C in the
United States—both finance a national stimulus package and
become a major nuclear power…and still have $157 billion
dollars left over!
At $1,000 a pill, this new deal is a bad deal indeed.
US and Canadian Institutions (without CME)
International Institutions (without CME)
EBP Electronic Archives
Jon O. Neher, MD
$500
Third Class postage paid at Columbia, MO 65202. The GST number for Canadian
subscribers is 124002536. Postmaster: Send address changes to FPIN, Inc.,
409 W. Vandiver Drive, Bldg 4, Suite 202, Columbia, MO 65202;
Attn: Kerri Reynolds. Kerri@fpin.org. 573-256-2066.
REFERENCE
1.Steinbrook R, Redberg RF. The high price of new hepatitis C virus drugs. JAMA Intern Med.
2014; 174(7):1172.
Statements and opinions expressed in abstracts and communications herein are those
ofthe author(s) and not necessarily those of the Publisher. The Publisher and editors of
EBP do not endorse any methods, product, or ideas mentioned in the newsletter, and
disclaim any liability, which may arise from any material herein.
Copyright © 2015 by Family Physicians Inquiries Network, Inc.
Evidence-Based Practice / Vol. 18, No. 5
3
Diving for PURLs
Not so sweet: Honey for acute cough in children
Oduwole O, Meremikwu MM, Oyo-Ita A, et al. Honey for acute cough in children.
Cochrane Database Syst Rev. 2014; (12):CD007094.
This Cochrane review compared the efficacy of honey
with diphenhydramine or dextromethorphan, as well as
no treatment or placebo in children with acute cough
from viral or bacterial sources. It included 3 trials with
568 children, ages 1–18 years, and evaluated cough
and related symptoms using a 7-point Likert scale with
caregiver responses ranging from “extremely” (6 points,
worst) to “not at all” (0 points, best). Secondary outcomes
included improvement in quality of sleep in children and
caregivers (also evaluated on a 7-point Likert scale, with
higher scores indicating better sleep), quality of life, and
adverse effects.
Pre- and postintervention Likert scale comparison
found honey was better at reducing cough frequency
than no treatment (mean difference [MD] –1.1; 95%
CI, –1.5 to –0.6) and placebo (MD –1.9; 95% CI,
–3.4 to –0.3). Honey was also likely better at reducing
cough frequency compared with diphenhydramine (MD
–0.57; 95% CI, –0.9 to –0.24), but was no better than
dextromethorphan (MD –0.14; 95% CI, –0.33 to 0.06).
Honey was inferior to dextromethorphan in
reducing cough severity (MD 0.61; 95% CI, 0.27–0.94),
but superior to diphenhydramine (MD –0.6; 95% CI,
–0.94 to –0.26), no treatment (MD –0.97; 95% CI,
–1.47 to –0.46), and placebo (MD –1.83; 95% CI, –3.3
to –0.34).
Similarly, there was no difference between honey
and dextromethorphan with regard to the children’s
sleep (MD 0.03; 95% CI, –1.1 to 1.2) or parents’ sleep
(MD –0.16; 95% CI, –0.84 to 0.53), but honey improved
sleep for both the children and parents compared with
diphenhydramine, no treatment, or placebo.
There was no difference in adverse events among
any of the treatment groups. Cough duration, cost,
changes in quality of life, and appetite were not studied.
RelevantYes
Medical care setting
Yes
ValidYes
Implementable
Yes
Change in practiceNo
Clinically meaningful
Yes
Additional information regarding the PURLs and Diving
for PURLs series can be found at: http://www.fpin.org/purls-faqs/
4
Evidence-Based Practice / May 2015
Bottom line: Honey may be better than no treatment,
placebo, or diphenhydramine for acute cough in
children aged 1–18 years, but is no better than
dextromethorphan.
Review and Summary Authors: Jennie Broders Jarrett, PharmD, BCPS,
and Jason Corbo, PharmD, BCPS,
UPMC St. Margaret FMRP, Pittsburgh, PA
Brief intervention reduces medication
overuse headaches
Kristoffersen ES, Straand J, Vetvik KG, et al. Brief intervention for medicationoveruse headache in primary care. The BIMOH study: a double-blind pragmatic
cluster randomised parallel controlled trial. J Neurol Neurosurg Psychiatry. 2014
Aug 11. [Epub ahead of print].
This cluster-randomized controlled study compared a
brief education-based intervention for 30 patients with
usual care for 45 patients diagnosed with medication
overuse headache. All patients were adults, 18 to 50
years old. Per design, the intervention was allocated by
practice not individual patient.
The intervention group received a presentation
on the topic, feedback on medication overuse headache,
and a plan for medication reduction. The usual-care
group received standard treatment that could include
a reduction plan, prophylactic medications, and other
abortive medications. The primary outcomes of the
study were number of headaches and medication days
per month.
Patients in the brief intervention group had 7.3
fewer headache days per month (95% CI, –11.3 to –3.2)
and 7.9 fewer days of medication use (95% CI, –12.5 to
–3.2), with chronic headache resolution in 50% of the
intervention group compared with 3% in the usual-care
group (P<.001; NNT=3).
RelevantYes
Medical care setting
Yes
ValidYes
Implementable
No
Change in practiceYes
Clinically meaningful
Yes
Bottom line: Although this brief intervention for
medication overuse headache demonstrates reduction
in headache and medication overuse days, it is not
implementable without more details about and access
to the training provided to the physicians in the study.
Review and Summary Author: Maryellen Schroeder, MD,
UPMC St. Margaret FMRP, Pittsburgh, PA
Diving for PURLs
Shoe advice for chronic foot pain
Underrecognized atrial fibrillation
in cryptogenic stroke
Gladstone DJ, Spring M, Dorian P, et al. Atrial fibrillation in patients with cryptogenic
stroke. N Engl J Med. 2014; 370(26):2467–2477.
This RCT compared 2 diagnostic methods for finding
new-onset atrial fibrillation among patients with recent
(within 6 months) ischemic stroke or transient ischemic
attack.
A 30-day event-triggered ECG (intervention group,
280 patients) was compared with a “usual” workup,
including a 24-hour ECG (control group, 277 patients).
The primary outcome was detection of atrial fibrillation
or flutter lasting ≥30 seconds within 90 days; secondary
outcome was atrial fibrillation lasting ≥2.5 minutes.
Atrial fibrillation was detected in 45 of 280 patients
(16.1%) in the intervention group compared with 9 of
277 (3.2%) in the control group, resulting in an absolute
difference of 12.9% (95% CI, 8.0–17.6; number needed
to screen [NNS]=8 to detect 1 additional case of atrial
fibrillation or flutter ≥30 seconds). Twenty-eight patients
in the intervention group (9.9%) had episodes of atrial
fibrillation lasting ≥2.5 minutes compared with 7 patients
in the control group (2.5%), with an absolute difference of
7.4% (95% CI, 3.4–11.3; NNS=14 to detect 1 additional
case of atrial fibrillation or flutter ≥2.5 minutes).
van der Zwaard BC, van der Horst HE, Knol DL, Vanwanseele B, Elders PJ. Treatment
of forefoot problems in older people: a randomized clinical trial comparing podiatric
treatment with standardized shoe advice. Ann Fam Med. 2014; 12(5):432–440.
This randomized trial compared shoe advice with
podiatry care in 205 adults aged ≥50 years with
nontraumatic musculoskeletal forefoot pain of
≥6 months’ duration. Patients with diabetes and
rheumatoid arthritis were excluded.
The shoe advice group received verbal guidance
and a standardized handout about appropriate shoe
fit at a primary care office. The podiatric care group
received usual podiatric care including, but not limited
to, recommended or custom footwear or orthotics at a
podiatry office.
Primary outcomes were foot pain and foot
function; secondary outcomes were general health, social
participation, and adherence. Data collection relied on
patient questionnaires every 3 months for 1 year.
Intention-to-treat analysis revealed no differences
in foot-related pain and dysfunction scores between the
2 study groups at all follow-up intervals. Foot-related
pain and dysfunction scores decreased significantly
from baseline in both study groups at 3 months and 12
months. General health and social participation did not
change.
RelevantYes
Medical care setting
Yes
ValidYes
Implementable
Yes
RelevantYes
Medical care setting
Yes
Change in practiceYes
Clinically meaningful
Yes
ValidNo
Implementable
No
Change in practiceYes
Clinically meaningful
No
Bottom line: Shoe advice with a standardized handout
from a primary care office may be as effective as podiatric
care for improvement in nontraumatic forefoot pain
and disability. However, adequate advice may require a
EBP
highly knowledgeable provider.
Review and Summary Author: Elizabeth Mohan, MD,
UPMC St. Margaret FMRP, Pittsburgh, PA
Review and Summary Author: Pooja Saigal, MD, NorthShore University
Health System/University of Chicago, Chicago, IL
Diving for PURLs Team
Bottom line: In patients who have had a cryptogenic
stroke, extended 30-day monitoring is superior to
24-hour ECG monitoring for detecting previously
undiagnosed atrial fibrillation. However, the precise
relationship (cause and effect) between atrial fibrillation
and cryptogenic stroke is unknown.
The University of Chicago, Department of Family Medicine
Goutham Rao, MD, Diving for PURLs Editor-in-Chief
Mari Egan, MHPE, MD
Katherine Kirley, MD
Debra Stulberg, MD
Kohar Jones, MD Sonia Oyola, MD
Kortnee Roberson, MD
Dionna Brown, MD Jennifer Bello, MD
Liz Nguyen, MD
Rush–Copley Medical Center, Department of Family Medicine
Kate Rowland, MD, PURLs Editor-in-Chief & Diving for PURLs Deputy Editor
Northshore University Health System, Department of Family Medicine
Janice Benson, MD Pooja Saigal, MD
University of Missouri, Department of Family Medicine
Jim Stevermer, MD, MSPH, PURLs Deputy Editor
Erik Lindbloom, MD, MSPH
University of North Carolina, Department of Family Medicine
Anne Mounsey, MD, PURLs Deputy Editor
University of Minnesota, Department of Family Medicine
Shailendra Prasad, MD, MPH
University of Pittsburgh Medical Center, St. Margaret Family Medicine
Jennie Broders Jarrett, PharmD, BCPS
Evidence-Based Practice / Vol. 18, No. 5
5
EBPediatrics
Do office-based literacy interventions promote
language development in children?
Bottom line
Reach Out and Read programs are associated
with an improvement in receptive language
development in preschool children (SOR: B,
heterogeneous RCTs), although they probably have
less effect on expressive language development
(SOR: C, heterogeneous RCTs).
Evidence summary
Reach Out and Read (ROR) is a literacy intervention
program adopted by many primary care offices in
which reading advice is given to parents and an ageappropriate book is given to patients during well-child
checks.1–5
In 2003, a case series reported language outcomes
of 64 children at their 3-year well-child check in an
urban pediatric clinic that had implemented ROR.1 A
Peabody Picture Vocabulary Test III and The Expressive
One Word Vocabulary Test measured receptive and
expressive language, respectively.
A multivariate analysis revealed higher receptive
language scores when increased anticipatory guidance
and more books purchased by parents were combined
(r2=0.025, P=.0006). The same analysis also revealed
higher expressive language scores when increased
number of visits and increased number of books
purchased were combined (r2=0.18; P<.001).1
A cross-sectional survey in 2002 compared
2 clinics in South Bronx, New York.2 Expressive and
Receptive One Word Picture Vocabulary Tests were
compared for 200 children aged 2–6 years in clinic A
with an established ROR program and clinic B with no
exposure to ROR. Children exposed to ROR scored
higher on receptive language testing (81.5 vs 74.3;
P=.005); however, expressive language testing was not
significantly different (79.5 vs 77.5; P=.26).
A similarly constructed 2001 cross-sectional survey
of 122 children aged 2–6 years compared Expressive
and Receptive One Word Picture Vocabulary Tests in
2 inner-city clinics.3 Again, 1 clinic had an established
ROR program and the other did not. Receptive
vocabulary was improved in the intervention group
(94.5 vs 84.8; P<.001), but expressive vocabulary was
not (84.3 vs 81.6; P=.23).
6
Evidence-Based Practice / May 2015
In 2000, an RCT with 205 infants aged 5–11
months placed patients in an ROR program or general
care during their well-child visits.4 After 3 well-child
checks or after the child reached 22 months, a Modified
MacArthur Communication Development Inventory
was performed for expressive and receptive language.
Higher receptive language (51.0 vs 39.3; P=.004)
and expressive (22.1 vs 15.9; P=.01) language was
noted in children tested at 18–25 months. However, in
children 13–25 months, no statistical differences were
EBP
observed.4
Christa Pittner-Smith, MD
A. Ildiko Martonffy, MD
University of Wisconsin–Madison
Madison, WI
REFERENCES
1. Theroit JA, Franco SM, Sisson BA, Metcalf SC, Kennedy MA, Bada HS. The impact
of early literacy guidance on language skills of 3-year-olds. Clin Pediatr (Phila). 2003;
42(2):165–172. [STEP 4]
2. Sharif I, Reiber S, Ozuah PO. Exposure to reach out and read and vocabulary outcomes in
inner city preschoolers. J Natl Med Assoc. 2002; 94(3):171–177. [STEP 3]
3. Mendelsohn AL, Mogilner LN, Dreyer BP, et al. The impact of a clinic-based literacy
intervention on language development in inner-city preschool children. Pediatrics. 2001;
107(1):130–134. [STEP 3]
4. High PC, LaGasee L, Becker S, Ahlgren I, Gardner A. Literacy promotion in primary care
pediatrics: can we make a difference? Pediatrics. 2000; 105(4 pt 2):927–934. [STEP 2]
5. Yeager Pelatti C, Pentimonti JM, Justice LM. Methodological review of the quality of reach
out and read: does it “work”? Clin Pediatr (Phila). 2014; 53(4):343–350. [STEP 2]
RECENT PUBLICATIONS IN CLINICAL INQUIRIES
• Khodaee M, Lombardo D, Montgomery L, et al. What’s the best test
for underlying osteomyelitis in patients with diabetic foot ulcers? J
Fam Pract. 2015; 64(5):309–310,321.
• Parker K, Dohr K, Neher J, et al. Do statins increase the risk of
developing diabetes? J Fam Pract. 2015; 64(4):245–246.
• Moreno L, Bonnell A, Neher J, et al. What therapies alleviate
symptoms of polycystic ovary syndrome? J Fam Pract. 2015;
64(4):247–249.
• Stephenson A, Kelsberg G, Neher J, St. Anna L. Treatments for
sciatica. Am Fam Physician. 2015; 91(9):612–615.
• Parad A, Leonard E, Handler L. Exercise and pregnancy loss. Am
Fam Physician. 2015; 91(7):437–438.
For more information regarding the Clinical Inquiries writing series,
please contact the Family Physicians Inquiries Network
at 573-256-2066 or email ci@fpin.org.
Topics in Maternity Care
How do the safety, efficacy, and patient satisfaction of uterine aspiration
for miscarriage management compare between the office and the
operating room?
Bottom line
Uterine aspiration in an office setting has
low rates of complications and high rates of
compared with the operating room (OR).
preferences and appropriate pain expectations
for patient satisfaction.
similar
efficacy
Patient
are key
Evidence summary
Safety
Uterine aspiration in the office has low rates of
complications. An RCT directly compared uterine
evacuation in the OR under general anesthesia with a
treatment room setting in the hospital using intravenous
sedation.1 In this study, 73 women received care in the
treatment room and 68 women were treated in the OR.
All women had a diagnosis of incomplete abortion
at <14 weeks, defined by a dilated cervical canal at
presentation and estimated uterine size on exam.
Severely anemic (Hgb <8) women and those with signs
of infection were excluded. Uterine evacuation was
completed with sharp curettage in both groups.
No uterine perforation or major complication
leading to hysterectomy occurred in either group.
Blood transfusion was significantly more common
for those treated in the OR under general anesthesia
(35% vs 18%; P<.03), including 2 cases of blood loss
>500 mL.1
A 2006, prospective, observational study enrolled
women who had already chosen surgical management
for first-trimester pregnancy loss.2 Clinicians gave
women a choice between having the procedure in the
office (n=115) using manual vacuum aspiration (MVA)
or in the OR (n=50).
Hemorrhage-related
complications
occurred
more commonly in the OR than in the office setting
(mean blood loss 311 vs 70 mL; P<.001). No major
complications were noted in either group.2
A 2007 nonrandomized study compared
157 women undergoing uterine aspiration in the
operative setting with electric vacuum aspiration under
anesthesia (n=68) or outpatient setting using MVA with
local anesthetic (n=89) for first-trimester pregnancy
loss.3 No significant differences were noted in objective
blood loss or fever between the 2 groups.
Efficacy
Efficacy of MVA in the office is high. A retrospective
study of 1,677 medical charts of women who had an
office MVA found that 99.5% had complete evacuation.4
Only 8 patients (0.5%) required a second procedure
because of retained products.
A similar study from 2009 in the United Kingdom
reviewed the charts of 245 women with early pregnancy
loss who were treated with MVA in a hospital unit
(not in the OR) and found a 95% success rate.5 Of
the 12 patients with retained products of conception,
8 underwent standard curettage and 4 were managed
expectantly with good result.
Satisfaction
Satisfaction is similar for MVA in the office and in
the OR. In the 2006 prospective study of 165 women,
89% of those who underwent aspiration in the office
would choose the same procedure again, while 93%
of those who underwent aspiration in the OR would
choose the procedure again (P=.11).2 Interestingly, the
study found that a disparity between expected and
experienced pain was negatively associated with patient
satisfaction, highlighting the importance of conferring
EBP
realistic expectations in patient education.
K. Hope Wilkinson, MS
Jessica Dalby, MD
University of Wisconsin
Madison, WI
REFERENCES
1.De Jonge ET, Pattinson RC, Makin JD, Venter CP. Is ward evacuation for uncomplicated
incomplete abortion under systemic analgesia safe and effective? A randomised clinical
trial. S Afr Med J. 1994; 84(8 pt 1):481–483. [STEP 2]
2.Dalton VK, Harris L, Weisman CS, Guire K, Castleman L, Lebovic D. Patient preferences,
satisfaction, and resource use in office evacuation of early pregnancy failure. Obstet
Gynecol. 2006; 108(1):103–110. [STEP 3]
3.Edwards S, Tureck R, Fredrick M, Huang X, Zhang J, Barnhart K. Patient acceptability
of manual versus electric vacuum aspiration for early pregnancy loss. J Womens Health
(Larchmt). 2007; 16(10):1429–1436. [STEP 2]
4.Westfall JM, Sophocles A, Burggraf H, Ellis S. Manual vacuum aspiration for first-trimester
abortion. Arch Fam Med. 1998; 7(6):559–562. [STEP 3]
5.Milingos DS, Mathur M, Smith NC, Ashok PW. Manual vacuum aspiration: a safe alternative
for the surgical management of early pregnancy loss. BJOG. 2009; 116(9):1268–1271.
[STEP 3]
Evidence-Based Practice / Vol. 18, No. 5
7
What is the best nonsurgical therapy
for carpal tunnel syndrome (CTS)?
Evidence-Based Answer
Local corticosteroids injections, splinting, and
therapeutic ultrasound all appear to be effective in the
short-term treatment of CTS compared with placebo
or no treatment. Ergonomic keyboards do not appear
to be effective. Among effective treatments, none
were clearly superior and no head-to-head trials were
evaluated (SOR: B, systematic reviews).
A 2009 Cochrane review of 12 randomized and
quasirandomized trials (N=671) assessed the
effectiveness of local corticosteroid injection compared
with placebo or other nonsurgical interventions
for treatment of CTS.1 Compared with injection of
placebo (saline or lignocaine), corticosteroid injections
improved subjective reporting of clinical severity on a
5-point ordinal scale (2 trials, N=141; risk ratio [RR]
2.6; 95% CI, 1.7–3.9).
Two trials compared local corticosteroid injection
with systemic corticosteroid. The first (N=60) showed
no improvement in symptoms (on a 50-point scale) at
2 weeks with local injection versus oral corticosteroid
(mean difference [MD] –4.2; 95% CI, –8.7 to 0.26), but
did show a difference at both 8 weeks (MD –7.2; 95%
CI, –11 to –2.9) and 12 weeks (MD –7.1; CI, –12 to
–2.5). The second trial (N=37) found improvement in
symptoms (by subjective report of clinical severity at
1 month, no further description) with local
corticosteroid injection compared with single systemic
corticosteroid injection (RR 3.2; 95% CI, 1.0–9.9).
Other single trials found no improvement with local
corticosteroid injection compared with other nonsurgical
treatments including splinting, high-dose corticosteroid
injection, or multiple corticosteroid injections.1
A 2012 Cochrane review of 19 RCTs with
1,190 patients compared wrist splinting for CTS.2 Only
2 trials directly compared splinting with no treatment.
Wrist splinting demonstrated significant improvement
using the Levine Questionnaire (1–5 scale) compared
with no treatment at 4 weeks (1 trial, N=80; MD –1.1;
95% CI, –1.3 to –0.85), 3 months (1 trial, N=48; MD
–0.90; 95% CI, –1.1 to –0.69), and 6 months (1 trial,
N=34; MD –0.9; 95% CI, –1.1 to –0.69).
A 2012 systematic review combined 2 RCTs
comparing therapeutic ultrasound versus placebo in
8
Evidence-Based Practice / May 2015
68 patients with CTS.3 A significant improvement,
defined as more patient ratings of overall symptoms
as “good or excellent,” was seen in ≤3 months in the
ultrasound group (RR 2.4; 95% CI, 1.4–4.0).
Another Cochrane review examined 2 RCTs of
CTS symptom relief with ergonomic versus standard
keyboards.4 In the first trial (N=25), CTS symptoms
were not statistically different at 6 weeks on a 10-point
symptom scale (MD –0.20; 95% CI, –1.5 to 1.1).
At 12 weeks, patients using ergonomic boards had
significantly lower symptom scores (MD –2.4; 95%
CI, –4.5 to –0.35). The second trial (N=80) evaluated
patients at 6 months and found no significant difference
on a 10-point symptom scale (MD 0.70; 95% CI, –0.97
to 2.4).
David Wyncott, MD
Theodore Neumann, MD
Broderick Howard, DO
Garrett Bentley, DO
St. Joseph Regional Medical Center FMR
Mishawaka, IN
1. Marshall S, et al. Cochrane Database Syst Rev. 2007; (2):CD001554. [STEP 1]
2. Page MJ, et al. Cochrane Database Syst Rev. 2012; (7):CD010003. [STEP 1]
3. Page MJ, et al. Cochrane Database of Syst Rev. 2013; (3):CD009601. [STEP 1]
4. O’Conner D, et al. Cochrane Database Syst Rev. 2012; (1):CD009600. [STEP 1]
In a patient with panic disorder, is a
combination of medication and psychotherapy
superior to either alone for long-term treatment
and prevention of symptoms?
Evidence-Based Answer
In patients with panic disorder, antidepressant
medication combined with psychotherapy is slightly
superior to either alone during active treatment. After
therapy is discontinued, psychotherapy and combination
treatment are equally effective at preventing future
symptoms, and superior to antidepressants alone (SOR:
A, systematic review of RCTs). Benzodiazepines do not
appear to add benefit to psychotherapy alone (SOR: C,
meta-analysis of heterogeneous RCTs).
A systematic review of 21 RCTs with 1,709 patients with
panic disorder compared treatment using antidepressant
medication or psychotherapy with the combination of
both strategies.1 Selective serotonin reuptake inhibitors
(SSRIs) or tricyclic antidepressants (TCAs) were used
in 22 of 23 comparisons and the monoamine oxidase
inhibitor phenelzine was used in 1 comparison.
Psychotherapy included behavioral therapy or cognitivebehavioral therapy (CBT) in 22 comparisons and
psychodynamic therapy in 1 comparison. The primary
outcome was significant improvement (“very much or
much improved”) from baseline on the Clinical Global
Impression Scale (CGIS), 40% or greater improvement
in the Panic Disorder Severity Scale (PDSS), or a 50%
or greater reduction in panic frequency on the Fear
Questionnaire–Agoraphobia subscale.
During the treatment phase (up to 36 weeks),
combination therapy had a better response rate than
medication alone (11 trials, N=669; RR 1.6; 95% CI,
1.2–2.2) and psychotherapy alone (19 trials, N=1,257;
RR 1.2; 95% CI, 1.0–1.5). After discontinuation of
therapy (measured 6–24 months later), patients who
underwent combination therapy had fewer symptoms
than patients in the antidepressant-alone group (5 trials,
N=376; RR 1.6; 95% CI, 1.2–2.1), but no difference in
response compared with patients in the psychotherapyalone group (9 trials, N=658; RR 0.96; 95% CI, 0.79–
1.2). The review was limited by the lack of validated
rating scales for panic disorder at the time of the RCTs,
and the lack of control over additional therapies after
discontinuation.1
Another systematic review of 3 RCTs involving
243 patients with panic disorder compared the
efficacy of benzodiazepines, psychotherapy, or both.2
Alprazolam was used in 2 studies and diazepam in the
third. Psychotherapy consisted of behavioral therapy
in 2 trials and CBT in 1 trial. Response was defined
as “much improved” or “very much improved” on the
CGIS or a score of 7 or below on the PDSS.
Two trials involving 166 patients showed no
difference between combination therapy compared
with psychotherapy alone at 8 weeks (RR 0.78; 95%
CI, 0.45–1.4) and after treatment had ended up to
7 months later (RR 0.62; 95% CI, 0.36–1.1). In 1 RCT,
combination therapy was superior to benzodiazepine at
the end of 8 weeks of treatment (N=77; RR 3.4; 95%
CI, 1–11). However, this finding barely met the level of
statistical significance (P=.05). After discontinuation of
treatment (7–12 months of naturalistic follow-up), no
significant difference was found (RR 2.3; 95% CI, 0.79–
6.7). The review authors noted that the small number of
studies and patients involved was a significant obstacle
to determining the superiority of one treatment over
another.2
The American Psychiatric Association recommends
SSRIs, serotonin-norepinephrine reuptake inhibitors
(SNRIs), TCAs, benzodiazepines, and CBT for panic
disorder with no preference given to any specific therapy
or combination.3
Jay Brieler, MD
Christine Jacobs, MD
St. Louis University Family and Community Medicine
St. Louis, MO
1. Furukawa TA, et al. Br J Psychiatry. 2006; 188:305–312. [STEP 1]
2. Watanabe N, et al. Cochrane Database Syst Rev. 2009; (1):CD005335. [STEP 1]
3.American Psychiatric Association. Practice Guideline for the Treatment of Patients with
Panic Disorder. 2nd ed. Washington, DC: APA; 2009. http://psychiatryonline.org/pb/
assets/raw/sitewide/practice_guidelines/guidelines/panicdisorder.pdf . Accessed March
20, 2015. [STEP 5]
How well does serum blood urea nitrogen-tocreatinine ratio correlate with prerenal etiology
of acute kidney injury (acute renal failure)?
Evidence-Based Answer
The blood urea nitrogen-to-creatinine ratio (BCR) does
not reliably correlate to a prerenal etiology of acute
kidney injury (AKI) (SOR: B, cohort studies).
A 2012, retrospective cohort trial compared 20,126
patients older than 14 years admitted to an Australian
medical center from 2000 through 2002 to assess if
BCR is a useful predictor of prerenal azotemia (PRA),
traditionally believed to be associated with a BCR >20,
or acute tubular necrosis (ATN), traditionally believed
to be associated with a BCR ≤20.1 AKI was defined
according to the Risk, Injury, Failure, Loss and EndStage (RIFLE) kidney disease classification system using
glomerular filtration rate criteria, which is a validated
tool.
A total of 3,641 patients met criteria for AKI by
RIFLE criteria who were further divided into those with
high BCR (>20) or low BCR (≤20). The authors did not
find a bimodal distribution of BCR, suggesting that
there was no useful diagnostic threshold for predicting
PRA versus ATN.1
In a 2002 prospective cohort trial of
102 hospitalized patients with AKI at Nassau County
Medical Center, NY, the BCR was compared with
various indices including fractional excretion of urea
nitrogen (FEUN) and fractional excretion of sodium
(FENa) for discerning PRA versus ATN.2 Three groups
Evidence-Based Practice / Vol. 18, No. 5
9
TABLE
Utility of laboratory findings for determining prerenal
azotemia as the etiology for acute kidney injury2
Laboratory finding
Sensitivity
Specificity
+LR
–LR
BCR >15
87%
80%
4.4
0.16
FEUN ≤35%
90%
96%
22
0.11
FENa <1%
77%
96%
19
0.24
were established: prerenal failure (50 patients; average
age 49 years), prerenal failure on diuretics (27 patients;
average age 51 years), and ATN (25 patients; average
age 47 years). The diagnosis was established by an
attending nephrologist based on expert guidelines
(rapidly increasing BUN [>30 mg/dL] and Cr
[>1.5 mg/dL], serum Cr increase >0.5 mg/dL in the
preceding 2 days) and thorough analysis of patient
history, physical exam findings, rate, and extent of
azotemia, urinalysis, and urinary and serum indices,
including sodium, urea, and creatinine. FEUN was
found to have 85% sensitivity and 92% specificity. The
findings support that a FEUN ≤35% and a FENa <1% are
better markers for PRA than a BCR >15 (TABLE ).
A retrospective outcomes trial of 191 patients (57%
women, mean age 42 years) hospitalized in Pakistan
for cholera examined the usefulness of the admission
BCR as a predictor of AKI.3 Patients had BCR levels
measured at admission and discharge (average stay
3.8 days). The mean BCR remained consistent at
12 (95% CI, 11–12) at admission and 12 (95% CI,
10.6–12.8) at discharge despite the fact that the
average amount of fluid resuscitation was 25 liters.
Additionally, of the 31% (60) patients who developed
AKI, the average BCR upon initial presentation was
lower than that of those who did not develop AKI (no
data provided; P=.04).
MD
MD
MD
MD
Naval Hospital Pensacola FMR
Pensacola, FL
“The views expressed in this article are those of the authors and do
not necessarily reflect the official policy or position of the US Department
of Navy, Department of Defense, or the US Government.
1. Uchino S, et al. Clin Kidney J. 2012; 5(2):187–191. [STEP 3]
2. Carvounis CP, et al. Kidney Int. 2002; 62(6):2223–2229. [STEP 3]
3. Tariq M, et al. PloS One. 2009; 4(10):e7552. [STEP 3]
10
Evidence-Based Practice / May 2015
Evidence-Based Answer
Dry needling or injections with botulinum toxin,
physiologic saline, lidocaine, or lidocaine with a steroid
all appear about equally effective in trigger point
treatment (SOR: B, RCTs). Clearly, dry needling would
be the least expensive.
BCR=blood urea nitrogen-to-creatinine ratio = [BUN (mg/dL)/Cr (mg/dL)].
FEUN=fractional excretion of urea nitrogen = [(UUN/BUN)/(UCr/PCr)] × 100.
FENa=fractional excretion of sodium = [(UNa/PNa)/(UCr/ PCr)] × 100.
David A. Moore,
Andrew J. McDermott,
Daniel E. Bradley,
Kathryn L. Bond,
What injection is most effective
for trigger points?
In a 2007 RCT, a total of 90 patients with myofascial
pain (66 women, 24 men; age range 25–40 years) were
randomly assigned to 1 of 5 study groups: Group 1
received botulinum toxin-A 10 U, group 2 received
lidocaine 5% 1 mL, and groups 3, 4, and 5 were treated
by other conservative measures without a trigger point
injection.1 A 10-cm visual analog scale (VAS) measured
pain intensity.
At week 1 posttreatment, all groups showed
statistically significant improvement compared with
baseline, with group 1 showing a VAS percent change of
–39% (P<.05), group 2, –40% (P<.05), group 3, –31%
(P<.05), group 4, –39% (P<.05), and group 5, –29%
(P<.05). However, none of the treatment methods
proved to be superior when intergroup comparisons
were made. At 1 month, statistically significant
improvements were detected in both the botulinum
toxin-A and lidocaine groups from baseline (group 1,
–71%; P<.05; group 2, –74%; P<.05). These 2 groups
were found to be equally effective.1
In 2001, a systematic review of 23 RCTs evaluated
whether needling therapies have specific efficacy in the
management of myofascial pain (ie, efficacy beyond
placebo).2 The 23 trials were divided into 4 categories:
(1) direct wet needling, (2) direct dry needling, (3)
indirect wet needling, and (4) indirect dry needling.
Fourteen trials (N=563) investigated direct wet needling
with different substances including bupivacaine,
lidocaine plus steroid, botulinum toxin-A, and isotonic
saline. Eight of the 10 studies compared wet needling
using different substances and found the effects were
independent of the substance used. Five trials (N=532)
directly compared dry with wet needling and found no
difference between the 2 groups. Data were not pooled
because the studies were of differing quality and design;
many different parts of the body were represented, and
were performed in 6 different countries. The authors
concluded that that no difference exists between trigger
point injections with different substances, or between
dry and wet needling.
A 2010 RCT of 80 patients with myofascial pain
syndrome (52 women, 28 men; age range 19–58 years)
compared the efficacy of local anesthetic injection and
dry needling methods on pain.3 Patients were randomly
assigned to receive either local anesthetic injection with
2 mL 1% lidocaine (group 1, n=40) or dry needling
(group 2, n=40). Both groups were also given stretching
exercises.
VAS scores (0–10) showed a statistically significant
decrease from baseline in both group 1 (from 5.8 to
2.3; P<.001) and group 2 (from 5.6 to 3.8; P<.001) at
4 weeks, but no difference between groups (P=.053).
The authors concluded that exercise associated with
local anesthetic and dry needling injections were equally
effective for myofascial pain.3
Andrew Saleh, MD, MPH
Adity Bhattacharyya, MD, FAAFP
Kelly Ussery-Kronhaus, MD
Rutgers-Robert Wood Johnson Medical School FMR at Capital Health
Trenton, NJ
1. Esenyel M, et al. J Back Musculoskelet Rehabil. 2007; 20(1):43–47. [STEP 2]
2. Cummings TM, et al. Arch Phys Med Rehabil. 2001; 82(7):986–992. [STEP 1]
3. Ay S, et al. Clin Rheumatol. 2010; 29(1):19–23. [STEP 2]
What are the most effective nonpharmacologic
therapies for irritable bowel syndrome (IBS)?
Evidence-Based Answer
Both probiotics and soluble fiber improve global
symptoms scores for IBS that are of unclear clinical
significance (SOR: B, meta-analyses of heterogeneous
RCTs). Exercise and mindfulness training appear to
produce clinically important improvement in IBS
symptoms (SOR: B, single RCTs).
A 2008 meta-analysis of 20 RCTs involving
1,404 patients with IBS compared treatment effects of
probiotics versus placebo.1 Probiotic therapy led to a
higher proportion of global symptom score responders
than placebo therapy (risk ratio [RR] 0.77; 95% CI,
0.62–0.94), although the definition of a responder
versus nonresponder had little uniformity across
studies. A similar magnitude of improvement was seen
with the secondary outcome of improved abdominal
pain (RR 0.78; 95% CI, 0.69–0.88). Again, definitions
of improvement in abdominal pain were not consistent
across studies. Not enough data were available to assess
any difference in other secondary outcomes such as
bloating or flatulence.
A 2004 systematic review of 17 RCTs including
1,363 patients with IBS assessed the treatment effect
of added fiber intake on global IBS symptom scores.2
This review found a general increase in dietary fiber led
to improvement in IBS symptom scores (not defined)
(RR 1.3; 95% CI, 1.2–1.5) compared with placebo.
Soluble fiber (eg, psyllium, ispaghula, calcium
polycarbophil) also showed improvement in global
IBS symptoms scores (RR 1.6; 95% CI, 1.4–1.8) when
compared with placebo. Insoluble fiber (eg, wheat bran,
corn) did not lead to significant clinical improvement
(RR 0.89; 95% CI, 0.72–1.1).
A 2011 RCT of 102 patients with IBS compared
the effect of increased physical activity with no change
in activity on IBS symptoms.3 The intervention group
received telephone advice from a physical therapist
once or twice a month aiming for a moderate level of
activity.
After 12 weeks, the 500-point IBS severity score
(IBS-SS) improved significantly from baseline when
comparing the physical activity group and the control
group (–51 vs –5; P=.003). A change of 50 points was
noted to be clinically significant.3
A 2011 RCT of 75 female patients with IBS
compared the efficacy of mindfulness training, a
cognitive-behavioral technique, to support group
involvement in reducing IBS symptom severity.4 The
mindfulness group had greater reductions in IBS-SS
(26% vs 6.2%; P=.006) immediately and 3 months after
intervention (38% vs 12%; P=.001) compared with the
group receiving IBS support group participation only.
The overall decrease in the IBS-SS for the mindfulness
group was well above the clinically significant mark of
50 points (75 points immediately and 108 points after
3 months).
Lisa Sun Rhodes,
Tom Hoke,
Josh Leavitt,
Justin Glass,
MD
MD
MD
MD
FMR of Idaho
Boise, ID
1. McFarland LV, et al. World J Gastroenterol. 2008; 14(17):2650–2661. [STEP 1]
2. Bijerk CJ, et al. Ailment Pharmacol Ther. 2004; 19(3):245–251. [STEP 1]
3. Johannesson E, et al. Am J Gastroenterol. 2011; 106(5):915–922. [STEP 2]
4. Gaylord SA, et al. Am J Gastroenterol. 2011; 106(9):1678–1688. [STEP 2]
Evidence-Based Practice / Vol. 18, No. 5
11
Is fennel tea an effective treatment
for colic in infants?
Evidence-Based Answer
Maybe. Fennel seed oil reduces crying time in infants
with colic (SOR: B, small RCTs). The optimal amount or
formulation of fennel is unclear.
Fennel seed oil has been suggested as a treatment for
colic in infants. In all studies listed here, colic was
diagnosed by Wessel’s criteria: agitation and crying
lasting longer than 3 hours per day for at least 3 days
per week for longer than 3 weeks.
A 2003, double-blinded RCT evaluated the effect of
fennel seed oil on colic in 125 infants 2 to 12 weeks old
who were either breastfed or formula-fed.1 Infants were
excluded if they were preterm, struggling with weight
gain, on medications, or ill. Parents administered 5 to
20 mL 0.1% fennel seed oil in polysorbate solution or
matched placebo at their discretion up to 4 times a day
for a total of 7 days, with a daily maximum dose of
12 mg/kg. Infants were followed after the intervention
for 7 days. The treatment was considered effective, or
colic eliminated, if cumulative crying was reduced to
less than 9 hours per week.
Colic was eliminated in 40 of 62 (65%) infants in
the treatment group and 14 of 59 (24%) infants in the
control group, for an absolute risk reduction (ARR)
of 41% (95% CI, 25–57) and an NNT of 2 (95% CI,
2–4). No significant difference was noted in efficacy
of treatment between the breastfed and formula-fed
infants.1
A 2005, double-blinded RCT studied the effect
of 0.1% fennel seed oil with German chamomile and
lemon balm on colic in 93 term, breastfed infants
21 to 60 days old.2 Infants had literate mothers and
were without current infections, gastrointestinal
disorders, metabolicdisease, or failure to thrive. They
received 1 ml/kg of treatment or matched placebo
twice daily at 5 pm and 8 pm for 7 days and followed
for 14 days after the intervention. Effect was measured
by reduction of crying time by 50%. Thirty-five of
41 (85%) infants responded in the treatment group and
23 of 47 (49%) infants responded in the control group
(ARR 36%; NNT=3).
A nonblinded RCT in 2008 examined the effect of
massage, sucrose, fennel tea, hydrolyzed formula, or
breastfeeding alone on crying time in 175 infants 4 to
12
Evidence-Based Practice / May 2015
12 weeks old.3 Infants had normal development, were
breastfed, and were without gastrointestinal disorders
or previous treatment for colic. Thirty-five infants were
given 35 mL of an undefined concentration of fennel
tea 3 times a day for 7 days. The primary outcome
was mean reduction in crying time after the treatment
period.
The fennel tea group had a baseline average crying
time of 5.1 h/d, which decreased to 3.2 h/d after 7 days
(P<.001). No significant reduction was noted in crying
time in breastfed-only infants, from a baseline of
4.6 h/d to 4.5 h/d after 7 days (P>.05). Major limitations
of the study included an undefined concentration
of fennel in the tea and the lack of a between-group
comparison.3
Jennifer Cook, MD
Crystal Pyrak, MD
Annie Valente, MD
FMR of Idaho
Boise, ID
1. Alexandrovich I, et al. Altern Ther Health Med. 2003; 9(4):58–61. [STEP 2]
2. Savino F, et al. Phytother Res. 2005; 19(4):335–340. [STEP 2]
3. Arikan D, et al. J Clin Nurs. 2008; 17(13):1754–1761. [STEP 2]
What is the most effective treatment
for alopecia areata?
Evidence-Based Answer
Intralesional and topical corticosteroids are moderately
effective in the short term (12 weeks) for patchy hair loss
in alopecia areata. There do not appear to be effective
long-term treatment options; however, spontaneous
remission is common (SOR: B, RCTs and evidencebased guideline).
A 2011 RCT compared the efficacy of 0.1% topical
betamethasone valerate foam twice daily, intralesional
triamcinolone acetonide 10 mg/mL every 3 weeks, and
0.1% tacrolimus ointment twice daily for the treatment
of alopecia areata in 78 patients aged 11 to 50 years
over 12 weeks.1 Response to treatment was measured
using a quantitative hair regrowth grade.
Compared with baseline after 12 weeks, 60%
of patients treated with intralesional triamcinolone
reported >75% hair regrowth (P<.05) and 43% of
patients treated with topical betamethasone reported
>75% hair regrowth (P<.05). No participants in the
tacrolimus group demonstrated >75% hair regrowth
after 12 weeks. There was no comparison between
groups.1
A 2011 randomized placebo-controlled trial
examined the effectiveness of twice-daily pimecrolimus
1% cream and clobetasol propionate 0.05% compared
with placebo for the treatment of alopecia areata in
100 patients (age range 3–65 years) referred to a
university dermatology department.2 Patients were
examined at 4, 8, and 12 weeks.
At week 12, the average area of alopecia regrowth
was 54% in the pimecrolimus group, 47% in the
clobetasol group, and 36% in the placebo group. No
statistically significant difference was noted among the
groups.2
A 2008 Cochrane review of 17 RCTs involving
540 adult and pediatric patients examined the longterm benefit of multiple treatments for patients with
alopecia areata.3 The primary endpoint examined was
>50% hair regrowth. The 17 trials compared a wide
variety of topical and oral treatments with placebo
including topical and systemic steroids, topical and
systemic immunomodulators, antidepressants, and
ultraviolet light therapy.
The review authors concluded that there was no
significant long-term treatment benefit specific to hair
growth compared with placebo for any intervention.
Due to heterogeneity of the 17 chosen trials, data were
not pooled. The authors noted that due to the possibility
GLOSSARY
NNT=number needed to treat
CDC=Centers for Disease
Control and Prevention
NSAID=nonsteroidal antiinflammatory drug
CI=confidence interval
OR=odds ratio
CT=computed tomography
RCT=randomized controlled trial
FDA=US Food and Drug
Administration
RR=relative risk
LOE=level of evidence
MRI=magnetic resonance
imaging
NNH=number needed to harm
Davis Yang, MD
Alina Merrill, DO
Edward Foley, MD
Carrie Holland, MD
MacNeal FMRP
Berwyn, IL
1. Kuldeep C, et al. Int J Trichology. 2011; 3(1):20–24. [STEP 3]
2. Ucak H, et al. J Dermatolog Treat. 2012; 23(6):410–420. [STEP 2]
3. Delamere FM, et al. Cochrane Database Syst Rev. 2008; (2):CD004413. [STEP 2]
4. Messenger AG, et al. Br J Dermatol. 2012; 166(5):916–926. [STEP 2]
EVIDENCE YOU CAN TRUST
ARR=absolute risk reduction
HR=hazard ratio
of spontaneous remission of alopecia areata, patients
may consider not being treated or using a wig.3
The 2012 evidence-based guidelines for the
treatment of alopecia areata from the British Association
of Dermatologists summarized current research and
updated the 2003 guidelines from the same group.4
The updated guidelines noted that few treatments have
been subjected to randomized controlled trials and
spontaneous remission is common (up to 80%).
For limited patchy hair loss, they recommended
potent topical steroids (based on “good quality
case control and cohort studies”) and intralesional
corticosteroids (based on “short-term evidence”).
For extensive patchy hair loss or alopecia totalis/
universalis, they recommended contact immunotherapy
(based on “systematic review of case control or cohort
studies”) and a wig or hairpiece (based on “expert
opinion and consensus”). They recommended contact
immunotherapy, but noted the response rate is low and
EBP
the treatment is not easily available.4
SOR=strength of
recommendation
SSRI=selective serotonin
reuptake inhibitor
WHO=World Health
Organization
The Family Physicians Inquiries Network Consortium
(FPIN) is a nonprofit 501(C)3 organization that operates
for the exclusive purpose of providing research and
education in the public interest. FPIN is the sole publisher
of Evidence-Based Practice, produced as an educational
service to members and subscribers worldwide.
The members of FPIN are deeply committed to providing
accurate, unbiased, and evidence-based information that
will help physicians provide better care to their patients.
You have our word on it.
Evidence-Based Practice / Vol. 18, No. 5
13
Spotlight on Pharmacy
Should a serum magnesium level be obtained in all patients newly
diagnosed with migraine headaches?
Bottom line
Prophylaxis with 600 mg oral magnesium oxide a day
may decrease migraine frequency or severity (SOR: C,
small inconsistent RCTs). It is unclear if knowing the
serum magnesium level would change management.
Evidence summary
Magnesium supplementation and reduction
in migraine frequency
A double-blind RCT of 81 patients diagnosed with
migraine by International Headache Society (IHS)
criteria (with or without aura) compared migraine
frequency between patients who underwent daily
prophylaxis with 600 mg oral magnesium oxide or
placebo over a 16-week period.1
Both study arms showed a decrease in migraine
frequency relative to baseline, but patients in the
magnesium arm experienced a significantly larger
decrease (42% vs 16%; P=.03). Of note, no significant
correlation between pre-prophylaxis serum magnesium
levels and attack frequency was observed.1
A double-blind RCT of 30 patients diagnosed
with migraine without aura by IHS criteria received
600 mg oral magnesium oxide and were compared
with 10 controls who received placebo over a 12-week
period.2 Both groups experienced a significant decrease
in median migraine attack frequency (from 3.0 to
2.0 attacks/month, P<.001 in the treatment arm vs from
3.5 to 3.0 attacks/month, P<.05 in the placebo arm).
With respect to attack severity (as recorded on a 1–10
visual analog scale), only patients in the treatment arm
experienced a significant improvement (from 7.6 to
4.0; P<.001 in the treatment arm vs 7.0 to 7.0 [ie, no
change]; P>.05 in the control arm).
Ionized magnesium concentration related to menses
and migraine status
A prospective cohort trial compared both ionized and
total serum magnesium concentrations of 61 patients
diagnosed with menstrual migraine with those of
66 controls.3 Mean total serum magnesium
concentration was normal among all patients tested
regardless of menses or migraine status. However,
14
Evidence-Based Practice / May 2015
ionized magnesium was lower during menses
(1.14 mEq/L), during menstrual-related attacks
(1.12 mEq/L), and during nonmenstrual-related attacks
(1.16 mEq/L) relative to control (1.2 mEq/L). However,
the only value that achieved statistical significance was
during menstrual-related attacks (P<.01).
Serum magnesium concentration related to migraine
status and headache frequency
A case-control study compared serum magnesium levels
between 140 migraineurs and 140 matched controls
without migraines.4 The average serum magnesium
concentration was significantly lower among migraine
sufferers (2.2 vs 2.6 mEq/L; P<.000). Sex, age, and the
presence of aura were not independently associated
with a significant difference in serum magnesium
concentrations. However, a linear relationship between
mean magnesium level and headache frequency was
EBP
noted (r=0.21; P=.016).
Adam Sprouse-Blum, MD
Anil Gogineni, MD
Bronx-Lebanon Hospital
Bronx, NY
REFERENCES
1.Peikert A, Wilimzig C, Köhne-Volland R. Prophylaxis of migraine with oral magnesium:
results from a prospective, multi-center, placebo-controlled and double-blind randomized
study. Cephalalgia. 1996; 16(4):257–263. [STEP 2]
2.Köseoglu E, Talaslioglu A, Gönül AS, Kula M. The effects of magnesium prophylaxis in
migraine without aura. Magnes Res. 2008; 21(2):101–108. [STEP 2]
3.Mauskop A, Altura BT, Altura BM. Serum ionized magnesium levels and serum ionized
calcium/ionized magnesium ratios in women with menstrual migraine. Headache. 2002;
42(4):242–248. [STEP 2]
4.Talebi M, Savadi-Oskouei D, Farhoudi M, et al. Relation between serum magnesium level
and migraine attacks. Neurosciences (Riyadh). 2011; 16(4):320–323. [STEP 2]
Evidence-Based Practice learning objectives
1To become knowledgeable about evidence-based
solutions to commonly encountered clinical problems.
2To understand how ground-breaking research is
changing the practice of family medicine.
3To become conversant with balanced appraisals of
drugs that are marketed to physicians and consumers.
CONTINUING MEDICAL EDUCATION TEST
MAY 2015
EBP CME Tests are online at www.fpin.org/cme
Each month CME subscribers may earn up to 4 AMA PRA Category 1 credits™ per test!
For each question, please mark the single best answer by checking the appropriate box.
To receive CME credit, a minimum score of 75% (6 out of 8 correct) is required.
1.Which of the following statements is true regarding differentiation between
type 1 and type 2 diabetes mellitus in adolescents or young adults?
o a.Differentiation can be established based on islet antibodies alone
o b.Islet autoantibodies and acanthosis nigricans cannot coexist in the
same individual
o c. Differentiation depends in part on family history
o d.Patients with a high body mass index will likely have an elevated level
of islet antibodies
2.Which of the following interventions has the least solid evidence
of consistently helping alleviate symptoms of carpal tunnel syndrome?
o a.Corticosteroid injection
o b. Therapeutic ultrasound
o c. Ergonomic keyboard use
o d. Wrist splinting
3.A 55-year-old woman with fibromyalgia presents with significant trigger point
pain. She is interested in injections for 3 trigger points that are particularly
bothersome. In counseling about injections, you can reliably state:
o a.Botulinum toxin-A injections are more effective than lidocaine and
steroid injections
o b.Lidocaine and steroid injections are more effective than
saline injections
o c.Dry needling will likely work as well as injecting any
available fluid
o d. Bupivacaine injections are not effective
4.With regard to determining prerenal etiology in acute kidney injury,
which of the following statements is true?
o a.The BCR (blood urea nitrogen-to-creatinine ratio) >15 is the gold
standard
o b.FENa (fractional excretion of sodium) <1% has a lower positive
likelihood ratio (+LR) for the diagnosis of prerenal azotemia than BCR
o c.FEUN (fractional excretion of urea nitrogen) ≤35% has a higher +LR for
the diagnosis of prerenal azotemia than BCR
o d.FEUN <35% has a lower specificity for the diagnosis of prerenal
azotemia than BCR
5.All of the following interventions have been shown to be helpful in reducing
symptom severity in irritable bowel syndrome except
o a.Soluble fiber
o b.Probiotics
o c. Mindfulness training
o d. Insoluble fiber
6.A mother brings in her 6-week-old infant for persistent crying. The history
is consistent with colic and there are no problems found on physical exam.
She has heard that fennel is good for colic. You can tell her:
o a.Fennel cannot be safely used in infants because only the whole seed is
effective
o b. Fennel might be effective, but not much data are available on the best
compound or dose
o c. Fennel teas have been proven effective in multiple well-designed trials
o d. Fennel oil is not effective and has significant toxicity
7.Which of the following statements is true regarding treatment
for alopecia areata?
o a. Intralesional corticosteroids are effective in the short term
o b.All cases should be treated aggressively, as spontaneous resolution
is rare
o c. Pimecrolimus 1% is more effective than placebo
o d. Topical corticosteroids have shown good long-term results
8.Which of the following statements is true about the treatment
of panic disorder?
o a.Selective serotonin reuptake inhibitors (SSRIs) are superior to the
combination of cognitive-behavioral therapy (CBT) and SSRIs for the
long-term treatment of panic disorder
o b.After the treatment phase is complete, benzodiazepines are more
effective in preventing symptoms than psychotherapy
o c.Current American Psychiatric Association guidelines suggest using
serotonin-norepinephrine reuptake inhibitors before other medication
classes
o d.Combining antidepressants with CBT may provide a small benefit over
either alone, during the active treatment phase
Through joint sponsorship by FPIN and the University of Colorado School of Medicine,
CME subscribers in 2015 are eligible to earn 4 AMA PRA Category 1 credits™ per month.
Answer key: 1. c; 2. c; 3. c; 4. c; 5. d; 6. b; 7. a; 8. d
This test must be received by March 31, 2016 to be accepted for credit
A maximum of 4 AMA PRA Category 1 credits™ per month may be earned by CME subscribers.
To ensure proper credit for your CME test, please provide the following information:
Title (MD, DO, etc)
Name (Please print)
SSN (last 4 digits)
Address
CityState
Zip Code
Daytime Phone Number
Ext.
Email address (to notify you of credits earned)
For CME credit, return this test to: FPIN, 409 W. Vandiver Drive, Bldg. #4, Ste 202, Columbia, MO 65202 or fax to 573-256-2078.
If you have questions, please contact Kerri Reynolds (ebp@fpin.org or call 573-256-2066).
Renew or Subscribe to EBP at www.fpin.org/subscribe or call 573-256-2066
Evidence-Based Practice
Family Physicians Inquiries Network, Inc.
409 West Vandiver Drive
Building 4, Suite 202
Columbia, MO 65202
Change Service Requested
HelpDesk Answers are now featured in
The Journal of Family Practice!
The Family Physicians Inquiries Network is thrilled to announce
that select HelpDesk Answers articles will now be featured in
The Journal of Family Practice!
Be sure to check out this new addition in the May print issue of JFP
or view online at http://www.jfponline.com/articles/helpdesk-answers.html.
Evidence-Based Practice
E L E C T R O N I C V E R S I O N •
Does screening patients with hypertension
for microalbuminuria improve outcomes?
Evidence-Based Answer
The presence of microalbuminuria is associated with
an increased risk of cardiovascular and cerebrovascular
disease (SOR: B, prospective cohort trials). However,
it is not clear if screening for microalbuminuria in
hypertensive patients will improve outcomes.
A 2000 prospective cohort trial observed 204 untreated
hypertensive patients for 9.5 years to evaluate if the
presence of microalbuminuria adds to the increased risk
of ischemic heart disease in this patient population.1
Microalbuminuria was found to be the strongest
predictor of developing ischemic heart disease regardless
of subsequent hypertension treatment (risk ratio [RR]
4.2; 95% CI, 1.5–12). When adjusted for the effects
of other risk factors of ischemic heart disease, the risk
increased (RR 5.6; 95% CI, 1.9–17).
A 2003 prospective cohort trial of 144 elderly
hypertensive patients, without previous cardiovascular
complications, evaluated if baseline values of urinary
albumin excretion and other cardiovascular risk factors
are predictive of cardiovascular complications.2 Patients
with microalbuminuria at baseline had a higher incidence
of cardiovascular events (myocardial infarction, angina
pectoris, cerebral infarction or hemorrhage, aortic
dissection) during the 8-year follow-up than patients
who had normoalbuminuria at baseline (18% vs 7.2%;
P=.05).
A prospective, multicenter, observational cohort
trial evaluated whether the incidence of cardiovascular
or cerebrovascular events or all-cause mortality was
higher in patients with microalbuminuria compared
to patients with normoalbuminuria among 3,529
nondiabetic, hypertensive patients receiving treatment
with an ACE inhibitor.3 Patients were followed for
an average of 43 months. At the beginning of the
study, 34% of the patients had microalbuminuria. All
patients, regardless of microalbinuria status, were
given ramipril (1.25–2.5 mg) as treatment. Ramipril
could be used alone or in conjunction with other
antihypertensive agents at the treating provider’s
discretion to reach target blood pressure goals. Patients
were then monitored prospectively for the development
V O L U M E 18
NUMBER 5
M AY 2015
of cardiovascular disease and cerebrovascular disease
as well as all-cause mortality as the endpoints.
The risk of cerebrovascular endpoints was higher in
patients with microalbuminuria at baseline than among
patients with normoalbuminuria (HR 1.5; 95% CI,
1.2–2.0).3
Lynsey Drew, DO
Sarah Daly, DO
Intermountain Healthcare, Utah Valley FMR
Provo, UT
1. Jensen JS, et al. Hypertension. 2000; 35(4):893–903. [STEP 2]
2. Nakamura S, et al. Hypertens Res. 2003; 26(8):603–608. [STEP 2]
3. Schrader J, et al. J Hypertens. 2006; 24(3):541–548. [STEP 2]
Should patients with asthma be routinely
screened for depression?
Evidence-Based Answer
The practice might be worth considering. Rates of
depression are about 2 times higher among adults and
adolescents with asthma than the general population.
Depression is associated with activity limitation in
adults with moderate to severe persistent asthma. It
is unknown if treatment of depression without other
intervention improves asthma control (SOR: C, crosssectional and meta-analysis of case-controlled studies).
A longitudinal cohort trial examined depressive
symptoms and risk factors for depression in 439 adult
patients with asthma.1 Data were collected via telephone
interview at baseline, 2 years, and 4 years. The 20-item
Center for Epidemiological Studies for Depression Scale
(CESD) was administered during each interview.
The prevalence of depression among the participants
was 17% at the first interview, 14% at the second
interview, and 15% at the time of the third interview.
By comparison, the estimated prevalence of depression
in the general adult population is approximately 7%.
The study dropout rate was 28% by the time of the
third interview. Individuals who scored positive for
depression at baseline were more likely to drop out (OR
1.8; 95% CI, 1.1–3.0).1
A meta-analysis and meta-regression compared
the prevalence of anxiety and depression among
adolescents with asthma with the prevalence of anxiety
Evidence-Based Practice / Vol. 18, No. 5
E-1
and depression among adolescents without asthma.2
Eligible studies included case-control, observational
cohort, or cross-sectional studies, with a minimum of
50 subjects, aged 13 to 18 years.
Analysis of the 8 eligible trials revealed a higher
prevalence of depression among adolescents with
asthma (n=3,564), compared with control adolescents
without asthma (n=24,884) (27% vs 13%; OR 2.1;
95% CI, 1.7–2.6).2
A retrospective cohort identified 1,812 adults
with moderate to severe persistent asthma (asthma
diagnosis for ≥12 months, and either 2 courses of oral
corticosteroids in the past 12 months or a shortacting beta-agonist and at least 2 additional longterm controllers) from the Greenfield Online Panel,
a web-based cohort of nearly 4 million individuals
weighted to match US Census demographics.3 The
individuals completed the 5-item Asthma Control
Test (an observational, cross-sectional questionnaire)
and self-reported additional information, including
comorbidities, healthcare use, and attitudes and
behaviors toward asthma. Each respondent also
completed a 14-question survey about type and degree
of activity limitations.
Twenty-five percent of the participants reported
depression. Among the survey participants, depression
was associated with outdoor activity limitation (OR
1.6; 95% CI, 1.1–2.2); physical activity limitation (OR
1.8; 95% CI, 1.3–2.5); and daily activity limitation
(OR 1.6; 95% CI, 1.0–2.4) when compared with study
patients who did not report depression.3
Jamie A. Ogden, MD
Laura Morris, MD MSPH
University of Missouri-Columbia
Columbia, MO
1. Katz PP, et al. Prim Care Respir J. 2010; 19(3):223–230. [STEP 1]
2. Lu Y, et al. Pediatr Allergy Immunol. 2012; 23(8):707–715. [STEP 2]
3. Haselkorn T, et al. Ann Allergy Asthma Immunol. 2010; 104(6):471–477. [STEP 1]
We invite your questions and feedback.
Email us at EBP@fpin.org.
E-2
Evidence-Based Practice / May 2015
Are cannabinoids safe and effective
for treatment of patients with rheumatoid
arthritis?
Evidence-Based Answer
Sativex, an oromucosal cannabinoid spray unavailable
in the United States, improves pain on movement, pain
at rest, and quality of sleep in patients with rheumatoid
arthritis; however, the improvements are small and may
not be clinically meaningful. Adverse effects including
dizziness, lightheadedness, dry mouth, nausea, and
falls are significantly more common with Sativex than
with placebo (SOR: C, small, low-quality RCT).
A Cochrane review identified 1 multicenter,
randomized, double-blind, parallel group trial that
examined the efficacy and safety of a cannabisbased medicine, Sativex (containing 2.7 mg
tetrahydrocannabinol and 2.5 mg cannabidiol per
activation), for 58 patients with rheumatoid arthritis
over 5 weeks of treatment.1 Patients were on a stable
regimen of prednisolone and NSAIDs for 1 month and
disease-modifying antirheumatic drugs for 3 months.
The average age of the patients was 63 years, and
79% (n=46) were female.
The daily dose in the last week of treatment was
5.4 sprays for the treatment group and 5.3 sprays for
the placebo group. Clinical assessments, conducted in
the morning, included pain on movement, pain at rest,
morning stiffness, and sleep quality assessed by a 0–10
numerical rating scale; the Short-Form McGill Pain
Questionnaire (SF-MPQ), which is 15 items on a 0–3
scale; and the 28-joint disease activity score (DAS28),
a calculated score based on sedimentation rate and
number of tender points. On all of these scales, a lower
score indicates fewer symptoms. The baseline score, the
average of the last 4 days of the 14-day baseline period
prior to treatment, was compared with the endpoint
score, the average of the last 14 days of the 5-week
treatment period.1
Statistically significant improvements were noted
for Sativex versus the placebo group in pain on
movement (median difference [MD] –0.95; 95% CI,
–1.8 to –0.02), pain at rest (MD –1.0; 95% CI, –1.9
to –0.18), quality of sleep (MD –1.2; 95% CI, –2.2 to
–0.14), DAS28 (MD –0.76; 95% CI, –1.2 to –0.28), and
SF-MPQ verbal pain at present (mean –0.72; 95% CI,
–1.3 to –0.14).2 No statistically significant improvement
was noted in morning stiffness, SF-MPQ total intensity
of pain, or SF-MPQ intensity of pain at present.
Adverse effects were more likely in the treatment
group (risk ratio [RR] 1.8; 95% CI, 1.1–3.0).2 The
most common adverse effects in the treatment group
were dizziness (26%, NNH=5), light headedness (10%,
NNH=17), dry mouth (13%, NNH=8), nausea (6%,
NNH=36), and falls (6%, NNH=16).1 There were no
withdrawals due to adverse effects in the treatment
group and 3 withdrawals in the placebo group (11%).
No serious adverse events were noted in the treatment
group. The type of placebo and blinding of placebo
was not described, contributing to the low quality of
the study. One of the authors was employed by the
pharmaceutical company that makes Sativex.
The NICE Clinical Guidelines, which is an evidencebased guideline for the management of rheumatoid
arthritis, does not mention cannibinoids.3
Nicholas Crowley, MD
Carin Reust, MD, MSPH
University of Missouri
Columbia, MO
1. Richards BL, et al. Cochrane Database Syst Rev. 2012; (1):CD008921. [STEP 1]
2. Blake DR, et al. Rheumatology (Oxford). 2006; 45(1):50–52. [STEP 2]
3.
National Collaborating Centre for Chronic Conditions. Rheumatoid Arthritis: National
Clinical Guideline for Management and Treatment in Adults. http://www.ncbi.nlm.nih.gov/
books/NBK51812/. Published February 2009. Accessed April 27, 2015. [STEP 1]
Is patient-controlled analgesia effective
for patients with sickle cell pain crisis?
Evidence-Based Answer
Patient-controlled analgesia (PCA) is no better than
continuous infusion or intermittent dosing of opioids
for pain control in patients with sickle cell disease
(SCD) experiencing vaso-occlusive crisis. PCA may lead
to lower cumulative doses of opioids than continuous
infusion (SOR: B, RCTs).
Two trials met inclusion criteria for a 2011 systematic
review of PCA versus intermittent dosing in the
treatment of SCD vaso-occlusive crisis, 1 looking at
meperidine and the other morphine.1
In one RCT (N=20 adults aged 17–39 years),
meperidine via PCA (25–30 mg/h) was compared with
intramuscular meperidine (75–100 mg) as needed every 3
to 4 hours. No significant difference in pain was noted.1
In the other RCT (N=45 adults aged 18–65 years),
2 different dosing regimens (low and high dose) were
used for 2 separate PCA versus intermittent intravenous
treatment groups. No significant difference was noted
in a 100-mm pain scale between either the low- or highdose PCA groups compared with intermittent dosing
(WMD –0.10 mm; 95% CI, –27 to 27 mm and WMD
9 mm; 95% CI, –18 to 36 mm, respectively).
Additionally, no significant difference was noted in the
amount of meperidine or morphine used between the
PCA and control groups in each of the studies.1
A 2007 RCT of 25 episodes of vaso-occlusive crisis
in 19 patients with SCD between 20 and 42 years old
compared intravenous morphine administration via
PCA versus continuous infusion.2 A significantly lower
mean and total cumulative morphine consumption was
noted in the PCA group (0.5 vs 2.4 mg/h; P<.001; and
33 vs 260 mg; P=.018). Based on a verbal response scale
(0=no pain, 10=worst pain), mean daily pain scores
were comparable (4.9 vs 5.3; P=.09).
Great Britain’s 2012 National Institute for Health
and Clinical Excellence (NICE) guideline included the
2007 trial above as well as one from 1991.3 NICE
guidelines are systematically developed based on the
best available evidence and expert consensus opinion
and undergo surveillance for updating every 2 years.
The guideline concluded that PCA should be
considered in patients who require repeat dosing of
opioids. PCA was also noted to have small additional
health improvements of between 0.002 and 0.003
quality-adjusted life-years per person as well as lower
hospital costs (between £170 and £1,329 [$110 and
$863 in US dollars, based on IRS average conversion
rates in 2011]), which seem correlated to shorter
hospital stays (between 0.39 and 2.8 days).3
Timothy Mott, MD, FAAFP
Timothy Algiers, MD
Carolyn Gosztyla, MD
Naval Hospital Pensacola FMR
Pensacola, FL
The views expressed in this article are those of the author and do not
necessarily reflect the official policy or position of the US Department
of Navy, Department of Defense, or the US Government.
1.Meremikwu MM, et al. Clin Evid (Online). 2011; 2:2402. [STEP 1]
2.van Beers EJ, et al. Am J Hematol. 2007; 82(11):955–960. [STEP 3]
3.National Institute for Health and Clinical Excellence. Sickle Cell Acute Painful Episode:
Management of an Acute Painful Sickle Cell Episode in Hospital. NICE Clinical Guidelines, no. 143. http://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0054575/. Published
June 2012. Accessed April 27, 2015. [STEP 5]
Evidence-Based Practice / Vol. 18, No. 5
E-3
Does routine testing for tuberculosis (TB) in lowrisk areas help to reduce the incidence of TB?
Evidence-Based Answer
Probably not. Routine screening of the general population
(including low-risk areas) to reduce the incidence of
TB has a number needed to screen (NNS) to prevent 1
case of active TB of 8,000 to 25,000 patients (SOR:
C, population models). Current guidelines recommend
targeted testing in elevated-risk populations in whom
the NNS is lower (SOR: B, national evidenced-based
guidelines).
A 2009 study created a population model to estimate
the benefits of screening for latent tuberculosis infection
(LTBI) with the tuberculin skin test (TST).1 The model
used data from the National Health and Nutrition
Examination Survey (NHANES) to determine the
prevalence of LTBI and reactivation rate to active TB
within the United States. The model used data from
other published reports to determine inputs for TST
sensitivity and specificity along with isoniazid therapy
effectiveness, adverse effects, and compliance. In
US-born residents, the NNS to prevent 1 case of active
TB was 8,333 to 25,000, whereas in foreign-born
residents the NNS was 319 to 1,216.
A 2011 study updated the population model
discussed above with data from the Centers for Disease
TABLE
Control and Prevention (CDC) TB surveillance reports,
NHANES III, and the US Census in order to estimate
the benefits of screening US residents with risk factors
for LTBI.2 The new model also evaluated screening with
interferon-gamma release assays (IGRA) and accounted
for the financial costs of screening and treatment. LTBI
prevalence, reactivation rate, and NNS for TST and
IGRA are shown in the TABLE .
A 2005 joint publication from the CDC, American
Thoracic Society, and the Infectious Diseases Society of
America recommends targeted testing for LTBI as shown
in the TABLE (Grade AII–strong recommendation based
on moderate-quality evidence).3 In their expert opinion,
generalized screening in low-risk areas allocates limited
resources away from areas of higher need.
Dennis J. Gerold, Jr, MD
Philip T. Dooley, MD
Eglin AFB Family Medicine Residency
Eglin AFB, FL
The opinions and assertions contained herein are those of the authors and
are not to be construed as official or as reflecting the views of the US Air Force
Medical Department, the US Air Force at large, or the Department of Defense.
1. Linas BP, et al. Am J Respir Crit Care Med. 2009; 179:A5283. [STEP 4]
2. Linas BP, et al. Am J Respir Crit Care Med. 2011; 184(5):590–601. [STEP 4]
3.American Thoracic Society, et al. Am J Respir Crit Care Med. 2005; 172(9):1169–1227.
[STEP 2]
4.ATS/CDC Statement Committee on Latent Tuberculosis Infection. Am J Respir Crit Care
Med. 2000; 161(4 pt 1):1376–1395. [STEP 2]
Number needed to screen to prevent 1 case of active TB1–4
TB reactivation
Targeted screening
(rate per 100 person-years)2
NNS with TST2
NNS with IGRA2
recommended2-4
Risk group
LTBI prevalence (%)2
Infected with HIV
5.3
2.1
71
67
Yes
Close contact adult
44
1.0
73
69
Yes
Close contact child
20
1.6
110
104
Yes
Recent immigranta adult
41
0.079
136
128
Yes
Foreign born living in the United States >5 years
6–14 years old
11
0.079
380
358
No
15–24 years old
12
0.079
450
424
No
25–44 years old
21
0.079
319
301
No
45–64 years old
27
0.079
363
343
No
≥65 years old
11
0.079
1,216
1,147
No
Homeless
290.079 436 343 Yes
Injection drug user
23
0.079
557
525
Yes
Recent immigranta child
7
0.079
617
582
Yes
Former prisoner
19
0.079
655
618
Yes
Immunosuppressive drugs
5.3
0.16
1,128
1,064
Yes
Underweight
2.8 0.12 2,0621,946 Yes
Gastrectomy
2.8 0.10 2,1762,054 Yes
Silicosis 2.8
0.10
2,962
2,795
Yes
Diabetic
2.8 0.13 3,6313,426 Yes
End-stage renal disease
2.8
0.19
4,904
4,628
Yes
US born resident
0.2–0.8
0.036
8,333 to 25,000
ND
No
1
with no risk factors
HIV=human immunodeficiency virus; IGRA=interferon-gamma release assay; LTBI=latent tuberculosis infection; ND=no data; NNS=number needed to screen; TB=tuberculosis; TST=tuberculin skin test.
aForeign born US residents living in the United States for ≤5 years.
E-4
Evidence-Based Practice / May 2015
Is biofeedback an effective treatment
for essential hypertension?
Evidence-Based Answer
Biofeedback has not been shown to significantly lower
systolic blood pressure (SBP) or diastolic blood pressure
(DBP) in patients with essential hypertension (SOR: C,
disease-oriented outcomes).
A 2007 systematic review and meta-analysis of
17 RCTs examined the effectiveness of stress reduction
programs in 960 patients with elevated SBP or DBP.1
Changes were compared from baseline measurements
to ≥8 weeks after each respective stress reduction
program began.
No difference was noted in SBP/DBP with simple
biofeedback (6 trials, N=300; –0.8/–2.0 mmHg;
P=not
significant
[NS]),
relaxation-assisted
biofeedback (4 trials, N=98; –1.9/–1.4 mm Hg;
P=NS), or stress management training (5 trials, N=207;
–2.3/–1.3; P=NS). Transcendental meditation programs
demonstrated a significant decrease in SBP, but
not DBP (6 trials, N=449; SBP –5 mmHg, 95% CI,
–9.4 to –0.8 mmHg; DBP –2.1 mmHg, 95% CI, –5.4
to 1.4).1
A 2010 systematic review of 36 RCTs (N=1,660)
examined the effectiveness of biofeedback for
the treatment of essential hypertension (SBP
>140 mmHg or DBP >90 mmHg) in adults.2 Twentyone trials used biofeedback with no adjunctive
antihypertensive therapy, whereas the other 15 studies
used biofeedback with antihypertensive therapy.
Patients with biofeedback training were compared with
groups undergoing antihypertensive therapy alone,
placebo (sham biofeedback), no treatment, or other
behavior treatments. These groups were studied for
<12 months.
Due to the overall poor quality of the included trials
and the heterogeneity of outcomes measured, only a
narrative summary using the trial authors’ conclusions
was reported. The authors of the review found no
results that consistently showed the effectiveness of
biofeedback for the treatment of essential hypertension
EBP
compared with the other therapies.2
Matthew C. Kelly, MD
Drew C. Baird, MD
Darnall Army Medical Center
Fort Hood, TX
The opinions and assertions contained herein are those of the
authors and are not to be construed as official or as reflecting the views of the
US Army Medical Department, the US Army at large, or the Department of Defense.
1. Rainforth MV, et al. Curr Hypertens Rep. 2007; 9(6):520–528. [STEP 1]
2. Greenhalgh J, et al. J Hypertens. 2010; 28(4):644–652. [STEP 1]
Evidence-Based Practice / Vol. 18, No. 5
E-5